Language, Knowledge and Pedagogy: Functional Linguistic and Sociological Perspectives

Language, Knowledge and Pedagogy

This page intentionally left blank

Language, Knowledge and Pedagogy Functional Linguistic and Sociological Perspectives

Edited by Frances Christie and J.R. Martin

Continuum The Tower Building 11 York Road London SE1 7NX

80 Maiden Lane Suite 704 New York, NY 10038

© Frances Christie, J.R. Martin and contributors 2007 All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. Published in paperback 2008 ISBN: 978-1-8470-6572-8 Library of Congress Cataloging-in-Publication Data To

Typeset by Kenneth Burnley, Wirral, Cheshire Printed and bound in Great Britain by MPG Books, Bodmin, Cornwall

For Basil Bernstein, Michael Halliday and Ruqaiya Hasan, in honour of the conversations they pursued.

This page intentionally left blank

Contents

List of tables List of figures Contributors

viii ix xi

Section 1: Theoretical Foundations 1 Ongoing dialogue: functional linguistic and Bernsteinian sociological perspectives on education Frances Christie

3

2 A sociology for the transmission of knowledges Karl Maton and Johan Muller

14

3 Construing knowledge: a functional linguistic perspective J.R. Martin

34

4 On splitting hairs: hierarchy, knowledge and the school curriculum Johan Muller

65

5 Knowledge–knower structures in intellectual and educational fields Karl Maton

87

6 Hierarchical knowledge structures and the canon: a preference for judgements Rob Moore

109

Section 2: Fields of Discourse – Disciplines of Discourse 7 Language for learning in early childhood Clare Painter

131

8 Building verticality in subject English Frances Christie and Mary Macken-Horarik

156

9 Vertical and horizontal discourse and the social sciences Peter Wignell

184

10 Mathematical and scientific forms of knowledge: a systemic functional multimodal grammatical approach Kay O’Halloran

205

Section 3: Research Prospects – Exploring Uncommon Sense 11 Taking stock: future directions in research in knowledge structure Frances Christie, J.R. Martin, Karl Maton and Johan Muller Index

239 259

List of tables

2.1 3.1 4.1 8.1 8.2

The arena of the pedagogic device Factoring modernist history as genre Maths content for grade 6 Models of school English in contemporary schools Semantic features with examples

18 58 81 168 178

List of figures

3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 4.1 5.1 5.2 5.3 5.4 5.5 7.1 7.2 7.3 7.4 7.5 9.1 9.2 10.1a 10.1b 10.1c 10.2 10.3 10.4 10.5 10.6a 10.6b 10.6c

Strata and metafunctions in SFL Social context as register and genre Classification taxonomy (types of rock) Compositional relations (volcanoes) Common and uncommon sense fields Genealogical organization of living things Cochlear ear implant Narrative image illustrating the cold front explanation Stratal harmony – grammar matching semantics Grammatical metaphor as stratal tension Projection relations in pedagogic discourse Typology of history genres Kepler’s star tower The two cultures as knowledge structures and knower structures Legitimation codes of specialization for the two cultures Legitimation codes of specialization University students’ survey University students’ perceptions of basis of achievement Context-bound language for categorizing ‘Self-contextualised’ language for categorizing Grammatically congruent expression of meaning Defining clause with a metaphorical participant A definition involving grammatical metaphor Taxonomy of value (Smith and Ricardo) Taxonomy of value (Marx) Bollmann map of New York Stable city landscape Horizontal and vertical lines Scientific experiment with frogs Newton: space, time, matter and prediction Newton: revisiting the moment mathematically Newton: suspending the instant Suspending the visual image: Newton Suspending the visual image: Prigent Suspending the visual image: Golovin

35 35 36 37 37 39 40 42 52 53 57 59 76 92 94 97 101 102 141 141 146 146 147 191 195 208 208 209 217 218 220 222 223 224 225

This page intentionally left blank

Contributors

Frances Christie is Emeritus Professor of Language and Literacy Education, the University of Melbourne, and Honorary Professor of Education, University of Sydney. She has research interests in Language and Literacy Education generally, including language curriculum theory and pedagogy, the history of English teaching, oral language and literacy development across the years of schooling, pedagogic grammar and classroom discourse analysis. Her publications have included (edited with J. R. Martin), Genre and Institutions: Social Processes in the Workplace and School, Cassell 1997; (edited with with Ray Misson), Literacy and Schooling, Routledge 1998; Pedagogy and the Shaping of Consciousness: Linguistic and Social Processes, Cassell 1999; Classroom Discourse Analysis. A Functional Perspective, Continuum 2002; Language Education in the Primary Years, University of NSW Press 2005. Mary Macken-Horarik is a Senior Lecturer in the School of Teacher Education within the Division of Communication and Education at the University of Canberra. She has worked for many years in the field of English and literacy education as a teacher, curriculum developer and researcher, with interests in teaching English as a mother tongue and as a second language. Mary has published widely in the field of systemic functional linguistics. Recent publications include ‘Negotiating Heteroglossia’ (a special issue of Text edited with J.R. Martin), Mouton de Gruyter 2003. Her other research interests include the analysis of curriculum in school English, academic literacy education and multimodal discourse analysis. She is currently writing a book for English teachers about systemic functional semiotics. J.R. Martin is Professor of Linguistics (Personal Chair) at the University of Sydney. His research interests include systemic theory, functional grammar, discourse semantics, register, genre, multimodality and critical discourse analysis, focusing on English and Tagalog – with special reference to the transdisciplinary fields of educational linguistics and social semiotics. Recent publications include Working with Discourse (with David Rose), Continuum 2003; Re/Reading the Past (edited with Ruth Wodak), Benjamins 2003; ’Negotiating Heteroglossia’ (a special issue of Text edited with Mary Macken-Horarik), Mouton de Gruyter 2003; Language Typology: A Functional Perspective (edited with A. Caffarel and C. Matthiessen), Benjamins 2004; and ‘Interpreting Tragedy: the Language of September 11th, 2001’ (a special double issue of Discourse and Society edited with John Edwards), Sage 2004; The Language of Evaluation (with Peter White), Palgrave 2005. He has recently completed a book on genre (with David Rose, in

xii

CONTRIBUTORS

press with Equinox). Professor Martin was elected a fellow the Australian Academy of the Humanities in 1998, and awarded a Centenary Medal for his services to Linguistics and Philology in 2003. Karl Maton is Senior Research Fellow in Sociology and Education at the University of Wollongong, Australia. He recently completed his doctoral thesis at the University of Cambridge, ‘The field of higher education: A sociology of reproduction, transformation, change and the conditions of emergence for cultural studies’, which develops Bernstein’s approach to create a sociology of higher education (http://www.KarlMaton.com). Karl has published in sociology, cultural studies, philosophy, linguistics and education, is co-editing a volume on cultural studies and global education, and is currently engaged in collaborative research into IT education, special educational needs, school music and the sociology of health. Rob Moore is Senior Lecturer in Sociology of Education in the Faculty of Education of the University of Cambridge, UK. He is a fellow of Homerton College and Director of Studies for Social and Political Sciences. His work includes research into youth and the labour market, educational decision-making, and issues of citizenship among school pupils and their parents and undergraduates (with John Beck and John Ahier). His current work is concerned with problems in the sociology of knowledge and the structuring of intellectual fields. His most recent book is Education and Society: Issues and Explanations in the Sociology of Education (Polity 2004). Johan Muller holds the chair of Curriculum at the University of Cape Town. He is currently the Director of the Graduate School of Humanities and Deputy Dean for Research and Postgraduate Affairs in the Faculty of Humanities. He has published in the areas of curriculum theory and policy, and in the sociology of knowledge. His book Reclaiming Knowledge was published by RoutledgeFalmer in 2000. Kay O’Halloran is Associate Professor in English linguistics and multimodality at the National University of Singapore. Her teaching and research interests include multimodality; systemic functional linguistics (SFL); mathematical and scientific discourse; multimodal approaches to hypertext, film and classroom discourse; digital media, visualization and technology. Her research interests extend to the development of software for linguistic and multimodal analysis. She has presented invited papers on many of these topics in international congresses in many parts of the world. She produced Systemics 1.0 software in 2002 with Kevin Judd, and recent publications include Multimodal Discourse Analysis (as editor), Continuum 2004, and Mathematical Discourse: Language, Symbolism and Visual Images, Continuum 2005. Clare Painter is a Senior Lecturer in the School of English at UNSW, Australia. She has a longstanding research interest in systemic-functional (SF) linguistics in relation to children’s language and literacy development and is the author of

CONTRIBUTORS

xiii

a number of articles and books in the field, including Learning Through Language in Early Childhood, Continuum 2000. Recent work in SF theory focuses on the semiotics of children’s picture books and informs her forthcoming book, Reading Visual Narratives, to be published by Equinox. Peter Wignell is Head of School of the Centre for Access and English as a Second Language at Charles Darwin University in the Northern Territory of Australia. He has researched in the fields of linguistics, language and literacy education for the past 20 years and is currently working on a book on the evolution of the language of the social sciences.

This page intentionally left blank

Section 1

Theoretical Foundations

This page intentionally left blank

1

Ongoing dialogue: functional linguistic and Bernsteinian sociological perspectives on education Frances Christie

Introduction Over the last 45 to 50 years, systemic functional linguistic (SFL) theory and Bernsteinian sociology have engaged in fruitful dialogue, exploring a number of educational questions, each enhancing what the other discipline has to offer. It was perhaps an unlikely dialogue between two rather different disciplines, though those most instrumental in forging what became a very productive alliance had more in common than they might first have realized. The two figures most fundamentally involved in the genesis of the dialogue that emerged were the linguist M.A.K. Halliday and the sociologist Basil Bernstein, though they were soon joined by the linguist Ruqaiya Hasan, who went to work with Bernstein in the 1960s. Halliday and Bernstein had both served in the forces in the Second World War, albeit in rather different capacities, and in the aftermath of the war both strove to make a difference in their respective areas of work by addressing issues of social justice, Bernstein initially as a school teacher, later a sociologist, Halliday as a teacher of Chinese, later of linguistics. While in the 1950s Bernstein was troubled by issues of disadvantage when observing how working-class children performed badly in London schools, Halliday, teaching at Cambridge, was an active participant in the Linguistics Group of the British Communist Party, and like his colleagues he had a passion to use his linguistics to make a difference in exploring the nature of social experience and in addressing questions of equity and social justice. In these ways Halliday and Bernstein thus had much in common, though there is at least one other sense in which the two were to share a common fate. It was that each was a remarkably original thinker and, as such, both were destined to be often misrepresented and misunderstood. Halliday (1986:1) once said that ‘one of the problems with Bernstein was that he never fitted people’s stereotypes, their ready-made categories into which all thinkers are supposed to fit. So the left branded him as right-wing and the right-wing branded him as left-wing.’ While Halliday never made any secret of his left-wing sympathies, he was, like Bernstein, often misrepresented and misunderstood, because his linguistics has never fitted what have been regarded as mainstream traditions of linguistics. Unlike linguists in more formalist traditions, he developed a linguistic paradigm said to be ‘extravagant’ principally because it engages with meaning and function as well as structure (thereby allowing a degree of messiness that formal traditions don’t accept), and builds a strong sense of the social into accounts of language.

4

LANGUAGE, KNOWLEDGE AND PEDAGOGY

By the time of his death in 2000, Bernstein’s work had begun to receive a much needed reassessment (e.g. Atkinson et al. 1995; Sadovnik 1995; Christie 1999; Morais et al. 2001; Muller et al. 2004; Hasan 2005), and numbers of scholars (some of whom contribute to this volume) continue to work with and pursue his theories and ideas. As for Halliday, he remains productive, while recent publications bear testimony to the major position he now occupies and the major contribution he has made (e.g. Hasan et al. 2005; The Collected Works of M.A.K. Halliday, edited Webster, from 2002). Thus, while Bernstein and Halliday both emerged as significant intellectuals in the often troubled years of the mid to later twentieth century, their scholarly contributions continue to be relevant and timely in the equally troubled years of the early twenty-first century. This book, then, seeks to maintain and extend the dialogue between two important traditions of scholarship by pursuing some significant educational issues. Of course both Bernstein and Halliday pursued many questions other than educational ones. Bernstein regarded himself primarily as a sociologist, though his was a Chair in the Sociology of Education, and Halliday was one of the two most important linguists of the second half of the twentieth century. At the same time, educational questions and issues have always been central to their respective concerns, even where teaching and learning are not at the forefront of attention. For Halliday, educational processes are part of the very warp and weft of life, and educational sites constitute major contexts for close analysis of language in use. In addition – a lifelong preoccupation of his – language is an essential resource in all teaching and learning. For Bernstein, pedagogic activity became over time a major preoccupation, causing him to propose a social theory regarding cultural transmission and the manner in which different social groups encode and communicate significant meanings and values. While both theorists thus worked ‘beyond the classroom’, as it were, their concerns as linguist and as sociologist have remained bound up with questions of educational activity in the broadest sense. The concerns of this volume are with the nature of knowledge, a topic that intrigued Bernstein in the last years of his life, though as Moore and Maton (2001: 154–5) have argued, this represented a necessary development upon the earlier work that Bernstein had undertaken over many years. For Bernstein (e.g. 1990: 165–218), as he reflected on sociology of education dating from the 1970s, what was missing in the various accounts available was any sense of the structuring of knowledge. Sociology of education, he argued, had become interested in matters to do with ‘relations (of educational knowledge) to’ other things, such as class, race or gender. It had lost sight of what he termed alternatively ‘relations within’ knowledge, and thus had little to say about the nature of the knowledge transmitted in pedagogic enterprises. What was needed was the development of a theory that would deal with the structuring of knowledge itself. The challenge he thus offered to sociologists was a bold one, and a very timely one, for reasons to be developed in this volume. To begin, however, it will be useful to sketch a little of the history of the emergence of the two traditions of scholarship involved here, and the sense in which they have joined together in common cause, while not overlooking the different methodologies each employs.

LINGUISTIC AND BERNSTEINIAN PERSPECTIVES

•

CHRISTIE

5

Developing a dialogue Halliday and Bernstein did not meet until the early 1960s, when Halliday was teaching linguistics at Edinburgh University and Bernstein was at the University of London. Halliday’s interest in linguistics developed from his work on Chinese, which he taught at Cambridge, moving to linguistics when he went to Edinburgh. While he acknowledged a great debt to Firth in developing his first accounts of linguistic theory, he later observed that the work he and his colleagues did with teachers while at Edinburgh was very significant (Halliday and Hasan, in press) in shaping his early thinking both about language, and about language education. In a subsequent move he made to the University of London he was invited to develop an English curriculum proposal for the Nuffield Foundation, and this led in time to the Nuffield/Schools Council Programme in Linguistics and English Teaching (1964–71). The latter programme produced extensive sets of curriculum materials for schools and resource books for teachers, all of which were influential in their day, and some of which remain in use. (See Hasan and Martin 1989 and Christie and Unsworth 2005.) Early work on a theory of grammar (e.g. Halliday, 1976, 2002) led to a model of grammar as systemic and functional. It is systemic, in that it offers sets of options or choices (not conscious) for making meaning; the choices provide a large network of systems of choices, each with an entry condition. In practice, simultaneous choices are said to apply when a language is used; in creating a clause, for example, one makes choices for transitivity, mood and theme. Language is functional, it was said, in that it serves basic human functions to represent experience and knowledge, to construct relationship, and to create meaningful messages. The functional aspect of the theory evolved into a theory of metafunctions which operate across all natural languages: the ideational, interpersonal and textual. The structures in which these metafunctions are realized are particular to any given language. Hence, in parsing and labelling of language, two kinds of labels apply, functional and class. Class labels in English, for example, identify nouns, verbs, adverbs and so on, while functional labels identify, say, for transitivity, participant, process and circumstance, and in each case, the type of participant, process and circumstance is also specified. And what of context? Working with notions of context of situation and later of context of culture, taken from Malinowski (discussed by Halliday in Halliday and Hasan 1985) Halliday and his colleagues (Halliday et al. 1964) pursued a theory of register variation: for any context of situation, language choices are in part a condition of the social activity (field), the nature of the relationships (tenor) and the role that language plays (mode). Text and context are mutually intelligible, for the one constructs the other. One of the many remarkable contributions that Halliday and his colleagues made in the developing theory of language was to propose a ‘hook up’ between grammar and register. Some sets of choices relate primarily to activity, some primarily to the nature of the relationship in construction, and others primarily to organization of the language as message. A theme to which Halliday has returned often is his interest in language as a social phenomenon. He once noted (Parret 1974) that when he studied with

6

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Firth, linguistics was ‘the study of language in society’, and he regretted the development of the term ‘sociolinguistics’, as he later regretted the emergence of ‘pragmatics’ (Halliday 2002: 8). For Halliday, the study of language necessarily involved the study of language in use, as a part of the social system. What mattered about language was what people did with it. It was this concern for language as social phenomenon, functioning to mediate social processes in profound ways, which aroused the interest of Halliday – and that of his colleague Hasan – in the early work of Bernstein. Bernstein, troubled, as we noted above, by the apparent educational failure of many working-class children, was seeking some explanation that would allow him to make a difference. While he was not sure what kind of language theory he needed, he early realized he must account for the apparent differences in uses of language that he and his colleagues observed between working-class and middle-class children; he cast about among the available linguistic theories, including various early formulations of sociolinguistics (see Bernstein 2000: 145–53). His early formulations (e.g. 1971) led him to propose the operation of restricted and elaborated language codes. Possession of the elaborated code enabled the speaker to select from a relatively wide range of meanings, allowing expression of personal attitude among other matters, as well as the capacity to expand upon and explain experience, and this code was rewarded for the purposes of school learning. The restricted code was associated with meanings that were more localized and immediate to situations than the elaborated code, and in consequence they were less amenable to expansion and explanation of experience of a kind that schooling tended to reward and encourage. The codes were acquired as part of socialization within one’s family, and while working-class children tended to use the restricted code, both codes were available to middleclass children. The learning of a code was profoundly important, for it involved learning the requirements of the social structure (ibid.: 124). Code theory, intimately linked as it was to theories of social class, led in time to a more extensive theory addressing cultural transmission. By 1981, Bernstein was to say that a code was ‘a regulative principle, tacitly acquired, which selects and integrates relevant meanings, forms of realizations and evoking contexts’. Hasan (1973) explored the differences between code, register and social dialect, making clear how code was to be distinguished from the other two. Dialects represented patterns of language spoken according to social and/or geographical location, while registers were distinguished by their use in different contexts of situation; but codes must be recognized for their semantic properties, for they had to do with orientations to meaning. Later on, in developing and enriching discussions of the coding orientations associated with different social groups, Hasan (1989) proposed the term ‘semantic variation’ to refer to ‘systematic variation in the meanings people select in similar contexts as a function of their social positioning’ (Williams 2005: 457). With colleagues who included Cloran (1994, 1999; Hasan 1986; Hasan and Cloran 1990) and Williams (1999, 2001) Hasan provided significant evidence to support Bernstein’s original claims about the manner in which meaning orientations functioned as a condition of social – more specifically family – positioning. Hasan collected natural discourse between mothers and their pre-school

LINGUISTIC AND BERNSTEINIAN PERSPECTIVES

•

CHRISTIE

7

children drawn from ‘high autonomy professionals’ (HAPs) and ‘low autonomy professionals’ (LAPs); using delicately developed semantic networks based on SF linguistics, she was able to show systematic differences in the coding orientations of mothers and children from the two social groups. Such detailed analysis using the functional grammar had not been possible at the time Bernstein had advanced his proposals in the late 1960s and 1970s, for the theory was insufficiently developed at that time. Bernstein (2002: 128–30) acknowledged a significant debt to Hasan and her colleagues. Code theory was accompanied in the 1970s by other and related developments in identifying models of school knowledge, such that distinctions were made between ‘collection’ and ‘integrated’ codes in the school curriculum, and between ‘visible’ and ‘invisible’ pedagogies (Bernstein 1975). Both distinctions served to extend and enrich his discussion of the nature of cultural transmission and the processes by which knowledge was shaped and pedagogic subject positions were formed. Indeed, code theory and its corollaries were subsumed into a more comprehensive theory of cultural transmission in which pedagogic activity achieved significance in the 1980s, when Bernstein turned his attention to the production of knowledge, and to the ‘symbolic control’ that was a feature of pedagogy. For Bernstein (1990: 134–41), pedagogic relationships embraced more than those found in educational settings, while there were many agents of symbolic control other than teachers, such as architects, scientists and psychologists. Pedagogy, he argued, functioned through the operation of a pedagogic device. This was instrumental in creating pedagogic discourse, which was defined as ‘a principle for appropriating other discourses and bringing them into a special relation with each other for the purposes of their selective transmission and acquisition’ (Bernstein, ibid.: 183–4). Pedagogic discourse takes the discourses of practice in identified sites (such as universities) and relocates or recontextualizes these for other purposes (such as teaching in schools). The pedagogic device works through distributive rules (which distribute forms of consciousness through the distribution of forms of knowledge); recontextualizing rules (which regulate the construction of specific instructional discourses for specific pedagogic purposes) and evaluative rules (which provide the criteria used to legitimate the forms of knowledge communicated). Overall, Bernstein argued, pedagogic discourse serves to shape consciousness, differentially distributing knowledge and experience. (See Maton and Muller, this volume.) Yet as Bernstein had also determined by the end of his life, the various discussions of pedagogy offered by sociologists (including his own) had had little to say of the nature of the knowledge communicated in pedagogic activities (Bernstein 2000: 155–74). He therefore began to develop his discussions of knowledge, proposing a theory of horizontal and vertical discourses, out of which a theory of knowledge structure emerged. Horizontal discourse is segmentally organized, in that its various forms are localized and particular to specific activities, creating what is often thought of as ‘common sense’ knowledge. Vertical discourse, by contrast, is not segmentally organized, for it ‘takes the form of a coherent, explicit and systematically principled structure’, and it is ‘hierarchically organized’ as in the sciences, or consists of ‘a series of specialized

8

LANGUAGE, KNOWLEDGE AND PEDAGOGY

languages’, of the kind found in the social sciences or humanities’ (ibid.: 157). Vertical discourse could be understood in terms of the operation of two knowledge structures: horizontal and hierarchical. Horizontal knowledge structures (e.g. English literature, philosophy) are apparent in the presence of a series of ‘languages’, possessed of ‘grammars’, some stronger or weaker than others; such structures by their nature are not easily related and they are often incommensurable. Hierarchical knowledge structures (e.g. physics) by contrast involve forms of knowledge which attempt to ‘create very general propositions and theories, which integrate knowledge at lower levels’ and which show ‘underlying uniformities across an expanding range of apparently different phenomena’. In this sense, they are produced using an integrating code (Bernstein, ibid.: 161). Overall, there is an internal logic which shapes the hierarchical knowledge structure, giving it a unity in terms of which new knowledge claims may be tested. (See Muller and Maton in this volume.) These discussions of vertical discourse are both powerful and interesting, but Bernstein was aware that the subject was by no means fully explored by the time he died. A great deal more work remains to be done in refining what is meant by the notion of vertical discourse, and in particular in refining the nature of horizontal and hierarchical knowledge structures. This is the case for at least two related reasons. Firstly, the twentieth century saw the appearance of many new forms of knowledge, especially at university level, having consequences for the ways knowledge in general was understood, though it also had consequences for the ways knowledge was later recontextualized in schools. Many once orthodox areas of knowledge were challenged, while the relationships of the various areas of inquiry available were often poorly defined. Not all new forms of knowledge were of equal merit, as Moore and Muller (1999) have argued with respect to sociology of education, and Maton (2000) has argued with respect to cultural studies, as well as sociology of education. The second and related reason why there is an urgent need to pursue further work on the nature of knowledge structures concerns the claims of the school curriculum. The steady advance of compulsory education led to a great deal of theorizing both about how knowledge should be conceived for school purposes, and about the most appropriate pedagogies to support its transmission. Curriculum theorizing throughout the twentieth century was significantly influenced by various progressivist and constructivist theories of knowledge and of the learner, whose effect was to diminish the status of knowledge structures, as well as the role of the teacher (e.g. Muller 2002; Christie 2002, 2004). Currently, in many parts of the English-speaking world there is considerable uncertainty, both about what forms of knowledge should be taught, and about the most appropriate teaching practices. Bernsteinian theory and SFL theory can together usefully address the problem. Just as the sociology of Bernstein and his colleagues was extended and enriched over the last several decades, so too was SFL theory. Halliday’s version of the functional grammar appeared in 1985, to be followed by a second edition in 1994, while the third edition (2004) was produced with his colleague Matthiessen. Various discussions of the SF grammar were in time produced by others, while several versions for teaching had appeared by the end of the

LINGUISTIC AND BERNSTEINIAN PERSPECTIVES

•

CHRISTIE

9

century (see Christie and Unsworth 2005, for some discussion). Alongside these developments Martin and his colleagues developed earlier work on cohesion (Halliday and Hasan 1976) as a stratum of semantics, mediating between grammar and context (Martin 1992; Martin and Rose 2003; Martin and White 2005). This contributed to the development of significant work on register theory, which had been outlined by Halliday, McIntosh and Strevens (ibid.) as referred to above, though many others later contributed, offering more than one framework (e.g. Benson and Greaves 1973; Hasan 1973; Halliday 1973; Gregory and Carroll 1978; Halliday and Hasan 1985; Hasan 1995; Matthiessen 1993; Martin 1992). The rival claims of the various accounts of register will not be addressed here. Suffice to note that probably the most influential of the models of register, at least for educational theory, was that of Martin (1992). His work, developed initially with Rothery (see Martin and Rose 2003, in press) offered an account of register and of genre. Text types or genres are said to differ, depending on the immediate context of situation (field, tenor, mode), and the context of culture (genre systems). Work in the tradition of genre-based research has led to many discussions of ‘subject specific literacies’ (e.g. MackenHorarik 1996, 1998; Unsworth 2000; Schleppegrell 2004; Martin 1999, this volume). Such work, stimulated in addition by the research of Halliday into differences between speech and writing (Halliday 1985) and grammatical metaphor (e.g. Simon-Vandenbergen, Taverniers and Ravelli 2003), has produced extensive accounts of different types of genres, spoken and written. An accompanying theory of pedagogy has also emerged, some representative accounts being given by Feez and Joyce (1998), Martin (1999), Christie (2002, 2005), and Martin and Rose (2005). Paralleling these developments has been important associated work on a language-based theory of learning (Halliday 1993, 2003) to which colleagues such as Painter (1999, this volume) and Torr (1998) have contributed. A significant additional development has been the entry into transdisciplinary discussion with scholars working on the human brain and on the emergence of language in the human species (e.g. Williams and Lukin 2004; Hasan 2005: 21–156). This promises to be a productive direction for future research.

About this book Systemic functional theory has had a strong tradition of working in dialogue with other areas of scholarship, and such transdisciplinary work has many merits, not least that many of the most significant problems worthy of examination require more than one scholarly tradition to do them justice. This volume aims to continue the transdisciplinary work that has been established involving both SFL theory and Bernstein’s sociology. The volume is transdiciplinary, in that each area of scholarship seeks to inform the other, in a common concern to pursue the nature of knowledge structures. Neither tradition tries to subsume the other, nor do the two overlook their methodological differences, including differences in uses of terms. A case in point is use of the term ‘grammar’, which, we have already noted, Bernstein used in referring to the ‘weak’ or ‘strong

10

LANGUAGE, KNOWLEDGE AND PEDAGOGY

grammars’ he associated with different horizontal knowledge structures. From a linguist’s point of view he used the term metaphorically, to refer to the underlying organizing principles that characterize knowledge structures. However, the issue points to a more general difference between the two traditions of scholarship. For the linguist, the object of study is language, though at least in the SF tradition language is always understood as part of social process. For the sociologist, the object of study is social structure, and though in the Bernsteinian tradition language is a very significant resource in the creation, reproduction and transformation of social structure, other resources, including education and knowledge, are also significant objects of study. The book is arranged in such a way that the transdisciplinary dialogue between the two traditions is early established and maintained. This chapter has sought to sketch in some major developments in the emergence of the two traditions, and something of the history of their previous partnerships. Theoretical Foundations are established on the one hand by the sociologists Maton and Muller, who examine Bernstein’s work on codes, the pedagogic device and knowledge structures, clarifying the connections between these areas of the theory. Then, on the other hand, Martin offers an SFL account, demonstrating how the theory is used to explore and identify specialist forms of knowledge. In the subsequent section devoted to Knowledge Structure, Muller, Maton and Moore all explore aspects of horizontal and vertical discourse, examining problems for conceptualizing knowledge, given the dichotomous nature of the theory, the notion of a ‘knower structure’ introduced to parallel that of the knowledge structure, and difficulties in dealing with hierarchical knowledge structures. The next section, devoted to Fields of Discourse – Disciplines of Discourse, commences with a chapter by Painter in which she examines the development of horizontal discourse in young children as an aspect of dealing with the ‘common-sense’ knowledge of daily life, also suggesting its limitations as the young push towards the need to master vertical discourse. Christie and Macken-Horarik, Wignell and O’Halloran all go on to address, respectively, knowledge structures in subject English, the social sciences, and mathematics and natural sciences. O’Halloran in particular looks at the impact of multimodality in the evolving nature of the sciences. All these chapters consider some of the problems as well as the strengths of characterizing knowledge structure in Bernstein’s terms. The concluding section, devoted to Research Prospects, explores some of the issues we have identified and suggests possible directions for future research.

References Atkinson, P., Davies, B. and Delamont, S. (eds) (1995), Discourse and Reproduction. Essays in Honor of Basil Bernstein. Kresskill, NY: Hampton Press. Benson, J.D. and Greaves, W.S. (1973), The Language People Really Use. Agincourt: The Book Society of Canada. Bernstein, B. (1971), Theoretical Studies towards a Sociology of Language. Class, Codes and Control, Vol. 1. London, Henley and Boston: Routledge & Kegan Paul.

LINGUISTIC AND BERNSTEINIAN PERSPECTIVES

•

CHRISTIE

11

Bernstein, B. (1975), Towards a Theory of Educational Transmissions. Class Codes and Control, Vol. 3. London and Boston: Routledge. Bernstein, B. (1981), ‘Code modalities and the process of reproduction: a model’, in Language and Society 10, pp. 327–63. Bernstein, B. (1990), The Structuring of Pedagogic Discourse. Class, Codes and Control, Vol. 4. London and NY: Routledge. Bernstein, B. (2000), Pedagogy, Symbolic Control and Identity. Theory, Research, Critique (rev. edn). Lanham, Boulder, NY and Oxford: Rowman & Littlefield. Christie, F. (ed.) (1999), Pedagogy and the Shaping of Consciousness. Linguistic and Social Processes. London and NY: Cassell. Christie, F. (2002), Classroom Discourse Analysis. A Functional Perspective. London and NY: Continuum. Christie, F. (2004), ‘Authority and its role in the pedagogic relationship of schooling’, in L. Young and C. Harrison (eds), Systemic Linguistics and Critical Discourse Analysis. Studies in Social Change. London and NY: Continuum, pp. 173–201. Christie, F. (2005), Language Education in the Primary Years. Sydney: University of NSW Press. Christie, F. and Martin, J.R. (eds) (1997), Genre and Institutions. Social Processes in the Workplace and School. London and NY: Continuum. Christie, F. and Unsworth, L. (2005), ‘Developing dimensions of an educational linguistics’, in R. Hasan, C. Matthiessen and J. Webster (eds), Continuing Discourse on Language. A Functional Perspective, Vol. 1. London and Oakville: Equinox, pp. 217–50. Cloran, C. (1994), ‘Rhetorical units and decontextualization: an inquiry into some relations of context, meaning and grammar’, in Monographs in Systemic Linguistics 6. Nottingham: Department of English Studies, University of Nottingham. Feez, S. with Joyce, H. (1998), Text-based Syllabus Design. Sydney: National Centre for English Language Teaching and Research, Macquarie University. Gregory, M. and Carroll, S. (1978), Language and Situation. Language Varieties and their Social Contexts (Language and Society Series). London: Routledge & Kegan Paul. Halliday, M.A.K. (ed. G. Kress) (1976), Halliday: System and Function in Language. London: Oxford University Press. Halliday, M.A.K. (1988), ‘Language and socialization: home and school’, in L. Gerot, J. Oldenburg and T. van Leeuwen (eds), Proceedings from the Working Conference on Language in Education, held at Macquarie University, 17th–21st November 1986, pp. 1–24. Halliday, M.A.K. (1985), An Introduction to Functional Grammar. London: Edward Arnold. Halliday, M.A.K. (1993), ‘Some grammatical problems in scientific English’, in M.A.K. Halliday and J.R. Martin (eds), Writing Science: Literacy and Discursive Power. London: Falmer, pp. 69–85. Halliday, M.A.K. (1993), ‘The Analysis of scientific texts in English and Chinese’, in M.A.K. Halliday and J.R. Martin, Writing Science: Literacy and Discursive Power. London: Falmer, pp. 124–32. Halliday, M.A.K. (1994), An Introduction to Functional Grammar. London: Edward Arnold. Halliday, M.A.K. (1973), Explorations in the Functions of Language (Explorations in Language Study). London: Edward Arnold. Halliday, M.A.K. (1975), Learning How to Mean. Explorations in the Development of Language. (Explorations in Language Study). London: Edward Arnold. Halliday, M.A.K (1985), Spoken and Written Language. Geelong, Victoria: Deakin University Press. Halliday, M.A.K. (1993), ‘Towards a language-based theory of learning’, in Linguistics and Education 5, 2, pp. 93–116. Halliday, M.A.K. (ed. J.J. Webster) (2002), On Grammar. Collected Works of M.A.K. Halliday, Vol. 1. London and NY: Continuum. Halliday, M.A.K. (ed. J.J. Webster) (2004), The Language of Early Childhood: M.A.K. Halliday, Vol. 4. London and NY: Continuum.

12

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Halliday, M.A.K. and Hasan, R. (1976), Cohesion in English. London: Longman. Halliday, M.A.K. and Hasan, R. (1985), Language, Context and Text: Aspects of Language in a Social Semiotic Perspective. Geelong, Victoria: Deakin University. Halliday, M.A.K and Hasan, R. (in press), ‘Retrospective on SFL and literacy’, in R. Whittaker, M. O’Donnell and A. McCabe (eds), Language and Literacy: Functional Approaches. Continuum: London. Halliday, M.A.K., McIntosh, A. and Strevens, P. (1964), The Linguistic Sciences and Language Teaching. London: Longman. Halliday, M.A.K. and Matthiessen, C. (3rd edn) (2004), An Introduction to Functional Grammar. London and NY: Arnold. Hasan, R. (1986), ‘The ontogenesis of ideology: an interpretation of mother child talk’, in T. Threadgold, E.A. Gross, G. Kress and M.A.K. Halliday (eds), Language, Ideology, Language. Sydney: Sydney Association for Studies in Society and Culture, pp. 125–46. Hasan, R. (1989), ‘Semantic variation and sociolinguistics’, in Australian Journal of Linguistics 9, pp. 221–75. Hasan, R. (1995), ‘The conception of context in text’, in P.H. Fries and M. Gregory (eds), Discourse in Society: Systemic Functional Perspectives. Norwood,NJ: Ablex, pp. 183–283. Hasan, R. (ed. J. Webster) (2005), Language, Society and Consciousness. The Collected Works of Ruqaiya Hasan, Vol. 1. London and Oakville: Equinox. Hasan, R. and Cloran, C. (1987), ‘A sociolinguistic study of everyday talk between mothers and children’, in M.A.K. Halliday, J. Gibbons and H. Nicholas (eds), Learning, Keeping and Using Language, Selected Papers from the 8th World Congress of Applied Linguistics, Sydney, 16–21st August, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 67–100. Hasan, R. and Martin, J.R. (eds) (1989), Language Development: Learning Language, Learning Culture. Meaning and Choice in Language. Studies for Michael Halliday. Norwood, NJ: Ablex. Hasan, R., Matthiessen, C. and Webster, J. (eds) (2005), Continuing Discourse on Language. A Functional Perspective, Vol. 1. London and Oakville: Equinox. Macken-Horarik, M. (1996), ‘Literacy and learning across the curriculum: towards a model of register for secondary school teachers’, in R. Hasan and G. Williams (eds), Literacy in Society. Harlow: Addison Wesley Longman, pp. 232–78. Macken-Horarik, M. (1998), ‘Exploring the requirements of critical literacy: a view from two classrooms’, in F. Christie and R. Misson (eds), Literacy and Schooling, London: Routledge, pp. 74–103. Martin, J.R. (1992), English Text. System and Structure. Philadelphia and Amsterdam: Benjamins. Martin, J.R. (1999), ‘Mentoring semogenesis: “genre-based” literacy pedagogy’, in F. Christie, pp. 123–55. Martin, J.R. and Rose, D. (2003), Working with Discourse. London and NY: Continuum. Martin, J.R. and Rose, D. (2005), ‘Designing literacy pedagogy: scaffolding democracy in the classroom’, in R. Hasan et al., pp. 251–80. Martin, J.R. and Rose, D. (in press), Genre Relations: Mapping Culture. London: Equinox. Martin, J.R. and White, P.R.R. (2005), The Language of Evaluation: Appraisal in English. NY: Palgrave Macmillan. Maton, K. (2000), ‘Recovering pedagogic discourse: a Bernsteinian approach to the sociology of educational knowledge’, in Linguistics and Education 11, 1, pp. 79–98. Maton, K. (2004), ‘The wrong kind of knower: education, expansion and the epistemic device’, in Muller, Davies (eds) (2004), pp. 218–31. Matthiessen, C. (1993), ‘Register in the round: diversity in a unified theory of register analysis’, in M. Ghadessy (ed.), Register Analysis: Theory and Practice. London: Pinter, pp. 221–92. Moore, R. and Muller, J. (1999), ‘The discourse of “voice” and the problem of knowledge

LINGUISTIC AND BERNSTEINIAN PERSPECTIVES

•

CHRISTIE

13

and identity in the sociology of education’, in British Journal of Sociology of Education 20, 2, pp. 189–206. Morais, A., Neves, I., Davies, B. and Daniels, H. (eds) (2001), Towards a Sociology of Pedagogy. The Contribution of Basil Bernstein to Research. NY: Peter Lang. Muller, J. (2002), Reclaiming Knowledge: Social Theory, Curriculum and Education Policy (Knowledge, Identity and School Life series). London and NY: RoutledgeFalmer. Muller, J., Davies, B.A. and Morais, A. (eds) (2004), Reading Bernstein, Researching Bernstein. London: RoutledgeFalmer. Painter, C. (1999), Learning Language in Early Childhood. London and NY: Cassell. Parret, H. (1974), Discussing Language. The Hague: Mouton, pp. 81–120. Sadovnik, A.R. (1995), Knowledge and Pedagogy: the Sociology of Basil Bernstein. Norwood, NJ: Ablex. Schleppegrell, M. (2004), The Language of Schooling: A Functional Linguistic Perspective. Mawah, NJ and London: Erlbaum. Simon-Vandenbergen, A.M., Taverniers, M. and Ravelli, L. (eds) (2003), Grammatical Metaphor. Views from Systemic Functional Linguistics (Amsterdam Studies in the Theory and History of Linguistic Science). Amsterdam and Philadelphia: Benjamins. Torr, J. (1998), From Child Tongue to Mother Tongue: Language Development in the First Two and a Half Years (Monographs in Systemic Linguistics). Nottingham: Nottingham University. Unsworth, L. (ed.) (2000), Researching Language in Schools and Communities. Functional Linguistic Perspectives. London and Washington: Cassell. Williams, G. (1999), ‘The pedagogic device and the production of pedagogic discourse: a case example in early literacy education’, in F. Christie (ed.), pp. 88–122. Williams, G. (2001), ‘Literacy pedagogy prior to schooling: relations between social positioning and semantic variation’, in Morais et al., pp. 17–45. Williams, G. (2005), ‘Semantic variation’, in Hasan, Matthiessen and Webster, pp. 457–80. Williams, G. and Lukin, A. (eds) (2004), The Development of Language. Functional Perspectives on Species and Individuals. London and NY: Continuum.

2

A sociology for the transmission of knowledges Karl Maton and John Muller

Why knowledge? In two of his final papers Basil Bernstein codified and extended a conceptualization of the different structures of knowledge associated with intellectual fields (1999a) and signalled a more general move in his theory ‘from pedagogies to knowledges’ (2001). In so doing he returned to a longstanding interest in discourse, a focus that has brought the approach associated with Bernstein’s sociology into regular and fruitful relations with systemic functional linguistics. Such cross-disciplinary dialogue has been ongoing since Bernstein in his early work adapted the linguistic notion of ‘code’ to his own sociological purposes, refining it over the years into a highly formal analytical concept: Thus a code is a regulative principle, tacitly acquired, which selects and integrates meanings, forms of realizations, and evoking contexts. (Bernstein 1990: 101; emphasis in original)

It may be useful to revisit those sociological purposes, addressing Bernstein’s particular appropriation of ‘code’, his analysis of the ‘pedagogic device’ and his subsequent move ‘from pedagogies to knowledges’. The intention is to make more comprehensible the distinctive nature of Bernstein’s preoccupation with forms of discourse of which knowledge is one. To enquire into Bernstein’s sociological roots, therefore, is to broach the question: why knowledge? In this chapter we briefly sketch a background to Bernstein’s theorization of knowledge. Such an account can be valuable not simply to place this later focus in the context of the unfolding of his wider theory. Many understandings of Bernstein’s work are stuck in two principal time warps, focusing on either his early interest in sociolinguistics (in particular the notions of elaborated and restricted codes) or his account of schooling in terms of pedagogic codes. Both freeze the theory to a time before knowledge itself became increasingly central to his thinking. In recounting its subsequent development, however, ours is of necessity a partial account, for several reasons immanent to the theory itself. One aspect of Bernstein’s method is that he was always reworking and recasting his ideas. Throughout his career theoretical developments made visible new objects of study for empirical research, which in turn required development of the theory and which then, in turn, raised further issues for research. A second aspect of the theory’s development is a form of excavation; Bernstein dug beneath the empirical features of education to explore their underlying struc-

SOCIOLOGY FOR TRANSMISSION OF KNOWLEDGES

•

MATON & MULLER

15

turing principles (most famously in terms of codes) and then excavated further to analyse what generates these principles. Bernstein was, therefore, always engaged in developing more general conceptual tools in the light of what was being revealed by both empirical research and theoretical excavation. Lastly, the resulting theory is driven by an abiding interest in social order and the nature of symbolic control, one reaching from the macro-structure of society to the microlevel of individual consciousness. To recount fully the development of this evolving and wide-ranging theory could lead to the temptation of drawing a map as big as the country. Instead, we shall simply trace one path through his work to illuminate how Bernstein came to conceptualize knowledge, explore the main aspects of these ideas, and thereby provide one context to their use in the studies collected in this volume.1 Our guiding thread is the questions raised by the developing theory as it unfolded over time. The term ‘social base’ provides an initial key, both to this trajectory and to relations between Bernstein’s approach and systemic functional linguistics. To ask about the sociological significance of any symbolic or linguistic ensemble – such as a curriculum, pedagogy or discourse – is to enquire after its social base, its grounding in a material social form of life. As Bernstein put it when describing the nature of his early interest in linguistics: Language was the structuring interface by means of which a complex set of ordering and disordering processes were specialised by the social base of its speakers. What was paramount for me was the identification of origins of these ordering and disordering processes, their maintenance and change. (2001: 363)

As a Durkheimian, Bernstein considered the principal features of this ‘social base’ to derive from the forms taken by the economic division of labour in society and the consequent forms of solidarity accompanying them.2 Both Durkheim and Bernstein described modern industrialized societies as having developed from a relatively unspecialized division of labour to a highly specialized division of labour. As a direct consequence, they tend also to develop from mechanical modes of solidarity to organic or more specialized and interdependent forms of solidarity. In tandem, societies develop specialized symbolic forms to give effect to the new specializations emerging from this growing division of labour. In a class-based society, these symbolic resources are not only differentially valued and resourced but also differentially regulated and distributed. The sociological task is to uncover how that differential valorization, regulation and distribution occurs and thereby to explore the means and mechanisms by which the underlying structures of a complex, specialized society such as ours perpetuates itself, develops and changes. As Bernstein stated: I think like Durkheim one can identify and make explicit the social base of the pedagogic relation, its various contingent realisations, the agencies and agents of its enactments. One can begin to formulate a language for the description of the production and reproduction of its discourses. At a more general level such a study connects with the maintenance and change of the knowledge base of society, and crucially with the maintenance and change of modalities of

16

LANGUAGE, KNOWLEDGE AND PEDAGOGY

symbolic control, especially those implicated in the process of cultural reproduction. (2001: 364)

This was Bernstein’s quest, and it provides insight into the nature of his evolving theory and the questions it raised. The principal concept Bernstein placed at the centre of his explanation of social and cultural reproduction, transformation and change was, at least initially, that of code.

Conceptualizing codes The concept of code has undergone several transmutations in Bernstein’s work although the underlying conceptual continuum has remained stable, tracing a range from ‘less specialised’ to ‘more specialised’. In his early work leading up to the restricted-elaborated code couple, Bernstein (1971) identified code differences in terms of more or less complex lexical, semantic and grammatical features. Soon, however, he came to consider code less as a linguistic repertoire and more abstractly in terms of an orientation to meaning. Bernstein argued that different positions within the social base, understood in terms of their degree of specialization, create, as he later put it, ‘different modalities of communication differentially valued by the school, and differentially effective in it, because of the school’s values, modes of practice and relations with its different communities’ (1996: 91). For Bernstein a restricted coding orientation, typical of someone in a relatively unspecialized context and with a relatively direct relation to the social base, predisposes that person to context-specific meanings; an elaborated code orientation, typical of someone in a more specialized context and with a more mediated relation to the social base, predisposes that person to universalistic, non-local, context-independent meanings. The principal attribute of an elaborated coding orientation is that it is able to realize more combinatorial possibilities than a restricted coding orientation. Crucially, in societies with a specialized division of labour, such as is found in modern, industrialized countries, to prosper requires the possession of an elaborated orientation to meaning. The principal sites for transmitting this privileged and privileging elaborated orientation are the home and the school, but not all homes and not all schools to the same degree. In this way, code becomes a key point of cleavage in class society. Though the concepts of elaborated code and (especially) restricted code have been the subject of considerable misunderstanding and criticism, Bernstein’s early work thereby laid out a basis for understanding how modern societies sustain themselves. In terms of the concept of code, much of this is common ground between sociologists using Bernstein’s approach and systemic functional linguists, who alone in the broader socio-linguistic community saw from the beginning that ‘code’ was not dialect and that code theory was neither a bourgeois alibi for middle-class speech nor a denigrating deficit account of working-class language, understandings Bernstein later somewhat mischievously described as ‘for people who not only won’t read but can’t read’ (2001: 371). Though quick to

SOCIOLOGY FOR TRANSMISSION OF KNOWLEDGES

•

MATON & MULLER

17

state that he never saw himself as working in sociolinguistics (1996: 147–56), Bernstein often highlighted this shared understanding, paying tribute to the ‘remarkable exception’ of a collection of papers edited by Frances Christie (1999) and to the ‘incalculable’ contribution of Michael Halliday and Ruqaiya Hasan both to the development of this phase of the code theory and his thinking more generally (Bernstein 2000: 146). As his work progressed, Bernstein continued to refine the theory of code towards greater levels of delicacy and generality. His interest shifted increasingly to exploring the elaborated code and the institutionalized sites of its dissemination, principally the school. Bernstein’s first step was to see code as a principle operating at a high level across a wide variety of contexts. The concept of code was thereby expanded beyond the possession of an individual to a more general principle or set of rules for the regulation and distribution of meaning. His initial codification of this idea lay in the algebra of the language of classification and framing, where classification conceptualizes relations of power that regulate relations between contexts or categories, and framing conceptualizes relations of control within these contexts or categories (1975). By distinguishing between power and control in this way, Bernstein opened up the possibility of exploring the different modalities an elaborated orientation to meaning might take (such as ‘visible’ and ‘invisible’ forms of pedagogy) and addressing why a particular modality was institutionalized for particular groups of pupils, and with what consequences for their educational experiences and outcomes. In the sociological work of his students and other researchers, the supercategories of classification and framing have been operationalized, into delicate observational instruments that have proved remarkably fecund in empirical studies of classrooms in many different national contexts, including Australia, Chile, Finland, Portugal, South Africa, the UK and the USA.3 This work, together with that conducted earlier at the Sociological Research Unit in London under Bernstein’s direction, more than gives the lie to a common misconception that the theory lacks empirical application. From this ongoing body of work we now have a remarkably nuanced view of the operation and effects of various pedagogic modalities with children from a variety of social backgrounds. As these studies show, Bernstein’s concepts of classification and framing enable not only the thick description prized in much educational research but also thick explanation. They offer a basis for researchers to address, for example, why particular social groups of pupils may do less well in particular classrooms or schools. Empirical research has been able to show that the group in question may not readily be able to recognize and/or realize the code required for achievement within that specific educational context. This in turn has direct implications for education policy. Simply put, one is effectively presented with a choice: change the underlying structuring principles of the school, curriculum or classroom to match the code already possessed by these pupils, or develop ways of providing those pupils with the key to the code enabling success within those contexts. Unfortunately, the former, often advanced by well-intentioned but misguided educationalists, would effectively relegate subordinate social groups to lowerstatus forms of educational knowledge and hence to the lower rungs of the division of labour; the latter is open to misreading as a deficit theory and requires

18

LANGUAGE, KNOWLEDGE AND PEDAGOGY

acknowledging the differential status of different forms of knowledge, an admission considered beyond the pale in much contemporary social science. This coding stage of Bernstein’s theory concentrated on the transmission and acquisition of elaborated orientations to meaning within schools and, to a lesser extent, within the home, and on relations between the two. However, they can in principle be applied to any ordered symbolic ensemble. Bernstein (1996: 188–91) gave the example of using code theory to describe the form taken by the layout and style of lavatories, their different modalities of usage, and their social base. Moreover, he argued that the pedagogic nature of social relations extended beyond the classroom to include, for example, doctor–patient, social worker–client and lawyer–client relations (1999b). Nonetheless, though code theory provided concepts that were highly applicable across a range of contexts from the macro to the micro, Bernstein’s principal ambitions also required the theory to be able to account for what it is that schools actually do in the broader realm of the circulation and advancement of culture in society. The question was how schools act as relays of society’s distribution of power and principles of control. Put another way, given the now much expanded understanding of the microphysics of school-based code transmission and acquisition, what is the Durkheimian big picture of society? Bernstein had, for example, argued that struggles between advocates of visible and invisible modalities of pedagogic practice represented an ideological conflict on the wider social stage between different fractions of the middle class: how might these different levels of analysis be brought together to explore further how the differentiation and regulation of symbolic forms shapes social structure? Bernstein’s answer to such questions was to attempt to relate the realms of knowledge production, pedagogic recontextualization and meaning acquisition. Conceptualizing codes thereby raised new questions that would lead Bernstein closer to a focus on knowledge: how does a society circulate its various forms of knowledge, and how is consciousness specialized in society’s image? Table 2.1

The arena of the pedagogic device (adapted)

Field of practice

Production

Recontextualization

Reproduction

Form of regulation

distributive rules

recontextualizing rules

evaluative rules

Kinds of symbolic structure

knowledge structure curriculum

pedagogy and evaluation

Principal types

hierarchical and horizontal knowledge structures

collection and integrated curricular codes

visible and invisible pedagogies

Typical sites

research papers, conferences, laboratories

curriculum policy, textbooks, learning aids

classrooms and examinations

SOCIOLOGY FOR TRANSMISSION OF KNOWLEDGES

•

MATON & MULLER

19

The pedagogic device Bernstein’s answer was by means of the pedagogic device (see Table 2.1). Named (potentially misleadingly) after the ‘language device’, the pedagogic device forms the basis of his account of: the ordered regulation and distribution of a society’s worthwhile store of knowledge, ordered by a specifiable set of distributive rules; the transformation of this store into a pedagogic discourse, a form amenable to pedagogic transmission, ordered by a specifiable set of recontextualizing rules; and the further transformation of this pedagogic discourse into a set of evaluative criteria to be attained, ordered by a specifiable set of evaluative rules.4 In Bernstein’s conceptualization each of these rules is associated with a specific field of activity: a field of production where ‘new’ knowledge is constructed and positioned; a field of recontextualization where discourses from the field of production are selected, appropriated and repositioned to become ‘educational’ knowledge; and a field of reproduction where pedagogic transmission and acquisition takes place. The specific activities of each field are, Bernstein suggests, primarily, though not exclusively, associated with specific sites. As the above implies, the three rules and fields constituting the device are said to be hierarchically related: production precedes recontextualization, which precedes reproduction. Together these three rules and their associated fields constitute an ‘arena’ created by the pedagogic device. Taking each in turn, the distributive rules distinguish between which knowledges are deemed more or less worthwhile and which of these forms of knowledge should be distributed to whom. Following Durkheim, Bernstein saw the fundamental division of labour – into mental and manual forms of labour – as related to a corresponding symbolic cleavage between sacred and profane symbolic orders, where the higher status, more rewarded, sacred symbolic orders are differentially distributed, in the past by various kinds of priestly castes and the Church and in more recent times by the education system. Distributive rules regulate relations between these symbolic orders and how they are to be distributed – who enjoys access to what forms of knowledge, and in particular, who enjoys access to the means of producing new knowledge. The recontextualizing rules construct pedagogic discourse, the ‘what’ and ‘how’ of schoolwork; they comprise principles for ‘delocating a discourse, for relocating it, for refocusing it’ (1996: 47), transforming knowledge into pedagogic communication. Evaluative rules complete the circle by establishing the evaluation nodal points that are to be acquired, stipulating the specialized consciousness that should result. These three sets of rules could thus be said to integrate analyses of power (distributive rules), knowledge (recontextualizing rules), and consciousness (evaluative rules). As Sadovnik argues, what is critical about the notion of the pedagogic device is that ‘Bernstein is concerned with more than the description of the production and transmission of knowledge; he is concerned with its consequences for different groups’ (1995a: 10). Moreover, as ever, Bernstein’s sociological concern lay with how the differential regulation and distribution of knowledge is related to the evolving structure of society. Thus the theory aimed not only to bring together power/knowledge/consciousness, but to place this within an account of

20

LANGUAGE, KNOWLEDGE AND PEDAGOGY

cultural and social reproduction, transformation and change. Bernstein argued that the three fields of production, recontextualization and reproduction together represent an ‘arena of struggle’ (1990: 206) in which groups attempt to appropriate or control the pedagogic device. To control the device is to have access to a ‘symbolic ruler of consciousness’, a ‘ruler’ in both senses of having power over consciousness and measuring the legitimacy of its realizations: Groups attempt to appropriate the device to impose their rule by the construction of particular code modalities. Thus the device or apparatus becomes the focus of challenge, resistance and conflict. (Bernstein 1996: 193; emphasis in original)

The code modality announces what should count as a marker of success or achievement and the pedagogic device is the means whereby this principle of hierarchization is created, reproduced, transformed and changed. Those in positions of power are, Bernstein suggests, able to metaphorically ‘set’ the device such that the dominant, higher-status code modality favours their own. Conversely, actors whose dispositions and practices are characterized by a different code may experience difficulty in recognizing and/or realizing practices that are rewarded within the specific context. The question Bernstein posits as crucial for research thereby became: ‘Whose ruler, what consciousness?’ (1996: 193); that is, who controls the pedagogic device and what kind of principle of hierarchization (code modality) are they attempting to impose as the measure? In Christie and Macken-Horarick (this volume) we can see a concrete effort to construct a counter-‘ruler for consciousness’, by means of constructing a linguistic metalanguage for the school English curriculum in a field currently dominated by invisible pedagogy. The ‘pedagogic device’ is an ambitious attempt to capture the role of education in the sociological big picture, reaching from social structure to individual consciousness. It represents a synoptic perspective on the orders of symbolic life; if classification and framing began from the micro-physics of the classroom, the device comprises an attempt at beginning a grand unified theory. It also works at a higher level of abstraction than codes. The rules regulated by the pedagogic device are resources for the construction, reproduction, transformation and change of codes rather than the codes themselves – one sees the effects of the device and not the device itself. At the same time, this stage of Bernstein’s work subsumed earlier understandings of codes at a higher level of abstraction, which he now defined as: O ±C

i/e

± Fi/e

where O refers to ‘orientation to meaning’ (elaborated or restricted) and the line refers to the embedding of this orientation in relative strengths (stronger and weaker) of classification and framing (1990: 43). Bernstein’s conceptualization thus operates at a high level of both generality and abstraction. It also, for that reason, lends itself less easily to empirical research. Indeed, Dowling (1999) argues that it cannot be operationalized, though Bernstein (2000:

SOCIOLOGY FOR TRANSMISSION OF KNOWLEDGES

•

MATON & MULLER

21

116–20) cites studies by Cox Donoso (1986), Swope (1992) and Singh (1993) which have deployed it as an overarching framework for empirical studies. The significance of the pedagogic device for this brief overview, however, lies more in the path it opens up to the issue of knowledge, a hitherto rather neglected focus in Bernstein’s otherwise ambitious sociology of education. For, in setting forth these ideas and making a case for the necessity of a theory of the pedagogic device, Bernstein highlighted the absence of an analysis of pedagogic discourse itself and raised questions of the forms taken by knowledge, issues he came to realize had not been answered by his account of the device.

Making knowledge visible It is interesting to observe that though knowledge is ostensibly the basis of education, and we are said to now live in ‘knowledge societies’, it is often left unremarked in studies of education. A potential barrier to seeing knowledge is a version of what the sociologist Dennis Wrong (1994) once called the ‘oversocialised’ image of people, which in education becomes what may be called an ‘over-ideologised’ image of knowledge and pedagogy. In its crudest forms this leads to treating knowledge as if it exists only to reproduce various forms of social inequality (which form depending on the perspective being advanced) or, in some more extreme positions, in claims that all knowledge is ideologized and can be understood wholly in terms of either domination or subordination. Here knowledge is reduced to the knower perspective, the ruling ideas of an age are the ideas of the ruling class (or gender or ethnicity and so forth), and nothing but (see Maton 2000; Moore and Muller 1999). To all of this there is some truth, but it is not the whole truth. Approaches operating with an overideologized image of knowledge provide much of value; they offer insightful accounts of the ways in which social relations of power pervade the conditions and contexts of the production, recontextualization and reproduction of knowledge, and stand as a corrective to any temptation to decouple power and knowledge. However, the barrier that an unnuanced image of knowledge and pedagogy creates is simply that, from these perspectives, we do not have to take the internal ordering of symbolic forms seriously. Education becomes a reflection or epiphenomenon of social structure, one without any intrinsic powers, properties or tendencies of its own. All questions of ‘what knowledge is at stake?’ give way to the question: ‘whose knowledge?’. This was the problem Bernstein felt was presented by most reproduction theory approaches to education, and which he intended his account of the pedagogic device to help rectify. He argued that such approaches analyse only ‘relations to’ education, such as relations of class, race or gender to pedagogic discourse: It is as if pedagogic discourse is itself no more than a relay for power relations external to itself; a relay whose form has no consequences for what is relayed. (1990: 166)

22

LANGUAGE, KNOWLEDGE AND PEDAGOGY

What was additionally required, he argued, was an analysis of ‘relations within’ pedagogic discourse, that is, an analysis of ‘the intrinsic features constituting and distinguishing the specialized form of communication realized by the pedagogic discourse of education’ (1990: 165). It was just such an analysis that Bernstein’s theory aimed to offer. However, to achieve such an aim required bringing knowledge more firmly to the centre of the theory’s focus, for several reasons. First, Bernstein came to believe that the concept of code ‘took for granted, and left unexamined, the form of the discourse’ (1999b: 273) and the pedagogic device ‘didn’t actually show the nature of pedagogic discourse. It showed how it was put together but it didn’t show its nature’ (2001: 373). In short, the pedagogic device was the condition for the construction of pedagogic discourse; what was still required was to address the forms this discourse might take, necessitating a focus more on the forms of knowledge than solely on the forms of pedagogic communication. Second, having analysed the transmission and acquisition of educational knowledge (in the field of reproduction) and its construction (in the field of recontextualization), the question remained as to the forms taken by the symbolic dimension of the field of production, the knowledge from which pedagogic discourse is created (see Table 2.1). Lastly, though arguing that we are now entering a new ‘Totally Pedagogised Society’ based on lifelong learning, for Bernstein a continuing focus on ‘pedagogy’ was paradoxically insufficient: I think now, looking forward, that a sociology of pedagogy does not indicate or suggest the conceptual development necessary to grasp the discursive culture for which we are being prepared. The term pedagogy has restrictive references, despite my attempt to expand its use. (2001: 367)

To understand the contemporary situation required a focus on what kinds of knowledges are being distributed to which social groups and to shape what forms of consciousness. With this, Bernstein signalled a move in his thinking from pedagogies to knowledges: ‘I have lately been attempting what could be called a sociology for the transmission of knowledges’ (2001: 368). In coming to address the lack in his approach of an understanding of the forms taken by discourse, Bernstein was keenly aware of Durkheim’s characterization of the internal properties of sacred symbolic ensembles, and how they differ from everyday modes of thought. How, asked Bernstein, do these differently patterned symbolic ensembles relate to the social base, and in what way do they differ in their specialization of consciousness? Bernstein had earlier asserted that elaborated codes differ from restricted codes in their ability to realize more combinatorial possibilities – could this be further stipulated? Moreover, his key notion of recontextualization highlighted how school subjects are not simply a reflection of their associated fields of knowledge: what forms might the latter take? In his last major contribution, Bernstein attempted to address these questions.

SOCIOLOGY FOR TRANSMISSION OF KNOWLEDGES

•

MATON & MULLER

23

Discourses and knowledge structures Bernstein’s conceptualization of the knowledges subject to pedagogic transformation (1999a) begins from the wider perspective of distinguishing two forms of discourse: horizontal discourse and vertical discourse. Horizontal discourse refers to everyday or ‘common-sense’ knowledge and ‘entails a set of strategies which are local, segmentally organized, context-specific and dependent’ (1999a: 159). The knowledges comprising this discourse ‘are related not by integration of their meanings by some co-ordinating principle, but through the functional relations of segments or contexts to the everyday life’ (1999a: 160). In contrast, vertical discourse ‘takes the form of a coherent, explicit, and systematically principled structure’ (1999a: 159). Where the knowledges of horizontal discourse are integrated at the level of relations between segments or contexts, the knowledges of vertical discourse are integrated at the level of meanings which are related hierarchically: ‘context specificity through “segmentation” in horizontal discourse, but context specificity through recontextualisation in vertical discourse’ (1999a: 161). It is common to conflate this distinction. For example, some cognitive psychologists will refer to knowledge as ‘everything that goes on in the head’ or as ‘everything that is a script for action’, and some linguists might consider knowledge to be ‘everything that is marshalled in a person’s linguistic repertoire’; this tempting conflationary gesture inheres in common sense itself. However, in making the distinction Bernstein is not suggesting the discursive practices of actors do not move between the two or that they represent an impassable fault line in the empirical world. Rather, the distinction is important for understanding social structure. If no such distinction is made, then relations between specialized symbolic forms and the specialized division of labour cannot be explored. The distinction is thus crucial for addressing the key sociological question: how are differently valorized and rewarded forms of knowledge differently distributed in society? Bernstein starts his account by recognizing that distinctions between forms of discourse had periodically been made in social science. However, he argued that a further attempt to account for knowledge is required because of a tendency within such discussions to offer a simple dichotomy – such as concrete/abstract thought, local/official knowledge or everyday/school knowledge – that obscures differences within vertical discourse (2000: 156). Accordingly, Bernstein then turns his attention to the different forms taken by vertical discourse, making a distinction between ‘hierarchical knowledge structures’ and ‘horizontal knowledge structures’. Bernstein defines a hierarchical knowledge structure as ‘a coherent, explicit and systematically principled structure, hierarchically organised’ which ‘attempts to create very general propositions and theories, which integrate knowledge at lower levels, and in this way shows underlying uniformities across an expanding range of apparently different phenomena’ (1999a: 161, 162). This form, exemplified by the natural sciences, Bernstein visually represents as a triangle of knowledge, one motivated towards building an apex of greater integrating propositions. In contrast, a horizontal knowledge structure is defined as ‘a series of

24

LANGUAGE, KNOWLEDGE AND PEDAGOGY

specialized languages with specialized modes of interrogation and criteria for the construction and circulation of texts’ (1999a: 162), such as the disciplines of the humanities and social sciences. For example, in sociology the languages refer to its wide array of competing theoretical approaches, such as functionalism, structuralism, Marxism, post-modernism, and so forth. A key difference between the two knowledge structures lies in the form taken by their development. According to Bernstein, a hierarchical knowledge structure develops by moves to widen the base and sharpen the tip of the triangle: theories are sought that embrace more empirical phenomena and comprise fewer axioms than existing theories. Intellectual progress is thus defined as the integration and subsumption of existing ideas within more overarching and generalizing propositions; for example, as the Nobel prize-winning physicist Richard Feymann put it: ‘Physics has a history of synthesising many phenomena into a few theories’ (1990: 4). In contrast, a horizontal knowledge structure develops through the addition of new languages. We thus have ‘integration of language in one case and accumulation of languages in the other’ (1999a: 163). These knowledge structures have ramifications for the intellectual shape of fields of production. Bernstein suggests, for example, that the segments of horizontal knowledge structures tend to be characterized by short-term obsolescence, only to reappear again some time in the future in a new guise. They are more vulnerable to the changing winds of intellectual fashion, but though names and styles may change, a similar account is likely to recur within each new language that emerges. Here, from the perspective of comparison with hierarchical knowledge structures, differences between these segments become akin to ‘fingerprints on identity cards which are otherwise exactly the same’ (Adorno and Horkheimer 1947: 154). The capacity to create knowledge that builds on and goes beyond existing knowledge is limited. These different knowledge structures, as a consequence, specialize consciousness differently. They also shape social practices and forms of pedagogy, and differently specialize consciousness within their intellectual fields. For example, Bernstein argues that in hierarchical knowledge structures acquirers do not have the problem of knowing whether they are speaking or writing physics: ‘the passage from one theory to another does not signal a break in the language; it is an extension of its explanatory/descriptive powers’ (1999a: 164). In horizontal knowledge structures acquirers are faced with an array of languages based on different, often opposed assumptions, making it less clear that one is indeed speaking or writing sociology. Given all this, the question then becomes: who has access to what form of knowledge? In conceptualizing these knowledge structures Bernstein is concerned with exploring the properties of what Karl Popper (1972) terms ‘World 3’, the products of our human minds or ‘objective knowledge’, rather than ‘World 2’, our mental states or subjective knowledge. For, as Popper put it, ‘no theory of subjective knowledge will be able to account for objective knowledge’ (1994a: 13). In contrast to the ‘over-ideologised’ image characterizing many approaches to education, Bernstein’s theory is thus an account of knowledge rather than of knowers. This is not to say that Bernstein’s approach diminishes human agency in favour of hypostasizing knowledge as an autonomous and freely floating

SOCIOLOGY FOR TRANSMISSION OF KNOWLEDGES

•

MATON & MULLER

25

entity separate from the social practices of actors. Rather, his conceptualization aims to make visible knowledge as an object, one with its own properties and powers that are emergent from, but irreducible to, social practices and which, indeed, help shape those practices. To explore these properties and powers, Bernstein could be said to abstract knowledge from practices and social contexts, ‘rather like a figure relieved of its ground’ (1975: 2), as he described how his earlier work abstracted transmission and acquisition in the school from its wider constraints and contexts in order to analyse those processes. In doing so, however, he brings to light the ways in which the structuring of knowledge itself works to shape social practices, identity, relation and consciousness.

Exploring questions raised by ‘knowledge structures’ As we have outlined, each stage of the theory’s development and empirical application raised questions which helped drive the theory forward. Bernstein’s conceptualization of knowledge structures is no exception; it raises questions, many addressed by the contributors to this volume. Here we shall explore two issues as a context for these engagements: the nature of differences between knowledge structures; and relations between knowledge structures and the curriculum. Differences between knowledge structures The first set of questions concern whether the distinctions in Bernstein’s conceptualization are too clear-cut: is there not a continuum between different knowledge structures? Put another way, we might ask: do not intellectual disciplines exhibit characteristics of both the forms Bernstein delineates? It can seem as if Bernstein is suggesting that the development of horizontal knowledge structures is characterized by permanent cultural revolutions that leave no trace of the past. This would make discovering even the smallest degree of continuity of problems, themes or terms within a discipline appear to be a sign of a hierarchical knowledge structure. However, such a discovery is only to be expected: for an intellectual field to exist, it must have a degree of continuity across time and space. Moreover, the addition of a new language within a horizontal knowledge structure does not necessarily eliminate, and is likely to incorporate, at least some of the terms of existing languages; when creating a new segmental theory authors are likely to start from established ideas. The issue for Bernstein is not whether over time actors within an intellectual field gnaw away at similar problems, in similar ways, using similar terms or referencing past authors. Rather, the question is whether new theories that emerge subsume and integrate past theories and aim for greater abstraction and generalizability, or are considered incommensurable with existing theories. It is a question not of whether authors use existing symbolic materials but of how they do so and with what epistemic outcomes. One would therefore expect to find the same themes, motifs, terms, styles and so forth recurring, being adapted, modified, recast and reworked over time within a horizontal knowledge structure, not only within a specific segment but also across segments. This in itself would not, for

26

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Bernstein, signal a hierarchical knowledge structure. Though theories in a horizontal knowledge structure may overlap in their use of common terms, they ‘make different and often opposing assumptions, with each language having its own criteria for legitimate texts, what counts as evidence and what counts as legitimate questions or a legitimate problematic’ (Bernstein 1999a: 163). So, though the segments of a horizontal knowledge structure may include the same terms, such as ‘patriarchy’ or ‘social class’, their authors are not speaking the same language – their assumptions and criteria for legitimate knowledge claims are different. Similarly, one would expect to discover a degree of integration and subsumption of past ideas within each language. However, the capacity for such development across languages is limited. Even if one were to describe a theory that develops in this way as a mini-triangle (see Wignell, this volume), in a horizontal knowledge structure each such mini-triangle does not subsume and integrate its predecessors and compatriots within the knowledge structure to form part of a bigger triangle. The serial character of development of the knowledge structure as a whole thereby remains accumulation rather than integration. If the two forms of knowledge structure are different, wherein then lie the bases of this difference? Put another way, is it possible for the social sciences to progress in the same way as the natural sciences and, if so, what would make such development possible? Muller (this volume) highlights two attributes of Bernstein’s conceptualization of knowledge structures as key foci for addressing these questions: ‘verticality’ and ‘grammaticality’. The first dimension refers to the degree to which the development of a knowledge structure is characterized by the integration and subsumption of knowledge into more overarching and generalizing propositions. What enables knowledge structures to develop in this manner is a crucial area for research, one in which systemic functional linguistics is providing valuable insights; see, for example, Martin (this volume) on the role played by linguistic technicality, especially grammatical metaphor, in enabling hierarchical integration of knowledge, and O’Halloran (this volume) on how mathematics enables progress in the natural sciences to develop in this way. The second dimension highlights the role played by what Bernstein terms the strength of ‘grammar’ or degree to which forms of knowledge exhibit ‘an explicit conceptual syntax capable of “relatively” precise empirical descriptions and/or of generating formal modelling of empirical relations’ (1999a: 164). Bernstein seems to suggest that, in any knowledge structure at any one time, more than one triangle or theory is likely to co-exist, suggesting a prima facie similarity to the segmented nature of horizontal knowledge structures. Both forms of knowledge structure are characterized by conflicts between advocates of alternative theories. However, in hierarchical knowledge structures choices between theories are, Bernstein argues, at least possible on the basis of recourse to empirical research because these theories have stronger grammars. The emergence of a new theory within a hierarchical knowledge structure is thereby both conflictual and integrative: to represent progress it must clash with its predecessor but also be able to explain that predecessor’s success. In other words, the new integrating theory includes but goes beyond its predecessor, or as Popper puts it:

SOCIOLOGY FOR TRANSMISSION OF KNOWLEDGES

•

MATON & MULLER

27

In all those cases in which its predecessor was successful it must yield results at least as good as those of its predecessor and, if possible, better results. Thus in these cases the predecessor theory must appear as a good approximation to the new theory, while there should be, preferably, other cases where the new theory yields different and better results than the old theory. (1994b: 12)

In contrast, the relatively weaker grammars of horizontal knowledge structures mean that relations between languages or segments cannot be settled by empirical research and are confined to critique. A key difference between knowledge structures is thus not one of stability–conflict, consensus–dissensus or orthodoxy–heterodoxy, but rather concerns the form taken by these conflicts and their outcomes, in which the strength of grammar plays a role. This issue of grammaticality may prove enlightening for understanding how some social science disciplines are more capable of sustained intellectual progress than others. Bernstein cites, for example, economics and linguistics as examples of horizontal knowledge structures with relatively stronger grammars, and sociology and cultural studies as representing relatively weaker grammars. Within a discipline with a stronger grammar, where languages purport to share the same empirical referents, there may be the possibility of something akin to the relations between theories that characterize hierarchical knowledge structures. These two interrelated dimensions are key to the differences Bernstein highlights between hierarchical and horizontal knowledge structures. Both play a role in determining the form taken by their development. In choosing between alternative theories, actors in hierarchical knowledge structures both critically examine their consistency and compatibility with others theories (their internal strengths of grammar) and their capacity to explain the results of empirical research (their external strengths of grammar). Together these provide the possibility of a rational basis for progress in hierarchical knowledge structures, to the extent that it can be decided whether a new theory represents an advance on existing theories by recourse to its integrative, subsumptive and explanatory power and to its fit with the facts.5 In contrast, significant changes in horizontal knowledge structures are all too often ideological rather than rational revolutions. Here alternative theories are in a war of hearts and minds, and choices between competing claims to insight are based more on a ‘knower code’, that is to say, on who is making knowledge claims rather than on what is being claimed and how. Though Bernstein returned to the issue a number of times (there are three versions of his paper in print [1996, 1999a, 2000] and at least one other longer version that was never published), this question of the basis of differences between knowledge structures could be clearer. There are two possible reasons for this. With Bernstein’s conceptualization we are, as Muller (this volume) highlights, ‘locked into an early (lexical) metaphorical stage of discussion, where the terms are more suggestive than they are explanatory’. Secondly, knowledge structures are not the only feature of intellectual fields of production; they represent the symbolic dimension of what are social fields of practice. To understand the development of any specific discipline one must also offer a

28

LANGUAGE, KNOWLEDGE AND PEDAGOGY

sociological account of ‘relations to’ knowledge, such as the roles played by the state, economy, social structure, and struggles between actors within the field.6 A discipline is more than just its structuring of knowledge; the concepts of ‘knowledge structures’, therefore, shed light on disciplinary development but are not the whole story. Knowledge structures and curriculum structures A second set of questions raised by Bernstein’s later ideas concerns relations between knowledge structures and educational knowledge. With the notion of ‘knowledge structures’ Bernstein is exploring the symbolic products of the field of production (see Table 2.1); the notion of recontextualization highlights that a knowledge structure is not necessarily a curriculum structure or pedagogic structure, and his theorization is not simply a recasting of pedagogic codes. Thus, in terms of Bernstein’s concepts as they currently stand, one would not describe a school curriculum in terms of exhibiting a horizontal or hierarchical knowledge structure.7 However, the caveat ‘as they currently stand’ highlights, of course, that Bernstein’s framework foresees its own reformulation and so whether such concepts will be extended and developed in future is open. Specifically, this raises questions of how the issues highlighted by these concepts (such as the degree to which knowledge structures develop through integrating and subsuming past ideas) can be explored in the school context. For example, Christie and Macken-Horarik (this volume) highlight how pupils experience a trajectory of schooling in subject English in ways that often debilitate the integration of already learned knowledge. Rather than explicitly building on past learning, the invisible pedagogy of the English curriculum renders the educational knowledge structure less visible with the effect that pupils often experience the equivalent of segmental rather than integrative acquisition of educational knowledge. If knowledge structures are not curriculum structures, this also raises the question of the degree to which the latter reflect the former. What are the relations between the knowledge structures of physics, mathematics or English literature and their respective curriculum structures? Bernstein argued that wherever there is recontextualization there is a space for the play of ideology. Even so, are there limits to recontextualization and thus limits on the degree to which ideology can shape the construction of pedagogic discourse? Bernstein, following Durkheim, saw that specialized divisions of labour demand specialized forms of consciousness. Since the specialized knowledges in the realm of production rest directly on the material base, there must surely be a limit to the amount of recontextualizing they can bear before defeating their purpose. This is made clear by the focus in Bernstein’s account of the pedagogic device on ‘evaluative rules’; these may be pedagogized artefacts, but if the criteria they construct bear no relation to their parent knowledges in the realm of production, then schooling will undermine its role as a relay of specialized knowledges. Relations between knowledge structures and their corresponding curriculum structures is, in short, a key area for future exploration. It is to this elaboration that existing work in systemic functional linguistics may have the most to con-

SOCIOLOGY FOR TRANSMISSION OF KNOWLEDGES

•

MATON & MULLER

29

tribute (see, for example, Christie and Martin 1997, Halliday and Martin 1993 and Unsworth 2000).

Conclusion A key advantage of Bernstein’s approach, we have argued, lies in its capacity to render knowledge visible as an object of study. Many approaches to education obscure the very thing that specializes education as a social field of practices. In terms of accounting for the forms taken by discourse and their relations to social structure, knowledge of knowledge has been in relatively short supply. With the concepts of codes, the pedagogic device and knowledge structures, Bernstein provides the basis for furthering that knowledge. In this chapter we outlined one way of understanding how Bernstein’s theory arrived at his conceptualization of knowledge, highlighting the questions raised by the theory’s development. We began by outlining how the issue of how differently valorized and rewarded forms of knowledge are differently distributed in society formed a key starting point of Bernstein’s problematic. We then traced how conceptualizing codes explored a key basis of social and cultural reproduction and change, raising inter alia the question of what gives rise to these codes; and how the pedagogic device modelled the construction of pedagogic discourse, raising the question of the nature of the knowledge subject to pedagogic transformation. Lastly, we outlined how Bernstein’s conceptualization of discourses and knowledge structures offers insights into the forms taken by knowledge, and sketched some of the questions these ideas have raised, issues opened up and addressed by many of the papers in this volume. At the outset of this chapter we asked the question: ‘Why knowledge?’. Why did Bernstein come to set out the basis of ‘a sociology for the transmission of knowledges’? One answer lies in the broadening of Bernstein’s focus from how educational knowledge is transmitted and acquired to how that educational knowledge is constructed, and thence to the forms of knowledge from which educational knowledge is recontextualized. Bernstein was thereby tracing knowledge from the school upstream towards its epistemic sources, exploring in turn the fields of reproduction, recontextualization and production (Table 2.1). However, an issue we repeatedly highlighted is the sociological nature of this quest. A second answer to the question ‘Why knowledge?’, therefore, lies in Bernstein’s belief that the differential valorization, regulation and distribution of forms of knowledge in society is a crucial aspect of how societies maintain themselves. Bernstein remained focused throughout his career on the social reproduction function of cultural production, transformation and reproduction. His unfolding account of knowledge sought to understand the role that symbolic forms play in the ordering of social life. Finally, a third answer to ‘Why knowledge?’ may lie with the nature of the discipline of sociology itself. Bernstein was repeatedly drawn to studying the sacred – whether understood as an elaborated orientation to meaning, vertical discourse or knowledge structures. Such forms of discourse have the singular

30

LANGUAGE, KNOWLEDGE AND PEDAGOGY

charm of being an accumulated sedimentation of symbolic extensions that has been constructed by innumerable co-operating knowers, extended in time as well as space, most of whom will remain unknown to one another. As Moore and Maton (2001: 172) put it when describing the community of mathematicians involved over many centuries and across the globe in unravelling Fermat’s Last Theorem: It represents an epistemic community with an extended existence in time and space, a community where the past is present, one in which the living members interact with the dead to produce contributions which, when they die, will be in turn the living concern of future members.

Such an endeavour is a pure form of communism, as Robert Merton (1973) pointed out, but also ‘a very individualistic kind of communitarianism’ as Arthur Stinchcombe (nd: 20) drily added. At its best, it comprises not only a knowledge structure where past insights are subsumed and integrated, standing on the shoulders of giants to see further (Muller 2006), but also a knower structure in which everyone in the scholarly community is potentially able to contribute (Maton, this volume). It is likely that Bernstein yearned for that form of community, one which he rarely found in the lower reaches of his own horizontal knowledge structure, sociology, and that he has done more than anyone else in the sociology of education to change. As we have outlined, Bernstein’s theory developed through the subsumption of past ideas within new, often more abstract, formulations as it sought to grasp an ever-widening range of phenomena. Bernstein’s own way of theorizing thereby attempted to help analyse and exemplify principles that could enable the building of knowledge reaching upwards, bringing together individuals within an epistemic community extended across time and space. The capacity of his ideas to do just that is illustrated by their continuing contribution to the ongoing, fruitful dialogue between sociologists and systemic functional linguists.

References Adorno, T.W. and Horkheimer, M. (1947), Dialectic of Enlightenment. London: Verso. Atkinson, P. (1985), Language, Structure and Reproduction: An introduction to the sociology of Basil Bernstein. London: Methuen. Atkinson, P., Delamont, S. and Davies, B. (eds) (1995), Discourse and Reproduction: Essays in honor of Basil Bernstein. Cresskill, NJ: Hampton Press. Bernstein, B. (1971), Class, Codes and Control, Volume I: Theoretical studies towards a sociology of language. London: Routledge & Kegan Paul. Bernstein, B. (1975), Class, Codes and Control, Volume III: Towards a theory of educational transmissions. London: Routledge & Kegan Paul. Bernstein, B. (1990), Class, Codes and Control, Volume IV: The structuring of pedagogic discourse. London: Routledge. Bernstein, B. (1995), ‘A response’, in Savodnik, A. (ed.), Knowledge and Pedagogy: The sociology of Basil Bernstein. Norwood, NJ: Ablex, pp. 385–424.

SOCIOLOGY FOR TRANSMISSION OF KNOWLEDGES

•

MATON & MULLER

31

Bernstein, B. (1996), Pedagogy, Symbolic Control and Identity: Theory, research, critique. London: Taylor & Francis. Bernstein, B. (1999a), ‘Vertical and horizontal discourse: An essay’, in British Journal of Sociology of Education 20(2): pp. 157–73. Bernstein, B. (1999b), ‘Pedagogy, identity and the construction of a theory of symbolic control: Basil Bernstein questioned by Joseph Solomon’, in British Journal of Sociology of Education, 20(2): pp. 265–79. Bernstein, B. (2000), Class, Codes and Control, Volume V: Pedagogy, symbolic control and identity. Oxford: Rowman & Littlefield. Bernstein, B. (2001), ‘From pedagogies to knowledges’, in Morais, A. et al. (eds), Towards a Sociology of Pedagogy: The contribution of Basil Bernstein to research. New York: Peter Lang, pp. 363–84. Christie, F. (ed.) (1999), Pedagogy and the Shaping of Consciousness: Linguistic and social processes. London: Continuum. Christie, F. and Martin, J.R. (eds) (1997), Genre and Institutions: Social processes in the workplace and school. London: Cassell. Cox Donoso, C. (1986), ‘Continuity, conflict and change in state education in Chile: A study of the pedagogic projects of the Christian Democrat and Popular Unity Governments’, in CORE 10(2). Dowling, P. (1999), ‘Basil Bernstein in frame: “Oh dear, is this a structuralist analysis?”’, http://www.ioe.ac.uk/ccs/dowling/kings1999. Feynmann, R. (1990), Q.E.D.: The Strange Theory of Light and Matter. London: Penguin. Halliday, M.A.K. and Martin, J.R. (1993), Writing Science: Literacy and discursive power. London: Falmer. Maton, K. (2000), ‘Recovering pedagogic discourse: A Bernsteinian approach to the sociology of educational knowledge’, in Linguistics and Education 11 (1), pp. 79–98. Maton, K. (2006), ‘On knowledge structures and knower structures’, in Moore, R., Arnot, M., Beck, J. and Daniels, H. (eds), Knowledge, Power and Educational Reform: Applying the sociology of Basil Bernstein. London: RoutledgeFalmer. Merton, R.K. (1973), The Sociology of Science: Theoretical and Empirical Investigations. Chicago: The University of Chicago Press. Moore, R., Arnot, M., Beck, J. and Daniels, H. (eds) (2006), Knowledge, Power and Educational Reform: Applying the sociology of Basil Bernstein. London: RoutledgeFalmer. Moore, R. and Maton, K. (2001), ‘Founding the sociology of knowledge: Basil Bernstein, intellectual fields and the epistemic device’, in Morais, A., Neves, I., Davies, B. and Daniels, H. (eds) Towards a Sociology of Pedagogy: The contribution of Basil Bernstein to research. New York: Peter Lang. Moore, R. and Muller, J. (1999), ‘The discourse of “voice” and the problem of knowledge and identity in the sociology of education’, in British Journal of Sociology of Education 20(2): pp. 189–206. Morais, A., Neves, I., Davies, B. and Daniels, H. (eds)(2001), Towards a Sociology of Pedagogy: The contribution of Basil Bernstein to research. New York: Peter Lang. Muller, J., Davies, B. and Morais, A. (eds), (2004), Reading Bernstein, Researching Bernstein. London: RoutledgeFalmer. Muller, J. (2006), ‘On the shoulders of giants: verticality of knowledge and the school curriculum’, in Moore, R. et al. (eds) Knowledge, Power and Educational Reform: Applying the Sociology of Basil Bernstein. London: RoutledgeFalmer. Popper, K. (1972), Objective Knowledge: An evolutionary approach. Oxford: Clarendon Press. Popper, K. (1994a), Knowledge and the Body-Mind Problem: In defence of interaction. London: Routledge. Popper, K. (1994b), The Myth of the Framework. London: Routledge. Sadovnik, A. (1995a), ‘Basil Bernstein’s theory of pedagogic practice: A structuralist approach’, in Sadovnik (ed.) Knowledge and Pedagogy. Norwood: Ablex.

32

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Sadovnik, A. (ed.) (1995b), Knowledge and Pedagogy: The sociology of Basil Bernstein. Norwood: Ablex. Singh, P. (1993), ‘Institutional discourse: A case study of the social construction of technical competence in the primary classroom’, in British Journal of Sociology of Education 14(1): pp. 39–58. Stinchcombe, A. (nd), ‘Universities and the communism of knowledge as the core of the world information economy’, Northwestern University, mimeo. Swope, J. (1992), ‘The production, recontextualising and popular transmission of religious discourse in eight basic Christian communities in Santiago, Chile’, in CORE 16(3). Unsworth, L. (ed.) (2000), Researching Language in Schools and Communities: Functional linguistic approaches. London: Cassell. Wrong, D. (1994), The Problem of Order: What unites and divides society. London: The Free Press.

Notes 1

2

3

4

5

6

7

For informative introductions to Bernstein’s ideas, see Atkinson (1985), papers collected in Sadovnik (1995b) and in British Journal of Sociology of Education 23(4), 2002, and the website of the Estudos Sociológicos da Sala de Aula centre in Lisbon (http://essa.fc.ul.pt). Given the state of the intellectual field of sociology (which Bernstein would describe as a horizontal knowledge structure with a weak grammar), offering an introductory way into, rather than a critique or development of Bernstein’s later ideas for a cross-disciplinary audience is not without its dangers: Bernstein himself may have described the following as ‘schizzing’ or ‘creative replacing’ (1990: 8–9); his critics, sympathetic or otherwise, may see only exegetic advocacy, the work of what Dowling (1999) describes as ‘disciples’. In highlighting the undoubtedly Durkheimian nature of Bernstein’s sociological enterprise we are, of course, not intending to obscure his wide-ranging theoretical interests and integration of insights from, among others, neo-Marxist, Weberian and post-structuralist approaches. Such intellectual ancestry is beyond the limited scope of this chapter. See, for example, studies discussed in Bernstein (2000) and those collected in Atkinson et al. (1995), Christie (1999), Moore et al. (2006), Morais et al. (2001), Muller et al. (2004) and Sadovnik (1995b). Bernstein’s use of the term ‘rules’ has led some commentators to suggest his theory argues that practices are deterministically rule-governed. However, for Bernstein, rules do not by themselves cause anything but rather direct our attention to the controls on the form taken by pedagogic discourse, i.e. to the principles which give rise to its structuring. This is not to say that this possibility is always taken up in practice. The history of science is replete with examples of overlooked or dismissed theories that, in retrospect, represented advances on existing theories. Whether this possibility is recognized and realized in any particular case is an empirical question in which the nature of an intellectual discipline as a social field of practice plays a key role. Conversely, one must beware the temptation to sociological reductionism whereby, for example, the astonishing expansion of science is explained solely in terms of the interests it serves. Bernstein’s approach suggests that such social power is an insufficient explanation; one must also take into account the epistemic power characterizing forms of knowledge. This point may appear nitpicking were it not for what could be termed pedagogic reductionism, the tendency in discussions of education for the school classroom to

SOCIOLOGY FOR TRANSMISSION OF KNOWLEDGES

•

MATON & MULLER

33

embody the gravity well of a black hole into which all other foci are drawn and distinctions crushed. Though we do not, of course, wish to obscure the significance of the chalkface, this reductionism can obscure the notion of recontextualization and issues specific to the production of knowledge. Similarly, analyses of textbooks or curriculum guidelines are studies of recontextualized pedagogic discourse rather than of knowledge structures.

3

Construing knowledge: a functional linguistic perspective J.R. Martin

1 Knowledge in school My interest in knowledge structure arose in the context of literacy development, around the evolving theory and practice of the ‘Sydney School’.1 As this work moved from primary into secondary school, and reconsidered secondary school literacy in relation to workplace and academic discourse (Christie and Martin 1997), it became increasingly important to focus on the knowledge encoded in various genres – from one discipline or profession to another. This was also critical by way of getting teachers focused on literacy across the curriculum, since by secondary school reading and writing is seen as a tool for learning.2 Our initial strategy was to focus on one key humanities discipline (history) and one key science (physical geography), and push on from there (Eggins et al. 1993; Wignell et al. 1990; Martin 2002a, b; for related work see Coffin 2000; Halliday 2004; Halliday and Martin 1993; Macken-Horarik 1998; Martin 1993b; Martin and Veel 1998; Martin and Wodak 2004; Schleppegrell 2004; Unsworth 1997a, b, 1998, 2000, 2001; van Leeuwen and Humphrey 1996; Veel 1992). Our informing theory for this work was systemic functional linguistics (hereafter SFL), drawing in particular on functional grammar (Halliday and Matthiessen, 2004), discourse semantics (Martin and Rose 2003; Martin and White 2005), register and genre theory (Ghadessy 1993, 1995; Martin and Rose 2005) and multimodality (Kress and van Leeuwen 1996, 2001). Accordingly our model of language mapped metafunctions (ideational, interpersonal and textual) across strata (graphology, lexicogrammar and discourse semantics) as in Figure 3.1 below; following Martin 1992, we deployed a stratified model of context, including register (field, tenor and mode) and genre, as in Figure 3.2 below. In a model of this kind, the register variable field provides a social semiotic perspective on knowledge structure; and knowledge is by and large realized through, construed by, and over time reconstrued through ideational meaning (via the modalities of language and image). So the rest of this paper can be read as an exploration of knowledge structure as field of discourse.

2 Field Following Martin, 1992, we treated field as a set of activity sequences oriented to some global institutional purpose, including the taxonomies of participants involved in these sequences (organized by both classification and composition).3 Taking science for example (geology in particular), we can recognize the

CONSTRUING KNOWLEDGE

•

35

MARTIN

textual

discourse semantics ideational lexicogrammar

phono-/graphology interpersonal

Figure 3.1

Strata and metafunctions in SFL

genre mode

field

textual ideational

tenor

interpersonal

Figure 3.2

Social context as register (field, tenor, mode) and genre

36

LANGUAGE, KNOWLEDGE AND PEDAGOGY

specialized activity sequence spelled out in the following text – explaining how volcanoes are formed: (1) In parts of the earth beneath the crust . . . heat accumulates to such an extent that it does cause local melting of rocks to form a molten mass called magma. This molten material is under such enormous pressure that some of it is forced into any cracks and crevices that might form in the upper solid crust of the earth, and in surrounding solid rock. Some of this molten material can actually cool and solidify without reaching the earth’s surface; in other cases molten material is pushed right through the earth’s surface and forms a volcano. When molten material is forced out to the earth’s surface it is called lava. (Messel 1963: 12.1, 40.4–8)

Note that the activity sequence4 gives rise to technical terms naming participants (in bold face above – magma, volcano and lava). From the perspective of taxonomy, such participants are additionally organized with respect to both classification (hyponymy) and composition (meronymy). The classification taxonomy complementing Text 1, which shows where igneous rocks come from, has to do with types of rock. An image for these relations is presented as Figure 3.3 below, with general classes to the left and subclasses to the right – the direction in which subclassification is portrayed in SFL (as opposed to biology, which prefers general classes on top and subclasses underneath). extrusions igneous intrusions clastic rocks

sedimentary non-clastic

metamorphic

Figure 3.3

Classification taxonomy (types of rock)

Turning to composition, the main relations relevant here have to do with the structure of volcanoes and the earth’s crust, as outlined in Figure 3.4 (from Messel 1963: 40.4). Whereas scientists (biologists excepted) seldom use images to model classification relations, schematic drawings with appropriate labelling are regularly used for compositional relation across sub-disciplines. In Bernstein’s terms, we’ve illustrated field here using examples from a hierarchical knowledge structure (vertical discourse). But the notions of activity sequence, classification and composition can be used to map all discourses (whether horizontal or vertical); what varies of course is the nature of the participants and activities at risk, and the level of technicality, which we’ll explore in sections 3 and 4 below.

CONSTRUING KNOWLEDGE

Figure 3.4

•

37

MARTIN

Compositional relations (volcanoes) (Messel 1963: 40.4) Common sense

domestic (guidance) oral transmission (doing)

sport recreational (‘coaching’) hobby

specialized (participation) trades (apprenticing) administration (co-operation) written transmission (studying)

humanities exploration (instruction)

social science

science Uncommon sense

Figure 3.5

Common and uncommon sense fields (Martin 1992: 544)

38

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Martin 1992 (pp. 544–5), inspired by Bernstein’s earlier notions of common and uncommon sense discourse, developed a crude mapping of fields according to the way he imagined they were learned5 and their degree of lexical specialization. His provisional outline of fields is reproduced as Figure 3.5 below, by way of illustrating the affinity of Sydney School research with Bernstein’s developing concern with knowledge structure. Martin’s common sense to uncommon sense cline parallels Bernstein’s horizontal discourse to hierarchical knowledge structure dimensionality (although the position of bureaucratic discourse in Bernstein’s model is unclear).

2 Science (hierarchical knowledge structure) Since we’ve begun with science, let’s focus more closely on its discourse here. Compared with other fields, scientific taxonomies are relatively comprehensive, deep and precise. Taking biological classification as a case in point, the taxonomies are comprehensive in that they set out to classify all of the phenomena in their gaze (all of life), deep in that they classify and subclassify at many levels of generality, and precise in that they ultimately depend on categorical genetic criteria. Classifying animals, for example, involves covering everything that is mobile, obtains food from other living things, has extensive movement of body parts, and so on; and it takes several steps to move from animals in general to human beings like us: Kingdom (animalia – obtain food, mobile . . .) Phylum (chordata – with hollow dorsal nervous system) Sub-phylum (vertebrata – with a backbone) Class (mammalia – with hair and mammary glands) Order (primates – with grasping hands) Family (hominidae – man-like reasoning) Genus (homo – . . .) Species (sapiens – . . .) At every step along the way, subclassification is categorical and exhaustive – for example, there are three different types of mammals, no less and no more, and every mammal belong to one group or another: – the egg-laying mammals called monotremes – the pouched mammals called marsupials – the higher mammals called placentals An overview is presented in Figure 3.6, based on genetic reasoning (as recontextualized for secondary school science by Haire et al. 2005). Scientific taxonomies contrast with everyday taxonomies which are relatively piecemeal, shallow and fuzzy. For one purpose or another we divide living things into groups such as food, pets, pests, garden plants and predators, including a relatively open-ended list of plants and animals in each group, and some-

CONSTRUING KNOWLEDGE

Figure 3.6

•

MARTIN

39

Genealogical organization of living things (Haire et al. 2005: 202)

times wondering whether something properly belongs to one group or another, or more than one (Is a tomato a fruit or a vegetable? Is a peanut a nut? Is a horse a food?). The classification sets are segmentally rather than hierarchically organized (i.e. lacking in higher level superordinates), and many living things belong to more than one group:

40

LANGUAGE, KNOWLEDGE AND PEDAGOGY

– food (meat, fish, fruit, vegetables . . .) – pets (cats, dogs, birds . . .) – gardens (trees, shrubs, flowers, grass, vegetables . . .) – pests (insects, vermin . . .) – predators (snake, spider, crocodile, shark, bear . . .) ... Another important feature of these everyday categories is that they can be learned ostensively, by having things we can see, touch, taste, hear and/or smell pointed out to us as examples (Painter 1998). We don’t learn them like technical terms, through definitions; we just learn them concretely, based on sensuous experience. The possibility of ostensive definition becomes an even more important criteria when we turn from classification to composition, and rely on uncommon sense images of various kinds to mimic ‘concrete’ learning. In Figure 3.7, for example, the composition of the human ear and Australia’s famous cochlear ear implant technology is outlined, and specialized terminology is presented as labels for the parts. But the image involved is not a photo of the human body and implant; rather it is a schematic idealization based on centuries of technologically assisted perception, by anatomists, using probing and recording instruments (including of course writing and images, to record results). Even if we had the stomach for it, most of us would be very hard pressed to recognize parts involved if a human body were made available to us. One needs the scientific understandings as part of finding one’s way around.

Figure 3.7

Cochlear ear implant (Haire et al. 2005: 236)

CONSTRUING KNOWLEDGE

•

MARTIN

41

At this point it may be useful to distinguish between technical and specialized lexis along the following lines. Technical lexis can be defined as lexis which has to be learned by definition, through language (typically in institutionalized learning); specialized lexis on the other hand is lexis that can be learned by observation, through gesture (it is ostensively ‘defined’). Specialized language is a feature of technology (White, 1998), and learned as part of apprenticeship in arts and crafts, trades, sports, hobbies and the like. As such it is not available to the community as a whole, but to people operating hands-on in particular spheres of activity (see Rose, 1997, 1998 on technological discourse in sciencebased industry, for example). Like everyday lexis, specialized lexis is concrete; but like technical lexis it belongs to specific walks of life (and similarly enacts the boundaries between common and less common sense discourses). Turning to activity sequences, scientific sequences focus more on cause than time, and lean towards a relatively deterministic view of causality. For this reason scientific activity sequences are referred to as implication sequences (in Martin, 1992, 1993c), and are designed to explain. In the following explanation of cold fronts (Messel 1963: 7.6–7.7), for example, a number of causal resources are deployed, including connections between clauses (if, how), verbs (causes, resulting), a noun (reasons) and preposition (as a result of). (2) Cold fronts. A stream of comparatively cold, dense air tends to move along close to the ground as it flows towards regions in which warmer, less dense, air is rising. This rising air becomes cooler for the reasons mentioned earlier, and if it is humid condensation of water vapour will take place. The resulting clouds are usually of the cumulus type. The front edge of the cold air mass is known as a cold front. Much of the rain that falls in Australia occurs as a result of cold front conditions. Figure 3.8 shows how a cold front causes uplift and condensation in a warmer, humid, air mass. The arrival of a cold front is marked by a sharp drop in temperature and a sudden change of the wind direction.

Text 2 is accompanied by a ‘narrative’ image (Figure 3.8 below), including two appropriately labelled vectors (cold air stream and boundary between air masses) which portray a cold air stream bumping into warm rising humid air and rain falling from the cumulus clouds which form. Note that the image vectors afford a temporal ‘before and after’ reading of this process, which the verbal text further specifies as cause and effect. The application of Kress and van Leeuwen’s 1996 work on images to vertical discourse of this kind is introduced in van Leeuwen and Humphrey 1996, and further developed in Unsworth 2001. Text 2 also demonstrates the way in which implication sequences in science participate in the definition of technical participants, in this case cold fronts (as opposed to warm fronts), and cumulus clouds (as opposed to cirrus, altostratus, etc.). Here, the classification of the terms depends in part on their role in the implication sequences in which they participate.

42

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Figure 3.8 Narrative image illustrating the cold front explanation (Messel 1963: 7.6–7.7)

3 History (horizontal knowledge structure) Unlike science, history is not a technical discipline. As far as participants are concerned, uncommon sense classification and composition is simply borrowed from other fields. So for history dealing with the politics of migration in Australia we find borrowed classification for kinds of police force (South Australian Police, Australian Federal Police . . .) or types of asylum seekers (Indochinese, Vietnamese, Afghans, Iraqis, Iranians . . .), and borrowed composition for government agencies: DIMIA DFAT DEET HREOC etc.

Dept of Immigration, Mutlicultural & Indigenous Affairs Dept of Foreign Affairs & Trade Dept of Employment, Education and Training Human Rights & Equal Opportunity Commission

Note in passing the use of acronyms adopted from the relevant administrative discourse here (something which is also a feature of the specialized lexis of technological fields introduced above; White 1997). Classification particular to history tends to be instantial, arising in the course of the development of a particular discussion, but not transcending this text into the field as a whole. In Text 3 below, for example, the historian reports on Ruddock’s competing visions for Australia’s population; but this classification does not become part of the technicality of the discipline, even for the specific discipline of Australian migration studies. The classification depends on reading this specific text, not on apprenticeship into the field.

CONSTRUING KNOWLEDGE

•

MARTIN

43

(3) The minister moves on to outline three competing visions for Australia’s population in the century ahead. The first scenario is the high-immigration model favoured by some business groups, which call for Australia’s net migration intake to be set at 1 per cent of existing population per year . . . The second scenario is net zero migration, the model pushed by sections of the environmental movement and by groups such as One Nation, which say that Australia should take just enough migrants to replace the number of people who permanently depart the country each year . . . The minister’s final forecast is reassuring – according to him, if we hold fast to the current government policy, Australia’s population will increase gradually for the next forty years before settling comfortably at around 23 million . . . (Mares, P. 2001 Borderline: Australia’s treatment of refugees and asylum seekers. Sydney: UNSW Press (Reportage Series) 141–2)

For technicality that is specific to history we need to look at activity sequencing, and even here it takes historians a while to get going. Of course for historians, chronology matters; but they construe different kinds of time. In Text 4, sequence in time is foregrounded – the relation of one event to another as happening beforehand or afterwards (as indexed by the left and right pointing arrows in the text below: ‘→’ for ‘subsequently’, ‘←’ for ‘previously’). (4) On 29 August the Tampa entered into Australian territorial waters approaching Christmas Island. → The prime minister told parliament that ← the captain had decided on this course of action because ← a spokesman for the asylum seekers ‘had indicated that they would begin jumping overboard if medical assistance was not provided quickly’. → Captain Rinnan gave a different reason for this decision: ‘We weren’t seaworthy to sail to Indonesia. There were lifejackets for only 40 people. The sanitary conditions were terrible.’ → The SAS came aboard and took over Tampa. → An Australian Defence Force doctor was given 43 minutes to make a medical assessment of the 433 asylum seekers. → He reported, ‘Four persons required IV (2 urgent including 1 woman 8 months pregnant).’ → Captain Rinnan was surprised at the prompt medical assessment, because ← his crew had already identified ten people who were barely conscious lying in the sun on the deck of the ship. → The prime minister then made a finely timed ministerial statement to parliament insisting that ← ‘nobody – and I repeat nobody – has presented as being in need of urgent medical assistance as would require their removal to the Australian mainland or to Christmas Island’. → One hundred and thirty-one fortunate asylum seekers were granted immediate asylum by the New Zealand government. → The rest, ← having been transported to Nauru, waited processing under the evolving Pacific Solution. (Brennan 2003: 42–3)

In this case the precise timing of events and what was said is crucial, since Brennan is keen to show that the government was acting illegally out of selfinterest for electoral purposes. In other cases, it is phases of time rather than individual events that matter, and chronology is organized through setting in time (realized through prepositional phrases of temporal location, typically sentence initial) rather than sequence (realized through conjunctions and one clause following another). In Text 5 below Brennan skips through time along these lines from 1978 to 1990.

44

LANGUAGE, KNOWLEDGE AND PEDAGOGY

(5) In 1978 the government set up a Determination of Refugee Status (DORS) Committee which would determine onshore refugee claims. A UNHCR representative joined this committee of pubic servants from the key departments. The committee made only recommendations to the minister. If it rejected an application, it could still recommend that the applicant be given temporary or permanent residence ‘on humanitarian or strong compassionate grounds’. In the early 1980s the committee considered fewer than 200 applications a year, with less than a one third approval rating. In 1982 the government decided that even offshore cases would be decided on a case by case basis. It would no longer accept the UNHCR’s blanket determination that anyone from Indochina was a refugee. It was now seven years since the end of the Vietnam War and it was more likely that some of those departing Vietnamese were economic migrants unimpressed by their economic prospects under a communist regime rather than refugees who were fleeing in fear of persecution. At the same time the government set up a Special Humanitarian Program to complement the offshore refugee program. In the first year, there were 20,216 offshore refugees and 1,701 applicants approved for migration to Australia under the Special Humanitarian Program. Within eight years there were only 1,537 under the offshore refugee category and 10,411 under the Special Humanitarian Program . . . (Brennan 2003: 29–31)

Once time is phased in this way it can be nominalized, a process of ‘thingification’ whereby activity is reconstrued as abstract things. In Text 6 below, for example, phases of Indochinese migration are enumerated as waves (first wave, second wave, third wave, etc.). (6) The first wave of 2,077 Indochinese boat people came to Australia in 54 boats between 1976 and 1981 . . . The first boatload of asylum seekers arrived in Darwin harbour on 28 July 1976 . . . At the end of that year another two boats arrived carrying 106 people . . . When the third Vietnamese boat of the first wave arrived, there was some media agitation . . . (p. 29) The second wave of boat people commenced with the arrival of a Cambodian boat at Pender Bay near Broome on 25 November 1989 . . . (p. 32) The third wave of boat people arrived between 1994 and 1998 . . . These Vietnamese and Chinese boat people were the last victims of the Comprehensive Plan of Action . . . (p. 40) The fourth and biggest wave of boat people in modern Australian history could not be so readily categorized as non-refugees or as refugees who had their claims determined elsewhere. In late 1999 boat people started arriving from Afghanistan, Iraq and Iran via Indonesia . . . (p. 40)

Once time is packaged as a thing it can be named; and where proper names become established for phases of history, they transcend the text/s which created them and enter into the field as technical terms. Examples include

CONSTRUING KNOWLEDGE

•

MARTIN

45

Tampa, Mabo, The Sharpeville Massacre, The Long March, The Depression, The First Gulf War. In short then, we can say that the technicality of history has to do with activity not participants, but that this activity is reconstrued as a participant that becomes named and potentially technicalized. ‘Thingification’ of this order is a recurrent feature of technicality in any vertical discourse. As well as chronicling, historians have to explain. More so than scientists, they have a strong preference for doing this within rather than between clauses. The contrast between explaining between clauses and inside them is well illustrated in Text 7 below. Manne begins explaining in the clause (unwillingness affecting response), and then unpacks this with three clarifications managing cause between clauses (each beginning with a because clause preceding its effect). (7) The Howard government’s unwillingness to apologise determined the nature of its response to other recommendations contained in Bringing them home. ↔ Because it refused to consider the present generation of Australians legally or morally responsible for the mistakes of the past, it refused altogether Bringing them home’s recommendation for financial compensation for members of the stolen generations. Because it thought the policies of child removal had been lawful and well-intentioned, it treated almost with contempt the arguments in Bringing them home . . . Because, nonetheless, it accepted . . . it was willing to allocate modest sums . . . (Manne 2001: 76)

More typical of historical explanation is a text like the following, in which the majority of cause and effect relations are construed inside a single clause. (8) In another part of Australia, Aboriginal people were themselves acting to assert their rights. On 23 August 1966 Vincent Lingiari, a Gurindji elder, led his people off the cattle station operated by the giant Vesteys pastoral organisation in protest against their wages and conditions. Their calls for Commonwealth involvement also strongly argued the case for land to establish their own cattle station. They subsequently sent a petition to the Governor-General, with no immediate result. Their stand against injustice, however, attracted national publicity for Aboriginal land rights grievances. The strike developed into a seven-year campaign by the Gurindji for the return of their traditional lands and became a cause célèbre across Australia. The campaign was strongly supported by the trade union movement and sparked a campaign for human rights, including land rights, by many Aboriginal people. It was a cry for Commonwealth leadership that would not be acted upon until the election of the Whitlam government. (Tickner 2001: 8)

In 8 there is only one explicitly temporal relation, realized through the conjunction subsequently. Otherwise Tickner draws on various clause-internal resources to explain what led to what. He uses causal circumstances (prepositional phrases) to explain why Lingiari led his people off (in protest against their wages and conditions, for the return of their traditional lands), and to explain what the Gurindji were appealing for (for Commonwealth involvement . . . leadership). These nominalized appeals (for involvement, leadership) are verbally connected

46

LANGUAGE, KNOWLEDGE AND PEDAGOGY

(argued, acted) to further abstractions (the case, the election). And processes are also used to causally connect the Gurindji walk-off to a burgeoning land rights movement (attracted national publicity, sparked a campaign). Why this drift towards explanation inside rather than between clauses? One thing the ‘cause in the clause’ motif does is enable historians to fine tune causality by deploying verbs which elsewhere literally construe material and verbal activity (e.g. argue, act, attract, spark), but here enact finely differentiated types of cause and effect relations. In horizontal discourse, processes tend to have concrete readings – with people acting and interacting, often involving other people and things: Frank argued with Mark. The pool attracted Mike. The lightning sparked a fire. The goal-keeper had to act. So we find sensuous participants materially affecting sensuous participants in examples such as these. Contrast the use of these processes in Tickner’s text, where abstractions affect abstractions – calls arguing the case, a stand attracting publicity, one campaign sparking another, a cry not acted upon: Their calls for Commonwealth involvement also strongly argued the case for land to establish their own cattle station. Their stand against injustice, however, attracted national publicity for Aboriginal land rights grievances. The campaign . . . sparked a campaign for human rights, including land rights, by many Aboriginal people. It was a cry for Commonwealth leadership that would not be acted upon until the election of the Whitlam government. This indefinitely enhances historians’ resources for explaining how one event affects another. Between clauses English provides for a simple model of cause – one thing leading to another via conjunctions such as so, because and therefore. Inside the clause on the other hand the nominal (cause, reason, basis, source, motive, etc.), prepositional (due to, owing to, because of, thanks to, etc.) and most importantly the verbal resources of the language open up for us. This means that within the clause historians can fine tune their explanations about the ways in which events give rise to one another. A simple inter-clausal relation is transformed into a finely nuanced clause-internal repertoire for interpretation as these resources are brought to bear. The possibilities are not endless, but indefinitely enlarged:

CONSTRUING KNOWLEDGE

•

MARTIN

47

Their stand against injustice, however, attracted national publicity for Aboriginal land rights grievances. Their stand against injustice, however, initiated national publicity for Aboriginal land rights grievances. Their stand against injustice, however, kindled national publicity for Aboriginal land rights grievances. Their stand against injustice, however, accelerated national publicity for Aboriginal land rights grievances. ‘Cause in the clause’ is also an important motif in science discourse, as illustrated in Text 2 above. But unlike historians, scientists prefer a simple model of cause and effect; they’re not interested in proliferating different kinds of cause (fuzzy cause would be a problem for scientists, not a solution). Even so, as Halliday has shown (Halliday, 1998, 2004), there is still an important pay-off inside the clause for scientists as far as managing argumentation and information flow is concerned. Note in (2) how the resulting clouds packages up the effect of the preceding sentence by way of participating in the definition of cumulus clouds; and in the next sentence as a result of cold front conditions parcels up the cause of Australian rainfall. The paragraph as a whole is then summed up in relation to the accompanying figure, which shows how a cold front causes uplift and condensation. The causality may be simple but the explanation is sweet. (2) . . . This rising air becomes cooler for the reasons mentioned earlier, and if it is humid condensation of water vapour will take place. The resulting clouds are usually of the cumulus type. Much of the rain that falls in Australia occurs as a result of cold front conditions. Figure 7.7 shows how a cold front causes uplift and condensation in a warmer, humid, air mass.

As Bernstein has pointed out, horizontal knowledge structures like history develop by engendering new languages of description, rather than by resolving difference by integrating new perspectives into a single over-arching theory. So far we’ve been looking at what we might refer to as modernist history – a segment of history discourse which works towards a grand narrative (Lyotard 1984) involving men (sic) of wisdom and courage dicing with fate (lady luck) to somehow get us where we are today. Thus Tickner immortalizes Lingiari’s historic walk-off as abstractions affecting abstractions in Text 8 above. An alternative Marxist stance is illustrated in Text 9 below, which focuses on the nationalist movement in the Philippines beginning in the late nineteenth century. (9) This local elite contributed to the growing intellectual ferment and for a time gave direction to the movement for nationhood. But because of their predisposition to compromise and their capitulationist tendencies dictated by their material aspirations, they ultimately became an impediment to the national struggle. The Philippine Revolution was the result of the conjuncture of the unarticulated

48

LANGUAGE, KNOWLEDGE AND PEDAGOGY

strivings of the people and the articulations of the ideologues of the emerging elite. The Revolution represented a temporary amalgam of the particular interests of the elite and the general demands of the masses which eventually broke down into its respective components during and after the attainment of a national state and the subsequent incorporation of this new state into the American colonial empire. (Constantino and Constantino 1978: 1–2)

As highlighted in Text 9, Marxist history is more technical than modernist history because of the terms it borrows from Marxist economics and political thought. The modernist ‘cause in the clause’ syndrome is sustained, and so these terms enter into explanations as agents of change. In Text 9, for example, we find a causative preposition (because of), a causative nominal group (result) and causative processes (e.g. contributed, dictated) all relating abstractions of Marxist thought to one another. Where modernist history had abstractions acting on abstractions, Marxist history technicalizes these abstractions (cf. Wignell, this volume) – with humanities leaning towards social science as one segment of the field of history moves a little way down the cline of ‘verticality’ (Muller, this volume) which distinguished modernist history from the science discourse reviewed above. prepositional phrase: because of their predisposition to compromise and their capitulationist tendencies. . . nominal group: The Philippine Revolution was the result of the conjuncture of the unarticulated strivings of the people and the articulations of the ideologues of the emerging elite . . . human agents affecting nominalized abstractions: This local elite contributed to the growing intellectual ferment . . . abstractions affecting abstractions: . . . and their capitulationist tendencies dictated by their material aspirations . . . Another important segment in contemporary history discourse is postcolonial readings of the past, illustrated in Text 10 below with a passage from Rafael’s reading of the colonization of large parts of the Philippines by Spanish Catholicism. (10) The Spanish demand is that nothing be held back in confession. One is to expend all that memory can hold in a discourse that will bring together both the self that recalls and that which is recalled. The present self that confronts the priest in confession is thus expected to have managed to control his or her past – to reduce it, as it were, to discursive submission. Whereas the examination of conscience requires the division of the self into one that knows the Law and seeks out the other self that deviates from it, a ‘good confession’ insists on the presentation

CONSTRUING KNOWLEDGE

•

MARTIN

49

of a self in total control of its past. It is in this sense that confessional discourse imposes on the individual penitent what Roland Barthes called a ‘totalitarian economy’ involving the complete recuperation and submission of the past to the present, and by extension of the penitent to the priest. (Barthes 1976: 39–75; Rafael 1988: 101–3)

Once again the segment operates with borrowed technicality, this time from contemporary critical theory (in this case, the French masters Barthes, Lacan and Foucault). And once again the borrowed technicality is agentive, only this time round what becomes agentive is not social categories but discursive ones: a discourse that will bring together both the self that recalls and that which is recalled the examination of conscience requires the division of the self into one that knows the Law . . . confessional discourse imposes on the individual penitent what Roland Barthes called a ‘totalitarian economy’. Put simply, what holds these ‘languages’ together as discourses of history is their focus on explaining what happened over time, using cause in the clause to do so. What distinguishes them is what they use to explain. Modernist history nominalizes activity and gets the resulting abstractions acting on one another; Marxist history takes this step but also technicalizes abstractions, drawing on social science to do so; post-colonial history technicalizes abstractions as discourse, drawing on an alternative ‘critical’ canon. The concern with agency remains; what differs is what ‘acts’ – modernist abstractions, Marxist technicality or critical discourse.

4 Grammatical metaphor (the gatekeeper) In one of his more scintillating appendices, Bernstein (1975: 153) writes: (11) Imagine four lavatories. The first is stark, bare, pristine, the walls are painted a sharp white; the washbowl is like the apparatus, a gleaming white. A square block of soap sits cleanly in an indentation in the sink. A white towel (or perhaps pink) is folded neatly on the chrome rail or hangs from a chrome ring. The lavatory paper is hidden in a cover, and peeps through its slit.

This is pretty clear; even a dumb linguist can understand. Because in this passage people and things are realized as nouns, descriptive qualities as adjectives and processes as verbs:

50

LANGUAGE, KNOWLEDGE AND PEDAGOGY

participants as nouns: lavatory, walls, washbowl, apparatus, soap, sink, towel, paper qualities as adjectives: stark, bare, pristine, sharp, white, square processes as verbs: imagine, are painted, sits, is folded, is hidden, peeps Here Bernstein is writing as we speak; we can read the meaning directly off the grammatical categories. There is no nominalization. The text could be understood by a pre-pubescent child. Bernstein moves on: (11 continued) In the second lavatory there are books on a shelf, pictures on the wall, and some relaxing of the rigours of the first. In the third lavatory there are books on the shelf, pictures on the wall, and perhaps a scattering of tiny objects. In the fourth lavatory the rigour is totally relaxed. The walls are covered with a motley array of postcards, there is a wide assortment of reading matter and curios. The lavatory roll is likely to be uncovered and the holder may well fall apart in use.

Meaning is less straightforward here. Processes are starting to be realized as nouns. We can’t trust the grammar any more. Not all nouns are realizing people, places and things; some are actually processes in disguise. some relaxing of the rigours of the first a scattering of tiny objects in use To get the meaning we have to unpack the nominalizations, translating what the grammar actually says into what it means. So some relaxing of the rigours of the first means that the second lavatory is not organized as strictly as the first; a scattering of tiny objects means that tiny objects have been scattered about; and in use means when used. some relaxing of the rigours of the first (= ‘not organized as strictly’) a scattering of tiny objects (= ‘tiny objects are scattered (about)’) in use (= ‘when used’) Now Bernstein is writing more like academic writing in general, with meaning of all kinds drifting towards the noun. And not surprisingly, as part of this drift to ‘verticality’, we find cause in the clause:

CONSTRUING KNOWLEDGE

•

MARTIN

51

(11 continued) We can say that as we move from the first to the fourth lavatory we are moving from a strongly classified to a weakly classified space; from a space regulated by strong rules of exclusion to a space regulated by weak rules of exclusion.

Bernstein’s sociological abstractions (strong and weak rules of exclusion) now govern space. And the text is getting much harder for outsiders to understand. The process enabling the drift from spoken to written discourse we are monitoring here is referred to by Halliday as grammatical metaphor (Halliday 1998, 2004; Halliday and Matthiessen 1999, 2004; Simon-Vandenbergen et al. 2003). Politically speaking, grammatical metaphor is Halliday’s most important concept, because it names the process which engenders vertical discourse. Why does he call it metaphor? What does this process do? In discourse, grammatical metaphor complements lexical metaphor, so let’s begin with the latter. Bernstein writes that The lavatory paper is hidden in a cover, and peeps through its slit. Here the lexical metaphor turns on the word peep, which is used in such a way that there are two meanings involved (‘peeps’ and ‘just visible’), the meanings are layered with one readily available on the surface and another lurking deeper behind (‘peeps’ literal and ‘just visible’ transferred), and one meaning implies the other (‘peep’, which we might define as ‘look quickly and secretly’ symbolizing ‘just visible’). Grammatical metaphor works in the same way, except that it involves structures instead of words. For example, a structure such as a space regulated by strong rules of exclusion is used in such a way that there are two meanings involved, an abstract clause and a more concrete clause complex: Medium

Process

Agent

a space

regulated

by strong rules of exclusion

x .β

the space was regulated

α

by using strong rules to exclude things

They are layered (literal/transferred; figure/ground; surface/deep). Medium process Agent – literal & enhancing clause complex – transferred One implies the other (symbolization) ‘written’ Medium process Agent implying ‘spoken’ enhancing clause complex

52

LANGUAGE, KNOWLEDGE AND PEDAGOGY

In other words, what might well have been realized congruently in spoken language as a sequence of clauses (the space was regulated by using strong rules to exclude things) comes out in the more written form as a nominal group with a post-modifier in which a nominalization (exclusion) is agentive. What’s at issue here is whether grammar and semantics match. In a tri-stratal model of language such as that introduced in Figure 3.1, we’re asking how discourse semantic categories such as processes, participants, qualities and logical relations are related to lexicogrammatical ones such as verb, noun, adjective and conjunction. Because of the way language has evolved and the way we all learn it, there is a natural pairing off of processes with verbs, participants with nouns, qualities with adjectives and logical relation with conjunctions, as outlined in Figure 3.9 below.

quality logical relation

participant adjective

discourse semantics

noun conjunction

process

verb lexicogrammar

phonology/ graphology

Figure 3.9

Stratal harmony – grammar matching semantics

But as we have already seen, in abstract discourse we find processes, qualities and logical relations realized as nouns, and logical relations realized as verbs – the skewed realization outlined in Figure 3.10. From the perspective of grammar, the key derivational process here is nominalization. Looking across strata, the result is stratal tension – a non-matching relation between grammar and semantics. The result is that meaning can’t be read directly off the wording. We have to read grammatical metaphors twice – once with respect to what the

CONSTRUING KNOWLEDGE

•

53

MARTIN

grammar is, and once again with respect to what lies behind. And the meaning of the metaphor is more than the sum of its parts (grammar ‘times’ semantics, if you will).

quality logical relation

participant adjective

discourse semantics

noun conjunction

process

verb lexicogrammar

phonology/ graphology

Figure 3.10

Grammatical metaphor as stratal tension

We have already touched on the pay-offs for the construction of vertical discourse, using examples from science and history. Let’s consolidate them here, using examples from Bernstein’s sociology. First, participants. In vertical discourse, uncommon sense classification and composition operates on entities. So whatever we choose to classify or decompose, we have to ‘thingify’ it first. Beyond this, in technical disciplines, terms have to be defined; and definitions also operate on entities. So technicality too depends on ‘thingification’. Here, for example, are Bernstein’s definitions of vertical and horizontal discourse, a key step in his classification of discourse and knowledge structures (2000: 157, his italics), annotated with Halliday’s labels for the technical term (Token) and its definition (Value) in identifying structures of this kind: . . . a Vertical discourse (Token) takes the form of (=)

54

LANGUAGE, KNOWLEDGE AND PEDAGOGY

a coherent, explicit and systematically principled structure, hierarchically organized as in the sciences, or it takes the form of a series of specialized languages with specialized modes of interrogation and specialized criteria for the production and circulation of texts as in the social sciences and humanities. (Value) A Horizontal discourse (Token) entails (=) a set of strategies which are local, segmentally organised, context specific and dependent, for maximising encounters with persons and habitats. (Value) In short then, the uncommon sense organization of participants in vertical discourse depends on grammatical metaphor. From a functional linguistic perspective, if no grammatical metaphor, then no verticality. Complementing this is the role of grammatical metaphor in explanation (i.e. vertical sequencing). In the following passage Bernstein focuses on strategies for the regulation of knowledge in a community (2000: 158): (12) Consider a situation where a small holder meets another and complains that what he/she had done every year with great success, this year failed completely. The other says that when this happened he/she finds that this ‘works’. He/she then outlines the successful strategy. Now any restriction to circulation and exchange reduces effectiveness. Any restriction specialises, classifies and privatises knowledge. Stratification procedures produce distributive rules which control the flow of procedures from reservoir to repertoire. Thus both Vertical and Horizontal discourses are likely to operate with distributive rules which set up positions of defence and challenge.

As with the history and science explanations reviewed above, we find abstractions acting on abstracts, including technical abstractions indexing the field as social science. As with history, a wide repertoire of verbs is deployed, connected cause to effect (with abstractions labelled as Agent and Medium, after Halliday): Now any restriction to circulation and exchange (Agent) reduces effectiveness (Medium) Any restriction (Agent) specializes, classifies and privatizes knowledge (Medium) Stratification procedures (Agent) produce distributive rules . . . (Medium)

CONSTRUING KNOWLEDGE

•

MARTIN

55

distributive rules which (Agent) control the flow of procedures from reservoir to repertoire (Medium) . . . distributive rules which (Agent) set up position of defence and challenge (Medium) Once again then, this time as far as explaining is concerned, if no grammatical metaphor, then no verticality. This means that from a functional linguistic perspective access to vertical discourse is bound up with control of grammatical metaphor, which in western societies students are expected to master in secondary school. Failure to access this recourse entails exclusion from hierarchical and horizontal knowledge structures. Here lies the social semiotic nub of institutionalized learning, educational failure and the distribution of knowledge in our expiring world (Martin 1993a, b, 1999, 2000; Muller 2000; Taylor et al. 2003).

5 Genre and field (and value) As pointed out in section 1, my interest in field developed out of my interest in genre, as part of designing literacy programmes for secondary school. And as outlined in Figure 3.2, we used genre to model the interaction of the register variables field, tenor and mode. In these terms genre theory is concerned with how a culture maps ideational, interpersonal and textual meaning onto one another in phases as a text unfolds. Part of this interpersonal meaning is value, which SFL approaches from the perspective of appraisal theory (MackenHorarik and Martin 2003; Martin and White 2005). So an important aspect of text analysis has to be the infusion of field with value. Returning to Text 3 above, Minister Ruddock doesn’t just classify Australia’s future into three scenarios; he evaluates each forecast as he goes. ‘Goldilocks’ Ruddock tries the business position, which is too fast (trends alarmingly upswards, betrays a certain lack of realism); the extreme left green cum extreme right nationalist position, which is too slow (population slump, sags in a depressing arc); and then the government position, which is just right (reassuring, favourable outlook, rises gently and unthreateningly)! (3) The minister moves on to outline three competing visions for Australia’s population in the century ahead. The first scenario is the high-immigration model favoured by some business groups, which call for Australia’s net migration intake to be set at 1 per cent of existing population per year. The red line on the graph trends alarmingly upwards, predicting an ‘inevitably rising’ population, which would hit 65 million in the year 2007. According to the minister, the business groups’ goal of 1 per cent net migration betrays ‘a certain lack of realism’. The second scenario is net zero migration, the model pushed by sections of the environmental movement and by groups such as One Nation, which say that Australia

56

LANGUAGE, KNOWLEDGE AND PEDAGOGY

should take just enough migrants to replace the number of people who permanently depart the country each year. With falling birthrates, this would see Australia’s population slump from 20 million to 14 million within the next century. The green line on the graph sags in a depressing downward arc, heading inevitably towards zero. The minister’s final forecast is reassuring – according to him, if we hold fast to the current government policy, Australia’s population will increase gradually for the next forty years before settling comfortably at around 23 million. This favourable outlook is represented on the graph by a blue line, which rises gently and unthreateningly before charting a stable course between the extremes of green and red. (Mares 2001: 141–2)

This coupling of knowledge with value is an important dimension for any field. Because of the metafunctional organization of language and register in Figure 3.2, the natural place to bring together ideational and interpersonal meaning is at the level of genre. We could argue for example that Ruddock’s classification (field) and ranking of scenarios (tenor) come together in an evaluative taxonomizing report which positions readers to align themselves with a neoconservative government migration and refugee agenda. Alternatively, it may be that we should revise the model to allow for the coupling of content and values within the register variable field, since apprenticeship to any field always involves not just learning how to do things but learning what to feel about them as well. With respect to Bernstein’s model, this makes me wonder whether values would be considered an essential dimension of vertical knowledge structures. As far as I can see, there is nothing in Bernstein’s programmatic discussion precluding this, although from a linguistic perspective his analysis and exemplification seems to be purely ideational.

6 Linguistics as a knowledge structure Bernstein (e.g. 1996: 46–50) defines pedagogic discourse as a rule embedding ‘a discourse of skills and their relation to each other (instructional discourse)’ in ‘a discourse of social order (regulative discourse)’, the purpose of the device being to ‘produce a symbolic ruler for consciousness’. Following Christie 2002, I prefer the term projection to embedding, with the regulative discourse projecting the instructional one. ‘Sydney School’ literacy programmes in fact involve a doubling of the instructional discourse, since disciplines are factored as systems of genres (and thus of field, tenor, mode constellations and of their realization in language and attendant non-verbal modalities of communication); so in classroom practice a discipline such as history is projected by regulative discourse via its linguistic deconstruction (actually its social semiotic deconstruction since all relevant modalities of communication are considered) – in Bernstein’s terms, history is ‘embedded’ in linguistics (so we have two instructional discourses), which is in turn ‘embedded’ in regulative discourse. As far as the design of our interventions is concerned, social semiotics is used to deconstruct the pedagogic discourse (not just instructional discourse); so in

CONSTRUING KNOWLEDGE

•

MARTIN

57

theory, history is projected by regulative discourse projected by social semiotic ‘instructional’ discourse; some critical theorists have argued that pedagogy and curriculum should be negotiated between teachers and students along these lines, by way of challenging the fundamental asymmetry of the teacher/learner relationship (a step we have not taken as far as I am aware). Bernstein’s conceptual notation for embedding in pedagogic discourse is adapted in Figure 3.11 below. ID

ID

SSID

RD

RD

SSID

ID RD

Key: ID *SSID RD Figure 3.11

instructional discourse social semiotic instructional discourse regulative discourse

Projection relations in pedagogic discourse

Factoring a discipline as a system of genres clearly recontextualizes that discipline. In secondary school, content which a teacher conceives of in purely ideational terms is reworked with value and texture into the text types students need to read and write. As part of this the genres need to be named, along with their stages; and then there is the question of how much register, discourse semantics, lexicogrammar, graphology and image grammar is brought into play. What knowledge about language (KAL) do secondary history students need to know to learn grammatical metaphor, for example, especially those students who can’t learn it by osmosis the way motivated middle class students apparently do? In our experience, outside of the typical English classroom where KAL is seemingly taboo, the names of key genres and their stages can be mobilized, but as a first step much more linguistic metalanguage than this is felt to be going too far too fast. An exemplary factoring of history as a set of genres is outlined in Table 3.1, which has been ordered as a learner pathway, beginning with genres similar to those students will be familiar with in the oral culture outside of school, and moving on through those particular to history, including various chronicling genres and also argumentation. The same genres are presented as a system network, SFL style, in Figure 3.12 – arranged vertically along a topological vector running from common through more and more uncommon sense. For further discussion and illustraton see Coffin 1997, 2000; Martin 2002a; Martin and Wodak 2004; Martin and Rose 2005.

58

Table 3.1

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Factoring modernist history as genre (learner pathway)

Genre [staging]

Informal description

personal recount [OrientationˆRecord]

agnate to story genres; what happened to me

autobiographical recount [OrientationˆRecord]

the story of my life [oral history]

biographical recount [OrientationˆRecord]

the story of someone else’s life

historical recount; [BackgroundˆRecord]

establishing the time line of ‘grand narrative’

historical account; [BackgroundˆAccount]

naturalizing linearization of ‘grand narrative’ as causal

factorial explanation [OutcomeˆFactors]

complexifying notion of what leads on to what

consequential explanation [InputˆConsequences]

complexifying notion of what leads on from what

exposition – one sides; promote [ThesisˆArguments]

problematic interpretation that needs justifying

challenge – one sided; rebut [PositionˆRebuttal]

problematic interpretation that needs demolishing

discussion – multi-sided; adjudicate [IssueˆSidesˆResolution]

more than one interpretation considered

Hurdling

As far as vertical discourse is concerned, what kind of recontextualization is this? One big transformation is the difference in the level of technicality. Linguistics is arguably the most technical discipline in the humanities, and possibly social sciences as well – the ‘science of language’ as it is often portrayed. There is also a fair degree of ‘verticality’ in its knowledge structure. As in science, networks for participants are deep, comprehensive and precise; and composition is an exacting discipline, as anyone trained in text generation and parsing can confirm. Explanation is perhaps less developed, since linguists are still unsure what their data is (intuitions about grammaticality, selected texts or corpora) and how to manage it (quantitatively, qualitatively); and all dimensions of language change are relatively poorly understood – logogenesis (unfolding text), ontogenesis (language development in the individual) and phylogenesis (language evolution and death). Teachers coming from horizontal discourses with relatively weak grammars (as Bernstein puts it) find this array of technicalization a shock; it’s hard to recognize one’s subject matter for the trees as it were. So the argument that literacy and thus content learning in fact depends for many students on projecting subject matter through SFL is an easy one to lose. The new knowledge about language required costs time and therefore money, teachers are busy, and stratified learning outcomes are blamed on the ability level of individual students who don’t learn grammatical metaphor

CONSTRUING KNOWLEDGE

•

first person biographical third person

temporal field time

historical causal

autobiographical recount

Common sense

biographical recount

historical recount

historical account inputs

explain

59

MARTIN

factorial explanation

outcomes

consequential explanation

text time

promote

exposition

one-sided rebut

argue multi-sided

Figure 3.12

discussion

challenge Uncommon sense

Typology of history genres

because they can’t (or perhaps won’t: Rose 2004). But democracy aside, are these stratified outcomes something we can still afford? In Bernstein’s terms, for all its scientific aspirationalism, linguistics is a horizontal knowledge structure, made up of competing languages of description – formal and functional schools for example, and many sub-varieties within, between and around blunt categorizations such as this. As Muller (this volume) comments, specific segments of a horizontal knowledge structure with strong grammars may exhibit a degree of verticality that approximates that idealized by philosophers of science for science proper (especially physics). And SFL is certainly a segment of this kind – not just in its technicality, but in its evolutionary path of development. The theory is robust enough to fit new perspectives in (for example, genre as a deeper stratum, appraisal in interpersonal discourse semantics, multimodality as mode, corpus counting in relation to probabilitistic systems, and so on). Bernstein (2000: 161) uses the image of a triangle as a model for hierarchical knowledge structures, to capture the sense in which they strive for general and abstract propositions integrating knowledge at lower levels. This metaphor is taken up in Wignell’s (this volume) characterization of social sciences as warring triangles. And this seems an apt characterization of SFL and its rival theories of language, each developing what they see as the kind of verticality they need to wipe out the others and become THE science6 of language linguists are bred to need. Lack of progress in this war confirms linguistics as precisely the kind of horizontal knowledge structure Bernstein recognizes it to be.

60

LANGUAGE, KNOWLEDGE AND PEDAGOGY

One comment I would make here, however, is that Bernstein’s typology of knowledge structures, with hard science at one end and soft humanities at the other, is a relatively modernist conceit. It may well be that a social semiotic theory is a different kind of knowledge structure from those dealing with the physical, biological and social world. In this regard, the notion of an apex in Bernstein’s triangle metaphor is a potentially misleading one, at least as far as SFL is concerned. The theory is more of a network of parameters than a pyramid with a broad bottom and a narrow tip – including several scales (e.g. delicacy, rank, realization, instantiation and individuation), and complementarities (e.g. metafunction, agnation, modality, axis, perspective, and of course the scales just reviewed). A rough characterization of the bits and pieces of this evolving theoretical web is offered below (including in parentheses some comparable conceptualizations from Bernstein’s work). SFL scales (magnification): rank delicacy realization instantiation individuation

clause–group/phrase–word–morpheme primary–secondary–tertiary . . . (forms of discourse: knowledge) phonology–lexicogrammar–discourse semantics– register–genre system–register–text type–text–reading (real/empirical) community–coding orientation–identity (reservoir/repertoire)

SFL complementarities (multinocular vision): metafunction ideational/interpersonal/textual (classification/framing) axis system/structure agnation typology/topology (knowledge map/pedagogy matrix) perspective synoptic/dynamic modality (MDA) language/image/music/architecture . . . (discourse/lavatory) (and scales above) For an overview of these parameters see Halliday and Matthissen 1999, which taken alongside Halliday 2003 suggests that knowledge structures dealing with social semiosis are much more sophisticated than anything developed by science because of the complexity of the phenomena in their gaze. Thus one day, paraphrasing Whorf, science (and I would extend this to include all vertical discourse) will have to look to linguistics to find out what it means.

7 Envoi In this chapter I have attempted to sketch out various parameters involved in a social semiotic interpretation of discourses and knowledge structures as field – using science and history as exemplars. Central to this overview was the introduction of grammatical metaphor as the critical linguistic resource which has evolved in written cultures for the construction of vertical discourse. Because of

CONSTRUING KNOWLEDGE

•

MARTIN

61

its ability to reconstrue any meaning as a thing, and thereby manage causality inside the clause, it is grammatical metaphor that enables uncommon sense classification, composition and explanation right across humanities, social science and science. The power of these discourses to consume our ecosystem (via technology) and manage populations (via bureaucracy) engenders both their prestige and the alienation their privileging verticality entails. Thanks to Bernstein and Halliday our legacy comprises a sociology with a place for language and a linguistics with a commitment to the social. These will never be one theory, one knowledge structure; nor would this loss of incommensurability be in any way desirable. But shared political commitment, especially with a focus on education, has maintained dialogue through to a second generation of research. I sense that if we can sustain this mutual interrogation, our understandings can grow; and more importantly, our interventions can prosper.

References Bernstein, B. (1975), Towards a Theory of Educational Transmissions. Class, Codes and Control Vol. 3. London: Routledge & Kegan Paul. Bernstein, B. (1990), The Structuring of Pedagogic Discourse. London: Routledge. Bernstein, B. (1996), Pedagogy, Symbolic Control and Identity: theory, research, critique (Revised edition, 2000). London: Taylor and Francis. Christie, F. (1992), ‘Literacy in Australia’, in ARAL 12, pp. 142–55. Christie, F. (ed.) (1999), Pedagogy and the Shaping of Consciousness: linguistic and social processes. London: Cassell (Open Linguistics Series). Christie, F. (2002), Classroom Discourse Analysis. London: Continuum. Christie, F. and Martin, J.R. (eds) (1997), Genre and Institutions: social processes in the workplace and school. London: Cassell. Coffin, C. (1997), ‘Constructing and giving value to the past: an investigation into secondary school history’, in Christie and Martin (eds), pp. 196–230. Coffin, C. (2000), ‘Defending and challenging interpretations of the past’, in Revista Canaria de Estudios Ingleses 40, pp. 135–54. Cope, W. and Kalantzis, M. (eds) (1993), The Powers of Literacy: a genre approach to teaching literacy. London: Falmer and Pittsburgh: University of Pittsburgh Press. Eggins, S., Martin, J.R. and Wignell, P. (1993), ‘The discourse of history: distancing the recoverable past’, in Ghadessy, M. (ed.), Register Analysis: theory and practice. London: Pinter (Open Linguistics Series), pp. 75–109. Feez, S. (1998), Text-based Syllabus Design. Sydney: National Centre for English Language Teaching and Research (NELTR), Macquarie University. Feez, S. (2002), ‘Heritage and innovation in second language education’, in Johns, A.M. (ed.), pp. 43–69. Ghadessy, M. (ed.) (1993), Register Analysis: theory and practice. London: Pinter. Ghadessy, M. (ed.) (1999), Text and Context in Functional Linguistics (CILT Series IV). Amsterdam: Benjamins. Haire, M., Kennedy, E., Lofts, G. and Evergreen, M. (2005), Core Science: Stage 5 Essential Content (2nd edition). Milton, Qld: John Wiley & Sons. Halliday, M.A.K. (1998), ‘Things and relations: regrammaticising experience as technical knowledge’, in Martin and Veel (eds), pp. 185–235. Halliday, M.A.K. (2003), On Language and Linguistics (Vol. 3 in The Collected Works of M.A.K. Halliday, ed. J. Webster). London: Continuum. Halliday, M.A.K. (2004), The Language of Science. (Vol. 5 in The Collected Works of M.A.K. Halliday, ed. J. Webster). London: Continuum.

62

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Halliday, M.A.K. and Martin, J.R. (1993), Writing Science: literacy and discursive power (Critical Perspectives on Literacy and Education). London: Falmer. Halliday, M.A.K. and Matthiessen, C. (1999), Construing Experience through Language: a language-based approach to cognition. London: Cassell. Halliday, M.A.K. and Matthiessen, C. (2004), An Introduction to Functional Grammar (3rd edition). London: Arnold. Hasan, R. and Williams, G. (eds) (1996), Literacy in Society. London: Longman. Johns, A.M. (ed.) (2002), Genres in the Classroom: applying theory and research to practice. Mahwah, NJ: Lawrence Erlbaum. Kress, G. and van Leeuwen, T. (1996), Reading Images: the grammar of visual design. London: Routledge. Kress, G. and van Leeuwen, T. (2001), Multimodal Discourse – The Modes and Media of Contemporary Communication. London: Arnold. Kress, G., Jewitt, C., Ogborn, J. and Tsatsarelis, C. (2001), Multimodal Teaching and Learning: the rhetorics of the classroom. London: Continuum. Lemke, J.L. (1990), Talking Science: language, learning and values (Language and Educational Processes). Norwood, NJ: Ablex. Lyotard, J. (1984), The Postmodern Condition. Minneapolis: University of Minnesota Press. Macken-Horarik, M. (2002), ‘Something to shoot for: a systemic functional approach to teaching genre in secondary school science’, in Johns (ed.), pp. 17–42. Macken-Horarik, M. (1998), ‘Exploring the requirements of critical school literacy: a view from two classrooms’, in F. Christie and R. Misson (eds), Literacy and Schooling. London: Routledge, pp. 74–103. Macken-Horarik, M. and Martin, J.R. (eds) (2003), ‘Negotiating heteroglossia: social perspectives on evaluation’, (Special Issue of Text 23.2). Martin, J.R. (1992), English Text: system & structure. Amsterdam: Benjamins. Martin, J.R. (1993a), ‘Genre and literacy – modelling context in educational linguistics’, in Annual Review of Applied Linguistics 13, pp. 141–72. Martin, J.R. (1993b), ‘Technology, bureaucracy and schooling: discursive resources and control’, in Cultural Dynamics 6.1, pp. 84–130. Martin, J.R. (1993c), ‘Life as a noun’, in Halliday and Martin, pp. 221–67. Martin, J.R. (1995), ‘Text and clause: fractal resonance’, in Text 15.1, pp. 5–42. Martin, J.R. (1998), ‘Practice into theory: catalysing change’, in S. Hunston (ed.), Language at work. Clevedon: Multilingual Matters (British Studies in Applied Linguistics 13 [BAAL 13]), pp. 151–67. Martin, J.R. (1999), ‘Linguistics and the consumer: theory in practice’, in Linguistics and Education 9.3, pp. 409–46. Martin, J.R. (2000), ‘Design and practice: enacting functional linguistics in Australia’, in Annual Review of Applied Linguistics 20 (20th Anniversary Volume: ‘Applied Linguistics as an Emerging Discipline’), pp. 116–26. Martin, J.R. (2001), ‘Giving the game away: explicitness, diversity and genre-based literacy in Australia’, in R. de Cilla, H. Krumm and R. Wodak et al. (eds), Functional Il/literacy. Vienna: Verlag der Osterreichischen Akadamie der Wissenschaften, pp. 155–74. Martin, J.R. (2002a), ‘Writing history: construing time and value in discourses of the past’, in C. Colombi and M. Schleppegrell (eds), Developing Advanced Literacy in First and Second Languages. Mahwah, NJ: Erlbaum, pp. 87–118. Martin, J.R. (2002b), ‘From little things big things grow: ecogenesis in school geography’, in R. Coe et al. (eds), The Rhetoric and Ideology of Genre: strategies for stability and change. Cresskill, NJ: Hampton Press, pp. 243–71. Martin, J.R. and Rose, D. (2003), Working with Discourse: meaning beyond the clause. London: Continuum. Martin, J.R. and Rose, D. (2006), Genre Relations: mapping culture. London: Equinox. Martin, J.R. and Veel, R. (eds) (1998), Reading Science: critical and functional perspectives on discourses of science. London: Routledge.

CONSTRUING KNOWLEDGE

•

MARTIN

63

Martin, J.R. and White, P.R. (2005), The Language of Evaluation: appraisal in English. London: Palgrave. Martin, J.R. and Wodak, R. (eds) (2004), Re/reading the past: critical and functional perspectives on discourses of history. Amsterdam: Benjamins. Muller, J. (2000), Reclaiming Knowledge: social theory, curriculum and education policy (Knowledge, Identity and School Life Series 8). London: Routledge. Muller, J., Davies, B. and Morais, A. (eds) (2004), Reading Bernstein, Researching Bernstein. London: RoutledgeFalmer. Ogburn, J., Kress, G., Martins, I. and McGillicuddy, K. (1996), Explaining Science in the Classroom. Buckingham: OUP. Painter, C. (1998), Learning through Language in Early Childhood. London: Cassell. Rose, D. (1997), ‘Science, technology and technical literacies’, in Christie and Martin, pp. 40–72. Rose, D. (1998), ‘Science discourse and industrial hierarchy’, in Martin and Veel, pp. 236–65. Rose, D. (2004), ‘Sequencing and Pacing of the Hidden Curriculum: how indigenous children are left out of the chain’, in Muller et al., pp. 91–107. Schleppegrell, M.J. (2004), The Language of Schooling: a functional linguistics perspective. Mahwah, NJ: Erlbaum. Simon-Vandenbergen, A.-M., Taverniers, M. and Ravelli, L.J. (eds) (2003), Metaphor: systemic and functional perspectives. Amsterdam: Benjamins. Taylor, N., Muller, J. and Vinjevold, P. (2003), Getting School Working: research and systemic school reform in South Africa. Cape Town: Peason Education. Unsworth, L. (1997a), ‘Scaffolding reading of science explanations: accessing the grammatical and visual forms of specialised knowledge’, in Reading 31.3, pp. 30–42. Unsworth, L. (1997b), ‘Explaining explanations: enhancing scientific learning and literacy development’, in Australian Science Teachers Journal 43.1, pp. 34–49. Unsworth, L. (1998), ‘“Sound” explanations in school science’, in Linguistics and Education 9.2, pp. 199–226. Unsworth, L. (ed.) (2000), Researching Language in Schools and Communities: functional linguistics approaches. London: Cassell. Unsworth, L. (2001), Teaching Multiliteracies across the Curriculum: changing contexts of text and image in classroom practice. Buckingham: Open University Press. White, P. (1998), ‘Extended reality, proto-nouns and the vernacular: distinguishing the technological from the scientific’, in Martin and Veel, pp. 266–96. van Leeuwen, T. and Humphrey, S. (1996), ‘On learning to look through a geographer’s eyes’, in Hasan and Williams, pp. 29–49. Veel, R. (1992), ‘Engaging with scientific language: A functional approach to the language of school science’, in Australian Science Teachers’ Journal 38.4, pp. 31–5. Wignell, P., Martin, J.R. and Eggins, S. (1990), ‘The discourse of geography: ordering and explaining the experiential world’, in Linguistics and Education 1.4, pp. 359–92. (Republished in Halliday and Martin 1993, pp. 136–65.)

WORKPLACE LITERACY REPORTS:

1

2

3

Rose, D., McInnes, D. and Korner, H. (1992), Scientific Literacy (Write it Right Literacy in Industry Research Project – Stage 1). Sydney: Metropolitan East Disadvantaged Schools Program. Iedema, R., Feez, S. and White, P. (1994), Media Literacy (Write it Right Literacy in Industry Research Project – Stage 2). Sydney: Metropolitan East Disadvantaged Schools Program. Iedema, R. (1995), Literacy of Administration (Write it Right Literacy in Industry Research Project – Stage 3). Sydney: Metropolitan East Disadvantaged Schools Program.

64

LANGUAGE, KNOWLEDGE AND PEDAGOGY

SECONDARY SCHOOL LITERACY PACKAGES:

English Rothery, J. (1994), Exploring Literacy in School English (Write it Right Resources for Literacy and Learning). Sydney: Metropolitan East Disadvantaged Schools Program. Rothery, J. and Stenglin, M. (1994), Spine-Chilling Stories: a unit of work for Junior Secondary English (Write it Right Resources for Literacy and Learning). Sydney: Metropolitan East Disadvantaged Schools Program. Rothery, J. and Stenglin, M. (1994), Exploring Narrative in Video: a unit of work for Junior Secondary English (Write it Right Resources for Literacy and Learning). Sydney: Metropolitan East Disadvantaged Schools Program. Rothery, J. and Stenglin, M. (1994), Writing a Book Review a unit of work for Junior Secondary English (Write it Right Resources for Literacy and Learning). Sydney: Metropolitan East Disadvantaged Schools Program. Geography 1 Humphrey, S. (1996), Exploring Literacy in School Geography. Sydney: Metropolitan East Disadvantaged Schools Program. 2 Humphrey, S. and Takans, P. (1996), Explaining the Weather: a unit of work for Junior Secondary Geography. Sydney: Metropolitan East Disadvantaged Schools Program. 3 Sikes, D. and Humphrey, S. (1996), Australia – Place and Space: a unit of work for Junior Secondary Geography. Sydney: Metropolitan East Disadvantaged Schools Program. History 1 Coffin, C. (1996), Exploring Literacy in School History. Sydney: Metropolitan East Disadvantaged Schools Program. 2 Brook, R., Coffin, C. and Humphrey, S. (1996), Australian Identity: a unit of work for Junior Secondary History. Sydney: Metropolitan East Disadvantaged Schools Program.

Notes 1

2. 3

4

5 6

For overviews of this genre-based literacy pedagogy see Christie 1992; Cope and Kalantzis 1993; Feez 1998, 2002; Macken-Horarik 2002; Martin 1993a, 1998, 1999, 2000, 2001. The Write it Right project workplace and secondary school literacy reports documenting this research are listed at the end of the references. The ways in which participants configure with participants and circumstances via processes as part of activity sequences is of course part of this picture, but we will pass over this dimension of analysis here; see Martin 1992 for discussion. Generally referred to as implication sequences in our work on science because of their ‘if a happens, then b ensues’ structure; see Halliday and Martin 1993 for discussion. For discussion of pre-school learning in the home, see Painter 1998. Somewhat ironically, it’s formal linguists, who try to present themselves as more scientific than others, who are most responsible for proliferating languages of description in linguistics, rolling their theories over in a post-Fordist carnival to ensure jobs for each new wave of graduate students; since these theories aren’t actually about very much of language, and not really used for anything outside the discipline, they are easy to change – and the power of their self-righteous exclusionary authority is maintained.

4

On splitting hairs: hierarchy, knowledge and the school curriculum Johan Muller

To me, wherever there is pedagogy there is hierarchy. What is interesting, it’s the language of description that we use, because the language of description masks hierarchy, whereas the language of description should attempt to sharpen its possibility of appearance. (Bernstein, in Morais et al. (eds) (2001), p. 375) In every hair there are an infinite number of lions, and in addition all the single hairs, together with their infinite number of lions, in turn enter into a single hair. In this way the progression is infinite, like the jewels in Celestial Lord Indra’s net. (Fa-Tsang, T’ang Dynasty, quoted in Collins 1998, p. v)

The trouble with hierarchy The set of papers in this volume can loosely be said to be about hierarchy in discourse – in language, knowledge, and pedagogy. The two disciplinary communities represented in this volume do not deal with hierarchy in the same way, but they are at least endeavouring to find common ground for representing it in the same discursive space. This is a large step forward. In fact, hierarchy is a word we generally shun in our lexicons, infused as they are with one or other kind of egalitarianism, which is generally speaking a good thing. Deeply embedded in our egalitarian zeitgeist is the unassailable assumption that hierarchy in discourse and hierarchy in society are connected, together with a corollary assumption that one inoculates oneself from the latter by avoiding talk about the former. Alas, sympathetic magic of this sort works about as well in social science as it does in folk medicine. Our two communities are on the face of it engaged in nicely complementary projects. The linguists are engaged in establishing what the building blocks of hierarchy are, while the sociologists of education are engaged in establishing how hierarchy is distributed. Both communities have made some progress, but neither community has settled the issue of what exactly discursive hierarchy derives from, or what knowledge hierarchy is. There are at least three obstacles that I can see. The first is that we are locked into an early (lexical) metaphorical stage of discussion, where the terms are more suggestive than they are explanatory, and where use of the same term does not guarantee equivalent meaning. Secondly, the egalitarian ethos seems to drive us to use terms that obdurately suggest variety without hierarchy – discovery, variation, and the like. Thirdly, a currently influential trend in social science that I call the ‘New

66

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Cynicism’ below, denies the very existence of discursive hierarchy, making all talk about it sacrilegious. Fortunately, the two communities gathered together in this volume seem united in the view that the devil you know is better than the one you don’t, and that understanding hierarchy in discourse is a necessary prelude to combating it in society. That is promising indeed. Though we routinely avoid addressing it directly, we willy-nilly imply hierarchy: to take an example almost at random, Hasan (2001) refers to ‘higher mental functions’, after Vygotsky. One infers that ‘higher’ functions enable greater abstraction, hence greater combinatorial power. But what about the levels in the parent discourse within which this higher abstraction is embedded? From a constructivist ‘knower’ perspective, it is almost as if each neophyte knower invents the levels ab initio. From a ‘knowledge’ perspective, where it is the distributive potential of different kinds of knowledge that is at stake, it is precisely knowledge hierarchy that conditions distributive potential. So there is no avoiding it. Martin (this volume) tackles it head on. Starting with Bernstein’s common sense-uncommon sense cline, Martin produces a branching diagram that, despite proceeding in the opposite direction to Bernstein’s (2000: 168) own branching diagram, anticipates its principal features in a quite remarkable way.1 The principal similarity is that both Martin and Bernstein use the branching tree device to complicate and fill in the continuum between the two ends of the spectrum. With this, the explanatory implication of the tree shifts from dichotomy to continuity. The question then becomes: how far do we want to push the implication? Do we want to imply that there is a complete continuum, or are there categorical differences between some of the symbolic ensembles unbranched by the tree? The answer depends upon whether one is primarily interested in knowledge as meaning or knowledge as distributed social goods. If one is interested in knowledge as meaning, then one is primarily interested in describing the universal semantic building blocks that enable transition from one form to the other. Martin goes on exemplarily to show how grammatical metaphor ‘engenders’ the ‘drift’ from spoken to written discourse, from horizontal to vertical discourse (Martin, this volume). If one is primarily interested in knowledge as distributed social goods, then one is interested in describing the way both forms have distributive rules which are in turn conditioned by discontinuities in semiotic structure that mirror, sustain and reproduce inequalities in society. For the first, the explanatory task is to uncover the most basic universal processes and hence to reveal the ideal underlying unity of semiosis; for the second project, the task is to delineate the social limits to distributive equality. The first is, if you like, a classical, ultimately optimistic enterprise; the second is a fundamentally tragic enterprise, no matter how optimistically driven. My intention is not to separate the enterprises. Quite the opposite. I hope to sharpen the difference in the projects and starting points the better to make visible where the projects at present overlap and where bridge-building should concentrate. Both Bernstein and Martin distinguish between discourse forms tied to empirical particulars (to ‘context’), and those which do not depend upon the world in order to make sense, which in fact float free from it. For Bernstein, a fundamental distinction between what he called horizontal and vertical

HIERARCHY, KNOWLEDGE & THE SCHOOL CURRICULUM

•

MULLER

67

discourse is that the latter is integrated at the level of meanings allowing decontextualization, the former at the level of (culturally specialized) segments binding the language to context. In his contrast between ‘everyday’ and ‘scientific’ taxonomies, Martin makes the same point. A key characteristic of everyday categories is that they are largely learnt ostensively or by modelling, that is, by modes of discursive action that require the coincidence of time and space, for Bernstein placing a constraint on their distribution. In middle class homes, as Painter (1999) shows, the transition from ostensive definition to decontextualized definition (to technical lexis), ‘allowing the speaker to attend to the meaning rather than the referent’ (Painter 1999: 82), is pretty well seamless, at least for young Stephen busy acquiring his semantic style (see also Painter, this volume). Bernstein would be in full agreement, but would want to make something more of the social gulf between the two forms of discourse. To see this, it is useful to take a small detour via Durkheim. In The Elementary Forms of the Religious Life Durkheim (1976) famously draws a distinction between two orders of existence which relate thought and practice in two fundamentally different ways. The first order is the profane world, the everyday world of ‘sensual representations’, the world of matter and sense, where meaning arises directly out of bodily encounters with the world, with other people, with reality. It is a world of flux and of particulars, and it is driven by the most practical and direct wisdom: proverbs, prudence, street lore, on-thejob knowledge, the rhythmic language and wisdom of the domestic community. The second order is the sacred (originally religious) world, one of prescriptions and interdicts that are not pragmatically modifiable but are ‘fixed and crystallised’ (Durkheim 1915: 433). This sacred world is an order of verities not originating in bodily hexis, and therefore arbitrary, in Pierce’s sense of unmotivated: taboos, for example, can be attached to any object. The religious world is thus a world of arbitrary conceptual relations, a symbolic order constructed by an accretion of ‘collective representations’ (op. cit.: 434) that are a collective accomplishment, the ‘work of the community’, in contrast to the ‘sensual representations’ of the everyday world that are the work of continually changing experiential particulars shared traditionally in face-to-face encounters (see Muller 2000). Religion is for Durkheim the primary cognitive classificatory scheme of the sacred, the primary form of ordering social representations in non-empirical, formal ways. The force of the ordering comes from ‘outside of the object in which it resides’ (quoted in Thompson 1994: 125), not from the object itself. It is the result of a process of ‘examination and elaboration’ (op. cit.: 126): it is the result of a cognitive process of idealization, or ‘schematic idealization’ in Martin’s terms (p. 8). Durkheim means at least two things with this faculty of idealization. The first is the purely cognitive or speculative sense of being able ‘to connect things with each other, to establish internal relations between them, to classify them and to systematize them’ (op. cit.: 133). The second is that of forward projection towards an order and a world more desirable, more felicitous, more powerful – in a word, better – than the one we have in hand at any specific point in history. Durkheim plays upon the double sense of ideal: ideal first as the facility to

68

LANGUAGE, KNOWLEDGE AND PEDAGOGY

manipulate objects and relations in non-empirical virtual space; ideal secondly as the projection into and towards that which is more desirable. Both together allow us to break with empirical facticity and to imagine an ordering of objects that is ‘logical’ and ‘hierarchical’ (op. cit.: 137). This is a key feature of virtual connections that allows, as Foucault (1981: 59) says in a related idiom when discussing disciplinarity, ‘the possibility of formulating new propositions, ad infinitum’, or as Hacking (1985: 156) says when discussing styles of reasoning, to ‘generate new classes of possibilities’. No surprise then that the exemplary form that the sacred takes in societies with a complex division of labour is science. The sacred and the profane are thus two fundamental orders present in all forms of society. The principal purpose of the strong classification between the two is to sustain the fundamental social cleavage between mental and manual labour, and to reproduce it. As easy as it is for Stephen with the help of Stephen’s mother to traverse the semiotic gap (Painter 1999), as difficult is it to close the social gap. For Bernstein, this is because of the bias in the distributive rules that regulate access to powerful discourses, to highly specialized forms of consciousness. The more differentiated the division of labour, the more differentiated will be the distribution of these sacred goods. This may seem to damn both Bernstein and Durkheim as incorrigible pessimists, yet, though they were both frequently derided as functionalists, they were both anything but pessimists. To say that power arises from the social base is not to say that power is monolithically and automatically reproduced. It is simply to say that symbolic configurations and their distributive potential are allied to socio-economic conditions. To understand how the symbolic configurations are reproduced is to enquire into the way these symbolic ensembles are recontextualized, and how the recontextualized discourses are acquired. This is to enquire into the workings of symbolic control and its agencies, which is to say, the education system, which is, as we used to say, a site of struggle within which much can be done. David Rose has shown this at the level of the classroom (Rose 2004), and Cuba has shown it at the level of the state, outperforming its fellow Latin neighbours for reasons that Durkheim and Bernstein (and Rose) would have applauded, namely, because the state has realized that reducing social inequality across the society was an important corollary to a large-scale equalizing of the distribution of educational competencies (Carnoy, Gove and Marshall 2004). Returning briefly to Martin’s cline, it is worth observing that the initial categories of his cline are oral transmission and written transmission. Bernstein almost seems to shy away from this contentious area,2 but it is noteworthy that his exemplification of the way everyday wisdom circulates has everything to do with its face-to-face (oral) nature (Bernstein 2000: 158). In both Martin and Bernstein thus we find echoes of the dual advantages of writing, both cognitive and social. On the one hand, as Collins (1998: 27) reminds us after Goody, Ong and Havelock, sustained writing is a ‘gateway to abstraction and generality’; on the other, writing breaks the strictures of time/space coincidence that mark oral cultures: ‘What is needed is a social arrangement for writing texts of some length and distributing them to readers at a distance . . .’ (op. cit.). I have been making the claim in this introductory section that while both sociologists and

HIERARCHY, KNOWLEDGE & THE SCHOOL CURRICULUM

•

MULLER

69

linguists share this dual interest, linguists concentrate on the former – how texts are specialized – while sociologists concentrate on the latter – how texts are distributed. Not all sociologists concentrate on distribution from the field of production to the fields of reproduction and acquisition, as Bernstein does. It may be interesting to examine briefly how an exemplary sociologist of science, Robert Merton, describes the distributional consequences of discursive structure for the discourse producers (rather than acquirers) – how disciplinary differences affect their practising scholars. Merton first establishes in a terse formulation his version of (a part of) the cline, the terms in which specialized disciplines differ: ‘Codification refers to the consolidation of empirical knowledge into succinct and interdependent theoretical formulations. The various sciences and specialities within them differ in the extent to which they are codified’ (Merton 1973: 507; see also Foray and Hargreaves 2003). Degree of codification has a series of consequences, three of which are the following:3 •

•

•

There is a higher rate of obsolescence in C+ than in C– disciplines, because they display a greater tendency to subsume past work. One consequence is that there is a greater percentage of references to recent rather than older work in C+ than is the case in C– disciplines. In C+ disciplines innovative work by young scholars is more easily recognized than in C– disciplines, where it is easier to be overlooked, leading to what Merton calls the Matthew Effect (Merton, op. cit.: 516), from St Matthew: to him who hath shall be given, etc. Young scholars find it difficult to break into C– disciplines. One consequence is that there are age differences in discovery patterns, summed up famously by Caius Asinius Pollio in Robert Graves’ I Claudius: ‘Science is a young man’s game’ while ‘history is an old man’s game’ (op. cit., fn. 39: 513), or Merton, with the irony for which he was famous: ‘This sort of thing can thus foster the illusion that good mathematicians die young, but that, say, good sociologists linger on forever’ (op. cit.). Induction into C+ entails grasping high-level propositions; into C–, into learning masses of particulars. Induction opportunity costs consequently differ.

Merton’s discussion of degrees of codification rings many bells in popular as well as esoteric literature: recall Bertrand Russell’s distinction between ‘hard’ and ‘soft’ disciplines, a distinction formalized by Tony Becher (Becher 1989; see also Becher and Parry 2005) using a version of Bernstein to aid him. It rings bells in the work of self-styled Bernsteinian heretic Paul Dowling (1999), who has coined the concept of discursive saturation to replace Bernstein’s code and discourse cline. Dowling’s objection to code/discourse involves an objection to the separation of the dimension of classification from that of framing. For Dowling, classification, the carrier of power and the distributive rules, is not to be separated from its semiotic enactments (framing), thus there is only high discursive saturation (DS+) and low discursive saturation (DS–); the former is specialized by generalizing strategies, the latter by localizing strategies. This is a stunningly original elaboration of the theory in its pedagogic dimension, that is,

70

LANGUAGE, KNOWLEDGE AND PEDAGOGY

in establishing which pedagogic strategies belong with which kinds of DS mode, although the core distinctions are arguably latent in Bernstein’s tacit and explicit transmission distinction. Yet in its predominant focus on texts and pedagogy, it takes us away, once again, from the field of production, from knowledge and how it grows, from hierarchy.4 The codification cline rings bells for Bernstein and Martin too, but I believe the affinity is greater with Martin than with Bernstein, the sociological congruence around distribution notwithstanding. This is because both Merton and Martin are pursuing a notion of textual specialization within knowledge structure across a single graded continuum, while Bernstein, as I hope to make clearer below, seeks to find the heart of the discontinuity between the way the two ends of the codified spectrum grow, and progress. The rest of the chapter is an inquiry into hierarchy from the point of view of the question of knowledge progress.

Knowledge and the dilemma of progress5 The reluctance to speak directly about hierarchy and its cognate, progress, referred to above, is an old one, and below I will trace its roots to the terms of a debate in the eighteenth century at the advent of the age of science, continuing to the present. This debate is about the idea of progress in general, and the idea of progress in knowledge in particular. We are, it would seem, exceedingly reluctant to speak about the social dimensions of knowledge hierarchy, not only in terms of relations between different knowledge forms, but particularly in terms of relations within knowledge forms. The idea of hierarchy haunts us, nowhere more so than in regard to the question of knowledge progression and growth. Bernstein has intervened decisively in the discussion about the forms of symbolic systems, setting out to delineate the ‘internal principles of their construction and their social base’ (Bernstein 2000: 155). As is by now well known, he distinguishes between two forms of discourse, horizontal and vertical, as the discussion above made clear. From here on, this chapter will not discuss the question of discourses further and will concentrate on the question of variation between knowledge structures within vertical discourse. Here Bernstein distinguishes between two kinds of knowledge structure, hierarchical and horizontal. For Bernstein, knowledge structures differ in two ways. The first way is in terms of what I call verticality. Verticality has to do with how theory develops. In hierarchical knowledge structures, it develops through integration, towards ever more integrative or general propositions, the trajectory of development of which lends hierarchical knowledge structures a unitary convergent shape. Horizontal knowledge structures, on the other hand, are not unitary but plural, consisting of a series of parallel incommensurable languages. Progress in horizontal knowledge structures occurs not (or at least not primarily) through theory integration but rather through the introduction of a new language which constructs a ‘fresh perspective, a new set of questions, a new set of connections, and an apparently new problematic, and most importantly a new set of speakers’ (ibid.: 162). Because these languages are incommensurable, they defy incorporation.

HIERARCHY, KNOWLEDGE & THE SCHOOL CURRICULUM

•

MULLER

71

The level of integration, and the possibility for knowledge progress in the sense of greater generality and hence explanatory reach, is thus strictly pegged. Before I proceed to discuss the second form of knowledge form variation, grammaticality, it is worth making a few observations on verticality. The first observation is that it artfully incorporates and recapitulates the fierce dispute in the philosophy and sociology of science between the logical positivists and the non-realists, a dispute I selectively revisit below. Bernstein is implicitly asserting that the logical positivists (or realists) were right, but only in respect of hierarchical knowledge structures; the non-realists (Kuhn and after) were likewise right, but only in respect of horizontal knowledge structures. In other words, encoded into Bernstein’s principle of verticality are the terms of debate in the philosophy of science since the romantic revolt of the eighteenth century. Secondly though, we should note that the category of horizontal knowledge structures span a very broad range, from mathematics to sociology and the humanities. Although there is more than one mathematical language, and mathematics is in this sense a ‘horizontal’ knowledge structure, this example makes clear that verticality is certainly possible within the discrete languages constituting horizontal knowledge structures, verticality of a kind approaching the triangular form obtained in hierarchical knowledge structures, as Wignell (this volume) argues. The germane question then becomes, not so much what constrains progression in horizontal knowledge structures, but rather, what internal characteristics fail to constrain it in those that proliferate languages compared to those where language proliferation is constrained. For Bernstein, this is the difference between ‘strong’ and ‘weak grammar’ horizontal knowledge structures. In this usage of the term ‘grammar’, Bernstein is referring to internal properties of the knowledge structure. This should not be confused with the external sense of grammaticality discussed below. If verticality has to do with how theory develops internally, with what Bernstein later called the internal language of description, grammaticality (in the external sense) has to do with how theory deals with the world, or how theoretical statements deal with their empirical predicates, the external language of description (Bernstein 2000). The stronger the (external) grammaticality of a language, the more stably it is able to generate empirical correlates and the more unambiguous because more restricted the field of referents; the weaker it is, the weaker is its capacity to stably identify empirical correlates and the more ambiguous because much broader is the field of referents, thus depriving such weak grammar knowledge structures of a principal means of generating progress, namely empirical disconfirmation: ‘Weak powers of empirical descriptions removes a crucial resource for either development or rejection of a particular language and so contribute to its stability as a frozen form’ (Bernstein 2000: 167–8). In other words, grammaticality determines the capacity of a theory or a language to progress through worldly corroboration; verticality determines the capacity of a theory or language to progress integratively through explanatory sophistication. Together, we may say that these two criteria determine the capacity of a particular knowledge structure to progress. The precise nature of the relation between verticality and grammaticality is unclear. A plausible surmise could be the following: that verticality is a

72

LANGUAGE, KNOWLEDGE AND PEDAGOGY

categorical principal, consigning knowledge structures to either a theoryintegrating or a theory-proliferating category. The latter can in turn be broken down into a constrained proliferation or an unconstrained proliferation category. Grammaticality on the other hand is an ordinal principal, constructing a continuum of grammaticality within each category of knowledge structure, or perhaps across the entire spectrum.6 In what follows, I will concentrate mainly though not exclusively on verticality, on the internal characteristics of the internal language of description. Why would one want to elaborate a theory of knowledge forms? After all, we seem to have got along reasonably well without one for a long time. Bernstein only turned to the issue towards the end of his work. The contention here is that this lacuna in the study of knowledge and education was not accidental. Rather, I suggest, it was produced by the failure of social thought to deal with the dilemma of progress and the distributive strictures of hierarchy. The failure to reckon with the material structural differences in knowledge forms has become something of an obstacle in educational thinking. This can briefly be illustrated in two domains of education practice: namely, curriculum planning and research administration. Curriculum planning has been thrust into the limelight by international learner performance comparisons, most vividly displayed by the Third International Mathemetics and Science Study (TIMSS).7 A central tenet of assessment is that the instrument measuring performance is valid to the degree that it assesses what has been made available for acquisition. TIMSS has made visible the fact that not all children of the same age cohort across the globe learn the same things in the same order at the same level of cognitive demand. This has put a spotlight on the stipulation, sequence and progression requirements of curricula, and has begun to suggest that not all subjects in the curriculum have the same requirements. Could this be because their parent knowledge forms are different and take different distributional forms which in turn lead to different recontextualizing requirements? As for the question of research administration, research assessments of individuals and bodies of work have made possible comparisons between individuals, faculties, universities and countries. As more and more comes to depend on assessments of innovation and novelty (‘Is this paper really a contribution to new knowledge, or a re-hash of the known?’), the question arises as to what exactly constitutes innovation in different areas of research endeavour, and whether they are at all comparable. This is only the tip of the iceberg: it soon becomes clear that there are different epistemic cultures, different kinds of collaboration, different publishing traditions, and so on. In short, the globally emergent audit culture compels us to reflect on our knowledge practices, at the centre of which sits the question of their likeness, their comparability, and their compatibility. Once again, we realize how little we really know about how they may be alike or different, and what difference this might make. At the centre of this conundrum lies the question of knowledge hierarchy and progression.

HIERARCHY, KNOWLEDGE & THE SCHOOL CURRICULUM

•

MULLER

73

Progress: the very idea, and its sceptics The foundation of the Cartesian revolution in the seventeenth century was an axiom that appeared to be radically new, namely, that ‘true’ knowledge was characterized by knowledge progression. As Berlin put it, that ‘which had once been established did not need to be proved again, that is to say, in which scientific progress, universally recognised as such by rational thinkers, was possible’ (Berlin 2000a: 28). The conventional account depicts this as the decisive moment in the emergence from the closed tautological world of antiquity, and the birth of the modern (Shapin 1996). There are a number of entailments to this view. First, Descartes believed that only in a bona fide branch of knowledge can we find ‘clear and distinct ideas’ (Berlin 2000a: op. cit.). The paradigm of true knowledge, according to the Cartesian school, consisted in beginning from truths so clear and distinct that they could be contradicted only on pain of falling into absurdities; and in proceeding thence, by strict deductive rules, to conclusions whose truth was guaranteed by the unbreakable rules of deduction . . . (ibid.)

This was indeed a lofty aim for knowledge, and it meant that Descartes viewed the knowledge array then available in a particular way. For example, the human sciences might generate edification and improvement, but were otherwise of little enduring social value because they could not produce ‘strict deductive rules’. Here lies the foundation of the distinction between science and all other symbolic ensembles, and it rests on the notion of what may be called strong progression, that is, the step-by-step accretion of certainties. No one today is a thorough-going Cartesian; no one today believes in strong progression. Challenges to Cartesian rationalism have come from both within and outside of science. One challenge to this idea of strong progression from within science has culminated in the generally accepted position in science today of what may be called weak progression, or what Haack (2003) would call progress ‘within reason’, which I will return to below. This is a revision which accepts the postulate of progression (and hence of the division of the field of representations into ‘true’ or progressive knowledge, and belief or mere narrative), but recognizing at the same time that the ‘true’ in true knowledge did not equal absolute knowledge, and that progress in knowledge, if based on the best certainty to hand at that time, could always and in principle be revised – hence, weak progression. The dominant challenge to strong progression from outside science has sought to overturn the distinction between knowledge that progresses (‘science’) and knowledge that does not. The first brilliantly original formulation can be traced back to Giambattista Vico who, with his 7th Inaugural lecture in 1708, and later with the publication of the first edition of Scienza nuova/New Science in 1725, rejected the fundamental premise of Cartesian rationalism, the distinction between the true (verum) and the artificial (factum). Vico begins by arguing their essential unity: ‘We demonstrate geometry because we make it’

74

LANGUAGE, KNOWLEDGE AND PEDAGOGY

(Berlin 2000a: 31). What he meant by this was that we can be said to fully know something, not only because we know what it is (i.e. through rational reconstruction) but because we know how it came to be (i.e. through historical or genetic reconstruction), which he called per caussas. By this logic, we only know what we create. If we did not create it, we cannot know it, because it then has no human history. ‘The true (verum) and the made (factum) are convertible’ (ibid.: 35), or, ‘The criterion of truth is to have made it’ (ibid.: 36). In other words, with this argument, truth becomes a human artefact, and Vico becomes the first constructivist. Thus, whereas Descartes with his criterion of ‘clear and distinct ideas’ fundamentally sundered verum from factum, Vico fundamentally subverts it by reuniting them. The ‘demarcation debate’ in the philosophy of science, as to whether there is or is not a significant distinction between ‘science’ and other knowledge forms, begins here. Vico’s revolt has come to be a mere dress rehearsal for the more thoroughgoing romantic revolt of the nineteenth century, and the anti-realist one of our own time. The European romantics took up Vico’s anti-demarcatory premise of the make-ability of truth and of the world: . . . ‘the common assumption of the romantics that runs counter to the philosophia perennis is that the answers to the great questions are not to be discovered so much as to be invented. They are not something found; they are something literally made’ (Berlin 2000b: 202, 203). Amongst the romantics and their contemporary successors there are strong and weak traditions of make-ability. Common to all, however, is the following: Hence that new emphasis on the subjective [the maker] and ideal rather than the objective and the real, on the process of creation rather than its effects, on motives rather than consequences; and, as a necessary corollary of this, on the quality of the vision, the state of mind or soul of the acting agent – purity of heart, innocence of intention, sincerity of purpose rather than getting the answer right, that is, accurate correspondence to the ‘given’. (ibid.: 203)

It is this tradition of ‘make-ability’ that I will trace below, first as it snakes its way through the debates on knowledge and then, in a homologous way, through the debates on curriculum. As we will see below, the tradition of ‘make-ability’ translates into a ‘knower’ as opposed to a ‘knowledge’ perspective (Moore and Maton 2001).

On the shoulders of giants What the scientific revolution of the seventeenth century accomplished was a decisive challenge by the self-styled Moderns to the ‘human-centred universe’ (Shapin 1996: 20) of the Ancients, as consecrated by Aristotelianism, which had become a hermetic dogma of a priori truth. The cultural shock should not be underestimated. By 1611, the clerical poet John Donne could write: ‘And New Philosophy calls all in doubt’, ending with:

HIERARCHY, KNOWLEDGE & THE SCHOOL CURRICULUM

•

MULLER

75

’Tis all in pieces, all coherence gone; All just supply, and all Relation. (ibid.: 28) A century before Vico, Donne was registering a commonly felt shock at the cultural displacement of a deeply cherished worldview. What was gone was the measure of man, man as the measure of nature, and with it, the dominance of the humanistic Trivium over the scientific Quadrivium was decisively broken (Durkheim 1977; Bernstein 2000). At the heart of it was the entirely novel notion of progression towards a not-yet-attained truth that was not determined by man, but could be discovered by him through rational methods and intellectual daring. With the future and man’s fate loosed from the comforting embrace of classical-Christian teleology, the classicists, Christians and humanists of every stripe discerned a cultural abyss. The threat was felt to every form of social authority that depended on that worldview. The humanist counterrevolution would be only a matter of time. In the meantime, the intoxication of the expanded temporal horizon that the idea of progress suddenly constituted can be graphically seen in drawings and sayings of the time. In the frontispiece to his 1620 Instauration magna/The Great Instauration, an already provocative title, Francis Bacon depicted a ship boldly sailing out beyond the Pillars of Hercules, symbolizing the traditional limits of knowledge, below which was written the biblical text from Daniel: ‘Many shall pass to and fro, and science shall be increased’ (cited by Shapin 1996: 21). But if the forward horizon was extended, so too was the horizon backwards into the past in like manner re-constituted. This expanded retrospect is given iconographical expression in Johannes Kepler’s tower (Figure 4.1). At the front of the tower are two modern columns named for Copernicus and Tycho Brahe. Further back are more rough-hewn columns representing the earlier knowledge of Ptolemy and Agrippa; at the back are crude columns representing ancient astronomy. Right at the front, seated, is Kepler himself: the more recent, hence, the more sophisticated. The Moderns are separated from the Ancients by two dimensions, thus. The first is time. But in order to express progress across time, a second dimension is crucial to the first, namely, greater differentiation. In order to express this, then, the further towers are less differentiated, the nearer ones more so. What evolves, or progresses, is differentiation. The condition of that progression is a progressive capitulation and building on previous knowledge, a greater differentiation of knowledge. This quintessentially ‘modern’ idea, progressive differentiation and dual temporality, is embodied in Newton’s famous aphorism, ‘if I have seen farther, it is by standing on the shoulders of giants’ (Merton 1993: 1). In one of the richest ironies in the history of science, it turns out that the aphorism, everywhere attributed to Newton and hence taken as emblematic of the modern view of progress, does not originate with him at all. Indeed, it appears to have been common currency in knowledge circles since at least Bernard of Chartres in 1126, who probably got it from his Priscian predecessors, and was used in various forms, at regular intervals. Merton (1993: 268, 269) records at least 27 usages in print before Newton. The aphorism continues to be

76

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Figure 4.1 The Astro-poecilo-pyrgium (the variegated star tower) from the title page of Johannes Kepler’s Tabulae Rudolphinae, 1627. (Shapin 1996: 77) used today in a wide variety of more or less appropriate contexts.8 The original point of the aphorism was to highlight inter-generational co-operation as the ground for scientific progress, usually expressed in humble or mock-humble terms: ‘A dwarf standing on the shoulders of a giant may see farther than a giant himself’ (ibid.: 4). As Merton himself puts it: When you come right down to it, the essential point is that the dwarfs-on-theshoulders-of-giants Aphorism is a rough equivalent to the twentieth-century sociological conception that scientific discoveries emerge from the existing social base and consequently become, under conditions that can be reasonably well defined, practically inevitable. (ibid.: 267)

What we learn from this pithy story is that scientists probably had a pragmatic view of themselves and the process of science from relatively early on – they certainly didn’t have to wait for social studies of science to discover messiness and

HIERARCHY, KNOWLEDGE & THE SCHOOL CURRICULUM

•

MULLER

77

serendipity in the twentieth century – but it was only in the seventeenth century that the threat to social order, up until then clearly contained by the hegemonic view, provoked the first in a series of backlashes that have continued with greater and greater sophistication to this day. The debate is variously figured, but it is probably fair to say that it has been between various kinds of realism and irrealism, those for a more or less strong view of progress in science (called by Haack 2003, the ‘Old Deferentialists’), and those against (the ‘New Cynics’). Before being sucked into the complexities of contemporary philosophy of science, I should immediately declare that my aim in what follows is solely to get a grip on the way that knowledge systems are internally differentiated, or are said to progress, and to answer the questions ‘what is it that progresses when they do progress? And is it only science that progresses? Has only science got hierarchy?’ The philosophy of science debate detracts from asking the question about the non-science disciplines. Yet the point must hold for the non-science disciplines too, at least in one form or another, if these disciplines are not to surrender all claims to progress and relevance in an information or knowledge age (Foray and Hargreaves 2003).

Kinds of verticality In answer to the question, ‘what is it that accumulates when knowledge progresses?’ the logical positivists gave an unequivocal answer: it was the piled-up structure of laws related to one another by strict definition, in strict order of explanatory integratedness, that is, in strict order of their approximation to the truth. Here is to be found the definition of Basil Bernstein’s verticality, namely, the degree of integratedness and ‘subsume-ability’ of theory. The logical positivist idea of progression has come under universal attack. The debate has been intricate and technical, a far cry from the ‘village or tabloid’ scapegoat of positivism (Matthews, 2004: 2) that has taken centre stage in the social sciences. The principal objection has been to the founding idea of progression, to the idea of a single convergent system of knowledge. There was no single progression path, went one criticism, not least because the explanatory reach and range of application of most covering laws was much more modest, leading to a notion of a cluster of ‘languages’ rather than a single converged-upon structure, an idea conceded already by leading logical positivist Otto Neurath, he of the modest title of ‘Director of the Agency for Full Social Planning’ in ante bellum socialist Vienna (Cartwright 2001). Nor is it a matter simply of ‘inductive scepticism’, of asserting the messiness of discovery against the seeming imperialism of the subsumptive structure: after all, the ‘Old Deferentialists’, following Reichenbach in the early twentieth century, maintained a distinction between the ‘context of discovery’ as distinct from the ‘context of justification’ to deal with just this objection (Phillips 2004), though the distinction proved difficult to sustain. The contemporary disavowal of progress by the ‘New Cynics’ cuts far deeper, against the heart of the impersonality of this vision of progress (Popper’s ‘epistemology without a knowing subject’), against the notion of a knowledge about the world that exists without man at its centre, in

78

LANGUAGE, KNOWLEDGE AND PEDAGOGY

its most extreme reduction, against the idea that there is a real independent world to be known at all, a disavowal that eliminates progress by disavowing the world and the possibility of real knowledge about it (Moore and Muller 1999). Just as all forms of realism have built into them some or other form of progress, thus, all forms of the ‘New Cynicism’ have built into them the idea that knowledge progress is incoherent. While there are many alternative accounts (meaning holism and constructive empiricism to name but two), the landmark account belongs to Thomas Kuhn and his account of innovation by ‘revolution’ or paradigm change (Kuhn 1962). This can be grasped by seeing that Kuhn turns the tables on the imperialism of the ‘Old Deferentialists’ who depicted all knowledge in the image of science, by depicting all knowledge, science included, as behaving like Basil Bernstein’s horizontal knowledge structures, advancing up to a point, only to break off into an alternative theoretical language or paradigm. For Kuhn, as for Bernstein, the crucial point was that the languages were incommensurable. One of Bernstein’s great contributions, as I suggested above, is to have recapitulated realism’s loss of innocence, as staged in the literature of the philosophy of science, and recast the terms of debate into a taxonomy of knowledge forms, with verticality, or ‘subsume-ability’, as one principal criterion. Earlier I suggested that verticality was a categorical principal. In the discussion above it is apparent that though this criterion has been considerably weakened, though it is now conceded that knowledge grows by virtue of different kinds and paths of conceptual change, these are still categorically distinguished as to whether they are commensurable or not. Or are there degrees of commensurability? Does incommensurability of a more absolute sort characterize what Bernstein called languages with weak internal grammars (and I called with ‘unconstrained proliferation’)? Is incommensurability less absolute among ‘constrained proliferation’ languages? We could put this to the following test: can SFL – a language by all accounts (including those of Bernstein, Dowling and Martin) that has a strong internal grammar, is theory-integrating, and is discursively saturated – commune with, say, other forms of sociolinguistics? Or is that conversation of the deaf as absolute as anything in the social sciences? I suspect that, in comparison with the hapless sociology, there is greater seepage between languages in linguistics than in sociology. The discussion may be summed up as follows. As far as the internal structure of internal languages of description is concerned, that is, as regards their verticality, languages vary as to their: • • •

capacity to subsume statements into logical types (syntactic/semantic axis); therefore their relative expressibility in terms of general and particular statements (general/particular axis); and therefore their relative expressibility in terms of propositional content and stylistic content (content/form axis).

One could go on. The question now for this paper is: how are different powers of subsumption, of verticality, expressed by pedagogy? Does hierarchy specify pedagogy, and if so how?

HIERARCHY, KNOWLEDGE & THE SCHOOL CURRICULUM

•

MULLER

79

The knowledge structure/pedagogic structure link Does knowledge structure constrain pedagogic structure, does it place any onus on the way that the ‘what is to be learnt’ is recontextualized? Do these internal characteristics of knowledge structures place limits on the form their curricular offspring optimally could and should take? We know that pedagogic structure has distributive potential, but does knowledge structure come with an already encoded distributive potential, placing structural limits on pedagogic form? We know that it takes a specialized language to specialize consciousness: but can we determine what kind of specialization a pedagogy must encode to effectively realize a specialized consciousness in a specialized language? All these different ways of posing the question presuppose answering the question one way or another. One affirmative response to these questions is given in the Review Committee’s (2000) report on a review of the South African grade 1 to 9 national curriculum, known as Curriculum 2005 (C2005). The Review found that the curricular form of C2005 was under-stipulated, under-sequenced and its pacing requirements under-signalled. It was a form of invisible or competence pedagogy (Bernstein, 2000) which provided minimal markers as to what should be learnt or evaluated at what level. Unsurprisingly, the Review found that poorly trained teachers, teaching largely disadvantaged children, fared worst with this pedagogy: they had not covered what the curriculum expected them to have covered by the end of each grade. Consequently children entered the next grade with knowledge gaps, elements of knowledge presupposed by the curriculum of the next grade. These knowledge gaps had more serious consequences in what the Review called ‘content-rich’ subjects (maths, science and language) than in more skills-based subjects like life-skills, because the former had content, sequence and progression requirements deriving from their parent disciplines. How does Bernstein answer the question? Equivocally, it turns out. In the paper on the pedagogic device, the answer is negative: As physics is appropriated by the recontextualising agents, the results cannot formally be derived from the logic of that discourse. Irrespective of the intrinsic logic which constitutes the specialised discourse and activities called physics, the recontextualising agents will select from the totality of practices which is called physics . . . But these selections cannot be derived from the logic of the discourse of physics . . . (Bernstein 2000: 34, emphasis added)

Here Bernstein appears to contradict the Review, but we should note the context, where Bernstein is asserting the theoretical priority of the regulative over the instructional, meaning that the internal order of school physics is wholly derived from normative social order. In one sense this is undoubtedly correct. Any state can, on the basis of its ideology, decide what pedagogic modality to impose. It can even appear to choose a de-specializing pedagogy in defiance or denial of the requirements of its specialized division of labour, at least for a while. This is indeed what happened in South Africa with C2005 in

80

LANGUAGE, KNOWLEDGE AND PEDAGOGY

the 1990s, as it had in the UK with the Plowden Report in the late 1960s (Fitz, Davies and Evans 2006). But in another sense, this is quite misleading. Indeed, the argument can be made that Bernstein came to the view, late in his career, that the instructional domain – or knowledge – has an internal determinative logic of its own, one which cannot simply be reduced to subordination to the regulative order. Although the sea change was presaged in the ‘pedagogic device’ paper, the strongest evidence for this view can be found in Bernstein’s ‘vertical and horizontal discourse’ paper (Bernstein 2000). The principal reason Bernstein opposed the idea of a link in his early work is because he maintained that the recontextualized discourse (e.g. school physics) was a wholly separate discourse from that of research-based physics. It was necessary to maintain this position in order to stay true to the postulate that all symbolic formations were specific to a context with its specializing practices. That context is conditioned by a society’s regulative or moral order. But if recontextualization totally severs any relation, then how are specialized knowledges ever reproduced? After all, school maths performance predicts (imperfectly to be sure) university maths performance; and that predicts in turn proper maths adeptness. The only way this can be intelligible is by conceiving that school maths competence ‘precurses’ (Gee 2001) university maths competence, which ‘precurses’ real maths adeptness. There has to be some form of specialization of consciousness continuum in play; this could be called a founding assumption of modern education, and it is strongly suggested by Bernstein’s view that the symbolic device is ‘condensed’ in the evaluative rules which binds the logic of the parent discipline to the attainment requirements for acquirers (Bernstein 2000). After all, this idea of the interpenetration of symbolic competence is built into Bernstein’s explanation of how the middle-class home code precurses its young into the school code better than does the working-class home code. So, a relation there must be. One might pursue the exact nature of the relation. A preliminary question for this chapter is: what effects the relation? What activates it? There are two typical answers: let us call them a knowledge answer, and a knower answer. A knowledge approach is epitomized by TIMSS and their principal intellectual construct, ‘opportunity to learn’ (OTL). OTL in its simplest form is defined as coverage of the curriculum, and the original TIMSS project defined coverage in terms of a serial list of topics only (Porter 2002). But is that sufficient to map coverage in a vertical curriculum? It certainly wasn’t adequate in TIMSS. A syntactical view will say: list the principal propositional steps in the knowledge hierarchy. Each propositional step will function like a rule with rules of combination, each cluster of which can generate an indefinite number of possibilities; for example, ‘odd numbers’ in arithmetic (see Pinker 1999: 318). A complete list will describe the internal grammar of the internal language of the subject. Table 4.1 shows one way of representing this, drawn from Reeves and Muller (2005). This table displays an extract from a finite list of maths content presumed to be learnt in grade 6, listed in order of cognitive complexity. The information in brackets (g4, g5, g6) indicates that these knowledge elements are considered essential at grade 4, 5 or 6 levels – in other words, they reflect work that learners

HIERARCHY, KNOWLEDGE & THE SCHOOL CURRICULUM

Table 4.1

•

MULLER

81

Extract of hierarchy of maths content for grade 6

Representing and comparing whole numbers including zero and fractions including: Whole numbers to •

4-digit numbers (g4)

•

6-digit numbers (g5)

•

9-digit numbers (g6)

Odd and even numbers to 1,000 (g4) Common fractions in diagrammatic form (g4) Common fractions with different denominators including •

halves (g4)

•

thirds (g4)

•

quarters (g4)

•

fifths (g4)

•

sixths (g4)

•

sevenths (g4)

•

eighths (g4)

•

tenths (g6)

•

twelfths (g5, 6)

•

hundreds (g6)

are, at a minimum, expected to cover at this level. However, although certain elements of topics or subtopics are considered essential for a particular grade level, there are other elements of topics or subtopics that are considered essential at all or more than one grade level. But here a pedagogical question arises. Does a list of topics in order of complexity, in other words in order of disciplinary progression, constitute an optimal learning path? The reconstituted logic of a discipline and the optimal pedagogical learning sequence might overlap only by default. The reason for this lies with the way subsumption works: the same semantic topics (the same particulars) play different roles in different generals. The upshot is that particular topics, even for the most hierarchical of subjects, are repeated across learning levels, but differently. In short, imperfect subsumption has so far stymied the linear representation of content in a curriculum, and the relation of curricular structure to disciplinary structure has remained an open, more usually an

82

LANGUAGE, KNOWLEDGE AND PEDAGOGY

avoided, question. The same question may be posed to the linguists: does a curriculum following a genre-sequential logic constitute an optimal learning path? How are these different cognitive logics to be braided into the artifice called curriculum and pedagogy? This is the nub of pedagogy. All in all, it is not too surprising that by far the most common way of representing verticality in the literature has been distilled from what teachers do or expect, in terms of an index of ‘cognitive demand’, usually depicted as a scale from ‘memorization’, through ‘routine procedures’, ‘communication of understanding’, ‘problem solving’, to ‘conjecture/generalise/prove’ (Porter 2002: 4). Each listed topic is given a ‘cognitive demand’ rating. Proportion of coverage by degree of demand yields a proportion of instructional time spent. This temporal proportion becomes a proxy for ‘opportunity to learn verticality’. There are clearly other ways of compiling a demand index (see Morais et al. 2004, for a good example). All of them, as far as I can see, shift the focus from the knowledge itself to the knower-actors, from a knowledge approach to a knower approach. They shift the focus from what knowledge is made available, to what levels of complexity teachers make available. Hierarchy moves back into the shadows. In the wake of the demise of logical positivism, and the discrediting of the distinction between the contexts of discovery and justification, a new orthodoxy under the aegis of the ‘New Cynics’ and their pedagogical confrères the educational progressives, took hold, to the effect that sequence and progression in pedagogy simply didn’t matter, that there was no one royal road to learning, and that only the most minimal stipulations were necessary to signal the learning end points as measured by common assessment instruments (Muller 2002; Labaree 2004). This turned out to be wrong, in South Africa and elsewhere. Such radically invisible pedagogies can work, but only for middle-class pupils and usually only in the lower grades, and exactly how they do that is still being unearthed. For the majority of poorer children, the evidence increasingly suggests, clear and explicit articulation of evaluation criteria is sine qua non (see the various papers in Muller et al. (eds) 2004). What are the evaluation criteria evaluations of? Of the knowledge steps to be traversed; it is hard to avoid this conclusion. To be sure, there is not only one set of steps per discipline, nor need we assume that these steps are always to be traversed in exactly the same order: in practical situations they simply won’t be. Nevertheless, insofar as the idea of theory integration means anything at all, it does, qua hierarchy, specify the formal, minimal steps to be acquired in order for sense to be made at all. So, making concessions to messiness and agreeing that we cannot stipulate a once-and-for-all-path, we would still have to concede, retrospectively considered, that there are a specifiable necessary minimum set of steps that must be pedagogically traversed. In a world of disadvantage, to lose sight of this is to lose sight of the target for the social equality of outcomes. Does this argument do away with the necessity of the teacher? Not at all. What it does is to emphasize the knowledge dimension of what makes a teacher a teacher. The condition for a teacher being an authoritative pedagogical agent is, at the minimum, an internalized map of the conceptual structure of the subject, acquired through disciplinary training (this is perhaps why Morais et al.

HIERARCHY, KNOWLEDGE & THE SCHOOL CURRICULUM

•

MULLER

83

2004, dub the index of the what ‘teacher competence’ in their study). In other words, the condition for teachers to be able to induct pupils into strong internal grammar subjects is that they themselves already stand on the shoulders of giants, that they can speak with the disciplinary grammar. But if they can’t? I am inclined to say: let us then train them so that they can. The difficulty is, it is often not clear what they don’t know. Two things stand out in the global literature on effective learning. The first is that teacher competence is by far the most important factor in learner attainment; the second is that in-service teacher training has had almost no effect. I fear this will continue unless we pay as much attention in future to knowledge as we have up until now to knowers.

Conclusion I have proceeded in this chapter on the hunch that the contemporary avoidance of knowledge structure, in this chapter principally of the question of hierarchy and progression, lies at the heart of many current pedagogical dilemmas, in particular those having to do with providing to poor children access to the tools of powerful knowledge. That this can be done has been incontrovertibly shown by Rose (2004) and Carnoy et al. (2004). I have not kept that in the foreground of the discussion, but it nevertheless remains a principal motive force. It remains plausible, perhaps even likely, that knowledge structure has distributional implications, which interact with distributional alignments of pedagogy. It has proved easier discussing verticality in hierarchical disciplines than in those with horizontal knowledge structures. I have suggested, nudging Bernstein and Martin, that not all horizontal structures have languages that are equally incommensurable. I have suggested that some, like linguistics perhaps, have languages whose boundaries are more permeable than those, like sociology perhaps, that are inclined to proliferate languages without constraint. Constraint must at least in part be promoted by hierarchy. This remains a task to be continued, not least because it seems likely to me that the exceptionalism that the social sciences and humanities have claimed for so long will prove increasingly implausible in the global networks of the knowledge society. When that happens, I fear that the ‘New Cynicism’ and its pedagogical analogues will be about as effective a cloak against the cold winds of global comparability as fashionable decadence was for the artistic fringe in ante bellum Berlin.

References Becher, T. (1989), Academic Tribes and Territories: Intellectual Enquiry and the Cultures of Disciplines. Buckinghamshire: SRHE and Open University Press. Becher, T. and Parry, S. (2005), ‘The endurance of the disciplines’, in I. Bleiklie and M. Henkel (eds), Governing Knowledge. Dordrecht: Springer. Berlin, I. (2000a), Three Critics of the Enlightenment: Vico, Hammann, Herder. Princeton: Princeton University Press. Berlin, I. (2000b), ‘The essence of European romanticism’, in I. Berlin, The Power of Ideas. London: Pimlico.

84

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Bernstein, B. (1994), ‘Discourse, knowledge structure and fields: some arbitrary considerations’. Mimeo. Bernstein, B. (1996), Pedagogy, Symbolic Control and Identity: Theory, Research, Critique. London: Taylor & Francis. Bernstein, B. (2000), Pedagogy, Symbolic Control and Identity: Theory, Research, Critique. Oxford: Rowman & Litttlefield. Carnoy, M., Gove, A. and Marshall, T. (2004), Why do some Students Achieve More in some Countries than in Others? A Comparative Study of Brazil, Chile and Cuba, in press. Cartwright, N. (2001), The Dappled World: A Study of the Boundaries of Science. Cambridge: Cambridge University Press. Collins, R. (1998), The Sociology of Philosophies: A Global Theory of Intellectual Change. Cambridge, MA: The Bellknap Press of Harvard University Press. Dowling, P. (1999), ‘Basil Bernstein in frame: “Oh dear, is this a structuralist analysis?”’ http://www.ioe.ac.uk/ccs/dowling/kings1999/index.html Dowling, P. (2005), ‘A timely utterance’. http://homepage.mac.com/paulcdowling/ ioe/publications/dowling2005/timely_utterance/index.htm Durkheim, E. (1976: first pub. 1915), The Elementary Forms of Religious Life: A Study in Religious Sociology. Trans. J. W. Swain. New York: The Free Press. Durkheim, E. (1977: first pub. 1938), The Evolution of Educational Thought: Lectures on the Formation and Development of Secondary Education in France. Trans. P. Collins. London: Routledge & Kegan Paul. Fitz, J., Davies, B. and Evans, J. (2006), Educational Policy and Social Reproduction: Class Inscription and Symbolic Control. London: Routledge. Foray, D. and Hargreaves, D. (2003), ‘The production of knowledge in different sectors’, in London Review of Education 1 (1), pp. 7–19. Foucault, M. (1981), ‘The order of discourse’, in R. Young (ed.), Untying the Text. London: Routledge & Kegan Paul. Gee, J. (2001), ‘Learning in semiotic domains: a social and situated account’, paper presented to the International Literacy Conference, Cape Town, 13–17 November, 2001. Goldman, A. (1999), Knowledge in a Social World. Oxford: Oxford University Press. Haack, S. (2003), Defending Science – Within Reason. New York: Prometheus Books. Hacking, I. (1985), ‘Styles of scientific reasoning’, in J. Rajchman and C. West (eds), PostAnalytic Philosophy. New York: Columbia University Press. Hasan, R. (2001), ‘The ontogenesis of decontextualised language: some achievements of classification and framing’, in A. Morais, I. Neves, B. Davies and H. Daniels (eds), Towards a Sociology of Pedagogy. New York: Peter Lang. Kuhn, T. (1962), The Structure of Scientific Revolutions. Chicago: Chicago University Press. Labaree, D. (2004), The Trouble with Ed Schools. New Haven: Yale University Press. Matthews, M.R. (2004), ‘Editorial’, Science and Education, 13, pp. 1–4. Merton, R.K. (1973), The Sociology of Science: Theoretical and Empirical Investigations. Chicago: The University of Chicago Press. Merton, R.K. (1993: first pub. 1965), On the Shoulders of Giants. Chicago: The University of Chicago Press. Moore, R. and Maton, K. (2001), ‘Founding the sociology of knowledge: Basil Bernstein, epistemic fields and the epistemic device’, in A. Morais, I. Neves, B. Davies and H. Daniels (eds), Towards a Sociology of Pedagogy: The Contribution of Basil Bernstein to Research. New York: Peter Lang. Moore, R. and Muller, J. (1999), ‘The discourse of “voice” and the problem of knowledge and identity in the sociology of education’, in British Journal of Sociology of Education 20 (2), pp. 189–206. Moore, R. and Muller, J. (2002), ‘The growth of knowledge and the discursive gap’, in British Journal of Sociology of Education 23 (4), pp. 627–37. Morais, A., Neves, I. and Pires, D. (2004), ‘The what and the how of teaching and

HIERARCHY, KNOWLEDGE & THE SCHOOL CURRICULUM

•

MULLER

85

learning’, in J. Muller, B. Davies and A. Morais (eds), Reading Bernstein, Researching Bernstein. London: RoutledgeFalmer. Muller, J. (2000), Reclaiming Knowledge: Social Theory, Curriculum and Education Policy. London: RoutledgeFalmer. Muller, J. (2002), ‘Progressivism redux: ethos, policy, pathos’, in A. Kraak and M. Young (eds), Education in Retrospect: Policy and Implementation Since 1990. Pretoria and London: HSRC Publishers and Institute of Education, University of London. Muller, J., Davies, B. and Morais, A. (eds) (2004), Reading Bernstein, Researching Bernstein. London: RoutledgeFalmer. Mullis, I., Martin, M., Gonzales, E., Gregory, K., Garden, R., O’Connor, K., Chrostowski, S. and Smith, T. (2000), TIMSS 1999 International Mathematics Report: Findings from IEA’s Repeat of the Third International Mathematics and Science Study at the Eighth Grade. Chestnut Hill: Boston College. Nagel, E. (1961), The Structure of Science: Problems in the Logic of Scientific Explanation. London: Routledge & Kegan Paul. Painter, C. (1999), ‘Preparing for school: developing a semantic style for educational knowledge’, in F. Christie (ed.), Pedagogy and the Shaping of Consciousness. London: Continuum. Phillips, D.C. (2004), ‘Two decades after “After the wake: postpositivistic educational thought”’, Science and Education 13, pp. 67–84. Popper, K. (2002: first pub. 1963), Conjectures and Refutations: The Growth of Scientific Knowledge. London: Routledge Classics. Porter, A. (2002), ‘Measuring the content of instruction: uses in research and practice’, in Educational Researcher 31 (7), pp. 3–14. Pinker, S. (1999), Words and Rules: The Ingredients of Language. Cambridge: Cambridge University Press. Reeves, C. and Muller, J. (2005), ‘Picking up the pace: variation in the structure and organisation of learning school mathematics’, in Journal of Education 37, pp. 103–130. Rose, D. (2004), ‘Sequencing and pacing of the hidden curriculum: how Indigenous learners are left out of the chain’, in J. Muller, B. Davies and A. Morais (eds), Reading Bernstein, Researching Bernstein. London: RoutledgeFalmer. Shapin, S. (1996), The Scientific Revolution. Chicago: The University of Chicago Press. Thompson, K. (ed.) (1994), Readings from Emile Durkheim. London: Routledge & Kegan Paul. Weinberg, S. (1993: first pub. 1992), Dreams of a Final Theory. New York: Vintage.

Notes 1

2

3 4

The first record in print of Bernstein’s tree is in a mimeo from March 1994. If Martin published his in 1992, as the text referred to above implies, then he is indeed responsible for a critical breakthrough. Wariness of the written-oral cline is explicit in the early forms of the ‘discourses’ paper (see Bernstein 1994), implicit in the later ones (see the version in Bernstein 2000). Highly codified disciplines are labelled C+; weakly codified disciplines C–. This does not stop him from commenting on disciplines, and although he professes relative ignorance, Dowling (2005: 6) is unusually generous about SFL: ‘Systemic Functional Linguistics, by contrast (again with the hapless sociology), is now well established, which is to say, strongly institutionalized, internationally and is clearly DS+ as is readily apparent from even an amateur’s perusal of Halliday and Mathiessen (2004). We may legitimately refer to SFL as a discourse in its own right . . .’

86

5

6

7

8

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Oddly enough he seems unaware of Kay O’Halloran’s work (see this volume), an SF linguist not only working in his own area, maths education, but who uses a term, symbolic saturation, close in spirit to his own. Sections of what follows will be published as ‘On the shoulders of giants: verticality of knowledge and the school curriculum’, in R. Moore, M. Arnot, J. Beck and H. Daniels (eds), Knowledge, Power and Educational Reform: Applying the Sociology of Basil Bernstein. London: RoutledgeFalmer 2006. Though Bernstein depicts grammaticality as a feature only of horizontal knowledge structures in the diagram on p. 168, he refers to physics as having a ‘strong grammar’ on p. 163. He is probably using it here in its internal sense of ‘explicit conceptual syntax’. TIMSS and the TIMSS Repeat studies are synonymous with international learner attainment comparisons, having generated comparable data from 38 countries for Maths and Science, and leading to the first international league tables (see Mullis et al. 2000). The cheekiest is its use as the heading on the opening page of the Google Scholar search engine.

5

Knowledge–knower structures in intellectual and educational fields Karl Maton

For every knowledge structure there is also a knower structure.

Introduction Why bother reading this chapter? On what grounds am I claiming insight or understanding? How can I claim to be a ‘sociologist’? These are the kinds of questions I shall focus on: the basis of achievement, status and membership; i.e. the issue of how knowledge and knowers are specialized. The work of Basil Bernstein is particularly valuable for addressing such questions. Where most approaches in the sociology of knowledge and education focus on relations to knowledge (of class, race, gender and so forth), Bernstein’s approach pays attention to relations within knowledge. Instead of simply showing how identity shapes knowledge, this approach also reveals how knowledge itself specializes identity, consciousness and relations. In this chapter I explore how Bernstein’s conceptual framework sheds light on these issues and how his ideas can be developed to create further insights.1 The reasons for developing the approach further are immanent in the form of its development. One trajectory that can be traced through Basil Bernstein’s sociology is from the analysis of the pedagogic practices of educational fields of reproduction (1977), through an account of the construction of educational knowledge (1990), to the study of the intellectual fields of production from which this knowledge is selected and recontextualized (1999). With the concepts of educational knowledge codes, the pedagogic device and knowledge structures, respectively, Bernstein showed how structurings of intellectual and educational knowledge specialize actors and discourses in ways that shape social relations, institutional organization, disciplinary and curricular change, identity, consciousness and habitus (Singh 2002; Moore 2004). This trajectory has been characterized by an unusually intimate dialectic between theory and research (Bernstein 1996). Thanks to the form taken by the theory (Moore and Muller 2002), in each case its development (in the light of previous research) created new ways of seeing existing objects of study and highlighted new objects for empirical research to explore. In turn, subsequent research spoke back to the theory, raising new questions and necessitating further conceptual development. Bernstein was fond of saying one should aim for ‘productive imperfection’ and stated that his concepts represented a provisional mapping of intellectual fields. One can thereby rest assured that the notion of ‘knowledge structures’ is not the end of the matter. In this chapter I elaborate on a means of continuing this neverending story.

88

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Specifically, I address two questions raised by the concepts of knowledge structures (see Maton and Muller, this volume). Bernstein’s language of description provides a means of systematically describing differences between the discursive practices of intellectual fields, raising the question of conceptualizing the generative principles underlying these fields of production. This formed a starting point for the development of legitimation codes (Maton 2000a, 2000b) and the epistemic device (Moore and Maton 2001). Legitimation codes provide a means of conceptualizing the structuring principles underlying intellectual fields; the epistemic device is the means whereby these codes (and so the form taken by intellectual fields) are created, maintained, reproduced, transformed and changed. The epistemic device was intended to complement rather than displace the pedagogic device; rather than being specific to intellectual and educational fields respectively, both devices form the basis for production, recontextualization and reproduction of knowledge. Though developed through studies of knowledge production, the concepts were thereby intended to illuminate educational knowledge and practice more generally, and in a number of studies are being used to analyse educational fields.2 However, in terms of Bernstein’s theory, a second question remains of relations between the concepts of knowledge structures and educational knowledge codes or, put another way, how intellectual fields of production and educational fields of reproduction can be analysed within the same conceptual framework.3 I shall argue that the concepts of legitimation codes and the epistemic device provide a means of addressing these questions, and that their answers are to be found by thinking in terms of Knowledge–knower structures. My basic argument is summarized in the opening motif: for every knowledge structure there is also a knower structure. I elaborate the implications of this claim in two stages. First, I introduce the concepts of knowledge structures and knower structures and show how they can be brought together and their underlying structuring principles analysed. I do so through considering fields of knowledge production, focusing on the example of the famous ‘two cultures’ debate on relations between science and the humanities. Secondly, I explore how these concepts relate to educational knowledge codes, elaborate on the forms taken by knowledge–knower structures, and show how the concepts can be applied to fields of reproduction, focusing on the example of studies addressing the marginalized position of Music in the English school curriculum.

Knowledge–knower structures in fields of production A useful way of introducing the notion of knowledge–knower structures is to consider the well-known ‘two cultures’ debate. This was sparked by C.P. Snow’s famous 1959 lecture in which he claimed the intellectual life of western society was being split into ‘two polar groups’ that ‘had almost ceased to communicate at all’ with ‘between the two a gulf of mutual incomprehension – sometimes . . . hostility and dislike, but most of all lack of understanding’. (Snow 1959: 3, 2, 4). Snow’s focus lay beyond the academy, but the ‘two cultures’ became associated with the humanities and science and the ensuing debate was constructed

KNOWLEDGE–KNOWER STRUCTURES

•

MATON

89

as a struggle over status within higher education. Though this picture of two cultures was already well established, the debate was ferocious, bitter, spread widely, and remains a source of contention. Why Snow’s lecture sparked such depth of feeling is clear from the ways in which participants portrayed science and the humanities as enjoying contrasting fortunes. On the one hand, what Snow termed ‘scientific culture’ was portrayed as enjoying a meteoric rise in stature; as one commentator tartly expressed: You cannot open a newspaper, let alone the ‘quality’ journals, without the importance of science and technology being trumpeted at you from the headlines. (Morris 1959: 374)

Feted by and enjoying massive funding from industry and the state, revered by the media and worshipped by the public, by the late 1950s scientists were said to be enjoying unprecedented prestige. In contrast, the humanities were portrayed as embattled, in decline and insecure. For example, an influential collection of essays entitled Crisis in the Humanities (Plumb 1964a) included accounts of proclaimed crises within Classics, history, philosophy, Divinity, literary education, sociology, the fine arts, and economics, as well as the humanities in schools. They were said to be unwanted by better quality students, considered irrelevant to a modern economy by industrialists, increasingly excluded from the corridors of power by politicians, no longer considered the repository of culture, and publicly ridiculed as offering little genuine knowledge. One historian, for example, claimed that ninety per cent of his colleagues believed their subject to be ‘meaningless in any ultimate sense’ (Plumb 1964b: 25). According to participants, the disciplinary map was undergoing a fundamental shift of power between humanist and scientific cultures in their longacknowledged struggle for status and resources. Two questions this raises are: what was the basis of their differences, and why was this shift of power occurring? A common contemporary explanation of their differences held that scientists and humanist intellectuals ‘speak different languages’ (Editorial, The Listener, 3 Sept. 1959: p. 344). Using Bernstein’s approach would suggest focusing instead on the underlying structuring principles of their languages. I shall explore these principles in terms of knowledge structures and then knower structures, before bringing them together to show how an analysis of knowledge–knower structures can shed light on the bases of intellectual fields. Knowledge structures Analysing the form taken by knowledge in intellectual fields, Bernstein (1996, 1999) distinguishes first between horizontal discourse (everyday or ‘commonsense’ knowledge) and vertical discourse (scholarly or professional knowledge), and secondly within vertical discourse between horizontal and hierarchical knowledge structures. The latter exemplify well one aspect of the way the two cultures were portrayed in the debate. Humanist culture was described by proponents as riven by competing claims for status between strongly bounded disciplines. Commentators argued that Classics had served as the basis of a ‘common culture’ or ‘unifying force’ (Lee 1955) and their decline had

90

LANGUAGE, KNOWLEDGE AND PEDAGOGY

fragmented a single, organic culture into a series of rival subcultures, with little dialogue between disciplines and no means of adjudicating between competing claims to be a new unifying centre. Humanist culture thereby resembled what Bernstein defines as a horizontal knowledge structure: a series of specialised languages, each with its own specialised modes of interrogation and specialised criteria . . . with non-comparable principles of description based on different, often opposed, assumptions. (Bernstein 1996: 172–3)

This comprises a series of segmented, strongly bounded languages which, developing Bernstein (1999: 162), can be visually represented as: L1

L2

L3

L4

L5

Bernstein further distinguished between horizontal knowledge structures with stronger grammars, ‘whose languages have an explicit conceptual syntax capable of “relatively” precise empirical descriptions and/or of generating formal modelling of empirical relations’ (1999: 164), such as mathematics, linguistics and economics, and those where these powers are weaker, such as anthropology, cultural studies and sociology. Humanist culture, as characterized by proponents, possessed a weaker grammar – its objects of and procedures for study were defined in ethereal, nebulous, even mystical terms, most famously and widely expressed as immersion in ‘the best that has been known and thought in the world’. In contrast to the segmentation of humanist culture, proponents of scientific culture claimed that scientists comprised an organic community; as Snow put it, they shared ‘common attitudes, common standards and patterns of behaviour, common approaches and assumptions’ (1959: 9). Unlike the pluralized humanities, science was often referred to in the singular and portrayed as integrated and whole, and though scientists were proliferating new knowledge and creating sub-disciplinary specialisms at a prolific rate, they were said to know how to integrate this knowledge. Scientific culture thereby resembled what Bernstein describes as a hierarchical knowledge structure: ‘an explicit, coherent, systematically principled and hierarchical organisation of knowledge’ which develops through the integration of knowledge at lower levels and across an expanding range of phenomena (1996: 172–3). This Bernstein represents as:

where the point of the pyramid represents the smallest number of axioms or theories and the base represents the maximal number of empirical phenomena explainable by these propositions.

KNOWLEDGE–KNOWER STRUCTURES

•

MATON

91

Knower structures Bernstein’s concepts enable the form taken by the knowledge structures characterizing the two cultures to be described. To reach an understanding of their underlying principles, I shall now turn to consider their knower structures. These reveal a different picture (see Figure 5.1). Taking humanist culture first, I described how the humanities were portrayed as having been a ‘common culture’ underpinned by the Classics. However, it was not Classics as skills, techniques and procedures that integrated the humanities into a culture, but rather the dispositions that a classical education was thought to guarantee. The ideal humanist intellectual was a gentleman amateur who pursued (usually) his studies ‘for the love of it’, viewing them as secondary to a clerisy role of cultivating the cultured sensibility of the ‘English gentleman’ among students selected on the basis of fitting the character of the university (Maton 2004). The humanities were said to humanize; underpinning such claims was an image of what it meant to be human – the sensibilities, character and personal attributes of an ideal knower. The basis of specialization in humanist culture was thus not knowledge (indeed, disciplinary specialization was strongly devalorized in favour of the all-round ‘generalist’) but the habitus of an ideal knower, and a classical education served as shorthand for these dispositions. This cultural focus was, moreover, a veneer for a tacit social hierarchy. To be educated in the Classics was (in the main) to have enjoyed a particular social and educational trajectory – typically male, higher social class, private school, ‘Oxbridge’ – against which other knowers were (tacitly) measured. In other words, humanist culture exhibited what I shall term a hierarchical knower structure: a systematically principled and hierarchical organisation of knowers based on the image of an ideal knower which develops through the integration of new knowers at lower levels and across an expanding range of different (innate and/or social) dispositions.4 This can be portrayed as a pyramid of knowers (Figure 5.1) with, in the case of humanist culture, the ideal of the ‘English gentleman’ at its pinnacle. The basis of the recontextualizing principle of humanist culture and its ruler (in both senses of measuring and dominating) was thus an idealized knower. We can here further distinguish between those hierarchical knower structures with (by way of analogy to Bernstein’s terms) stronger ‘knower-grammars’, where the biological and/or social bases of the ideal knower(s) are articulated relatively explicitly (such as the Great Chain of Being underpinning papal and monarchical hierarchies and, more recently, various standpoint theories), and those with weaker ‘knower-grammars’, where the biological/social bases of knowers are more tacit and, as Bourdieu would put it (Bourdieu and Passeron 1990), ‘misrecognized’ as cultural (as was the case for the ‘English gentleman’).5 In terms of knowers, scientific culture was portrayed differently. Where the humanist intellectual’s ‘ability is a personal matter, which on the whole he does not owe to his advanced training’, scientific knowledge was widely said to be ‘fairly independent of the personal merits of its possessor’ (Gellner 1964: 75–6). Snow, for example, compared the class-bound basis of humanist culture to the democratic and meritocratic nature of scientific culture. He claimed science was

92

LANGUAGE, KNOWLEDGE AND PEDAGOGY

blind to colour, race, creed; it cut ‘across other mental patterns, such as those of religion or politics or class’ (1959: 9). In short, the basis of specialization in science was knowledge of scientific procedures, regardless of biological or social background. Science was thus portrayed as a horizontal knower structure: a series of strongly bounded knowers, each with its own specialized modes of being and acting, with non-comparable habituses or embodied dispositions based on different biological and /or social backgrounds and histories. In terms of their dispositions, scientists could represent a series of segmented knowers (Figure 5.1), each strongly bounded from one another in terms of their (non-scientific) ‘gaze’ and capable of being based on very different, even opposed, assumptions. Exploring knower structures highlights something not immediately obvious from studying knowledge structures alone: as illustrated in Figure 5.1, it is not only hierarchical knowledge structures that are characterized by a hierarchy. The location of the ‘hierarchical’ in an intellectual field could be described as the site or basis of its recontextualizing principle. Hierarchical knower structures thereby also possess a systematic principle for selecting and arranging actors and discourses into a hierarchy. That is to say, within intellectual fields actors and discourses are selected and recontextualized into positions within the field on the basis of a principle emanating from the knowledge structure, knower structure or, as I shall discuss, neither or both. In the case of the portrayal of the two cultures, the basis of their recontextualizing principles lay in the knowledge structure for science and in the knower structure for humanities. The difference between intellectual fields may thus be less whether they are hierarchical or not and more where their hierarchizing principle lies. Humanist culture

Scientific culture

Knowledge structures (horizontal) (hierarchical)

Knower structures (hierarchical)

Figure 5.1

(horizontal)

The two cultures as knowledge structures and knower structures

KNOWLEDGE–KNOWER STRUCTURES

•

MATON

93

Analysing knowledge–knower structures Having described differences between the two cultures in terms of their knowledge structures and knower structures, we are now in a position to analyse the structuring principles underlying fields of production. If we understand the discursive practices of intellectual fields as knowledge–knower structures that specialize actors and discourses in different ways, then the principles underlying these practices can be addressed in terms of their legitimation codes of specialization. This notion is based on the simple idea that actors and discourses are not only positioned in both a structure of knowledge and in a structure of knowers, but also establish different forms of relations to these two structures. One can thereby analytically distinguish between an epistemic relation (ER) to the knowledge structure and a social relation (SR) to the knower structure. Each of these relations can exhibit relatively stronger (+) or weaker (–) classification and framing. Varying their strengths for each relation independently generates four principal codes: ER+/–, SR+/–, where ‘ER+’, for example, condenses ‘ER(+C, +F)’. In other words, actors may emphasize the knowledge structure, the knower structure, neither or both as the basis of distinctiveness, authority and status; conversely, their identity, relations and consciousness are shaped in different ways by these two structures. These legitimation codes represent different ‘settings’ of the epistemic device, the means whereby intellectual and educational fields are maintained, reproduced, transformed and changed (Moore and Maton 2001). Whoever controls the epistemic device possesses the means to set the shape of the field in their favour, making what characterizes their own practices (in terms of legitimation codes) the basis of status and achievement in the field. This brief and somewhat formal definition can be fleshed out by considering the different ways in which the two cultures established relations to their knowledge–knower structures. Perhaps the most controversial claim Snow made in his lecture was that science was the basis of a true ‘common culture’: ‘the scientific culture really is a culture . . . Without thinking about it, they respond alike. That is what a culture means’ (1959: 9, 10). The basis of this culture was scientists’ ‘sense of loyalty to an abstraction called “knowledge”’ (Mackerness 1960: 15), commitment to ‘truth’ (Bronowski, 1961) and allegiance to their discipline (Pakenham, 1963), which specialized their identity and claims to insight. In other words, for science the epistemic relation to its knowledge structure was central to the field; this structure strongly classifies and frames actors and discourses within the field (ER+), while the social relation to its knower structure was less significant (SR–): what is defined as a knowledge code (see Figure 5.2). In the case of humanist culture, knowledge itself mattered a lot less; possession of procedures and skills was relatively unimportant in defining identity and achievement, so the epistemic relation to its knowledge structure was weakly classified and framed (ER–). Instead, the basis of specialization was possessing the right kind of dispositions or character. In other words, the field strongly classifies and frames knowers (SR+); for the humanities, the social relation to its knower structure was the key to the field – a knower code. Comparing the two cultures in Figure 5.1 shows it is that which is hierarchical (the pyramids) that

94

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Humanist culture

Scientific culture

Epistemic relation

–C, –F

+C, +F

Social relation

+C, +F

–C, –F

knower code (ER–, SR+)

knowledge code (ER+, SR–)

Legitimation code

Figure 5.2

Legitimation codes of specialization for the two cultures

Note: Classification (C) refers to relative strength of boundaries between categories or contexts; framing (F) refers to relative strength of control within these categories or contexts; ER refers to epistemic relation and SR to social relation; ‘+/–’ indicates relatively stronger/weaker. The notation for legitimation codes condenses, for example, ‘ER (+/–C, +/–F)’ to become ‘ER+/–’.

strongly classifies and frames actors and discourses within the intellectual field (bold type in Figure 5.2): the epistemic relation to the knowledge structure for scientific culture and the social relation to the knower structure for humanist culture. Having described the two cultures in terms of their knowledge structures and knower structures and analysed the role they play in specializing insight and identity in terms of legitimation codes, we can now return to the two questions raised earlier: the basis of difference between the two cultures and reasons for the shift of power between them. First, the debate can be redescribed as a struggle for control of the epistemic device between intellectual fields characterized by contrasting legitimation codes. These different codes characterize the kind of resources actors bring to the struggle. This is clearly illustrated by the two most famous protagonists in the debate: C.P. Snow and F.R. Leavis. Snow repeatedly emphasized: On these issues [of relations between the two cultures] our personalities mean nothing: but the issues themselves mean a great deal . . . The important thing is to take the personalities, so far as we are able, out of the discussion.(1964: 56, 59)

In contrast, Leavis was concerned with Snow as a legitimate knower: It is not any challenge he thinks of himself as uttering, but the challenge he is, that demands our attention.(1962: 10–11)

For humanists, as Leavis put it, a ‘judgement is personal or it is nothing; you cannot take over someone else’s’ (1962: 28). This represents a struggle between ‘what you know’ (knowledge code) and ‘who you are’ (knower code) as measures of status, identity and insight.

KNOWLEDGE–KNOWER STRUCTURES

•

MATON

95

Given this code clash it is little wonder that between the two cultures was said to lie, as Snow put it, ‘a gulf of mutual incomprehension’. Leavis could be speaking for both sides when he exclaimed: ‘He doesn’t know what he means, and doesn’t know he doesn’t know’ (1962: 10). Moreover, the rise of science and the proclaimed crisis in humanities were intimately interrelated: rising status for science threatened to change the basis of the distribution of resources and status within the field and relegate humanists to second-class citizens. If scientists controlled the epistemic device, then the field would tilt in their favour by making a knowledge code the basis of achievement. Secondly, the difference in codes also suggests reasons for why this shift in power seemed imminent. One reason lies in the different relationships the codes establish between their knowledge formations and horizontal discourse (or everyday knowledge). As discussed, science was portrayed as specialized by its language rather than its speakers: who was speaking was said to be less important than what they were talking about and how. The mathematization of science from the seventeenth century onwards had made this language progressively different to commonsense understanding, making discursive distinction from the contents and form of horizontal discourse the basis of the specialization of science. The scientist B.C. Brookes, for example, claimed ‘it will never be possible’ to translate between the two and that ‘the learning of science is the learning of a first, not a foreign, language’ that needed ‘lengthy and ruthless indoctrination’ (1959a: 502–21, 1959b: 783–4). Measured in terms of its knowledge code, science was thereby becoming ever more specialized in relation to horizontal discourse. In contrast, the knower code basis of identity and status in the humanities made dispositional distinction the basis of status; i.e. distinction between the dispositions of humanist knowers and those of the laity, rather than the possession of specialized knowledge and skills. In these terms the position of humanists was being undermined on two fronts. First, expansion was bringing more varied knowers into higher education presenting challenges to its hierarchical knower structure (and so its basis of specialization) under wider social conditions where belief in the integrating knower was waning; by the 1960s the ‘English gentleman’ was becoming viewed as outdated in what was being portrayed as an emergent ‘meritocratic’ society. Second, when judged by the discursive distinction of science’s knowledge code, the humanities were becoming less special. The extension of literacy under educational expansion was giving birth to ‘the articulate society’ where everyone felt entitled to speak and in which the ‘clerk is a nobody not merely because he is not a scientist, but also because in the developed societies everyone is now a clerk’ (Gellner 1964: 78). The humanities did not involve learning specialized procedures – there ‘is no enormous discontinuity, a yawning gap, bridgeable only by prolonged training’; instead one could pick up a discipline ‘simply by soaking in the ambience’ (Gellner 1964: 70) – and so humanists were vulnerable to being viewed as speaking little more than a jargon-ridden form of everyday language. In short, what threatened humanist culture was the entry of new knowers into a field of higher education increasingly dominated by a knowledge code. To recap, thus far I have addressed fields of knowledge production and the

96

LANGUAGE, KNOWLEDGE AND PEDAGOGY

first question I raised at the outset of the chapter of their generative principles. Alongside Bernstein’s ‘knowledge structures’ I introduced the notion of ‘knower structures’ to more fully describe differences between intellectual fields. I argued that analysing these knowledge–knower structures in terms of legitimation codes of specialization provides a means of conceptualizing their underlying structuring principles. I also briefly illustrated the kinds of insights this can provide into the form taken by different intellectual fields. The second question raised concerned relations between the concepts of knowledge structures and educational knowledge codes, or how Bernstein’s languages of description of fields of production and fields of reproduction can be integrated. It is, therefore, to educational fields of reproduction I now turn.

Knowledge–knower structures in fields of reproduction In his paper outlining the concepts of educational knowledge codes, Bernstein (1971, 1977) identifies two principal modalities as dominating educational systems: a collection code of relatively stronger classification and stronger framing, and an integrated code of weaker classification and weaker framing. These code modalities, he argues, help shape educational identity and consciousness in different ways. A collection code emphasizes educational knowledge, producing what he calls a ‘clear-cut and bounded’ educational identity based on one’s academic subject (+C, +F). Specialization is thus based on the possession of knowledge; it ‘makes of educational knowledge something not ordinary or mundane, but something esoteric, which gives a special significance to those who possess it’ (1977: 99). In contrast, under integrated codes the role of educational knowledge is weakened and one’s educational identity is (on this basis) less certain and must be negotiated constantly (–C, –F). This analysis of educational knowledge codes can be understood as homologous to that provided by knowledge structures such that one can say Bernstein is analysing the way educational knowledge structures (such as a curriculum) specialize actors and discursive practices.6 In these terms, Bernstein’s analysis is coding the epistemic relation of educational knowledge (ER+ and ER– for collection and integrated codes, respectively). For every educational knowledge structure there is also an educational knower structure. So, in addition to Bernstein’s analysis, we can also code the role in specialization of the social relation to the educational knower structure. This strength depends on the particular empirical case being examined, but for simplicity of illustration one can say it is likely that under collection codes the dispositions of knowers play a lesser role (–C, –F). When emphasizing the possession and transmission of their academic subject knowledge as the basis of professional identity and practice, teachers are likely to comparatively downplay the significance of their (and their pupils’) biological and/or social dispositions. In contrast, under integrated codes there is more space for knowers’ dispositions to play a greater role in identity and consciousness (+C, +F); for example, more emphasis may be given to the capacity to develop the dispositions of the ‘whole child’. These classification and framing strengths, which

KNOWLEDGE–KNOWER STRUCTURES

•

MATON

97

invert those normally associated with collection and integrated codes, refer to the social relation to educational knower structures (in these cases, SR– and SR+). Bringing the above together to consider educational knowledge–knower structures, the examples represent a knowledge code (ER+, SR–) and a knower code (ER–, SR+), respectively. Thus far in this chapter I have focused on instances where coding orientations for the epistemic relation to (educational) knowledge structures and the social relation to (educational) knower structures are inverted: ER+, SR– (knowledge code) and ER–, SR+ (knower code). However, this is not necessarily always the case. The strengths of these two relations may vary independently of each other; knowledge structures and knower structures can independently play a stronger or weaker role in specializing actors and discourses. Varying their relative strengths generates four principal legitimation codes of specialization, as represented by Figure 5.3. As I shall show when using the concepts, this is not a set of dichotomized or binary ideal types: strengths for relations are relative and represent a continuum; the four legitimation codes are akin to naming directions created by points on a compass to help orientate oneself within the terrain. epistemic relation ER+ knowledge

social relation

elite

SR–

SR+

relativist

knower ER–

Figure 5.3

Legitimation codes of specialization

Of the four legitimation codes highlighted in Figure 5.3, I have already discussed: • •

a knowledge code (ER+, SR–), where possession of knowledge (procedures, skills, techniques) is emphasized as the basis of specialization; and a knower code (ER–, SR+), where the dispositions or ‘gaze’ of knowers are emphasized, whether these are described as innate or natural (such as notions of genius), inculcated (such as an artistic or literary sensibility

98

LANGUAGE, KNOWLEDGE AND PEDAGOGY

cultivated through prolonged immersion in great works) or resulting from the knower’s social position (such as standpoint theory based on class, race, gender, sexuality, religion, age, and so forth). In addition one can also highlight two further coding orientations: •

•

a relativist code (ER–, SR–), where legitimate identity and insight is ostensibly determined by neither knowledge nor dispositions – thoroughgoing relativism; and an élite code (ER+, SR+), where legitimacy is based not only on possessing specialist knowledge but also being the right kind of knower.7

The élite code is exemplified in intellectual fields by science during the early Enlightenment period, when it was not enough to follow scientific procedures to be considered a legitimate scientist, but one also had to be a gentleman (I discuss an élite code in educational fields, below). In short, to think in terms of educational knowledge–knower structures is to ask what makes actors, discourses and practices special or legitimate: knowledge (knowledge code), dispositions (knower code), neither (relativist code) or both (élite code)? To illustrate how these concepts can be used to investigate educational fields I shall briefly discuss some empirical research that is addressing the problematic position of music in the English school curriculum. School music: an élite code qualification In the current English school system there are a number of Key Stages (henceforth ‘KS’) at which children are tested: • • • •

KS1: school years 1–2 (ages 5–7) KS2: years 3–6 (ages 7–11) KS3: years 7–9 (ages 11–14) KS4: years 10–11 (ages 14–16)

Pupils study a compulsory curriculum of ten academic subjects for KS1–3. At this point they can choose, from a wider range of available subjects, which ones they wish to study for GCSE qualifications (comprising a combination of coursework and examination) completed by the end of year 11. Music is popular among pupils up to the end of KS3 (Lamont et al. 2003), but there is very low uptake for GCSE qualifications: approximately 7 per cent of pupils choose to take GCSE Music, compared to 38 per cent for History, 38 per cent for Art and Design, and 15 per cent for Drama. This unpopularity has not gone unnoticed; in July 2004 a ‘Music Manifesto’ was launched by the British government, aiming to champion the status of the subject and encourage young people to remain involved in music making. However, the question of why music is so comparatively unpopular remains unanswered. Most studies of music focus on the learning and playing of musical instruments in formal and informal settings outside school, and music in the curriculum is typically described as simply ‘out of touch’ or perceived as irrelevant

KNOWLEDGE–KNOWER STRUCTURES

•

MATON

99

by most children (e.g. Green 2001; Sloboda 2001). Studies of school music itself, including the few studies addressing the issue of low uptake at GCSE level, typically offer speculation or ad hoc, piecemeal and largely descriptive accounts of best practice in teaching (e.g. Bray 2000; Harland et al. 2000; see Lamont 2004; QCA 2004). This question of why the GCSE qualification in music is so unpopular forms the starting point for a collaborative, interdisciplinary project (with Alexandra Lamont, a music psychologist at Keele University, UK) using the concepts of legitimation codes to investigate the ways achievement and educational identities in music are constructed. Our developing hypothesis is that GCSE Music represents an élite code and that this code plays a role in its low take-up rate. To illustrate how these concepts can be used to analyse educational fields I shall selectively report on the early stages of this research, focusing on three pilot studies that address: 1. 2. 3.

definitions of achievement in National Curriculum documents and syllabi; school-pupils’ perceptions of self-ability in, the significance of, and the basis of achievement in a range of academic subjects including music; and perceptions of university students of significance of and success in various school subjects.8

1. Curriculum documents The first study addresses levels of achievement expected of pupils at different Key Stages expressed in National Curriculum attainment targets and programmes of study (for KS1–3), and in the GCSE syllabi of major examination boards. The documents were analysed in terms of whether they emphasized the assessment of: skills, procedures, techniques and knowledge or dispositions of the learner, such as aptitude, attitude and personal expression. This analysis suggests the official requirements for music embody different legitimation codes for different stages of the curriculum. In KS1–2 the National Curriculum defines achievement in terms of the capacity of pupils to express themselves rather than demonstrate skills or knowledge. At the end of KS2, for example, pupils are expected to be able to ‘develop their own compositions . . . with increasing personal involvement, independence and creativity’ (DfES/QCA 1999: 18): a knower code. At KS3 (ages 11–14) attainment targets downplay aptitude, attitude and personal engagement in favour of the demonstration of musical skills and knowledge; pupils should show an ‘increasing ability to discriminate, think critically and make connections between different areas of knowledge’ (DfES/QCA 1999: 20): a knowledge code. At GCSE level the code changes again. Examination syllabi for GCSE Music require pupils to demonstrate both their capacity for personal expression and ability with technical skills. The syllabus of the examination board Edexcel, for example, includes a solo musical performance assessed for being both ‘accurate and fluent’ and ‘an expressive performance that is generally stylish’, with equal emphasis given to ‘Accuracy’ and ‘Interpretation’ (Edexcel 2002: 21, 22): an élite code. This suggests one possible reason for low uptake may be a shift in legitimation code underlying prescribed definitions of achievement in music: from knowledge code at KS3 to élite code at GCSE.

100

LANGUAGE, KNOWLEDGE AND PEDAGOGY

2. Perceptions of pupils Having analysed the National Curriculum, the next pilot study focused on whether these definitions of achievement are reflected in the perceptions of pupils. A questionnaire was completed by 912 pupils aged 8–14 at four comprehensive schools of average size and achievement rating. The survey included three main questions about music, the core curriculum subjects of English, mathematics and science (which are compulsory subjects for study in KS4), and history (for comparison). For each subject pupils were asked to rate the importance of being good at the subject, rate their self-ability, and describe the basis of success at the subject. I shall focus on the third question here: ‘What do you think makes someone good at [the subject]?’. Respondents were offered a forced choice of one of four options, representing our first attempt at capturing relativist, knowledge, knower and élite codes, respectively: (A) Anyone can do it, nothing special is needed; (B) You need to learn special skills or knowledge; (C) You need to have ‘natural ability’ or a ‘feel’ for it; (D) Only people with ‘natural ability’ can learn the special skills needed. Analysis of the data for all pupils across all years suggests the basis of success is viewed differently for science and the humanities: modal responses were knowledge option B for science, maths and music, and relativist option A for English (marginally, over B) and history. (As I discuss below, the latter two may result from our wording of options C and D.) However, this global picture conceals significant differences in results for different subjects and different pupil ages. One such result of interest here is that among pupils who have chosen their GCSE subjects in year 9, Music was far more often characterized as embodying an élite code than other subjects: 19 per cent chose option D for music, compared to a maximum of 3.6 per cent for the other subjects. This figure almost doubles to 35 per cent among those pupils who chose to study music at GCSE. I shall return to consider the implications of these results shortly. 3. Perceptions of university entrants on school subjects The third pilot study explored, through surveys and focus groups, the perceptions on school subjects of older students who have already made a number of subject choices and are starting their university studies. The survey I shall discuss comprised 93 new entrants, first-year students at a middle-ranking English university. This included questions about the same three issues as outlined above for the same four subjects, as well as psychology (in which all the students were taking at least a module). For this study we redesigned the question of the basis of success in academic subjects for three main reasons. First, our previous dispositional option C offered only ‘natural ability’ or ‘feel’, neglecting the notion of cultivated sensibilities or refined judgement, such as are often emphasized in literary and art criticism (which may account for the previous low response rate for options including ‘natural ability’). Secondly, the phrasing of the élite option D made ‘natural ability’ the basis for access to ‘special skills’ rather than bringing together both dispositions and knowledge.

KNOWLEDGE–KNOWER STRUCTURES

•

101

MATON

Thirdly, the pupil survey was our first attempt at using coding concepts in quantitative research; our choice of a forced-choice design began from the four code modalities rather than from their basis in the strengths of epistemic and social relations. Such a categorical scale design suits ideal typical groupings, whereas the theory emphasizes the relative strengths of the two relations in determining coding orientations, requiring a more continuous scaling approach. To address these issues, ‘taste, judgement or feel’ was added as an option alongside ‘natural talent’ (separately because though both are dispositional they are often opposed in well-known debates over, for example, ‘nature versus nurture’), and the forced-choice design was replaced with a sliding scale of importance (see Figure 5.4). The new design thus asked respondents to rate the significance of a subject’s knowledge structure (‘skills’) and of its knower structure (‘talent’ and ‘taste’). In your opinion, how important are these things for being good at [the subject]? Not at all

Not very

Quite

Very

Skills, techniques and specialist knowledge

[ ]

[ ]

[ ]

[ ]

Natural-born talent

[ ]

[ ]

[ ]

[ ]

Taste, judgement or a developed ‘feel’ for it

[ ]

[ ]

[ ]

[ ]

Figure 5.4

University students’ survey; question of basis of achievement

The theory’s emphasis on relative strengths was also reflected by the analysis. The ratings were coded numerically as 1–4 and mean scores calculated across all subjects for the ‘skills’ scale and for the ‘talent’ and ‘taste’ scales taken together, to give baseline scores of significance for the epistemic and social relations, respectively. The scores of each subject for these two relations were then compared to these two overall mean scores. From this analysis (see Figure 5.5): •

• • • •

natural science (and, to a lesser extent, psychology) scored higher for ‘skills’ (stronger epistemic relation) and lower for ‘talent / taste’ (weaker social relation) – a knowledge code; for English these results were reversed, with the epistemic relation being weaker and the social relation being stronger – a knower code; maths was average for both, a result requiring further investigation; history scored lower for both skills and ‘talent / taste’ – a relativist code; and music scored higher for both – an élite code.

In other words, comparative to other subjects one requires both skills and taste or talent to succeed in music.

102

LANGUAGE, KNOWLEDGE AND PEDAGOGY

–

Natural Science

Maths

Psychology

English Literature

History

Music

■ Skills, techniques and specialist knowledge ■ Talent, taste, judgement or a developed ‘feel’

Figure 5.5

University students’ perceptions of basis of achievement

We also conducted a series of focus groups with members of the same population. The above coding orientations were reflected in how participants discussed achievement. Group discussions of science and English literature, for example, illustrate their respective knowledge code and knower code. Participants tended to state that for English the knower is the source of a legitimate gaze which generates insights, whereas in science it is educational knowledge that forms the basis of insight and identity; for example: Moderator: What does it take to be good at English? Participant 1: I learnt to have my own opinion and back it up with my own evidence but then use evidence from other people that have the same opinion as me, so you’re still using other opinions but you’re finding them after you’ve made your own. Moderator: Is that different from science or maths? P2: Yeah, definitely. You can’t really say ‘Well, my theory of evolution is . . .’ It’s not like you can make up your own theory. P1: You’re given theories and you choose one rather than having your own opinion and then finding someone who agrees with you. The élite coding orientation of music was reflected in how participants would often shift between talent or natural ability, developing skills and such issues as portraying emotion when discussing achievement in the subject:

KNOWLEDGE–KNOWER STRUCTURES

•

MATON

103

P3: It’s more talent-based, you have to have a natural ability. P4: You can’t just throw anyone in there and teach them, they have to have that ability before they start. Everyone can learn the basics but to get to the top . . . P3: Music takes lots of practice . . . you have to practise every day to get better at it. P5: You can never say you’ve done all the work for it. You can always do a bit better. Whereas in science if you learn it there is a point where you’ve learned everything that you need to know. P3: Even someone with natural talent that’s very good at music still has to practise. P5: It’s talent and skills and hard work. P3: You need to be able to portray emotion too. Implications of music’s élite code If Music is portrayed as embodying an élite code in curriculum documents and the perceptions of school-pupils and university entrants, the question is how that coding orientation might relate to its unpopularity at GCSE level. The above discussion summarizes only part of the analysis of the data, which includes age differences, social variables such as gender, differences between ‘taste’ and ‘talent’, rating of self-ability and the significance of different results. The research is also ongoing (for example, the redesigned survey is being used within an intervention study in English secondary schools). In addition, further studies are required for a fuller picture, including analysis of: the social distribution of legitimation codes among different groups of pupils; constructions of achievement within the intellectual field of music education research; the formulation of curriculum in the field of recontextualization; and the ways school music is taught at different Key Stages. However, the results I have discussed are themselves suggestive in several ways. First, the shift to élite code at GCSE level is not simply a code shift (as happens between KS2 and KS3, from knower code to knowledge code) but to one that is doubly demanding: pupils must both demonstrate possession of musical knowledge and skills, and express musical dispositions. In other words, not only are the rules of the game changed, but it becomes harder to play – an élite code has two hierarchies (one of which, the knower structure, it may not be possible to climb). Second, this may make music an unenticing option, particularly if one considers its perceived significance. When asked to rate the importance of being good at a subject, music was the least important subject for both school-pupils and university entrants. Its élite code thereby does not seem to be reflected by status; as one focus group described it: P6: I don’t think if you were going to apply to be a doctor they’d say, ‘Have you got your grade 9 piano’ or whatever. P7: I think if I told people I was doing a music degree everyone would be like ‘What’s the point? Waste of time!’ kind of thing. P8: Yeah, everyone thinks doing music at university is learning to play ‘Three Blind Mice’ on the recorder.

104

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Lastly, if music’s élite code is not widely distributed socially and the keys to the code not made visible in pedagogic practices, then school qualifications are likely to remain restricted. The central point here, however, is less the basis of music’s position and more the issue of how analysing knowledge–knower structures in terms of legitimation codes offers fruitful ways forward for empirical research into educational fields. It reveals not only contexts exhibiting stronger or weaker classification and framing, but also those with both; such contexts may appear contradictory or confusing if one considers educational knowledge structures on their own. Élite schools, for example, may operate with selection criteria based not only on qualifications but also on issues of character and disposition. Integrating knower structures into the analysis may show that such contexts exhibit an élite legitimation code. By being anchored on the concepts of classification and framing, the strong external language of description of legitimation code theory also enables both qualitative and quantitative analysis of the underlying principles structuring curriculum guidelines, teaching practices, pupils’ perceptions, school organization, and so forth in a manner enabling systematic comparisons within and between these contexts, something currently lacking from existing research on music in the curriculum.

Conclusion In this chapter I have argued that one can analytically distinguish two message systems within the discursive practices of intellectual and educational fields. The first is that addressed by Bernstein’s conceptualization of fields of production in terms of knowledge structures, and fields of reproduction in terms of educational knowledge codes. These (educational) knowledge structures announce how knowledge should be created, distributed, recontextualized and evaluated. The second is to be found in their knower structures. I introduced the notions of hierarchical knower structures and horizontal knower structures with stronger and weaker ‘knower-grammars’; these proclaim how legitimate knowers are created, distributed, recontextualized and evaluated. This is, though, an analytical distinction. To understand fully intellectual and educational fields one must, I argued, bring these together to think in terms of knowledge–knower structures. Together they define the basis of specialization of actors and discourses within fields and so help shape relation, identity and consciousness. These concepts enable two questions raised by Bernstein’s ideas to be addressed. First, I showed how analysing knowledge–knower structures in terms of legitimation codes and the epistemic device provides a means of exploring the structuring principles underlying intellectual fields. This also highlights new issues of interest; for example, it recasts the question of hierarchies in intellectual fields from ‘whether’ to ‘where’ (in their knowledge structure, knower structure, neither or both) – horizontal knowledge structures may be characterized by hierarchical knower structures. Doing so further clarifies the recontextualizing principles of intellectual fields – this principle may be not only discursively based but also dispositionally based, with implications for the structure and develop-

KNOWLEDGE–KNOWER STRUCTURES

•

MATON

105

ment of the field. The notion of knowledge–knower structures also expands the reach of the analysis, subsuming and integrating (rather than displacing) the existing conceptualizations of knowledge structures and educational knowledge codes. Integrating the analysis of knower structures with that of knowledge structures within the concept of legitimation code not only enables their different insights to be brought together but also enables us to generatively conceptualize new possibilities, such as relativist and élite codes. Secondly, having considered intellectual fields (with the example of the ‘two cultures’ debate) I illustrated how these concepts can also be applied to educational fields by discussing studies into school music. Preliminary results from these studies suggest that the unpopularity of school qualifications in music may be related to its élite code. The concepts thereby provide a means of bringing analyses of fields of intellectual production and of educational reproduction within the same framework. The use of these concepts is at a relatively early stage; as more empirical questions are addressed, their tacit potential and limitations will become increasingly evident, necessitating further development of the theory and opening up the next chapter of the story.

References Bernstein, B. (1971), ‘On the classification and framing of educational knowledge’, in M.F.D. Young (ed.), Knowledge and Control. London: Collier Macmillan. Bernstein, B. (1977), Class, Codes and Control, Volume III: Towards a theory of educational transmissions. London: Routledge & Kegan Paul. Bernstein, B. (1990), Class, Codes and Control, Volume IV: The structuring of pedagogic discourse. London: Routledge. Bernstein, B. (1996), Pedagogy, Symbolic Control and Identity: Theory, research, critique. London: Taylor & Francis. Bernstein, B. (1999), ‘Vertical and horizontal discourse: An essay’, in British Journal of Sociology of Education 20(2), pp. 157–73. Bourdieu, P. and Passeron, J-C. (1990), Reproduction in Education, Society and Culture. London: Sage. Bray, D. (2000), ‘An examination of GCSE music uptake rates’, in British Journal of Music Education 17(1), pp. 79–89. Bronowski, J. (1961), Science and Human Values. London: Hutchinson. Brookes, B.C. (1959a), ‘The difficulty of interpreting science’, in The Listener LXII, pp. 519–21. Brookes, B.C. (1959b), Letter to the editor, in The Listener LXII, pp. 783–4. DfES/QCA [Department for Education and Science/Qualifications & Curriculum Authority] (1999), The National Curriculum for England: Music. London: HMSO. Doherty, C. (2004), ‘Student subsidy of the internationalized curriculum: Knowing, voicing and producing the Other’. Paper presented at Reclaiming Knowledge: Registers of discourse in the community and school conference, University of Sydney, December. Edexcel (2002), Specification: Edexcel Advanced Subsidiary GCE in Music (8501) and Advanced GCE in Music (9501). London: Edexcel Foundation. Gellner, E. (1964), ‘The crisis in the humanities and the mainstream of philosophy’, in Plumb, J.H. (ed.), Crisis in the Humanities. Harmondsworth: Penguin. Green, L. (2001), How Popular Musicians Learn: A way ahead for music education. Burlington: Ashgate.

106

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Harland, J. et al. (2000), Arts Education in Secondary Schools: Effects and Effectiveness. Slough: National Foundation for Educational Research. Hickox, M. and Moore, R. (1995), ‘Liberal-humanist education: The vocationalist challenge’, in Curriculum Studies 3(1), pp. 45–59. Hood, S. and Maton, K. (2005), ‘The languages of disciplinarity: Knowledge, knowers and recontextualisation’. Paper presented at Discourses of Hope: International Systemic Functional Congress 32, University of Sydney, July. Lamont, A. (2004), What are the possible reasons for the low take-up of music GCSE? Music in Key Stage 4: QCA Music Development Group working paper. Lamont, A., Hargreaves, D.J., Marshall, N.A. and Tarrant, M. (2003), ‘Young people’s music in and out of school’, in British Journal of Music Education 20(3), pp. 1–13. Leavis, F.R. (1962), Two Cultures: The significance of C. P. Snow. London: Chatto & Windus. Lee, H.D.P. (1955), ‘The position of Classics in schools’, in Universities Quarterly 9(2), pp. 135–44. Mackerness, E.D. (1960), ‘Ignorant armies’, in The Universities Review 33(1), pp. 14–17. Maton, K. (2000a), ‘Recovering pedagogic discourse: A Bernsteinian approach to the sociology of educational knowledge’, in Linguistics and Education 11(1), pp. 79–98. Maton, K. (2000b), ‘Languages of legitimation: The structuring significance for intellectual fields of strategic knowledge claims’, in British Journal of Sociology of Education 21(2), pp. 147–67. Maton, K. (2002), ‘Popes, kings and cultural studies: Placing the commitment to nondisciplinarity in historical context’, in S. Herbrechter (ed.), Cultural Studies: Interdisciplinarity and translation. Amsterdam: Rodopi. Maton, K. (2004), ‘The wrong kind of knower: Education, expansion and the epistemic device’, in Muller, J. Davies, B. and Morais, A. (eds), Reading Bernstein, Researching Bernstein. London: RoutledgeFalmer. Maton, K. (2006), ‘On knowledge structures and knower structures’, in Moore, R., Arnot, M., Beck, J. and Daniels, H. (eds), Knowledge, Power and Educational Reform: Applying the sociology of Basil Bernstein. London: Routledge. Moore, R. (2004), Education and Society: Issues and explanations in the sociology of education. Cambridge: Polity Press. Moore, R. and Maton, K. (2001), ‘Founding the sociology of knowledge: Basil Bernstein, intellectual fields and the epistemic device’, in Morais, A. Neves, I., Davies, B. and Daniels, H. (eds), Towards a Sociology of Pedagogy: The contribution of Basil Bernstein to research. New York: Peter Lang. Moore, R. and Muller, J. (2002), ‘The growth of knowledge and the discursive gap’, in British Journal of Sociology of Education 23(4), pp. 627–38. Morris, M. (1959), ‘The two cultures and the scientific revolution’, in Marxism Today 3(12), pp. 374–80. Muller, J. (2004), ‘On the shoulders of giants: A digression on knowledge, curriculum, and finally, the teacher’. Paper presented at the Third International Basil Bernstein Symposium, University of Cambridge, July. Pakenham, T. (1963), ‘The study of the university teacher: The Gulbenkian Educational Discussion 1962’, in Universities Quarterly 17(2), pp. 149–64. Plumb, J.H. (ed.) (1964a), Crisis in the Humanities. Harmondsworth: Penguin. Plumb, J.H. (1964b), ‘The historian’s dilemma’, in J.H. Plumb (ed.), Crisis in the Humanities. Harmondsworth: Penguin. QCA [Qualifications and Curriculum Authority] (2004), Music in Key Stage 4: Developing qualifications within the vision of a holistic and progressive music education for all. QCA Music Development Working Paper. Singh, P. (2002), ‘Pedagogising knowledge: Bernstein’s theory of the pedagogic device’, in British Journal of Sociology of Education 23(4), pp. 571–82. Sloboda, J.A. (2001), ‘Conference Keynote: Emotion, Functionality and the Everyday Experience of Music’, in Music Education Research 3(2), pp. 243–53.

KNOWLEDGE–KNOWER STRUCTURES

•

MATON

107

Snow, C.P. (1959), The Two Cultures and the Scientific Revolution. Cambridge: Cambridge University Press. Snow, C.P. (1964), The Two Cultures and a Second Look. Cambridge: Cambridge University Press. Wheelahan, L. (2005), ‘Comparing realist and rationalist defences of propositional knowledge in the curriculum’. Paper presented at Engaging Realist Alternatives: IACR Conference, Sydney, July.

Notes 1

2 3

4

5

6

7

8

9

This paper expands on Maton (2006), in particular elaborating on grammars of knower structures and developing the account of knowledge–knower structures in educational fields. See Doherty (2004), Hood and Maton (2005), Lamont (2004), Maton (2004) and Wheelahan (2005) for examples of educational studies using legitimation codes. This is not the same as asking how specific kinds of knowledge structures and curriculum/pedagogic structures are related; for a perceptive account exploring this issue, see Muller (2004). Integration of new knowers may be through resocialization (such as that attempted by the creation of new campus universities as resocializing institutions in 1960s English higher education; Maton 2004) or through a mixture of indoctrination and coercion (such as underlay medieval monarchical and papal hierarchies; Maton 2002). Educational expansion has typically accommodated new knowers through offering a choice of resocialization as the condition of entry into higher status institutional and disciplinary positions, or relegation into lower levels of these status hierarchies (cf. Hickox and Moore 1995). There may be more than one idealized knower and pyramid of knowers in an intellectual field characterized by a hierarchical knower structure. This has effects for the possibility and means of transmission/acquisition of the legitimate ‘gaze’ conferred by knower status. In the case of stronger ‘knower-grammars’ acquisition may be explicitly restricted to those already possessing knower status and transmission may focus on attempting to raise to consciousness what is proclaimed to be a pre-existing gaze; with weaker ‘knower-grammars’, the possibility of successful acquisition of the legitimate gaze is claimed to be more widely available and transmission may take the form of attempting to resocialize potential knowers, restructuring their habituses. I should emphasize that Bernstein’s concepts of knowledge structures refer to intellectual fields of production and not educational fields – they are not describing curriculum or pedagogic structures. I use ‘educational knowledge structures’ by way of analogy as a step towards offering a means of integrating analyses of intellectual and educational fields. One can expand the language of description to provide a more subtle theorization by considering differing strengths of classification and framing. This generates sixteen modalities: ER(+/–C, +/–F), SR(+/–C, +/–F). Here I restrict discussion of coding orientations to where classification and framing strengths are both stronger or weaker for conceptual economy and because Bernstein’s theorization and applications of the concepts suggest they are the most commonly found orientations. My principal focus is illustrating the application of the concepts rather than the music question per se. I thus discuss only selected aspects of the studies. We shall report the results of this ongoing research and their implications for school music more fully in future publications. It is tempting to account for the low uptake rate of music in terms of value in the occupational marketplace alone; however, this would not easily account for the uptake rate for drama being double that of music.

108

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Acknowledgements I am extremely grateful to Alexandra Lamont, with whom I am conducting the music research; Jim Martin and Fran Christie, for inviting me to give a keynote paper at the Reclaiming Knowledge conference; Geoff Whitty and Brian Davies for enlightening discussions of my work as a whole; and Joe Muller and Rob Moore for critical encouragement in my ideas.

6

Hierarchical knowledge structures and the canon: a preference for judgements Rob Moore

As part of society, the individual naturally transcends himself, both when he thinks and when he acts. (Emile Durkheim) ‘Truth’ is the reigning sacred object of the scholarly community, as ‘art’ is for the literary/artistic communities; these are simultaneously their highest cognitive and moral categories, the locus of highest value, by which all else is judged. (Randall Collins)

Introduction This chapter addresses from the perspective of aesthetic concerns a set of issues more usually approached in terms of epistemology. Essentially, the concern is with the production of judgements – the evaluation of the relative merits of art objects and the discrimination of some as being ‘better’ than others. It must be stressed that this is an exercise in sociology, not in aesthetics or criticism. No claims are made about any particular art object relative to any other. To say that it is sociological is to indicate that the focus is upon the sociality of the production of aesthetic judgements: that the production of aesthetic judgement is something that people do within a special type of socio-historical context (a field or arena) that can be described in terms of its structural features, its generative principles (its ‘powers’) and understood at the level of the individual in terms of the mediation of habitus. This type of context is conventionally termed a ‘canon’ and its distinctive feature is that it is extensive in space and time. In his later writings on hierarchical and horizontal knowledge structures, Basil Bernstein (Bernstein 2000) began to develop a new conceptual language through which to theorize the structurings of intellectual fields and the conditions of their productivity. The basic difference between these two types of knowledge structure (each of which can vary in strength within itself as well as differing between each other) is the capacity of hierarchical knowledge structures to produce knowledge at increasingly higher levels of abstraction, generality and integration (they are strong in grammaticality – see Muller, this volume) and, hence, are able to produce progression in knowledge. By contrast, horizontal knowledge structures specialize knowledge to knowers by constructing a high level of segmentation between social categories of ‘knowers’ and grounding knowledge in their distinctive and incommensurable sets of experience – in particular, experiences of unequal power relations in society. Horizontal knowledge structures legitimate themselves in terms of who knows

110

LANGUAGE, KNOWLEDGE AND PEDAGOGY

rather than what is known. They authorize themselves through the ‘voice’ of those whose experiences they claim to represent (Moore and Muller 1999). As Karl Maton has demonstrated (2000), these two types of knowledge structure operate with different ‘languages of legitimation’ associated with different orientations of the ‘epistemic device’ (Maton, this volume; Moore and Maton 2000). In an early formulation before he had developed the concept of ‘knowledge structures’, Bernstein wrote as follows: In a subject where theories and methods are weak, intellectual shifts are likely to arise out of conflict between approaches rather than conflict between explanations, for, by definition, most explanations will be weak and often non-comparable, because they are approach-specific. The weakness of the explanation is likely to be attributed to the approach, which is analysed in terms of ideological stance. Once the ideological stance is exposed, then all the work may be written off. Every new approach becomes a social movement or sect which immediately defines the nature of the subject by re-defining what is to be admitted, and what is beyond the pale, so that with every new approach the subject almost starts from scratch. (Bernstein 1977: 167–8)

This is from a ‘brief account’ of the sociology of education from the early 1970s. It is of interest because the distinction between debates at the level of ‘approaches’ or ‘explanations’ points forward to the concepts of horizontal and hierarchical knowledge structures, and also because, some thirty years later, under the influence of post-modernist, post-structuralist and feminist standpoint approaches, intellectual fields within the social sciences and humanities display these features of horizontal segmentation (low grammaticality) even more strongly than when Bernstein was originally writing! The intention of this paper is to indicate ways in which Bernstein’s later concepts (and also the work of others) might enable solutions to problems in the sociology of knowledge through a new theory of the structuring of intellectual fields and, in so doing, suggest the conditions for intellectual productivity and progress in knowledge in the social sciences and humanities. Specifically, this chapter will address the ‘sociality of judgement’ through a sociological defence of the concept of the canon.

The canon under fire For some time, the idea of a canon, and its associated concepts, have been challenged from a number of quarters (by various currents of post-modernism, for instance). The essence of this ‘critical’ challenge is that the basic principle of a canon – that it enshrines those things (art objects, ideas, etc.) that are of intrinsic, superior worth – is fundamentally false. Canons are seen as arbitrary constructions reflecting no more than the tastes and fashions of dominant social groups or, at worse, as ideological forces that legitimate and reproduce the position and power of dominant groups. Whereas in the traditional manner a canon is taken as inclusive and universal, in the critical approach it is seen as

HIERARCHICAL KNOWLEDGE STRUCTURES

•

MOORE

111

exclusive and partial. Critical approaches tend to assume two forms: reductionism and relationalism. Reductionism has its origins in Marxist ideology analysis and standpoint theory. It operates by identifying the social base of a work of art or movement, the standpoint that it reflects and the interest it serves. In the traditional Marxist form artistic movements might be classified as ‘bourgeois art’; more recently, feminist examples of this approach might describe the literary canon as the ‘male literary canon’. Relationalism is best represented in the work of Pierre Bourdieu (e.g. 1993). In this case, items are distributed in a ‘field’ (e.g. the cultural field) and are related to each other in terms of their synchronistic positioning. Fields are structured between high and low (that which is deemed to have distinction and confers ‘cultural capital’, and that which is vulgar). Aesthetic value is generated through the relationships between positions rather than through the art object in itself. To attribute intrinsic aesthetic value to an art object is ‘essentialism’ that involves ‘misrecognition’ and entails ‘symbolic violence’. Values are ultimately arbitrary in that they reflect no more than the social distribution of taste and the structure of the field; no more than the system of power relations and economic interests of society (Moore 2004a). Although relationalism is conceptually more complex and sophisticated than reductionism, ultimately, they both say the same things: cultural and intellectual fields are basically arbitrary and have no intrinsic values that can be considered independently of external social relations, interests and power. Once the social base has been revealed and the interest it serves and the standpoint it reflects exposed, there is no more to say – nothing in terms of truth or beauty in their own right. In recent years these debates have been fiercely expressed in the so-called culture and science ‘wars’ (Graff 1992; Brown 2001). Richard Hoggart has asked: Why are the arguments so angry? Why are so many people so violently disinclined to admit any differences in the value of different works of art; or between human choices as to activities? This is the most revealing of our multiple cultural hangups. It involves many people – the conclusion is inescapable – almost entirely rejecting ‘great’ works of art in any form (people who dismiss George Eliot as ‘merely a reflection of nineteenth century bourgeois values’ and clearly have not read her work). Such people as these are well-versed in one or the other art, but unwilling any longer to make value-judgements between them. By extension, they are uneasy about talk of art’s possible relation to ‘meaning’. They avoid any vertical judgements, in favour of the endlessly horizontal. By the Nineties a senior official with Radio 3 [the BBC’s classical music station] could announce: ‘There is no art; only culture’. A spiritual mate adds: ‘Each man is his own culture’. (Hoggart 1995: 57–8)

Basically, the problem is that of relativism – cognitive, moral and aesthetic. The argument in this paper is that relativism is the consequence of a mistaken preoccupation with absolutes in relation to values, and with a failure to properly understand the character of the sociality of knowledge.

112

LANGUAGE, KNOWLEDGE AND PEDAGOGY

It is interesting that Hoggart also employs a distinction between the ‘horizontal’ and the ‘vertical’. For Bernstein, the origin of these ideas is in Durkheim. In the later part of the twentieth century Durkheimian scholars (notably Basil Bernstein and Randall Collins) began to develop the conceptual apparatus whereby the structures, principles, powers and possibilities of these forms of sociality could be theorized and explored systematically. Others (e.g. John Beck, Karl Maton, Rob Moore, Johan Muller, Michael Young, individually and in various combinations: see the bibliography to this chapter) extended the range of this effort. This chapter draws upon that language and indicates ways in which longstanding philosophical issues might be addressed in terms of sociological understandings of how symbolic production ‘works’ in terms of the sociality of judgement.

The mind itself – the endlessly horizontal Because the field of debate is so broad, the focus will be upon one particular recent text in order to provide a central reference point. The emeritus professor of English Literature at Oxford University, John Carey, published a book provocatively entitled: What Good are the Arts? (Carey 2005). This book has been extensively reviewed and, as intended, no doubt, caused considerable controversy. Carey explicitly endorses the view cited above, that ‘Each man is his own culture’. His central argument is: ‘Anything can be a work of art. What makes it a work of art is that someone thinks of it as a work of art’ (Carey 2005: 29). In more substance, there is a distinction between those who . . . assume the existence of a separate category of things called works of art . . . which are intrinsically more valuable than things which are not works of art, and which accordingly deserve universal respect and admiration. These assumptions, we can now see, belong to the late 18th century, and are no longer valid. The question ‘Is this a work of art? – asked in anger or indignation or mere puzzlement – can now receive only the answer ‘Yes, if you think it is; no, if not.’ If this seems to plunge us into the abyss of relativism, then I can only say that the abyss of relativism is where we have always been in reality – if it is an abyss. (Carey 2005: 30)

Carey’s position is allied with post-modern relativism. For Carey, and others, relativism simply is not a problem. However, Carey does have a problem because his book contains two arguments, the second of which is a defence of literature on the grounds that it is ‘superior to the other arts, and can do things they cannot do’ (ibid., p. 173). Carey is not unaware of the fact that this argument seems at odds with the relativism that precedes it (loc. cit.), but he does not in any satisfactory manner resolve this tension. The reason for this is that it is not just that Carey presents two contradictory arguments, but that he draws upon or implies two quite different theories of knowledge in doing so. In the first instance there is the subjectivist relativism that he explicitly espouses, but in his second argument there are the elements of a radically different approach. In

HIERARCHICAL KNOWLEDGE STRUCTURES

•

MOORE

113

Bernsteinian terms, Carey number one argues in ‘horizontal’ mode (who knows), whereas Carey two argues in ‘vertical’ mode (what is known). The first task, though, is to demonstrate the wider implications of Carey’s position for the issues being addressed. The argument will begin by drawing attention to how Carey believes the arts would have to be if they were indeed to be special: ‘a separate category of things called works of art . . . which are intrinsically more valuable than things which are not works of art’. He grounds this, as he sees it, elitist and mistaken approach to the arts, in the ideas of Kant: It is easy to identify the dictates of Kant and his followers in the notions about art that are still in circulation today. That art is somehow sacred, that it is ‘deeper’ or ‘higher’ than science and reveals ‘truths’ beyond science’s scope, that it refines our sensibilities and makes us better people, that it is produced by geniuses who must not be expected to obey the same moral codes as the rest of us, that it should not arouse sexual desire, or it will become pornography, which is bad – these and other superstitions belong to the Kantian inheritance. So does belief in the special nature of artworks. For Kantians, the question ‘What is a work of art?’ makes sense and is answerable. Works of art belong to a separate category of things, recognized and attested by certain highly gifted individuals who view them in a state of pure contemplation, and their status as works of art is absolute, universal and eternal.(Carey 2005: 14)

Compare Carey’s statement with this example of a post-modernist account of science: The ascendancy of scientific method as the means of establishing knowledge has resulted in a consistent failure to examine science as a social practice and as a historical and cultural product. Science has instead been seen as transcendental and decontextualised. Knowledge, as well as the knowing subject, therefore, becomes context free. Rationality is cast as universal and transcendental, operating across all historical and social contexts and practices but independent of all of them. The result is an individualistic epistemology where the solitary individual confronts an independent reality of objects. (Usher and Edwards 1994: 36)

Carey himself does not set out to present a critique of science. Indeed, at a number of points, his argument is precisely that aesthetics cannot do what science does (and in this respect Carey does not appear to be an out-and-out cognitive relativist). The point is that in both cases these writers present the same basic logic of argument for rejecting truth claims in the areas of aesthetics and science. To put it succinctly, essentialism is to aesthetics what foundationalism is to epistemology. In both cases, it is being argued that truth claims or value judgements can only be forwarded if they are ‘universal and transcendental, operating across all historical and social contexts and practices but independent of all of them’, and are ‘absolute, universal and eternal’. And, similarly, there is the (basically Cartesian) model of the ‘privileged knower’ as ‘certain highly gifted individuals who view them [artworks] in a state of pure

114

LANGUAGE, KNOWLEDGE AND PEDAGOGY

contemplation’ and where ‘the solitary individual confronts an independent reality of objects’. In both cases these writers present absolutist or infallibilist definitions of truth whether aesthetic or epistemological and atomistic, contemplative models of the self. They both provide strong demolitions of absolutist claims. The problem is that they then conclude that because truth claims cannot be absolute – ‘universal and transcendental, operating across all historical and social contexts and practices but independent of all of them’ – then we cannot have truth claims at all. In this crucial respect, both sets of writers are simply wrong. The fact that we cannot produce truths that are absolutely infallible in the areas of epistemology, aesthetics or morals does not mean that we cannot intelligibly produce judgements that some things or practices are better than others. The All-or-Nothing Fallacy It is important to note that Carey advances his position on the basis of a particular set of alternatives: ‘The champion of high art would have to mean not just that his experience were more valuable to him, for that would not prove the superiority of high art, only his preference for it. He would have to mean that the experiences he derived from high art were in some absolute and intrinsic sense more valuable than anything the other person could get from low art.’ We are allowed only the choice between a preference or an absolute. This is a form of what J.C. Alexander has called the ‘epistemological dilemma’ – i.e. that the only choices available to us are those of relativism and absolutism (1995). For Carey, if we cannot provide absolute (‘divinely decreed’) support for our judgements then we cannot have any judgments at all, but merely preferences and tastes. Niiniluoto has made an observation similar to Alexander when describing ‘dogmatic scepticism’: ‘This mode of thinking could be called the All-or-NothingFallacy. Its different forms propose a strong or absolute standard for some category . . . and interpret the failure or impossibility of satisfying this standard as a proof that the category is empty and should be rejected’ (2002: 81). This is precisely Carey’s approach. Carey’s thinking is an example of a wider mode of thought that has been identified and critiqued by a number of commentators (Moore 2004b: ch. 6). Preferences and judgements In a key statement, Carey displays not only the manner in which his approach parallels a number of other kinds of relativism in the social sciences and cultural studies, but also indicates the kernel of the problem at the centre of such perspectives – essentially, a confusion over the difference between a preference and a judgement. Carey declares himself as follows: I have suggested that those who proclaim the superiority of high art are saying, in effect, to those who get their pleasure from low art, ‘What I feel is more valuable than what you feel.’ We can see now that such a claim is nonsense psychologically, because other people’s feeling cannot be accessed. But even if they could, would it be meaningful to assert that your experiences were more valuable than someone else’s? The champion of high art would have to mean not just that his

HIERARCHICAL KNOWLEDGE STRUCTURES

•

MOORE

115

experience were more valuable to him, for that would not prove the superiority of high art, only his preference for it. He would have to mean that the experiences he derived from high art were in some absolute and intrinsic sense more valuable than anything the other person could get from low art. How could such a claim make sense? What could ‘valuable’ mean in such a claim? It could have meaning only in a world of divinely decreed absolutes – a world in which God decides which kinds of feelings are valuable and which are not – and this, as I have said, is not the world in which I am conducting my argument. (Carey 2005: 25)

There are four symptomatic points to take from this that apply equally across the range of relativisms. •

The first is its subjectivism: claims are specialized to the experience or feelings of the knowing subject.

•

The second is that, as a consequence, knowledge relations are rewritten as relations between social groups – groups of ‘knowers’ or, indeed ‘feelers’.

•

The third is that statements of judgement are no more than statements of group tastes or preferences (Carey devotes three pages, 117 to 120, to Bourdieu’s classic study, Distinction – albeit with some reservations).

•

The fourth is that the only alternative to taste is divinely decreed absolutes.

In these respects, Carey’s position in aesthetics reproduces the basic propositions that underpin post-modern scepticism, feminist standpoint theory and a wide range of other forms of contemporary relativist perspectivism. Knowledge is no more than what different knowers know. This results in a kind of social embarrassment in which it is taken that to judge some things, values or practices to be better than others is to imply that some people are better than others. Because we cannot detach judgements about knowledge from judgements about knowers we cannot judge the one without judging the other (or ‘Other’ in the post-structuralist, semiotic version of this). This answers Hoggart’s question as to why the arguments are so angry – because people are taking them personally. What is to one an aesthetic judgement is to another a cultural snub. Carey’s view is that we are limited only to preferences or absolutes in the area of judgement, and because we cannot have absolutes we are left with nothing but preferences – assertions of taste alone. However, it is a commonsense observation that there is no major problem in distinguishing between the exercising of a preference and the production of a judgement. The claim by any individual that they prefer to read P.D. James’ murder mysteries, but judge the novels of Henry James to be of superior literary significance is perfectly intelligible, as would be the statement that ‘I prefer to listen to Chuck Berry, but judge Beethoven to be musically superior’. Preferences and judgements are different things that operate in different ways. In the main, preferences are acquired in an ad hoc and contingent kind of way whereas judgements are actively produced according to publicly shared rules of various kinds. For any particular individual,

116

LANGUAGE, KNOWLEDGE AND PEDAGOGY

it is simply a matter of life-history whether they prefer beer to wine, and, if beer, traditional English ales to Continental lagers, or Chinese cuisine to Indian or French to Italian. The world of preferences is the world of relativism, of taste, because in many areas of life there is no good reason why anyone should do any one thing rather than any other. But, as Carey makes very clear in his second argument in defence of literature, it is not the same in the world of judgements. Carey acknowledges that his attempt to demonstrate the ‘superiority’ of Literature as a form of art sits strangely with his argument that art is simply what anyone wants it to be. He presents his intention in this way: In the rest of this book I intend to make out the case for valuing literature, taking examples for the most part, but not exclusively, from English literature – a branch of knowledge which, in recent years, has been progressively devalued in schools and universities, and regarded as rather shamefully parochial and old-world compared to, say, media studies or cultural history. In opposition to this, I shall also try to show why literature is superior to the other arts, and can do things they cannot do. Just in case anyone should seize on these aims as inconsistent with the relativist cast of the first part of my book, let me emphasize that all the judgements made in this part, including the judgement of what ‘literature’ is, are inevitably subjective. (Carey 2005: 173)

The crucial point at the end of this is the reference to the ‘inevitably subjective’ nature of Carey’s judgements. He immediately goes on to say that his definition of literature is ‘writing that I want to remember – not for its content alone, as one might want to remember a computer manual, but for itself: those particular words in that particular manner’ (loc. cit.). But even in this, Carey cannot help but introduce a distinction between an ordering of words that is intrinsically of value and others (such as those in a computer manual) that are merely instrumental in their purpose, and whose end is extrinsic. At this early stage of the argument, Carey could retort that somewhere there could be someone for whom the computer manual does in fact produce ‘artlike’ experiences, and, hence, for them it is an artwork – they prefer reading computer manuals to reading Keats’ odes and, indeed, experience the same things as do poetically inclined ode readers when reading Keats (though, according to Carey, we could never actually know this). Though it would be consistent with his first argument, this is not the path that Carey follows in his second. Instead, he hopes that his argument will ‘persuade some or all of my readers’ (ibid., p. 174, my emphasis). The point is that Carey’s readers should not be persuaded simply by reading a list of Carey’s preferences because, as he himself notes, the simple fact that someone prefers one thing to another is in itself neither here nor there. A statement of preference is a statement about the person preferring (i.e. ‘I am the kind of person who likes this’), not about the thing preferred (i.e. ‘Given that this is an instance of a thing of this kind, it is better than that other instance of its kind, but not as good as this one of its kind’; in the way that hunters in a hunter-gather society might say, ‘This is a particular spear and, given how we understand spears to be the kind of thing they are, we judge that it is a better spear than this one, but not as good as that one’).

HIERARCHICAL KNOWLEDGE STRUCTURES

•

MOORE

117

Carey’s preferences in life might be relevant to that small circle of people who buy him birthday presents, but not to the majority of his readers who don’t actually know him – our interest is in his judgements. The ‘subjectivity’ in producing his argument in support of literature is not simply, ‘I just happen to prefer it’, but, ‘I judge it to support my argument’. If Carey’s wish is to ‘persuade’ his readers, then he does so by producing reasoned judgements, not merely by telling us about his preferences (i.e. telling us things about him, Carey, as opposed to providing us with good reasons why we might come to share his judgements). And this is precisely Carey’s own argument for the superiority of literature.

The mind outside itself – the possibility of verticality A central dimension of Carey’s argument throughout the book is that we cannot compare artworks because we cannot compare the experiences that people have of them because we cannot access the minds of others. Carey suggests, first, that to compare artworks is invidious because it is to compare people and, secondly, that it is impossible anyway because the minds of others are closed to us. The latter is a peculiar argument from a teacher, critic and public intellectual. At least a certain aspect of Carey’s consciousness is available to us – that which is intentionally expressed in his book in order to engage our consciousness. We must assume that Professor Carey, as a teacher at Oxford University, made aspects of his consciousness available in an intelligible and comprehensible way to his students. Perhaps we are seeing again a problem of dogmatic scepticism – of course we cannot access Carey’s consciousness absolutely; not even he, he admits, can do that. But we don’t have to. We only have to access each other’s consciousness well enough, from the most intimate to the most fleeting encounter, to maintain normal human social interaction. A problem with Carey’s extreme subjectivism is its neglect of the structured character of intersubjectivity and its variety. Carey quotes the metaphysical poet John Donne and goes on to say: To take a proper, diagnostic look at your own mind, Donne reasons, you would need to get outside your own mind – and that cannot be done. The instrument you must use to probe your mind is already bent, for it is your mind. (Carey 2005: 182, my emphasis)

This is the very heart of the problem. Emile Durkheim took the opposite view to that of Donne and Carey. He said when defining a ‘thing’: A thing is any object of knowledge which is not naturally controlled by the intellect, which cannot be adequately grasped by a simple process of mental activity. It can only be understood by the mind on condition that the mind goes outside itself by means of observation and experiments, which move progressively from the more external and immediately accessible characteristics to the less visible and more deep-lying. (Durkheim 1956: 58–9, my emphasis)

118

LANGUAGE, KNOWLEDGE AND PEDAGOGY

At one point, Carey quotes Durkheim approvingly: Philosophers have often speculated that, beyond the bounds of human understanding, there is a kind of universal and impersonal understanding in which individual minds seek to participate by mystical means; well, this kind of understanding exists, and it exists not in any transcendental world but in this world itself . . . (Durkheim, cited in Carey 2005: 252)

Carey appears unaware, when invoking Durkheim, that the ‘kind of universal and impersonal understanding in which individual minds seek to participate by mystical means’, that Durkheim wishes to relocate in ‘this world itself’, is precisely the Kantian ‘. . . mysterious realm of truth, which he [Kant] called the “supersensible substrate of nature”, where all such absolutes and universal resided’ (Carey 2005: 9). Carey’s reductive subjectivist relativism could not be further removed from Durkheim’s emergent historical materialism and the manner in which he translates Kant’s ‘mysterious realm of truth’ into ‘this world itself’. Durkheim did this by arguing for the sociality of knowledge and for the social origins of Kant’s transcendental categories (Moore 2004b: ch. 5). It is in this manner that for Durkheim the mind can go ‘outside itself’, and it is in Carey’s second argument in defence of literature that we can develop a sense of how this is so. In support of literature How, then, does Carey advance the cause of literature, and what are the broader implications of his second argument? The essence of his case is that: . . . literature gives you ideas to think with. It stocks your mind. It does not indoctrinate, because diversity, counter-argument, reappraisal and qualification are its essence. But it supplies the materials for thought. Also, because it is the only art capable of criticism, it encourages questioning and self-questioning. Its function as a mind-developing agency gives it especial relevance in our present culture. (Carey 2005: 208–9)

The language employed here is quite different from that of Carey’s first argument. In the extract preceding the above, he refers to literature as ‘a branch of knowledge’ and in the above this is explicated in terms of ‘ideas to think with’, ‘counter-argument, reappraisal and qualification’, ‘materials for thought’, ‘criticism . . . questioning and self-questioning’, ‘a mind-developing agency’. Furthermore: Literature is not just the only art that can criticize itself, it is the only art, I would argue, that can criticize anything, because it is the only art capable of reasoning. Of course, paintings can convey implicit criticism – Hogarth’s The Gate of Calais, say, or Ford Madox Brown’s Work. But they cannot make out a coherent critical case. They are locked in inarticulacy. Operas and films can criticize, but only because they steal words from literature, which allows them to enter the rational world. When literature criticizes other arts its target is often their irrationality. (Carey 2005: 177, my emphases)

HIERARCHICAL KNOWLEDGE STRUCTURES

•

MOORE

119

This passage contains the key terms: literature can criticize itself (it is reflexive), it is capable of reasoning, it can make out a coherent critical case, it is part of the rational world and can target the irrationality of other arts. Once again, Carey’s argument comes up against the problem of his prior assumption that truth claims can only ever be absolute, and he assumes that this is the case in the natural sciences and that it is in this respect that they differ radically from the humanities and social sciences: . . . once belief in a God is removed, moral questions, like aesthetic questions, become endlessly disputable. Indeed, moral questions could be defined as questions to which no answers are available. Consequently, agreement about them is not to be expected. In this they differ from scientific or mathematical questions. Disagreement is, in other words, a necessary condition for the existence of ethics as an area of discourse. (Carey 2005: 172)

But moral and aesthetic questions do not differ from scientific and mathematical questions in this way (though they do differ in other ways) – they also are ‘endlessly disputable’. It is the possibility of falsification and paradigm change that drives knowledge in these areas. Carey appears to be wedded to a strangely anachronistic, positivistic model of the natural sciences. He might well be correct in claiming that ‘moral questions could be defined as questions to which no answers are available’, but, given the absolutist way in which he defines an ‘answer’, this is not the same as saying that these are questions about which no judgements can be made. Carey is much closer to what is really significant in these areas when he says that . . . it is not being right or wrong that makes a scientist. It is respect for proof and freedom from prejudice. (Carey 2005: 184)

What he says, here, about the scientist is as true for the critic. Certainly, what counts as ‘proof’ will differ in that the critic cannot draw upon empirical evidence in quite the same way, but being able to do so does not provide the scientist with infallible truths that are absolute and certain and the critic must provide textual evidence to support his or her argument (as Carey, of course, does throughout his book and we take for granted that he does so in an honest and truthful and rigorous scholarly way rather than just inventing writers and quotations as it suits him). Although the mathematician, the scientist and the critic will employ different models of ‘proof’, they do not employ different models in the respecting of proof, and each must be equally free from prejudice (as far as that is ever humanly possible). In any college bar anywhere in the world there will be groups of mathematicians, scientists and critics sitting around tables arguing about problems in their respective fields. In key respects, they will be doing things differently, but in the most fundamental respect they will all be doing the same thing – producing, not necessarily agreements, but informed, rigorous judgements grounded in ‘respect for proof’. Crucially, ‘respect for truth’ is not simply a

120

LANGUAGE, KNOWLEDGE AND PEDAGOGY

slogan. It is an internalized value, acquired through a methodology of ‘education’ (the systematic formation of a habitus), embedded within and manifested through shared, collective procedures, principles and criteria. And in doing this they are doing something qualitatively different from the first thing they all did when entering the bar; namely, ordering their drinks – the exercise simply of their personal preferences. In the course of their discussions over their drinks, certain of these individuals might judge that they should ‘change their minds’ on particular issues, but they could not in the same way be reasoned into changing their preferences about what to drink. Preferences are simply personal, but judgements are intrinsically collective in character and presuppose collectivities of certain kinds. In support of knowledge Carey’s second argument can now be the starting point for the exploration of an alternative theory of knowledge that draws upon a different understanding of the sociality of knowledge that sustains the variety of relativisms of which Carey’s first argument is an instance. The key question is: how does the mind ‘go outside itself’? Consider Carey comparing aesthetics with ethical theory: In aesthetics, likewise, there are no absolutes, we have to choose. Even in choosing to have no interest in the arts at all is a choice. But though preferences between arts, and decisions about what a work of art is, are personal choices, that does not mean they are unimportant. On the contrary, like ethical choices, they shape our lives. Nor does it mean that they are unalterable. Just as we can be argued out of or into moral convictions (as, for example, in cases of religious conversion), so our aesthetic preferences may change. This may be sudden and dramatic . . . Or it may be the result of gradual discovery and persuasion – a process we generally call education. (Carey 2005: 172, my emphases)

We see the fatal muddle concerning preferences and judgements and the way in which that is associated with the preoccupation with absolutes. We have, on the one hand, ‘preferences that are personal choices’, but on the other it is acknowledged that we ‘can be argued out of or into moral convictions’, but there is no rigorous conceptual distinction between these two things in Carey’s book. The reason being, that in the absence of absolutes in the areas of ethics and aesthetics, then, for Carey, there is no difference. What is important for Carey is that in both ethics and the aesthetic, ‘we have to choose’. He is correct, but changing preferences is not the same thing as being ‘argued out of or into’ something. To change a preference is simply to decide to opt for one thing rather than another. Someone might decide that they are bored with Indian food and will start eating Chinese instead, but they need to tell their friends no more than ‘I just happen to feel like doing it’. If, by contrast, someone declares that they have been persuaded (‘argued into’) changing their judgement about the relative literary merits of P.D. James and Henry James, and have come to accept that P.D. is the better writer, they would be required to provide reasons – that is, to do all those things that Carey believes

HIERARCHICAL KNOWLEDGE STRUCTURES

•

MOORE

121

make literature ‘superior’ and which he himself, of course, as a critic, does in his book. Furthermore, there would be nothing in principle unintelligible in this person also declaring that in the course of being persuaded in this judgement, they have also found that they now prefer to read Henry rather than P.D. and opt to do so. Preferences and judgements and opting and choosing are different things gone about in different ways. Carey’s first argument collapses judgements into preferences and choosing into opting. His second argument does the opposite – judgements regulate and have priority over preferences and we are persuaded rather than merely opting. There could well be an argument to the effect that, in the liberal humanist tradition, the purpose of education is to articulate preferences with judgements (and this could be implicit in what Carey says above), but this is a big question for another time. To proceed, then, in the very different direction indicated by Carey’s second argument and his recognition of a process of ‘gradual discovery and persuasion – a process we generally call education’. Implicit in this can only be an acknowledgement of the fact that in certain areas our preferences can be refined by judgements (and in others operate independently), and that this process (‘education’) occurs when we interpolate our personal preferences into public arenas of collective judgement (though this does not necessarily imply that in all cases we have to – we may choose to be left to our own devices as to our choices of football team, footwear, sexual orientation or alcoholic beverages). We interpolate our preferences and opinions in order to actively engage with explanations and reasons in a field of ‘counter-argument, reappraisal and qualification’. Fields of judgement are not those of the purely personal subjectivity of preferences, of opting. They are fields of inter-subjectivity of a special kind. They are essentially interlocutory in character and their purpose is the formation of judgements according to publicly shared procedures and criteria. Coalitions of the mind Unlike the exercise of preference, the formation of judgement is a particular kind of inter-subjectivity (of sociality) structured within specialized communities of interlocutory interpolation. Interlocutory in that they are grounded in dialogue, and interpolationary in that these are dialogues concerning reasons and explanations – ‘why?’ discourses: precisely those procedures that Carey identifies as constituting the ‘superiority’ of literature among the arts. The problem that this presents for relativist thinkers such as Carey (in aesthetics) and Usher and Edwards (in epistemology) has been well summarized by Randall Collins: That ideas are not rooted in individuals is hard to accept because it seems to offend against a key epistemological point . . . It is assumed that objective truth itself depends on the existence of a pure observer or thinker, untrammelled by anything but insight into truth. The notion is that the social is necessarily a distortion, an alien intrusion in epistemology; if ideas are determined by social interaction, then they cannot be determined by truth. This objection comes so naturally that it is hard to think except within this dichotomy: either there is truth that is independent of society, or truth is social and not objectively true. There are

122

LANGUAGE, KNOWLEDGE AND PEDAGOGY

two prejudices here. One is the assumption that constructing an idealised individual, outside of the social, provides a vantage point that social networks cannot provide just as well. On the contrary: there is even more difficulty in connecting such a disembodied individual to the world than there is in connecting a social group to the world, since a group is already to some degree extended in the world of time and space. The second prejudice or tacit assumption is that the criterion of truth exists in free-floating reality, along with the free-floating thinker-observer. But the very concept of truth has developed within social networks, and has changed with the history of intellectual communities. (Collins, 2000: 7–8)

The ‘prejudices’ described by Collins are those exhibited by Carey and Usher and Edwards, and are symptomatic of relativism in general – they are the prejudices of ‘dogmatic scepticism’, of the ‘All-or-Nothing Fallacy’. The core problem is the assumption that ‘if ideas are determined by social interaction, then they cannot be determined by truth’. The irony is that this proposition, assumed dogmatically by relativists, is precisely the foundationalism that underpins the positivism that they are out to refute, but which they themselves presuppose in their sceptical conclusions (Moore 2006). In the Durkheimian tradition, Collins’ argument is that truth is social in that it is the product of a distinctive form of structured and enduring sociality: ‘ . . . thinking consists in making “coalitions of the mind”, internalised from social networks, motivated by the energies of social interactions’ (Collins 2000: 7). These ‘coalitions of the mind’, as Collins’ work demonstrates, as canons of various kinds, are extensive in time and space (Moore and Maton 2002). The philosophical canons of Europe, Islam, India and China, for instance, have endured, and interacted, over many centuries, and intellectual networks today are global and transcultural. Collins opens his argument in a manner that indicates how things that for Carey and others like him are fatally problematical, can be approached in a different way: Intellectuals are people who produce decontextualised ideas. These ideas are meant to be true or significant apart from any locality, and apart from anyone concretely putting them into practice. A mathematical formula claims to be true in and of itself, whether or not it is useful, and apart from whoever believes [it]. A work of literature, or of history, claims the same sort of status, insofar as it is conceived as art or scholarship: part of a realm that is higher, more valid, less constrained by particular circumstances of human actions than ordinary kinds of thoughts and things. Philosophy has the peculiarity of periodically shifting its own grounds, but always in the direction of claiming or at least seeking the standpoint of greatest generality and importance. This continues to be the case when the content of philosophy is to assert that everything is transient, historically situated, of local value only; for the relativist statement itself is asserted as if it were valid. This is an old conundrum of the skeptical tradition, discussed at great length in Hellenistic philosophy. Skeptics in attempting to avoid making assertions implicitly stand on a meta-distinction among levels of assertion of varying force. This illustrates the sociological point admirably, for only the intellectual community has the kind of detachment from ordinary concerns in which statements of this sort are meaningful. (Collins 2000: 19)

HIERARCHICAL KNOWLEDGE STRUCTURES

•

MOORE

123

This statement contains much of the kind of thing that Carey dislikes: ‘A work of literature, or of history, claims the same sort of status, insofar as it is conceived as art or scholarship: part of a realm that is higher, more valid, less constrained by particular circumstances of human actions than ordinary kinds of thoughts and things.’ Collins also points to the unavoidably self-refuting logic of the relativist position. But what is significant, here, is his view that ‘Intellectuals are people who produce decontextualized ideas’. And what is important is that it is the ideas, not the intellectuals that are ‘decontextualised’. The key relationship is that between ‘particular circumstances’ and the ‘higher realm’. It is the idea of a ‘higher realm’ that Carey’s first argument rejects (but which is smuggled back within the second argument). How does Collins underpin this ‘higher realm’, and how might the way in which he does so connect with Carey’s second argument? Collins notes that ‘. . . if one refuses to admit anything beyond the local, one arrives at some version of scepticism or relativism; if one idealizes what happens in situations as the following of rules and uses these inferred rules as a tool for constructing the rest of the world, one arrives at a type of idealism’ (20–1). Carey’s world of preferences as the only alternative to absolutes is the kind of relativistic localism described and entails the parochialism that his second argument rejects and which, he argues, literature can take us beyond. Collins’ purpose is to understand the sociality of knowledge in a way that is neither relativistic nor idealist. Let us begin at the site of all action: the local situation. All events take place in a here-and-now as concrete and particular. The perspective of micro-sociology, which analyzes the structures and dynamics of situations, is all too easily interpreted as a focus on the individual actor or agent. But a situation is just the interaction of conscious human bodies, for a few hours, minutes, or even micro-seconds; the actor is both less than the whole situation and larger, as a unit in time which stretches across situations. The detached agent who makes events happen is as artificial a construction as the non-social observer, who represents the idealized vantage point of classical epistemology. The self, the person, is more macro than the situation (strictly speaking, the person is meso); and it is analytically derivative because the self or agent is constructed by the dynamics of the situation. (Collins 2000: 20)

The fact that all human embeddedness, consciousness and action is, in the first instance, local does not mean that it is nothing but local: ‘The local situation is the starting point of the analysis, not the ending point. The micro-situation is not the individual, but it penetrates the individual, and its consequences extend outward through social networks to as macro a scale as one might wish’ (Collins, loc. cit.). It is through ‘extension’ (as with Bernstein’s ‘verticality’) that Collins moves from the particular, the local, to the ‘higher’ realm. However, he makes a crucial conceptual adjustment to the spatial metaphor: To deny that anything exists other than the local is true in one sense, misleading in another. It is true that nothing exists that is not thoroughly local; if it did not

124

LANGUAGE, KNOWLEDGE AND PEDAGOGY

exist locally, where possibly could it be found? But no local situation stands alone; situations surround one another in time and space. The macro-level of society should be conceived not as a vertical layer above the micro, as if it were in a different place, but as the unfurling of the scroll of micro-situations. Micro-situations are embedded in macro-patterns, which are just the ways that situations are linked to one another; causality – agency, if you like – flows inward as well as outward. What happens here and now depends on what happened there and then. We can understand macro-patterns, without reifying them as if they were self-subsisting objects, by seeing the macro as the dynamics of networks, the meshing chains of local encounters that I call interaction ritual chains. (Collins 2000: 21)

The sociality of judgement Carey’s second argument is not, actually, that far removed from that of Collins (in fact, for him to intelligibly do what he does in his book presupposes a Collinstype situation). The crucial difference lies in Carey’s preoccupation with individual subjectivism as opposed to Collins’ sociological concern with the sociality of inter-subjectivism. This difference can be illustrated by contrasting Collins above with Carey below: Since every reader’s record of reading is different, this means that every reader brings a new imagination to each book or poem. It also means that every reader makes new connections between texts, and puts together, in the course of time, personal networks of association. This is another way in which what we read seems to be our creation. It seems to belong to us because we assemble our own literary canon, held together by our preferences. The networks of association we build up will not depend on spotting allusions or echoes, though sometimes we may notice these, but on imaginative connections that may exist only for us. (Carey 2005: 242, my emphases)

As the added emphases illustrate, Carey’s view is unremittingly egocentric. He is in a certain sense quite correct in what he says, but what he says is not all that can be said. Carey’s description would apply only to a completely noncommunicative reader who keeps his or her views entirely to themselves and assiduously avoids encountering or engaging with the views of others. The canon of ‘personal preference’ that exists ‘only for us’ applies only to that detached, asocial, contemplative self that Carey and the epistemological relativists reject. In principle there is no reason why any particular individual might not opt to adopt this position (it is not uncommon to encounter committed individual readers who resolutely proclaim that they do not want to ‘study’ literature because it would destroy the pleasure they derive from it). It can also be the case, however, that other readers enjoy talking to each other about what they have read (as the current enthusiasm for readers’ groups demonstrates). There is a distinctive form of sociality in which individuals come together precisely in order to trade and exchange those ‘personal networks of association’ – in just the way in which Carey in his book does with those who read his book. The

HIERARCHICAL KNOWLEDGE STRUCTURES

•

MOORE

125

fundamental point in this is that these ‘networks of association’ are intersubjective, not solipsistically or paradigmatically sealed and incommensurable. The imaginative nexus There are two important points here. The first is one that Carey himself makes and extensively illustrates in his book: writers operate intertextually within a tradition, a canon. They write against the background of what Harold Bloom (1995) calls the ‘anxiety of influence’ – they want to be judged original, as speaking in their own voice, and not merely repeating that which others have said before. This is the force that drives all modern fields of intellectual production: the quest for novelty and originality (it is, historically, their radically distinctive feature (Whitley 2000)). The second is that readers’ preferences flow into each other. Carey’s ‘personal networks of association’ translate, communicatively, into Collins’ ‘coalitions of minds’, into participation in enduring, structured forms of sociality: ‘The focus is on a particular kind of speech act: the carrying out of a situation-transcending dialogue, linking past and future texts’ (Collins 2000: 28, my emphasis). The novelist Ian McEwan effectively conveys the sense of this ‘situationtranscending dialogue’ in the following: Those who love literature rather take for granted the idea of a literary tradition. In part, it is a temporal map, a means of negotiating the centuries and the connections between writers. It helps to know that Shakespeare preceded Keats who preceded Wilfred Owen because lines of influence might be traced. And, in part, a tradition implies a hierarchy, a canon; most conventionally it has Shakespeare dominant, like a lonely figurine on top of a wedding cake, and all the other writers arranged on descending tiers. In recent years, the canon has been attacked for being too male, too middle class, too Eurocentric; what remains untouched is the value of a canon itself: clearly, if it did not exist, it could not be challenged. But above all, a literary tradition implies an active historical sense of the past, living in and shaping the present. And reciprocally, a work of literature produced now infinitesimally shifts our understanding of what has gone before . . . Ideally, having read our contemporaries, we return to re-read the dead poets with fresh understanding. In a living artistic tradition, the dead never quite lie down. (McEwan 2006)

As this suggests, what Carey calls ‘the imaginative nexus’ (Carey 2005: 243) is not simply an aspect of individual consciousness, but in the Durkheimian sense, of a collective consciousness extended in time and space, and that embraces both reader and writer – ‘the unfurling of the scroll of micro-situations’. According to Collins, although all intellectual exchanges are grounded in the local: Texts do not merely transcend the immediate particulars of the here-and-now and push toward abstraction and generality. To be oriented towards the writings of intellectuals is to be conscious of the community itself, stretching both backwards and forwards in time. Intellectual events in the present – lectures, debates,

126

LANGUAGE, KNOWLEDGE AND PEDAGOGY

discussions – take place against an explicit backdrop of past texts, whether building upon them or critiquing them. Intellectuals are peculiarly conscious of their predecessors. And their own productions are directed toward unseen audiences. Even when they lecture to an immediate group, perhaps of personal students, disciples or colleagues, the message is implicitly part of an ongoing chain, which will be further repeated, discussed, or augmented in the future. (Collins 2000: 27)

Carey tells us that the ‘essence’ of literature is: ‘diversity, counter-argument, reappraisal and qualification . . . questioning and self-questioning’. But these things entail community, they entail, as Collins says, ‘a particular kind of speech act’ – the interpolation of ‘why?’ within a collective arena. The problem in Carey’s book is not with where he ends up, but with where he begins. The problem being that he cannot actually get to the one via the other – he needs to observe the ancient advice to the traveller asking directions in a strange land: ‘Don’t start from here’. Carey, in his first argument, wishes to avoid snubbing people by making cultural judgements and retreats into a critical diffidence that says: in matters of literature, your feelings are just as good as mine (the professor of English Literature at Oxford University). But it is not Carey’s feelings that interest us, it is his judgements. However, in his second argument, Carey advances the view that ‘Like drugs, drink and antidepressants, literature is a mind-changer and an escape, but unlike them it develops and enlarges the mind as well as changing it’ (Carey 2005: 210, my emphasis). In order to sustain this view, he needs to start from somewhere else. Carey says of literature: ‘Once its words are lodged in your mind they are indistinguishable from the way you think’ (Carey 2005: 245) – they contribute to the habitus that links the individual to society and the present to the living past. As Collins stresses, it is the activity that is important and, on that basis, the lesson of Carey’s book, perhaps, is to do as Carey does, not as he says.

Conclusion It was stated in the Introduction that this essay is a work of sociology, not of criticism or aesthetic theory. It is sociological in that it takes as its focus what I have termed the ‘sociality of judgement’ – that is, a particular form of structured inter-subjectivity associated with a distinctive form of activity that occurs within a distinctive type of supra-individual social arena that is extended in time and space: a canon. Bernstein’s theory of knowledge structures allows us to begin the modelling of the forms, principles and possibilities of such fields of symbolic production. The canon enables a particular type of activity and its principles are generative – it is the arena of endless dispute. Understood in this way, Bernstein’s approach avoids the false extremes of absolutism and relativism by enabling a new focus on the sociality of judgement. Judgements are less than absolutes in that they acknowledge their fallibility. They are more than preferences in that they submit themselves to historically evolved rules of collective evaluation. It is the

HIERARCHICAL KNOWLEDGE STRUCTURES

•

MOORE

127

knowledge not the knower that counts. It would be wrong, however, to restrict this distinctive mode of sociality to the realms of academic scholarship, criticism and science where it assumes its most rigorous forms in what Bernstein calls ‘hierarchical knowledge structures with strong grammars’. Its principles extend further into the public sphere and constitute the discursive rules of civil society in modern, secular liberal democracy (Ahier et al., 2003). More formally, a canon is the type of thing that might be called a ‘Durkheimian emergent social kind’. Durkheim describes such ‘things’ in the following way: Collective representations are the product of an immense cooperation that extends not only in space but also through time; to make them, a multitude of different minds have associated, intermixed, and combined their ideas and feelings; long generations have accumulated their experience and knowledge. A very special intellectuality that is infinitely richer and more complex than that of the individual is distilled in them. (Durkheim 1995: 15)

It is this ‘very special intellectuality’ (an enduring structured form of sociality with generative powers, habitus) that has been the focus of this chapter. The major work of thinkers such as Basil Bernstein and Randall Collins (among others) has opened up new avenues for the sociology of knowledge and the promise of breaking the deadlock that has vitiated theoretical advance in that area for so long.

References Ahier, J. Beck, J. and Moore, R. (2003), Graduate Citizens? Issues of citizenship and higher education. London: RoutledgeFalmer. Alexander, J.C. (1995), Fin de Siècle Social Theory: Relativism, Reduction and the Problem of Reason. London: Verso. Beck, J. (1999), ‘Makeover or Takeover? The strange death of educational autonomy in neo-liberal England’, in British Journal of Sociology of Education 20 (2), pp. 223–38. Bernstein, B. (1977), Towards a Theory of Educational Transmission: Class, Codes and Control, Vol. 3 (second edn). London: Routledge & Kegan Paul. Bernstein, B. (2000), Pedagogy, Symbolic Control and Identity: Theory, Research, Critique, (revised edition). Oxford: Rowman & Littlefield. Bloom, H. (1995), The Western Canon: the books and schools of the ages. London: Papermac. Bourdieu, P. (1993), The Field of Cultural Production: essays on art and literature, R. Johnson (ed.). Cambridge: Polity. Brown, J. (2001), Who Rules in Science? An opinionated guide to the wars. Cambridge, MA: Harvard University Press. Carey, J. (2005), What Good are the Arts? London: Faber & Faber. Collins, R. (2000), The Sociology of Philosophies: A Global Theory of Intellectual Change. Cambridge, MA: Harvard University Press. Durkheim, E. (1956), Education and Sociology. Toronto: The Free Press. Durkheim, E. (1995), The Elementary Forms of Religious Life. New York: The Free Press. Graff, G. (1992), Beyond the Culture Wars: how teaching the conflicts can revitalize American education. New York: W.W. Norton. Hoggart, R. (1995), The Way We Live Now. London: Chatto & Windus.

128

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Maton, K. (2000), ‘Languages of Legitimation: the structuring significance for intellectual fields of strategic knowledge claims’, in British Journal of Sociology of Education 21 (2), pp. 147–67. Maton: this volume. McEwan, I. (2006), ‘A Parallel Tradition’, in The Guardian Saturday Review, 1 April 2006, p. 4. Moore (2004a), ‘Cultural Capital: objective probability and the cultural arbitrary’, in British Journal of Sociology of Education 25 (4), pp. 445–56. Moore, R. (2004b), Education and Society: issues and explanations in the sociology of education. Cambridge: Polity. Moore, R. and Muller, J. (1999), ‘The Discourse of “Voice” and the Problem of Knowledge and Identity in the Sociology of Education’, in British Journal of Sociology of Education 20 (2), pp. 189–206. Moore, R. and Maton, K. (2002), ‘Founding the sociology of knowledge: Basil Bernstein, intellectual fields and the epistemic device’, in Morais, A., Neves, I., Davies, B. and Daniels, H. (eds), Towards a Sociology of Pedagogy: the contribution of Basil Bernstein to research. New York, Peter Lang. Muller. J (2000), Reclaiming Knowledge. London: RoutledgeFalmer. Niiniluoto, I. ( 2002), Critical Scientific Realism. Oxford: Oxford University Press. Usher, R. and Edwards, R. (1994), Postmodernism and Education. London, RoutledgeFalmer. Whitley, R. (2000), The Intellectual and Social Organisation of the Sciences. Oxford: Oxford University Press. Young, M.F.D. (2000), ‘Rescuing the sociology of knowledge from the extremes of voice discourse: towards a new theoretical basis for the sociology of the curriculum’, in British Journal of Sociology of Education 21 (4), pp. 523–36.

Section 2

Fields of Discourse – Disciplines of Discourse

This page intentionally left blank

7

Language for learning in early childhood Clare Painter

Basil Bernstein (1999) has described as ‘horizontal discourse’ the forms and uses of language that realize the particular form of knowledge he characterizes as ‘common-sense’ in character. Such knowledge stands in contrast to specialized and institutional forms of knowledge accessible only via specialist or ‘vertical’ forms of discourse. It is common (sense) in being accessible to all, in applying to all and in ‘arising out of common problems of living and dying’ (Bernstein 1999: 159). Following from this, the most obvious case of the construction of common-sense knowledge must be in an individual’s very early life, where the child learns how to manage the routines of everyday life, and builds up the skills and information that is common to all members of that community. It is therefore in the informal pedagogic contexts and processes of the home, family and peer group that the nature of horizontal discourse may perhaps best be exemplified. In this chapter, I shall use data from informal conversations in the home between parent and child to provide such exemplification with a threefold aim. One goal will be to clarify the pedagogical and linguistic nature of horizontal discourse, a second will be to show how the discourse can change over time in the process of construing the world, while a third will be to demonstrate its essential limitations for the developing child straining to gain information beyond everyday experience. The data on which I shall base my discussion come from longitudinal studies of the language development of my own sons, Hal and Stephen. Data from the elder child, Hal, was collected up to the age of thirty months and from his brother (41/2 years younger) from thirty months to five years. In each case, a naturalistic methodology was adopted whereby spontaneously occurring conversations of up to 45 minutes’ duration were audio-taped at least once a week, supplemented on a more frequent basis by the pen and paper recording of novel utterances, brief conversational exchanges and contextual information. At no point were the children tested; nor were situations deliberately contrived in the process of data collection, and the children remained unaware throughout that their language was under any scrutiny. The aim was to collect material that reflected as accurately as possible spontaneous and unselfconscious conversations between family members. (See Painter 1984, 1999, for further details and discussion of data collection methods.) It should be noted that the children’s family is one in which an ‘elaborated’ coding orientation predominates (Bernstein 1987) so that the forms of horizontal discourse described are those most likely to harmonize with later forms of vertical discourse. At the same time, the data described can support the very general principle that for any child, as language develops and new forms of knowledge are confronted, both

132

LANGUAGE, KNOWLEDGE AND PEDAGOGY

pedagogy and language use will necessarily be subject to change. Moreover it is possible that an elaborated coding orientation will put horizontal discourse under earlier pressure in the child’s life, and thus amplify its nature and limitations. To pursue these points, the remainder of this chapter will be organized into three sections. The first will very briefly outline the systemic-functional orientation to language and learning that underpins the case-study research. The second will describe and exemplify Bernstein’s notion of horizontal discourse using data from the first case study. The third section will use data from the second case study to focus on problems encountered by the child in the continuing construal of everyday knowledge when using the linguistic resources and forms of discourse typical of the first two or three years of life. It will be suggested that in meeting these challenges, the child’s language comes under pressure to develop in ways that provide greater potential for accessing less common-sense forms of knowledge. In the process, there are indications of the limitations of horizontal discourse for building these new forms of knowledge.

Language and learning: the centrality of dialogue The orientation towards language and learning taken here is best summarized by M.A.K. Halliday’s claim that there are three aspects to learning one’s mother tongue: learning language, learning through language, and learning about language. In a classic paper, Halliday (1980/2004) emphasizes that these are in fact three different facets of the same process: as a child learns the language, so he or she construes the social and informational knowledge that is realized in language, including knowledge of language itself. To study the individual’s development of language, then, is necessarily to study the development of knowledge (construction). The notion that learning language is at the same time a matter of learning (other things) through language clearly resonates with Bernstein’s statement that ‘horizontal discourse, in its acquisition, is the major cultural relay’(1999, p. 160). However, Halliday prefers to speak of development or learning rather than acquisition, since to speak of language or its usage as something that is ‘acquired’ can suggest that it is something finite, monolithic and unchanging rather than being – as Halliday would view it – an infinite variable, dynamic resource, constructed and maintained interactively. The interactive construction of language is supported by psychological research into the very beginnings of life and semiotic experience. According to neonatal researchers such as Trevarthen (1992), Bateson (1975), Meltzoff and Moore (1998), and others, it is emotional empathy between mother and baby that provides the basis for all later developments. This is evident not only in infants’ displayed innate preference for attending to people over other stimuli (Field and Fox 1985; Messer 1994), but in their pre-linguistic communicative exchanges. We now know that well before language – indeed as early as the second month of life – infants exchange attention with care-givers, using gaze, facial, vocal and bodily movements in emotionally charged encounters with other persons. In these encounters, ‘complex, well-organized repertoires of

LANGUAGE FOR LEARNING IN EARLY CHILDHHOOD

•

PAINTER

133

action and facial expression are displayed that appear responsive to the form and timing of adult communication’ (Murray 1998: 127). In other words there is a ‘precise interplay of address and reply in time’ (Trevarthen 1992: 108)) being achieved between infant and care-giver (see further Bateson 1975; Trevarthen 1977, 1998). It appears then that dialogic exchange is at the heart of human experience (and human learning). It is not only the human world that attracts the child’s attention of course, and after a few months the infant directs reaching and grasping movements towards inanimate objects in the environment. Halliday (2004) has suggested that the attention to objects and other external phenomena creates a kind of puzzle for the child in that what is ‘out there’ in the world beyond impinges (when seen and touched and reacted to) on what is ‘in here’ – the world of the infant’s consciousness. But the infant’s consciousness is a world that has already begun to be set up as a social domain of ‘you and me’. Certainly there is evidence from developmental psychology to suggest that there may for a time be a competition between the infant’s fascination with the world of objects and events (to which reaching movements are directed) and his or her engagement with other persons (where attention is coordinated) (Trevarthen 1987: 191). This conflict between the claims of the material and social worlds is resolved when the child is able to extend the dialogic exchange to include reference to the world beyond the interpersonal dyad. This occurs from around nine months of age when a child can achieve joint attention to objects or happenings in the environment by creating an idiosyncratic vocal or gestural sign with which to address the other person. This is the ‘proto-language’ phase of language development documented in Halliday (1975/2004), Painter (1984, 2005) and Torr (1997). It can be exemplified by the following two examples from Hal (H) as a baby in interaction with his mother (M) and father (F). Here and throughout this chapter the age of the child is given in years, months and days following the text example. Example 1 H: (reaching to biscuit tin) amamama’ma [gloss ‘I want that, give it to me’] M: Okay you can have one. (opening lid) H: amamama’ma (M hands over biscuit) (0;9;12) Example 2 H: (playing with hairbrush, touches bristles to hair) a’da H: (Looks up at F and brandishes brush) a’da! F: (smiles at H) H: (looks across at M and holds brush aloft) a’da! M: (smiles) Yes you’ve got a brush for your hair; is that fun?

(1;1;1)

We can see here that through the use of a sign, some phenomenon external to the dialogue – some phenomenon from the ‘objective’ domain of experience – is brought into the intersubjective context of mutual address. In this way engagement with the world beyond the speaking interactants begins to be

134

LANGUAGE, KNOWLEDGE AND PEDAGOGY

mediated through other people. Thus even at this proto-language stage, objective reality comes to be acted on, but also reflected on, or ‘known’ through symbolic interaction, not through private cognitive apprehension. This means that the orthodox psycholinguistic position that language is ‘a system of expression’ designed ‘for taking the internal, personal, private mental meanings of individuals and making them external and public’ (Bloom 1993: 19) is simply not tenable. Instead we must acknowledge that symbols arise in the context of dialogue – not as an externalizing of the private inner, as Bloom suggests, but as a way of making sure that the inner is developed socially and that the world outside the child comes to be known through the mediation of symbols. This is an orientation entirely in harmony with Bernstein’s emphasis on the centrality of dialogue in the construction of identity and consciousness. With an orientation of this kind, it is evident that the child develops protolanguage and then language in the course of using it to make meanings with others. And the fact that at the same time as the system of language is built up – and the child is learning how to form questions and answers, how to indicate past and present time, how to make causal or contrastive connections, how to refer to self, addressee and third parties through the pronoun system, and so on – at the same time as all the various features of the grammar are being constructed, the child is experiencing these features, not as abstract paradigms, but as ways to realize meaning on particular occasions of use with particular other persons. This can be readily demonstrated if we skip ahead in the developmental story to the point where the child is first learning words, and the knowledge they instantiate, as seen in Example 3 below: Example 3 (H is naming different animals in a picture book about a train ride) H: Bukkersefy; nother bukkersefy. M: Butterfly; butterfly. H: Butterfy. It’s got horns. M: No, they’re not horns, they- theyH: (turns page and points at horse) It’s got horns. M: It’s got ears. H: (points at rabbit) It’s got horns. M: No, it’s a rabbit! It’s got ears; big ears. H: (points at squirrel’s tail curved behind its head) It’s got horns. M: It hasn’t; it’s got a tail; it’s a squirrel. This one’s [=cow] got horns. H: It’s got horns. M: Yeah. (turns page) H: Doggy. M: (points at train and makes choo choo train noises) H: (points at lion) Lion. (makes growly noises) M: Lion. And where’s the lion? H: (sees cat and climbs off lap) Katy, Katy. M: No, leave her; you’re reading a book with Mummy. H: Hal stroke Katy; Hal stroke Katy; stroke Katy. M: Now you stroke her gently. (1;11;18)

LANGUAGE FOR LEARNING IN EARLY CHILDHHOOD

•

PAINTER

135

It is clear that Hal is here learning language. But it is equally clear that in the process he is doing more. He is also construing knowledge about animals and their body parts, he is enacting (and thus learning to engage in) a particular social relationship with the mother and also gaining experience in the social process (or ‘genre’) of bedtime book reading. He is learning all these things through linguistic interaction at the same time as he is learning the words and structures of the language itself. A Hallidayan perspective on language development, then, is one that sees learning through language as an inextricable part of the process of learning the language itself.

The characteristics of horizontal discourse in the first two and a half years Bernstein describes the kind of knowledge realized in horizontal discourse as being typically ‘oral, local, context dependent and specific, tacit, multi-layered, contradictory across but not within contexts’ and, most crucially, ‘segmentally organised’ (1999: 159). In this section a further small selection of texts from Hal’s early childhood will be used to illustrate some of these points and to discuss the kind of pedagogy and the forms of language involved. Each text cited is representative of dozens of comparable ones that share the same features. In presenting examples I will focus initially on the variable nature of pedagogy in horizontal discourse of this kind, which is said by Bernstein to vary from ‘segment’ to ‘segment.’ For example, there are occasions when the pedagogy is quite explicit: that is to say both parties know what is to be learned, how it is to happen, who is teacher and who is taught, and what counts as success. This is most obvious when the child is being instructed in a specific skill, as exemplified in Examples 4a, b and c. Example 4a (M instructs Hal how to pull elastic for cat to chase) H: Katy, Katy. (pushes a bundle of elastic tape at her) M: No, she likes it when it moves; it’s got to move, you see. H: There you are Katy. M: No, she won’t take it like that. Come here and I’ll show you. You hold this here; hold it there; and then you dangle this bit like that; like that; and then you say ‘Here you are Katy’. H: Here Katy! Here Katy! M: Just throw this and wheeeee! (tosses elastic) Chase it Katy! (H runs up to cat) M: No, she’s a bit frightened of you. (2;3;25) Example 4b (M instructing H to buckle up sandal) M: Pull up; pull it up; that’s a good boy. No, got to pull it up some more. That’s it. Pull it hard as you can. That’s the way. Now put this [=metal prong on buckle] in. Good boy! Want to put this [=end of strap] in? Let

136

LANGUAGE, KNOWLEDGE AND PEDAGOGY

me start it and then you can push it in. That’s the way! Oh, you’re a clever little boy aren’t you? (2;3;25) Example 4c (H treads heavily on M’s toe, while Grandmother (G) observes) M: Ow!!! (crossly) Oh Hal, be careful! G: It was an accident. H: Sorry Mummy. M: Darling. (kisses) (M and G talk) H: I tread on Penny too. M: You trod on Penny. Did you say sorry to Penny? H: No I didn’t. M: Did Penny shout at you? H: No; she say ‘Ow!’ M: Well, next time say ‘Sorry Penny’. (2;6;16) While the pedgagogy is explicit in relation to the physical or social skills being taught in each case, it should also be noted that there is tacit learning taking place here. Neither party is conscious for example that such conversations are also teaching the child that the appropriate way to learn new skills is through talk (rather than say simply by observation), which is something that we know varies from one social group to another and has consequence for the transition into school (see Heath, 1983). What is also tacitly conveyed of course is that a pet cat is a possible playmate whose feelings have to be taken into account; that managing your own dressing is a praiseworthy goal and that apologies defuse anger and deflect criticism. As well as the tacit knowledge that is gained incidentally during instances of explicit pedagogy, there are many occasions where the instruction is itself implicit: that is to say, the learner is unaware of the goal of the lesson or the route to its achievement. Here we can consider text 5, which is typical of those that took place before age 2 when Hal’s father returned home from work: Example 5 M: Tell Daddy where we went today. H: (looks at M) M: Where did we go, Hal? H: Park. M: What did we do? H: Play football in park. M: Mm, we played football; it was fun, wasn’t it?

(1;8;1)

Here, through modelling, there is implicit teaching of aspects of grammar in the mother’s elaborated final response; for example of the past tense of play and the use of the subject pronoun we. In addition, through the use of questioning and elaborated responses Hal’s mother is also implicitly teaching him how to structure a recount of experience. Even more importantly, such an exchange

LANGUAGE FOR LEARNING IN EARLY CHILDHHOOD

•

PAINTER

137

demonstrates to the child that language can be used to convey information to someone who didn’t share the experience, a realization that may take some time for a young child to appreciate. (At first, as here, the child may only understand that it is possible to re-share in language an experience that was shared in the first place, rather than reporting to a third party. See Halliday 1975; Painter 1984, for further discussion.) Such conversations, even without the adult being conscious of the fact, alert the child to the potential of language for true information-giving and also to the kind of information (and its attitudinal appraisal) that it is appropriate to share when a family member returns to the home. On still other occasions of horizontal discourse, the pedagogy may be further submerged. That is to say, neither party may even be conscious of the discourse as having any informative or instructive character. Text 6 exemplifies: Example 6 (M and F have entered kitchen with H mid-morning to make coffee) M: (to F) Kettle’s boiling. H: I want my lamb chop Mummy. M: Um it’s a little bit early for lamb chops darling; would you like to have half a banana? H: No no no no! A big banana! M: All right; you get one from up there. (indicating fruit bowl on counter) H: Can’t. M: Yes you can. H: (moves chair to reach) There! (2;6;0) Even though the context here appears to the interactants to be one of action rather than teaching and learning, the contributions by the mother incidentally convey cultural information about when it is appropriate to eat which kind of food, as well as the fact that acting independently is valued in their social group. Another and very different context for tacit learning is provided by playful interactions. Text 7 below is one of very many examples in the corpus where parent and child collaborate in a ‘misnaming’ game, initially introduced to the child by the adult. It too looks at first to be a poor example of a pedagogic text, but is again significant for the tacit understandings gained or rehearsed: Example 7 (Looking at picture book which displays a picture of a tiger and of an elephant) M: Which one d’you like best? The tiger or the elephant? H: (points at tiger) That’s tiger. M: (confirming) That’s the tiger. H: (points at elephant) Tiger! M: (laughs) H: No. M: That’s not a tiger; that’s a gorilla. H: No! M: It’s a giraffe.

138

LANGUAGE, KNOWLEDGE AND PEDAGOGY

H: No! M: That’s a deer. H: No! M: That’s a zebra. H: No. M: What is it? H: Some more zebra? [i.e. request to continue misnaming] M: (laughing) That’s . . . what is it? H: Pussy cat! M: Pussy cat, is it? miaow! H: No. M: Is it a donkey? H: No. M: Is it a giraffe? . . . [further exchanges omitted as they cycle through six further wrong names] M: Is it an elephant? H: Yes! M: Yes! (2;1;27) Since the child is well aware of the words tiger and elephant and to which animals they refer, this is not an explicitly pedagogic text in the sense of being instructional or informative in the same way as Example 3. And as far as the interactants are concerned, the object is simply playful enjoyment. But tacitly, the child is learning here about the possibility of unhinging meaning and wording. This, it can be argued, constitutes the first step towards a more conscious understanding that a name is a social symbol and that knowledge itself is a representation and therefore can be misrepresented. Summary: horizontal discourse up to age two and a half From these very few examples cited from the much larger corpus we can get a better sense of what Bernstein intends by the term horizontal discourse. In particular, the following characteristics are prominent. First of all, we can see that each fragment of conversation is, in a certain sense, self-contained. That is, the skills or knowledge gained in one episode aren’t required before the learner can tackle something new. It is not necessary for example to have mastered shoe buckling before learning to apologize; it is not necessary to have mastered the difference between horns and antennae before being able to play a misnaming game in relation to elephants and tigers. This quality of the learning encounters is a key feature of what Bernstein refers to as the ‘segmental organization’ of horizontal discourse. Two further features associated with this segmental organization are the context-specific nature of the knowledge being learned, and the fact that learning is a matter of repeating or accumulating successive instances. Learning is context-specific due to the isolation of segments from one another and the lack of explicit provision of more generalized information. For example, Hal is instructed in what to say when treading on someone’s toe rather than being provided with an account of intentional and accidental behaviour as a basis for

LANGUAGE FOR LEARNING IN EARLY CHILDHHOOD

•

PAINTER

139

the application of general moral rules. At the same time, the segmented, context-specific organization means that a key feature of horizontal discourse is that learning takes place by continually accumulating or repeating instances, whether it be of buckling a shoe, practising a name or relating news. A fourth feature of horizontal discourse that Bernstein draws attention to is that the success of any segment is inherent in the specific occasion. Whether it is a meaning or a physical accomplishment that is being essayed, the child does not need to be told by someone else whether s/he has succeeded or failed. Whether or not Hal received praise, he knew whether the shoe got done up or whether the apology worked or whether he got a banana, or whether he figured out to his satisfaction the difference between ears and horns. Fifthly, there is the question of the tacit nature of many of the understandings construed through horizontal discourse. The data show clearly that the nature of the teaching and learning may be more or less explicit on different occasions. However, it is also clear that even where the pedagogic intention is explicitly directed at a particular goal, there will be other tacit learning going on. Moreover, it is the combination of segmental organization and tacit learning that makes it possible for contradictory or inconsistent knowledge to be held relatively unproblematically. This means, for example, it can be taken for granted in one segment that one small animal can be a playmate and family member whose affects are important, and in another segment that another animal can be a commodity for consumption. Finally, when it comes to a consideration of the language drawn on in horizontal discourse at this time, two additional and related features are important: one is the context-bound nature of the language, and the other is that the meanings all pertain to concrete experience. The characteristic of contextdependence relates to the fact that the language of a child under three will generally either accompany/request action in the immediate context (e.g. texts 4a, 4b and 6), provide a commentary on the shared visual context (e.g. texts 3 and 7), or reconstruct a recalled context (as in texts 4c and 5). The meanings are ‘concrete’ in that they relate to the observable and tangible rather than the general or abstract. As will be shown in the next section, these are limitations that begin to be transcended after about age 3. All the general characteristics of horizontal discourse summarized above have been observed by Bernstein (1999), and are likely to be common features of any child’s early language use. What is worth emphasizing though is that despite the segmental nature of horizontal discourse, there are aspects which cut across the boundaries of segments. I would suggest that the more tacit aspects – like the value of language to mediate physical learning (as seen in e.g. 4a and 4b), the potential to unhinge meaning and form (apparent in the misnaming game), and the value of being linguistically explicit and of conveying unshared information to others – are actually relevant across different segments. And it is these particular tacit aspects that may be specific features of elaborated code users. So even at this early stage, there are likely to be differences in the nature of horizontal discourse in different social groups.

140

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Problems and resolutions: developments in the linguistic system and its use after age three So far, I have exemplified the characteristics of horizontal discourse typical of a child’s use of language up to the age of about 3 years. Now I wish to consider the limitations of this kind of discourse for construing knowledge. One way to address this is to look at examples of talk where the child appears to struggle with meaning or to get things wrong. This throws light on the limitations of the language and also the aspects of knowledge that prove problematic. In this section, therefore, I shall identify moments from the second case study where Stephen has difficulty with construing common-sense knowledge, and outline the new developments in his linguistic system and its use that follow from these. (In these examples, because the child seeks help in construing his world, the pedagogy is more consistently of a relatively explicit kind, although of course many other conversations had a different character.) The developments described involve on the one hand a move away from the necessity of contextdependence and specificity, and on the other a movement towards more abstract meanings and a greater reliance on attention to textual rather than purely observational information sources. New linguistic resources and new possibilities for discourse will be shown to equip the child to gain glimpses of educational knowledge, but the lack of technicality, and the unsystematic and incidental character of learning through conversation, limit the child’s access to such knowledge. Problem 1: Limitations of ‘context-bound’ talk and observation for building semantic taxonomies In Examples 8a and 8b below, Stephen experienced difficulties tracking information that was detached from the observational context, and in organizing names of everyday things into semantic hierarchies. Example 8a (M and S’s brother have used the term pet in their talk together) S: (having overheard) What’s a pet? M: A pet is an animal who lives in your house; Katy’s our pet. (later same day) S: What’s a pet called? (2;11;15) Example 8a above is still very much in the realm of common-sense knowledge relevant to the domestic sphere, but the use of language by the parent is different. It is different from other texts where things are named in being no longer context-dependent. Instead of pointing (linguistically and physically) at something directly observable in the context by saying that’s a pet, Mum offers Stephen a definition (a pet is an animal who lives in your house); in other words, both terms in the relationship X = Y are constructed in a linguistic form that holds good regardless of the context of situation. Such a form can be termed ‘self-contextualized’. The difference between the two forms is set out below in Figures 7.1 and 7.2:

LANGUAGE FOR LEARNING IN EARLY CHILDHHOOD

Figure 7.1

•

that

is

a pet

reference to material entity

relator

name/category

PAINTER

Context-bound language for categorizing

a pet

is

an animal who lives in your house

name/category

relator

category

Figure 7.2

141

‘Self-contextualised’ language for categorizing

Because Stephen was used to learning categories through contextually supported language and not by processing definitions, he had difficulty with his mother’s statement and so he returned to the issue again later in the day, asking what’s a pet called? At this time call was Stephen’s only metalinguistic verb and his apparently inappropriate use of it here was doubtless an attempt to use his current linguistic resources to make the new move to talking about meaning rather than talking directly about observable reality. We see the same problem arising in Example 8b: Example 8b (Traffic passes in the street where S and M are walking) S: A bus isn’t a truck, is it Mum? M: No. S: What is it? M: A bus is for carrying people. S: No, what is a bus called? I said, a bus, what is it called? M: It’s called a bus; that thing (points) is just a bus and it’s – S: (plaintively) Oh, why is it just a bus? (3;6;25) In Example 8b, Stephen wanted a superordinate term in order to better understand the relationship between buses and trucks, but his mother failed to recognize this. As his enquiry was about names and categories, Stephen attempted to reframe it by using his only metalinguistic verb (in what is it called?), but failed to make himself understood. From grappling with these problems of categorization, Stephen was impelled to develop his own language further, and to learn in a less context-bound mode. He took up the metalinguistic verb mean and the use of self-contextualized defining and classifying clauses, together with reference to generic categories. In Examples 9 and 10 below, it is apparent that using this kind of language has advantages for extending knowledge both more quickly, and beyond here-and-

142

LANGUAGE, KNOWLEDGE AND PEDAGOGY

now experience: it shortcuts learning by discussing meanings rather than just tangible experiential reality. (In doing so, learning language and learning through language become also a matter of learning about language.) Development: metalanguage Example 9 (S looks up to window on hearing noise of gale blowing) S: It’s blowing and blowing out there. Can you hear it? M: Yes, I can. S: Hear it isn’t mean a talk is it? Means listen to it, isn’t it? Means listen to it, isn’t it? It’s two words, two words. (3;4;5) By focusing on synonymy and antonymy relations between the words hear and talk, or hear and listen to, Stephen in Text 9 could address meaning outside the confines of a specific context. At the same time, the meanings at stake are still very much of the world of lived experience and the everyday. Development: generic reference and ‘self-contextualized’ defining and classifying clauses In Example 10 below, Stephen shows a new ability to refer to generic categories (e.g. dogs), and can also manage defining and classifying clauses so as to refurbish a taxonomy without needing to see and refer to any actual specific new instances. (Here and elsewhere, uncertain transcription is indicated by (? ).) Example 10 (M asks S if he knows a word in the book they are looking at) S: No. M: It’s an animal. S: Rabbit? M: No, it’s [the word] dog. S: Dog’s not an animal! M: Yes it is . . . [further talk omitted] What is it then? S: It’s, it’s just a dog. M: Yes, but dogs are animals. S: No, they aren’t. M: Well, what’s an animal then? S: Um (?a) giraffe’s an animal. M: Oh, I see, you think animal is only for zoo animals. S: Yeah. M: Dogs are animals too, they’re tame animals. And cats, cats are animals too. Did you know that? H: (chipping in) And people, we’re animals. S: We’re not. (3;8;1) This text, as it progresses, is self-contextualized; that is, the language itself does not depend on reference to the immediate observable context of situation. It

LANGUAGE FOR LEARNING IN EARLY CHILDHHOOD

•

PAINTER

143

construes categories linguistically and then refers to those linguistic categories rather than to material reality. These linguistic developments are important for solving the problem of building taxonomies through everyday talk, and enable a more explicit instruction in meaning by the conversational partner. Problem 2: Observational understandings contradicted by textual information As Stephen continued to use his language for learning during his fourth year, he constantly discovered that his own understandings, gained from lived experience, might be contradicted by the more knowledgeable adult. Interactions where this was the case oriented him to see text as the key source for learning. In excerpts 11a and 11b for example, the adult responses he received implicitly taught him that a linguistic context provided by a conditional clause (if you bang your head; if he was trying) carries more weight than the observable here and now context of the talk: Example 11a (M warns S that he will smash his head if he has accident on hard floor) S: No, ’cause it’s got bone in, see? (taps head) M: But bones can break. S: No, it’s hard. M: Yes, but if you bang your head on the hard floor, it can still break, the bones can smash. (3;5;20) Example 11b (In car at traffic lights discussing sports cars) F: And they [i.e. sports cars] go fast ’cause they’ve got a big engine. S: But that doesn’t go faster than us (pointing at car parallel to them) See? We will go faster (as the light turns green) F: He’s not trying; if he was really trying he could go much faster than us. S: If he goes very fast he can – if he goes very fast he can beat us. (3;6;30) In the last utterance, Stephen tries out his new understanding, mirroring the conditional clause structure modelled by the adult. Development: privileging text over lived experience The fact that this kind of interaction takes the child in a new direction is apparent when we see Stephen puzzling over the status of his own observational experience and privileging textual information as a knowledge source. Text 12 is an interesting example of this: Example 12 (In car, S talks about seeing a spotted horse) S: You didn’t see it Mum. M: No, I didn’t. Did it have black spots? S: Yes, (?it didn’t have white spots). Really dogs have spots. M: Yes there are spotted dogs. But horses can have spots too. S: No, only dogs have spots.

144

LANGUAGE, KNOWLEDGE AND PEDAGOGY

M: But you just saw a horse. And giraffes have spots. And leopards have spots. S: Yes. Tell me some more animals. (3;5;2) Here Stephen appears uncertain about the status of an observed instance of a class and expresses a desire to establish the category of spotted animals through talk. Further unexpected examples of the tension between material and textual experience took place just after Stephen’s fourth birthday, when he began to ask whether or not he could swim. This was despite the fact that he was regularly at the beach and pool and should have known perfectly well that he could not. Excerpt 13 is from such a conversation: Example 13 (M driving S to his pre-school) S: Mummy, can I s- Mummy can I swim by myself now? M: We’ll have to try, won’t we? You nearly can. (?You just have to) practise a bit more. S: um I – I’m better tell Frank if he can swim or not – M: Ask Frank. (correcting) S: because, because you might be swimming when you’re five. M: (? . . . . . . ) It doesn’t really matter how old you are; it’s just whether you’ve had some practice . . . [long further discussion on 4 or 5 different topics omitted] M: (now talking about new person at pre-school) Oh, is he a little boy? S: Yup, he’s three. I’m four. I’m bigger. M: You are, yes. S: Can I swim? M: No, not quite. You don’t swim just because you’re four. S: (?) Swim when you’re 5? M: You swim after you’ve practised – some people swim when they’re three, some when they’re four, some when they’re five, some when they’re ten. S: Some when they’re six. M: Yeah. S: Or some when they’re seven. (4;0;8) At the time, Stephen’s questions seemed very strange, but it is likely that he was puzzling about his status as a swimmer because of the contradiction between his lived experience and verbal assurances he had received that he would be able to manage swimming when he was older. The fact that he had been being guided to orient more and more to text as a source of knowledge meant that he felt the need to check up on this as soon as his fourth birthday arrived. Development: reasoning from text All the discursive features so far illustrated – the use of definitions, generic categories, taxonomizing, orientating to text – come together in very occasional examples where Stephen shows the ability to use logical reasoning where he

LANGUAGE FOR LEARNING IN EARLY CHILDHHOOD

•

PAINTER

145

draws an inference from a linguistically presented premise. Example 14 is such a case: Example 14 S: Do whales eat people? M: No. S: Can they kill people? M: I don’t think they usually do. S: Can they sometimes? M: There may be one kind of whale that can; but most whales are nice creatures. S: They’re not creatures, Mum; they’re whales. M: Yes, creature is anything that’s alive. S: Are we creatures? M: Yeah. S: (laughing) No, we’re not! (4;4;10) This text is a far cry from those produced by Hal in his second and third year and described as fitting the descriptors of horizontal discourse. Here the language is no longer context-specific and context-bound, but rather evidences reasoning from linguistic propositions as Stephen draws the necessary (but to him surprising) conclusion that people must be creatures given that a creature is anything that’s alive. This text brings together many features that constitute one major strand in the development and use of language for learning after age three. These features include generic categories, definitions, the use of self-contextualizing language and a cognitive orientation to attending to text. Problem 3: The inaccessibility of (proto-)grammatical metaphor and abstraction A second major strand of development after age three is the move towards less concrete meanings, which in turn builds on the less context-dependent language that has already been discussed. In particular, the ability to use defining clauses helped Stephen to begin understanding an early form of what Halliday (1994) refers to as ‘grammatical metaphor’ (Taverniers 2003). Grammatical metaphor most frequently occurs when a meaning that is congruently expressed as a verb (as with laugh or he laughed) is instead expressed by a noun (as with laughter). When this kind of ‘nominalization’ happens, the action (here ‘laugh’) takes on the quality of a thing, since nouns congruently name things or participants in processes. Occasionally coming up against such grammatical metaphor was a third source of problems for Stephen as he attempted to make sense of his world through talk. As well as nominal expressions relating to time, such as week or year, a particular expression that caused difficulty for Stephen was the term traffic jam. We can understand this by considering both the congruent and (grammatically) metaphorical ways this meaning might be expressed, as shown in Figures 7.3 and 7.4 below:

146

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Example

This traffic

is jamming up

close together

Grammatical function

Participant

Process

Manner

Meaning

concrete things

action

manner

Figure 7.3

Grammatically congruent expression of meaning

Example

This

is

a traffic jam

Grammatical function

Participant

=

Participant

Meaning

observable context

Figure 7.4

action and manner

Defining clause with a (grammatically) metaphorical participant

Examples 15a, b and c below illustrate Stephen’s difficulty at age 31/2 in making sense of the term traffic jam. Example 15a (S and M in car waiting to turn right at traffic lights) S: Is this a traffic jam? M: No, it’s (pause) just waiting at the lights. S: No, is this called a traffic jam? (3;5;8) Example 15b (S and M in car stationary at red traffic light) S: Is this a traffic jam, Mum? M: Well, (pause) S: Is this a traffic jam? M: No, not really. (3;5;13) Example 15c (S and M in car stationary at red traffic light) S: Is this a traffic jam? M: (weary of the question) Oh, it’s a little jam. S: No, it’s a big jam, a big traffic jam; there’s all cars (pause). Is it a traffic jam? (3;5;14) Development: steps towards abstraction These examples indicate the difficulty of interpreting an expression like traffic jam purely by reference to the perceptual context. Stephen attempts repeatedly to test out whether a particular instance of experience constitutes an example

LANGUAGE FOR LEARNING IN EARLY CHILDHHOOD

•

PAINTER

147

of the category he is grappling with. But it proves an ineffective way of understanding the concept. Indeed, it is only when he receives a definition of sorts (see Example 15d below) that he makes any progress: Example 15d (M explains that now they are in a real traffic jam, because when the light goes green they still can’t go because there are too many cars. She then continues explaining) M: A traffic jam is when you can’t go even when the light is green. (The car then stops again as the light ahead goes red) S: (? ) It means when it’s green it’s called a traffic jam, but this is not a traffic jam cause it’s red; a green one is a traffic jam. (3;6;23) The definition Stephen is offered involves another instance of grammatical metaphor because a complex of meaning involving different processes (when you can’t go, even when the light is green) is construed as a single participant in the grammatical structure, as shown in Figure 7.5, below:

Example

A traffic jam

is

when you can’t go, even when the light is green

Grammatical function

Participant

=

Participant

Meaning

action and manner

=

action of participant + conceded condition

Figure 7.5

A definition involving grammatical metaphor

For a contemporary urban child, the phenomenon of a traffic jam doubtless constitutes common-sense knowledge, and it is certainly related to his everyday experience of travel. At the same time, as far as the language is concerned, some degree of grammatical metaphor1 and therefore a movement towards abstraction is involved. The construal of abstract concepts is not what we expect of horizontal discourse, but it is a necessary development for the achievement of literacy and other forms of school(ed) knowledge. As Halliday explains: The patterns of writing create systemic properties which are then named as abstract objects, like the beginning and end of a page or a line, spaces between words, and letters, capital (or big) and small. (Halliday 1996: 342)

In the following text we can see Stephen attempting to come to grips with the number system rather than the writing system, where a similar kind of abstraction is at play:

148

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Example 16 (S overhears father (F) mention the word fifty) S: Is fifty a number? F: Yeah. S: How does it go? F: It comes after forty-nine. S: A hundred comes after forty-nine. F: A hundred comes after ninety-nine. (3;11;7) Abstraction is involved because although one can point to a material inscription of 50 on the page and name it as an object with that’s 50, Stephen treats ‘a number’ as an abstract category with no material referent when he says is fifty a number. A form of grammatical metaphor is also evident in the use of come after to signal a relationship in a series rather than a material action of movement. These examples (15 a, b, c, d and 16) indicate that when children like Stephen come into contact with grammatically metaphorical expressions and especially with the semiotic forms of educational knowledge, familiar forms of language and the ways these have typically manifested in horizontal discourse prove inadequate, and the newer resources of self-contextualized talk are called into play. Through these means the child can then begin to reflect on the abstract forms of literacy and numeracy. The problem of building systematic knowledge It is not only in dealing with non-congruent grammatical expressions and the abstractions of numeracy/literacy that the limitations of horizontal discourse become apparent. Stephen’s parents sometimes struggled to respond to his questions about everyday phenomena, such as the function of power points in the house: Example 17 (M warns S not to play with wall socket) M: It’s a switch for electricity. S: (?But what is it?) M: It’s for electricity. S: But I don’t know what it looks like. M: No, well, you can’t see it; the electricity makes things go. In this case the invisibility of the named ‘thing’ (i.e. electricity) is problematic due to its verbalization being a nominalized form, but also because a satisfactory explanation seems to call for a groundwork of concepts the child has not yet developed. Stephen was able to use electricity in his everyday life, switching on lights and televisions etc., but not to understand it more ‘theoretically’ in terms of a definition. Similarly, the following conversation created puzzlement for Stephen:

LANGUAGE FOR LEARNING IN EARLY CHILDHHOOD

•

PAINTER

149

Example 18 (S and M are returning barefoot from the beach. M guides him off the scorching bitumen) M: Walk on the white bit [i.e. the pale paving slabs]; the white is always so much cooler than the black [i.e. bitumen]. S: Why? M: Because when the sun shines down on the black, the black keeps the hot, but when it shines on the white, the white – the white throws it off. S: (after a pause) You can’t feel it throw off. M: What? S: You can’t feel it throw the hot off. M: No, no you can’t. The adult here grapples with the difficulty of explaining matters without using nominalized or technical words, compromising with the black keeps the hot . . . the white throws it off, while the child tries to make sense of this by attempting to observe the phenomenon that has been described. While he can feel for himself the difference in temperature, he can only glimpse the nature of explanation, which draws on uncommon-sense knowledge. A further series of examples of the limitations of horizontal discourse for accessing new knowledge can come from one area of consistent interest to Stephen, that of relationships within the animal kingdom. He engaged in many conversations discussing the eating habits of animals and humans which essentially involved information about food chains and food webs. In the attempt to extend knowledge into such a domain, the segmental nature of the discourse and the fragmented, unsystematic nature of the learning proved to be a constraint. Texts 19–22 demonstrate well the nature of pedagogic conversations in Stephen’s fifth year. Example 19 below focuses on the eating habits of humans: Example 19 (S, H and M are reading a picture book where a boa constrictor has squeezed its prey) S: And what’s this? (pointing at mangled fawn) M: That’s a deer. S: Oh! (upset) I wish a snake weren’t coming and eat it. H: You know the – you eat animals. S: No I don’t! H: You do. You eat – fish, don’t you? Salmon. S: Yeah, but just – (? . . . the little) fishes at the beach but – but – but I don’t eat crocodiles, (M laughs) and that’s – that’s a animal. . . . [long discussion ensues about animals that people eat, including frogs] S: Yeah but what – we don’t – how can we eat the skeleton? [conversation continues for several minutes about the bones of various foods: frog, chicken, salmon] S: (loudly, touching his back) This is backbone. M: Your backbone, yes . . . (4;0;2)

150

LANGUAGE, KNOWLEDGE AND PEDAGOGY

This snippet can be related to a number of others that took place over succeeding months and which address questions of animal anatomy, particularly the function and appearance of bones. Stephen initiated such conversations with questions such as Mummy, how can your bones move? Mummy how can you move when something’s hard inside you? How can something move your bones? Why do your gums hold your teeth a long time? and so on. At the same time, Example 19 links to other conversations about animal predation. For example, Text 20 provides more of the conversation about whales that was presented earlier as Example 14: Example 20 (M and S are at home together) S: How do dolphins squirt water up their head? M: (pauses for thought) S: How do they? M: I think it’s whales that do that. S: Yeah, I meant (?that). M: I don’t really know how they do it, darling. S: Hey Mum, can dolphins eat boats? M: No. S: Why? M: Well – S: Why don’t they? M: They don’t want to eat boats; they eat fish. Are you thinking of that movie? S: Yes. M: Oh, that was a whale, but they don’t really swallow boats. S: Do whales eat people? M: No. S: Can they kill people? M: I don’t think they usually do. S: Can they sometimes? M: There may be one kind of whale that can; but most whales are nice creatures. S: They’re not creatures, Mum; they’re whales. M: Yes, creature is anything that’s alive. S: Are we creatures? M: Yeah. S: (laughing) No, we’re not! (4;4;5) A few weeks later, a moment of play led to a further conversation that touched on the same question of what animals eat and on their relations to each other and to people (Example 21). Then, the following month, these issues are pursued at the zoo (Example 22). The questions Stephen asks in these conversations suggest that he is concerned to construct some hierarchy of predation as well as to accumulate more knowledge about creatures that were not accessible to his own perceptual explorations:

LANGUAGE FOR LEARNING IN EARLY CHILDHHOOD

•

PAINTER

151

Example 21 (S leaps out at M from behind the curtain, roaring) S: Were you frightened? M: (teasingly) No, I’m not scared of lions. (grabs S) I just give them a cuddle. (hugs S) S: (seriously) But it might scratch you. M: Mm, it might eat me up. S: Can lions eat people? M: Yes. S: Oh! Can sharks eat lions? M: Lions don’t go in the sea. S: Can lions eat sharks? M: Sharks never go in the jungle, so they can’t, they never meet. S: In the river, the lion might swim in the river. M: But sharks don’t go in the rivers, they live in salty water – in the sea. Anyway, I don’t think lions swim in the rivers much; tigers do, but I don’t know about lions. S: That’s what I’m talking about – tigers. M: Oh, we’re talking about tigers now are we? S: Why don’t lions swim? M: I don’t know; maybe they do swim, but I don’t think they like the water. (4;4;25) Example 22 (The family is at the zoo watching a hippo) S: Does every animal eat hay? M: No. S: Horses do. They can’t eat people, can they? . . . [further talk omitted] (In another part of the zoo, F lifts S to see the fox) S: Can foxes eat people? Can foxes eat people, Mummy? M: Oh no; they eat chickens sometimes. (4;5;16) I would suggest that although these texts exemplify horizontal discourse in some respects, they also show its limitations as a pedagogical discourse when the knowledge at stake edges into the realm of the un-common-sense. It is true that some of the individual pieces of information contained in these texts are in principle accessible to observation and lived experience, and might therefore count as ‘common’ knowledge. Moreover the discourse is segmental in that the contexts in which these texts arise are varied – initiated by a movie narrative, a peepbo game, a visit to the zoo, looking at a picture, respectively – and in the fact that information is not systematically sequenced in its presentation nor finalized on any one occasion. Indeed Stephen’s continual return to similar themes indicates his desire to engage in further ‘goes’ at meaning so as to better integrate his knowledge fragments. In certain respects, then, the talk has continuities with what has gone before. However, just as contact with the semiotic systems of school(ed) knowledge

152

LANGUAGE, KNOWLEDGE AND PEDAGOGY

(numbers and written words) creates pressures on the informal pedagogy of the home and stretches the child’s language towards greater abstraction, so I would argue that the attempt to construe knowledge whose basis is less accessible to observation does the same. When we consider the language of Texts 19–22, we can observe that it exemplifies generalizations, is largely independent of the immediate context, and may even use logical reasoning from a textual premise. The language has thus moved quite a long way from the context-specific and dependent language said to characterize horizontal discourse. Yet even with these more developed linguistic resources, the knowledge is somewhat elusive since it is being gained in a piecemeal fashion that does not allow all the fragments to be brought into a relationship with one another. It is enlightening here to look at how at these topics get expressed in a more overtly ‘vertical’ discourse. Any selection of texts from the World Wide Web on the topics of food chains and food webs reveals both similarities and differences between the conversational Examples 19–22. One such text explains a food chain as shown in Example 23: Example 23 FOOD CHAINS AND FOOD WEBS ‘In the living world, every form of life is food for another. Food chains and webs show how food and energy are passed between species.’ FOOD CHAIN ‘A food chain is a food pathway that links different species in a community . . . ’ (The Balance of Nature, accessed 12/12/05) Already it is obvious that the language here is self-contextualized, involves definitions of terms, and uses generalizations, metaphor and abstraction (e.g. food chain, food pathway). The text continues with explanations of ‘food web’, ‘trophic level’, ‘producers’, ‘consumers’ and subcategories of these provided in rapid succession. The information in such a text thus involves explicit taxonomies of concepts, systematicity, rapidity and comprehensiveness. Accessing such a text depends on the kind of language that Stephen has been developing in response to problems encountered in the process of construing common-sense knowledge, and in beginning to encounter more specialized forms of knowledge. Without the ability to learn from definitions, to manage generic categories, to attend closely to text, and to handle abstraction, the meaning of such a text cannot be unpacked. Even so, the differences from the discourse of the home are striking. First there is the speed at which new concepts are provided and built on, something very different from the child’s conversational experience; secondly, there is the technicality that is a necessary part of organizing the knowledge in a systematic way. Such texts may also combine technical language with a visual display of some of the relationships being systematized. This again achieves something different from the talk in the home – it enables information that occurred at different points in the oral texts (but was never put together there) to be brought into explicit relationships and integrated into a coherent whole. Thus in the process of building up their everyday knowledge, children like

LANGUAGE FOR LEARNING IN EARLY CHILDHHOOD

•

PAINTER

153

Stephen will engage in talk that is not restricted to the context-specific, contextdependent forms of horizontal discourse but will still not be able to build a hierarchical knowledge structure without a more radical departure from the segmentally organized and unsystematic conversational pedagogy of the home.

Conclusion My object in this chapter has been to bring into clearer focus the nature of the language used for learning in early childhood, demonstrating the close relation between ‘learning language’ and ‘learning through language’, while simultaneously clarifying the nature, limitations and potential of children’s experience with ‘horizontal discourse’. Evidence from the two case studies indicates that Bernstein’s characterization of this discourse is an apt one for the first few years of life and for much of the learning that takes place before formal education begins. However, this should not obscure trans-segmental aspects of the tacit learning that occur from the start, nor the significance of developments that gradually take place under the pressure of making sense of everyday experience and in response to the child’s own inclination to extend his knowledge through talk – developments that change the character of the talk in significant ways. I have provided examples of how one child grappled with a range of problems in meaning between the ages of 3 and 5 years. The first of these led to an increasing ability for self-contextualized talk. The problem of building deeper semantic taxonomies when limited to context-dependent classifications was met by making use of everyday metalanguage (e.g. hear it means listen to it) and by using classifying and defining clauses which make no direct reference out to the immediate context (e.g. what’s a pet?, (a) giraffe’s an animal), together with the construal of generic categories (dogs have spots). All of these developments enabled the discussion of meanings without reference to observed instances of the phenomena under attention. This in turn shortcut the gradual process of building taxonomic relations and allowed for a single text to set up new relations or revise an existing classification (M: Cats are animals too, did you know that?). In these ways the existing roles of language for accompanying, monitoring and reconstructing specific contexts could be extended to generalizing across contexts. At the same time, the possibility of creating a context linguistically through a hypothetical scenario (if he goes very fast he can beat us; if you bang your head . . . it can still break) was made salient in talk within Stephen’s family, perhaps due to the elaborated coding orientation favoured, and alerted him to another problem: conflicts between material and semiotic experience. A privileging of the latter enables exploratory talk and logical reasoning, and thus has the potential to further facilitate the building of knowledge beyond observed and lived experience. Paying close attention to text is in fact important both for building common-sense knowledge more quickly and for later learning through the written medium. The other ‘meaning problems’ encountered by Stephen related to forms of language or knowledge that are less common-sense. On the one hand there were occasional encounters with nominalized language, both within the realm

154

LANGUAGE, KNOWLEDGE AND PEDAGOGY

of the everyday (traffic jam) and in the exploration of other semiotics (number). These led to brief forays into the abstract and metaphorical language that is more characteristic of vertical discourse. On the other hand, there was the difficulty of satisfying his curiosity about matters inaccessible to observation (what’s electricity? why is the white [ground] always cooler than the black? how can your bones move?), as well as difficulty in integrating different pieces of less problematic information (e.g. that some animals eat plants, others eat other animals, humans eat animals, but not all animals eat humans, etc.) These problems were not readily overcome because their resolution requires further development of language through forms of pedagogy specifically designed to build knowledge in a systematic and coherently sequenced manner. While all the linguistic developments outlined above will eventually help the child to engage in earnest with knowledge beyond the everyday, they can only be capitalized on in a limited way in the fragmented pedagogic encounters of everyday life. The systematic organization of knowledge and the need to master it in a particular sequence so that one concept can build on and relate to another is typical of certain kinds of ‘educational’ knowledge. It calls for a different form of discourse (e.g. Text 23) and for forms of pedagogy that will help the learner to access it. The efficacy of horizontal discourse for creating our common and often tacit understandings should not be assumed to make it well designed for the more conscious creation of the kind of knowledge to which Stephen appeared to be striving in his fifth year.

References The Balance of Nature: Food Chains and Food Webs, www.cas.psu.edu/DOCS/ WEBCOURSE/WETLAND/WET1/balnat.html Bateson, M.C. (1975), ‘Mother-infant exchanges: the epigenesis of the conversational interaction’, in D. Aaronson and R.W. Reiber (eds), Developmental psycholinguistics and communication disorders; Annals of the New York Academy of Sciences, 263, pp. 101–13. Bernstein, B. (1987), ‘Elaborated and resticted codes: an overview 1958–85’, in U. Ammon, N. Dittmar and K.J. Mattheier (eds), Sociolinguistics: An International Handbook of the Science of Society Vol. 1. Berlin: De Gruyter. Bernstein, B. (1999), ‘Vertical and horizontal discourse: an essay’, in British Journal of Sociology of Education 20.2, pp. 157–73. Bloom, L. (1993), The Transition from infancy to Language: Acquiring the Power of Expression. Cambridge: Cambridge University Press. Field, T.M. and Fox, N.A. (eds) (1985), Social Perception in Infants. Norwood, NJ: Ablex. Halliday, M.A.K. (1975/2004), Learning How to Mean. London: Arnold. Reprinted in his Collected Works Vol. 4. London: Continuum, 2004. Halliday, M.A.K. (1980/2004), ‘Three aspects of children’s language development: learning language, learning through language and learning about language’, in Y.M. Goodman, K.K. Haussler and D.S. Strickland (eds), Oral and Written Language Development: Impact on Schools (Proceedings from the 1979–80 Impact Conferences). International Reading Association and National Council of Teachers of English, pp. 7–19. Reprinted in his Collected Works Vol. 4: The Language of Early Childhood. London: Continuum, 2004, pp. 308–26. Halliday, M.A.K. (1994), Introduction to Functional Grammar, 2nd edn. London: Arnold.

LANGUAGE FOR LEARNING IN EARLY CHILDHHOOD

•

PAINTER

155

Halliday, M.A.K. (1996), ‘Literacy and linguistics: a functional perspective’, in R. Hasan and G. Williams (eds), Literacy in Society. London: Longman, pp. 339–76. Halliday, M.A.K. (2004), ‘On grammar as the driving force from primary to higher-order consciousness’, in G. Williams and A. Lukin (eds), The Development of Language: Functional Perspectives on Species and Individuals. London: Continuum, pp. 15–44. Heath, S.B. (1983), Ways with Words: Language, Life and Work in Communities and Classrooms. Camridge: Cambridge University Press. Meltzoff, A.N. and Moore, M.K. (1998), ‘Infant intersubjectivity: imitation, identity and intention’, in S. Braten (ed.), Intersubjective Communication and Emotion in Early Ontogeny. Cambridge: Cambridge University Press, pp. 47–62. Messer, D. (1994), The Development of Communication: From Social Interaction to Language. Chichester: Wiley. Murray, L. (1998), ‘Contributions of experimental and clinical perturbations of motherinfant communication to the understanding of infant intersubjectivity’, in S. Braten (ed.), Intersubjective Communication and Emotion in Early Ontogeny. Cambridge: Cambridge University Press, pp. 127–43. Painter, C. (1984), Into the Mother Tongue. London: Pinter. Painter, C. (1999), Learning through Language in Early Childhood. London and New York: Continuum. Painter, C. (2005), ‘The concept of protolanguage in language development’, Linguistics and the Human Sciences 1.2, pp. 177–96. Taverniers, M. (2003), ‘Grammatical metaphor in SFL: a historiography of the introduction and initial study of the concept’, in A-M. Simon-Vandenbergen, M. Taverniers, L. Ravelli (eds), Grammatical Metaphor: Views from Systemic Functional Linguistics. Amsterdam and Philadelphia: John Benjamins, pp. 5–33. Torr, J. (1997), From Child Tongue to Mother Tongue: A Case Study of Language Development in the First Two and a Half Years. Nottingham: Dept of English Studies, University of Nottingham. Trevarthen, C. (1977), ‘Descriptive analyses of infant communicative behavior’, in H.R. Schaffer (ed.), Studies in Mother-Infant Interaction. London: Academic, pp. 227–70. Trevarthen, C. (1987), ‘Sharing makes sense: intersubjectivity and the making of an infant’s meaning’, in R. Steele and T. Threadgold (eds), Language Topics: Essays in Honour of Michael Halliday Vol. 1. Amsterdam and Philadelphia: John Benjamins, pp. 177–99. Trevarthen, C. (1992), ‘An infant’s motives for speaking and thinking in the culture’, in A.H. Wood (ed.), The Dialogical Alternative: Towards a Theory of Language and Mind. Oslo: Scandinavian University Press, pp. 99–137. Trevarthen, C. (1998), ‘The concept and foundations of infant intersubjectivity’, in S. Braten (ed.), Intersubjective Communication and Emotion in Early Ontogeny. Cambridge: Cambridge University Press, pp. 15–46.

Notes 1

This is a ‘proto’ grammatical metaphor in that it is only a single isolated routinized expression, rather than a more creative manipulation of the grammar in the course of organizing a complex text field of knowledge.

8

Building verticality in subject English1 Frances Christie and Mary Macken-Horarik

Introduction Subject English is just over a century old. The first university chairs in English were created in the second half of the nineteenth century, while by the turn of the century, subject English (as distinct from discrete areas of knowledge, such as ‘spelling’ or ‘reading’, Christie 2004a), had appeared in the programmes of the elementary schools. As the twentieth century unfolded, English was to become the most important subject in the school curriculum, its status rivalled only by the claims of mathematics, though arguments about both the contents of mathematics and its pedagogy have never rivalled those about subject English.2 In fact, subject English has always been a highly contested site in the school curriculum, not least because discussion about education in the national language has been intimately bound up with discussion of matters to do with the national psyche and identity, as well as with notions of the economic and social good of English-speaking countries. It was, for example, no accident that in the aftermath of the First World War, the English held their first inquiry into the teaching of English (the Newbolt Report, 1921).3 The war had exposed problems to do with the literacy skills of the armed forces and there was an interest in improving citizens’ capacities to read and write. In the post-war period, secondary education was extended, especially in the English-speaking world, and this led to a renewed significance for subject English. Before the end of the twentieth century, the English held two more major inquiries into the teaching of English, leading to the Bullock Report (1975) and the Kingman Report (1988). The work of the latter, developed in various ways by the Cox Working Group (1989), led to the adoption of the National English Curriculum (1995). Developments in Australia were different, though by the 1990s several federally co-ordinated initiatives led to major policy statements on language and literacy. These included a federal Government White Paper, Australia’s Language. The Australian Language and Literacy Policy (1991), a nationally endorsed statement, English: a Statement for Australian Schools (1994), and a later national policy statement, Literacy for all, the Challenge for Australian Schools (1998).4 Thus, over the twentieth century, subject English assumed an increasingly important role in schooling, attracting considerable interest at governmental levels and also, by the end of the century, ushering in what have become well-established testing regimes in the English-speaking world, intended to monitor and report on the progress children make in mastering English literacy. Today, so significant is subject English, success in it is now an important passport to many avenues of privileged life and education. Yet, ironically, given

VERTICALITY IN SUBJECT ENGLISH

•

CHRISTIE & MACKEN-HORARIK

157

its increased importance, the nature of English is increasingly elusive, its mastery not available to many students. A powerful invisible pedagogy often applies, such that what is evaluated as success is tacitly understood, rather than clearly articulated. In the contemporary school curriculum, there are in fact many ‘Englishes’, even if the singular form is retained in the name of the subject. A review of its history (e.g. Christie 1993, 2003, 2004a) would suggest that it has been of the nature of the subject that it has proliferated many forms, or ‘languages’, each possessed of its own preoccupations, its own questions, its own value positions. Bernstein (2000: 155–74) drew a distinction between horizontal and hierarchical knowledge structures, where each of these, he argued, constituted one of the two possible modalities of vertical discourse. There is no complete fit between Bernstein’s use of the word ‘grammar’ and that in the systemic functional (SF) tradition. However, we shall suggest that in Bernstein’s terms school subject English has horizontal knowledge structures with ‘weak grammars’, meaning that it is constructed in a series of specialized languages, segmentally organized, wherein what counts as achievement is adoption of a particular pedagogic position or ‘gaze’ (Bernstein ibid.) rather than any strongly defined theoretical position which might confer a degree of commensurability across the languages. Furthermore, with the passage of time, we shall suggest, while various ‘languages’ have thus emerged, the content of subject English has steadily become more invisible, the exception being some functional approaches (Macken-Horarik 1998, 2002) which have been notable for their claims to a visible content and a visible pedagogy in all subjects, including English. We shall also argue that, despite what has been historically an overall drift towards invisibility of content, it is possible to identify certain apparent ideologies that apply in English, and that privilege certain value positions over others. Among other things, what is at issue for the learner is tacit acquisition of certain values and ‘truths’ of the knowledge of English, to do with expression of sensibilities or what we might term ‘finer feelings’. For obvious reasons, such tacit acquisition creates a situation that advantages some students rather more than others. Important questions for us concern how to address problems to do with the tacit acquisition of the values of English, and how to develop a model of the discipline that is more theoretically robust and more transparent for students. Such a model would, among other things, be internally coherent, based on well-theorized organizing principles and articulated in a (meta)language that allows for progression up the years of schooling. Like Muller (this volume) we shall use the term verticality to refer to the characteristics of such a model. We shall suggest that the model of English we are seeking will be marked by greater verticality than presently applies. That is to say, it will show an enhanced capacity for integration of the elements of its knowledge structure, providing an expanding sense of a coherent knowledge base as students move through their schooling. This chapter has two parts, in the first of which we take up the challenge issued by Bernstein in discussing horizontal knowledge structures and their emergence, ‘to look at changes in the development of specialized languages

158

LANGUAGE, KNOWLEDGE AND PEDAGOGY

across time’ and ‘to plot the increase in the number of languages’ found (Bernstein 2000: 166). Where Bernstein suggested this for sociology, we do it for subject English. To do this, we explore the origins of different specialized languages (or models) of subject English and propose that the proliferation of languages/models has created difficulties for students not already well positioned to manage successfully its increasingly tacit requirements. This is particularly evident in contexts where students are asked – often in examinations – to respond to open questions about unfamiliar texts, or to write responses to rather general questions, and where the basis of assessors’ judgements is left implicit. We argue that careful attention to the semantic qualities of successful students’ responses in these contexts enables us to probe the linguistic dimensions of the relevant discourse in English, thereby exposing the criteria for successful performance which otherwise remain unclear. In the second part of the chapter, we develop the argument by investigating the linguistic features of two highly valued texts produced in formal examinations by students in England and in Australia. Drawing on metalinguistic tools available in SFL, we highlight the orientations to meaning that are rewarded at the key transitional school stages that are represented in public examinations. We identify the linguistic choices which realize these successful orientations and conclude by arguing the potential of SFL to increase the ‘verticality quotient’ (Muller 2004) of English.

The building of the languages of English teaching It is possible to interpret the development of school English as the unfolding genesis of knowledge structures (models of curriculum) from its inception in the nineteenth century till the early twenty-first century. The nineteenthcentury concern for provision of a basic education for the poorer classes led to the emergence of Basic Skills, and this gave way with the opening of the twentieth century to the Cultural Heritage model, whose preoccupations were with the civilizing influence of ‘great literature’. In the period after the Second World War, two different models emerged in the same period. On the one hand there was a strong functional tradition we have termed Functional Language Studies, and on the other hand, a strong personalist tradition, referred to as Personal Growth. Where the former focused on socially conceived functions of language and proposed overt teaching of knowledge about language, the latter focused on personal growth in learning language, and its tendency was to resist teaching of knowledge about language. While Functional Language Studies were influential in many places, Personal Growth tended to dominate in many versions of English curriculum across the western world. In the last two decades of the century into the early twenty-first century Cultural Studies became influential, presently constituting what is often the dominant model of subject English at the senior levels. Unlike Cultural Heritage, Cultural Studies eschews the literary canon, stressing the value of studying popular and multimedia texts, and encouraging students to make multiple readings of texts. Unlike Functional Studies this tradition has no overt commitment to teaching knowledge about language, and in

VERTICALITY IN SUBJECT ENGLISH

•

CHRISTIE & MACKEN-HORARIK

159

fact it has some affinities with Personal Growth, in its emphasis on personal responses to life and learning. Two further models of English emerged in the late twentieth century: Multiliteracies and New Literacy Studies. The latter studies emphasize the ‘situated’ nature of texts and meanings, asserting the irrelevance of much traditional language study, while Multiliteracies stress the importance of multimodality and sociocultural diversity in contemporary communication, and like Cultural Studies they propose the study of a wide variety of texts. In the following sections, we turn our attention to a more detailed account of the genesis of each model of subject English. Basic Skills English in the last years of the nineteenth century amalgamated what had been hitherto some ‘discrete language skills’. The teaching of reading had had a long history (see Davies 1973, and Christie in press), while the teaching of writing, once an aspect of the teaching of rhetoric, had also had a reasonably long history, though by the nineteenth century it had survived rather better in Scotland than in England. English school grammar, derived originally from the study of Latin, had been taught at least as far back as the sixteenth century, though it was in the eighteenth century that several accounts of English grammar for students were written, the most influential being that of Murray (1795/1968; Christie 1993). What made the nineteenth century significant for subject English was that this was the century in which school education was regularized and the state became involved in providing elementary education. After the 1830s in England, and the 1850s in the Australian colonies, steps were taken to provide state-supported elementary schooling (Christie 1976), and teacher training became an increasingly important enterprise. Not surprisingly, it was English language, especially literacy, skills, that loomed large in the thinking of those who devised the elementary school programmes. Large numbers of graded school readers, spelling books, grammar books, dictionaries, and some composition books, appeared. Theories about the most appropriate methods for teaching the various English language skills abounded, and textbooks appeared for teacher training. One, for example, was that by Fitch (1880), Her Majesty’s Inspector of Schools, which focused on the values of teaching the English language, grammar and grammatical analysis, paraphrasing and classification of words (Fitch ibid.: 256–85). Writing in 1880, Fitch noted the need for the English programme ‘to make some reference to English Literature as a branch of school instruction’, also observing that ‘this is a comparatively new ingredient introduced of late into the school course’ (Fitch, ibid.: 279–80). The ideal pedagogic subject position associated with the nineteenth-century elementary school involved someone with basic literacy competence. Considerable importance attached to learning to read (Gosden 1969; Goldstrom 1972; Cathcart Borer 1976; Christie 1976), for this was part of acquiring a little useful ‘general knowledge’, including knowledge of the Bible, while it was also held to create sober, upright workers and citizens. The nineteenth-century elementary schools brought a degree of ‘social governance and spiritual guidance’ to the children of the working class and the poor, and this regulatory function was

160

LANGUAGE, KNOWLEDGE AND PEDAGOGY

central to the creation of social order by emergent nation-states of the nineteenth century (Hunter, 1994). The operation of what Bernstein later called ‘the pedagogic device’ (e.g. Bernstein 1990) helped the young to forge important inner and outer disciplines related to Christian values, personal comportment and civic responsibility; basic literacy skills were important to the development of these things. The various English literacy skills were taught using a visible pedagogy, in that children were drilled in such things as the alphabet, spelling and the parts of speech, and they were taught to read with graded readers. Composition writing came much later, after painstaking drill in formation of letters, as well as writing simple sentences. As many children did not stay at school for long, relatively few learned to write more than their names or perhaps simple letters. The visible, albeit conservative, pedagogy involved had some merit in that it explicitly identified such things as spelling rules and taught children to read simple texts, while it also taught parts of speech and some sentence analysis for those who learned to write. However, the language content taught was quite meagre, and the linguistic practices it taught were unrelated to the rhetorical concerns of some earlier traditions of language study (Christie 1993). Cultural Heritage It was Arnold, poet and inspector of schools (1896; Sutherland 1973), who first proposed teaching poetry to children because of its civilizing influence. This would, he argued, contribute to achieving the order and harmony that a modern society required. He advocated having children commit poems to memory, and regarded the reading of ‘great literature’ (by which he normally meant poetry) as an important palliative to the ills that beset a world undergoing rapid change and social disorder. His general philosophy was part of a broader commitment to social regulation (Hunter, ibid.) that emerged with nineteenth-century elementary schooling, and Arnold contributed to the emergence of the Cultural Heritage tradition. F.R. Leavis, influential at Cambridge after the First World War, articulated more fully the language of the Cultural Heritage, drawing on the work of Arnold, but also that of colleagues such as T.S. Eliot, I.A. Richards, and Denys Thompson. The emergence of the so-called Literary Canon is most associated with their names, though Maybin (2000) suggests that its origins were much older (also see Peel et al. 2000). For Leavis, like Arnold, the function of literary studies was to redress the significant social ills that he saw in the contemporary world. His works (e.g. with Thompson 1933; 1933, 1934, 1965) express a deep concern for the poverty of the modern world, destroyed by ‘that great agent of change . . . the machine-applied power’ (Leavis and Thompson 1934/1964: 3), which had robbed society of culture and community. Culture had always been in the keeping of a minority, Leavis believed, and the function of an English programme was to cultivate ‘taste and sensibility’ in people (Leavis and Thompson, ibid.: 1). Having proposed numbers of great works of literature for study (e.g. Chaucer, Shakespeare, Keats, Austen, George Eliot and T.S. Eliot), Leavis proposed teaching ‘close reading’, seeing literary texts as the expressions of

VERTICALITY IN SUBJECT ENGLISH

•

CHRISTIE & MACKEN-HORARIK

161

universal truths which the discerning reader could establish, thereby achieving ‘culture’ and ‘sensibility’. However, even though Leavis and his colleagues espoused close reading of literary texts, they did not engage substantively with the language involved. Later in the century, students taught in the traditions of the New Critics, such as Northrop Frye, also addressed what were deemed great literary works, seeking to focus in rather decontextualized ways on such texts, where the intention was to develop taste and discrimination. Again, however, there was no direct engagement with the language used. The relatively visible pedagogy associated with the learning of the Basic Skills (limited though its goals were) was buried beneath the weight of the apparent imperative to enter into a culturally valued sensibility (Cultural Heritage) or a refined taste (New Criticism). As a result, the analytical skills that could be used for critical exploration of values in literary texts were not foregrounded. As the twentieth century proceeded, many literary critics rejected the élitist and ethnocentric exclusions of Leavis and others, while they challenged the idiosyncratic nature of his selection of literary texts, as well as his indifference to important post-colonial and feminist writers. Despite the fact that Leavis and his colleagues fell into some disfavour among academic teachers of English, the influence of the Cultural Heritage in school English persisted well into the late twentieth century, often sitting uneasily alongside what remained of the older tradition of Basic Skills. This meant that the typical school English programme that had emerged by the 1950s involved the study of ‘great literature’ and some amalgam of basic skills in school grammar, sentence parsing and analysis and written composition. The various elements of the English programme often did not relate easily to each other, for in practice the ‘gazes’ required for each differed. The skills taught were often unrelated to students’ reading or writing of texts and the study of the great works was rarely used to illuminate work on grammar, or vice versa. Personal Growth The Personal Growth model of English, like Functional Language Studies (discussed below) emerged in the 1960s. Both were born in the aftermath of the Second World War, and the need felt at that time actively to rebuild schooling, though in practice the two offered very different models of the pedagogic subject. Personal Growth was a response to the appearance of various progressivist models of educational theory (after Dewey) and constructivist models (after confused readings of Piaget and Vygotsky) (see Muller 2000 and Christie 2004b), whose effect was to propose an ideology of the individual learner as embarked on some personal journey of discovery. Constructivism and related theories have had many names: ‘pupil-centredness’, ‘learner empowerment’, ‘activity-based learning’, ‘inquiry learning’, ‘process approaches to learning’, ‘problem-based learning’, to name a number. Stenhouse (1975) provided one influential account of process models of curriculum, while the Plowden Report on Children and their Primary Schools (1967) proposed a similar ‘process’ approach to the primary curriculum, reducing the role of the teacher and stressing that children should learn in their ‘own’ ways. Several English-teaching specialists met at Dartmouth to discuss the teaching

162

LANGUAGE, KNOWLEDGE AND PEDAGOGY

of English in 1966, and Dixon’s Growth through English (1967) was an outcome of the meeting. As the title suggests, the classroom was envisaged as a site in which children could be encouraged to grow in independent ways, using their ‘own’ language for self-expression and personal discovery. The curriculum theorists in the constructivist tradition, and the English-teaching theorists after Dartmouth, shared a commitment to teacher facilitation rather than instruction, and they held an associated tendency to diminish the notion of any organized form of knowledge, on the grounds that persons should fashion their ‘own’ knowledge. Correspondingly, principles of scope and sequence in the design of the English programme, which had been very strong, for example in the Basic Skills tradition, (and which were also a feature of much Cultural Heritage teaching) were actively rejected, on the grounds that students should forge their own understandings, unfettered by external ordering of knowledge and experience. The skills of spelling, grammar, sentence construction and reading foregrounded in the Basic Skills tradition were greatly reduced in significance. Grammar was to be taught ‘at the point of need’ rather than formally introduced. (Christie 2004b, offers a critique of this view.) The values of ‘close reading’ of literary texts in the manner of the Cultural Heritage were similarly downplayed, as children were encouraged to read widely and choose texts on the basis of personal, rather than institutional, interests. With the movement to Growth, the language of English became increasingly invisible, while the pedagogic position was that of one actively involved in ‘self expression’. Subject English now embraced no less than ‘life itself’ (Britton and Squires, in Dixon 1975: xviii), and children should discover their ‘own’ truths, their own values and ‘competences’. Bernstein (2000: 41–63) offered a very interesting critique of the emergence of ‘competence’ in pedagogic models, arguing that though various, often inimical, scholarly traditions had contributed to it, all had much in common. They imagined ‘an in-built creativity, an in-built virtuous self-regulation’ in pedagogic subjects, such that all could develop in essentially benign ways, pursuing their own interests and acquiring similar competences. They reflected what Bernstein termed ‘an announcement of a universal democracy of acquisition’. Such a model of the pedagogic subject was a very unsatisfactory one, because it deflected attention from ‘the analysis of power and principles of control which selectively specialise modes of acquisition and realisations’ (Bernstein, ibid.: 43). Thus, for the purposes of learning English, the Personal Growth model encourages the expression of sensitivities about life but the facility with language necessary to achieve this remains at best tacitly understood, because a theory prevails that persons are best left to ‘work things out for themselves’. Those students well equipped by life experience and opportunity to intuit the desired skills are thus rewarded, while those unable to intuit these are denied an opportunity to learn. Neither the content to be mastered nor the criteria that apply for evaluation of students’ efforts are made clear. In this context Bernstein’s observations about evaluation are significant. In discussing the operation of the pedagogic discourse (Bernstein 1990: 165–218) he observed that three sets of ‘rules’ were in play, all of them involved in the shaping of consciousness. Distributive rules concern the identification and specialization of ‘different orders of meanings’; recontextualizing rules concern

VERTICALITY IN SUBJECT ENGLISH

•

CHRISTIE & MACKEN-HORARIK

163

the manner in which the specialist meanings are taken and relocated for pedagogic purposes; evaluative rules involve evaluation of the pedagogic practices. Hence for school subject English, certain meanings – often to do with self expression and/or sensibility – are variously taken from literary and other sites and recontextualized for pedagogic purposes. Success in control of these is certainly evaluated, as we shall demonstrate by reference to students’ texts below, yet the terms of the task given the student do not on the whole make clear the nature of the pedagogic position that is rewarded, nor the linguistic practices that will embody the ‘right response’. Functional Language Studies The tradition of Functional Language Studies owes most to the work of Halliday, who in the 1960s led two major projects in London devoted to the teaching of English. (See Hasan and Martin 1989 and Halliday and Hasan, in press). A theory of Functional Language Studies was initially articulated by Halliday, McIntosh and Strevens (1964) to complement the developing language projects, though over the years the theory was significantly expanded and extended in the light of research. (See Christie and Unsworth 2005, for a review of developments in SFL language education from the 1960s to the present.) Many curriculum materials for schools as well as resources for teachers emerged from the work of the 1960s, while many more materials have appeared since, in the light of developments especially in Australia. After the publication of the first edition of Halliday’s Introduction to Functional Grammar (1985), numbers of books for use in schools appeared (e.g. Christie et al. 1990, 1992a, 1992b; Derewianka 1990), while a little later, several pedagogic versions of the functional grammar were published (e.g. Derewianka 1998; Christie and Soosai, 2000, 2001; Droga and Humphrey 2002, 2003), and volumes to support educational research and teacher education have also appeared (e.g. Unsworth, 2000; Christie 2005). As the name suggests, Functional Language Studies adopts a functional orientation, proposing that language is a social semiotic, involved in the negotiation, ordering and structuring of experience, and the functional grammar is the fundamental tool for analysis of texts. Like Personal Growth, Functional Language Studies have always espoused an interest in notions of language varieties, though the Functional tradition formalized this interest in studies of both dialect and register, and as a result it proposed a much more rigorous examination of language than did the Personal Growth proponents. In addition, and unlike Personal Growth, Functional Language Studies recognized grammatical differences between speech and writing, and argued for the value of teaching explicit knowledge about language (‘KAL’), the better to equip students for informed use of the language system. In other words, in this tradition learning about language does not just have an abstract analytical importance, as it often did in the Basic Skills tradition for example. On the contrary, in this tradition, language is itself the object of study for teaching and learning, and an intimate association of text and context is established, while the relationship of text and the grammatical resources in which it is realized is a major focus of attention.

164

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Functional Language Studies has been far more interventionist in its pedagogical principles than Personal Growth, eschewing what Halliday once referred to as its policy of ‘benevolent inertia’, and adopting a model of the teacher as actively guiding and directing students’ learning. Furthermore, Functional Language Studies has greatly increased the visibility of language in the curriculum, while it has also pointed directions for pedagogic practice. After the late 1970s and early 1980s, probably the best known of its pedagogic interventions has been genrebased pedagogy with which the name of Martin (1999) is associated, though many of his colleagues have by now contributed to the ongoing development of this pedagogy (see Macken-Horarik 2000). The ideal pedagogic subject position in this model is a linguistically informed one, capable of the analysis, production and critique of texts and contexts. A knowledge about language, it is held, enables students to interpret the linguistic bases of value positions and ideologies embodied in texts. An understanding of the power of language supplements an understanding of the language of power (Macken-Horarik 1998; Martin and Rose 2003, 2006). For all these reasons, we argue that use of the systemic functional grammar has made important gains in developing a model of language studies which has the potential to inform the development of an enhanced discourse for subject English, one which can point directions for the design, scope and sequence of the language curriculum across the years of schooling. Among other things, the functional theory provides an understanding of the nature of the language system, an awareness of the grammatical differences of speech and writing, and a sense of the developmental changes that need to occur as students mature, achieving greater mastery of language and the meanings it realizes. Below, by reference to two texts written by school students, we illustrate something of the power of the functional grammar in unpacking what it is to be successful for the purposes of school learning in subject English. To do this, we demonstrate some of the linguistic resources students deploy when they are evaluated as successful, and we suggest that an understanding of these resources should inform the teaching of English, providing knowledge about language that should constitute part of the overt objects of study for the future. First, however, we review three remaining models of subject English. Cultural Studies Later in the twentieth century, the specialized languages of English expanded further with post-modernism and a powerful relativizing of the dominant narratives of mainstream western literary traditions. Cultural Studies incorporates and valorizes the texts of popular culture, elevating them to the status of literary texts, and this model of subject English offers students a range of reading and subject positions. In her work on Critical Practice, for example, Belsey (2002) explores the notion of subjectivity in textual study and foregrounds the importance of interrogating texts from a variety of discursive reading positions. Here, the gaze implicitly constructed for learners is one of responding to and interrogating the subjective positions (values and ideologies) embodied in texts. Multiple readings of texts are invited, with an emphasis on critical analysis of texts for their dominant discourses and sociocultural effects. Like Personal

VERTICALITY IN SUBJECT ENGLISH

•

CHRISTIE & MACKEN-HORARIK

165

Growth, Cultural Studies takes the tools for interrogating texts largely for granted. There is no detailed attention to the linguistic tools necessary to produce and substantiate particular readings, and as we noted above for Cultural Heritage, a concern for the language in which meanings and values are realized is buried – in this case in the drive to tease out ideological positionings. Despite the continued significance attaching to the Cultural Heritage model already referred to, it is apparent that, at least at senior secondary levels of schooling, the Cultural Studies model is proving increasingly influential. In New South Wales, for example, there is a curious disjuncture between the junior secondary English curriculum whose influences appear to be mainly Cultural Heritage, and the senior secondary English curriculum, whose expectations for the ideal pedagogic subject position are quite other than those for the junior school, revealing the influence of Cultural Studies. Thus, the Higher School Certificate (HSC) in English for 2005–7 provides for a major Area of Study (NSW Board of Studies 2005), in which students are asked to study a disparate range of texts yoked together under the diffuse, abstract heading of Journeys. They can, for example, focus on Physical Journeys, Imaginative Journeys or Inner Journeys, and they receive a Stimulus Booklet (involving some loosely connected short texts and images about journeys) as well as a set of prescribed texts, both literary and nonliterary. These include Shakespeare’s The Tempest and Coleridge’s poems such as The Rime of the Ancient Mariner. In the case of Imaginative Journeys, for example, students must complete a written examination task in which they ‘compose a narrative which reflects the important ideas (they) have learnt from studying Imaginative Journeys’. Moreover, students must ‘demonstrate an understanding of the concept of journeys in the context of (their) studies’ and go on to write a response that ‘reflects the important ideas’ learnt; they must also use language ‘appropriate to audience, purpose and context’ and show ‘use of appropriate narrative techniques’. A study of texts produced in one privileged selective high school in New South Wales (Christie and Derewianka, in progress) reveals considerable confusion in the Year 12 students (who are regarded as talented) in attempting to deal with the task; what they write bears no discernible relationship to the texts identified, and it is difficult to establish how notions of ‘appropriate narrative techniques’ are understood, or indeed what are the ‘important ideas’ regarding journeys students might have learned. In addition, and compounding these problems, there is no evidence of any close attention to the language of the texts studied, for the curriculum guidelines given to teachers and students offer no advice about these matters. Overall, the tools needed to analyse and interpret the visual and verbal texts involved are not addressed, and Cultural Studies thus carries forward the by now well-established tradition of rendering invisible the linguistic skills students need to succeed in English. The pedagogic subject position is in one sense not unlike that associated with Personal Growth, since students are intended to respond in a ‘personally felt’ manner to the ‘themes’ selected for study, while the selection of texts used, though including some valued in the Cultural Heritage tradition, also reflects the general commitment of Cultural Studies to the view that a very catholic range of texts should be selected.

166

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Multiliteracies Multiliteracies is a development of the late twentieth century, and in some formulations it bears a close relationship to Cultural Studies, in that it proposes an interest in a wide variety of texts, verbal and visual, popular and otherwise, spoken and written, and its tendency is often to downplay the significance of a clear model of language. Its emergence is often associated with the manifesto of the New London Group, which appeared in the Harvard Educational Review in 1996, and the ideas involved later led to a volume on Multiliteracies (Cope and Kalantzis 2000). Some members of the New London Group have espoused a commitment to use of the functional grammar, while also arguing that the world of communication requires more than a theory of language (e.g. Kress 2002, 2003) to explain and teach about it. In practice, among many proponents of Multiliteracies the use of the functional grammar receives a very limited acknowledgement (e.g. Hawisher and Selfe 2000). Unsworth (2001) has offered the most comprehensive discussion of Multilieracies, which uses the functional grammar. Like others of his colleagues in the SF tradition (see Unsworth 2002) he refers to a range of ‘subject specific literacies’, a term intended to capture the particular claims on literacy represented by the different subjects found in the school curriculum. Together with colleagues (Unsworth et al. 2005) he has recently discussed the teaching of children’s literature and computer-based learning. Overall we suggest that for some proponents of Multiliteracies, this is best understood as a ‘dialect’ of Cultural Studies, while for others, like Unsworth, it is part of the Functional Language Studies model. Hence, two somewhat different pedagogic subject positions are possible, the one closer to that of Cultural Studies, the other to that of Functional Language Studies. New Literacy Studies The term New Literacy Studies (NLS) is one most associated with the work of Street (1984, 1993, 1997, 2001) who, working as an anthropologist, observed the uses of literacy, especially in non-western societies, and who argued that unless one understood the range of social practices that surrounded a use of literacy, it was impossible fully to understand it or to teach it effectively. Taking the term ‘literacy event’ from Heath (1983), he went on to argue that people bring to a literacy event ‘concepts, social models regarding what the nature of the event is, that make it work and give it meaning’ (Street 2001:11). He drew a distinction between what he termed conventional models of literacy that are ‘autonomous’ (involving teaching basic literacy skills in an unproblematic manner, to be simply recycled throughout life), and ‘ideological’ models (which see literacy as part of social practice and, as such, intimately bound up with values and meaning). According to Street and his colleagues, ‘autonomous’ models of literacy are unsatisfactory because they imply a singular or unitary model of literacy, whereas ‘ideological’ models recognize a diversity of literacy practices, which are ‘situated’. NLS attracted others including Gee (1996) and Barton (1994), and led to proposals that since ‘literacy practices are situated in broader social relations’ it was necessary ‘to describe the social setting of literacy events, including the ways in which social institutions support particular literacies’ (Barton, ibid.: 35). Anyone working in the SFL tradition, at least, has no difficulty acknowledging the

VERTICALITY IN SUBJECT ENGLISH

•

CHRISTIE & MACKEN-HORARIK

167

significance attaching to social processes, though the importance attaching to ‘literacy practices’ in the work of NLS theorists is nonetheless unfortunate, since it serves to deflect attention away from the language system, and its effect for pedagogic purposes is to marginalize language study. Hence, like both Personal Growth and Cultural Studies, the NLS tradition has served to bury the study of language itself, in this case because of the apparent imperative to focus on the social practices that surround the uses of literacy. The pedagogic subject in this tradition is one preoccupied with the immediate literacy practice, such that relatively little interest attaches to any sense of the linguistic system with which one engages in order to construct that practice. Muller (2001) offers a useful critique of NLS. Table 8.1 presents the models of English synoptically, though as noted, it does not present Multiliteracies separately. It distinguishes among the knowledge dimensions of English (the objects of study in the different models), knower or subject position dimensions (related to the ‘gaze’) and the linguistic practices by which these are created. All these models are present in contemporary school English, though some in rather shadowy ways. The difficulty of reading the requirements of the discipline is multiplied in this context, for teachers and students alike are often left uncertain of what is valued. Recognizing and realizing ‘what counts’ in examination English becomes an increasingly daunting task (Macken-Horarik 2006). Bernstein pointed to the difficulties of acquisition in horizontal knowledge structures such as English: Because a horizontal knowledge structure consists of an array of languages, any one transmission necessarily entails some selection, and some privileging within the set recontextualized for the transmission of the horizontal knowledge structure. The social basis of the principle of the recontextualizing constructs the perspective of the horizontal knowledge structure. Whose perspective is it? How is it generated and legitimated? (Bernstein 2000: 164, emphasis added)

We now turn our attention to the semantic features of students’ written texts and what a functional linguistic analysis of these can teach us about the kinds of knowledge, gaze and linguistic practices which are actually valued in examination English. Using the evidence of our analyses, we argue both that it is possible to make explicit a great deal of the linguistic resources needed to succeed, and that in making it explicit we take a critical step towards enhancing the verticality of subject English.

Unpacking the linguistic underpinnings of ‘finer feelings’ Despite the diversity and apparent openness of knowledge structures in subject English, careful attention both to the examination texts produced by students and their evaluation reveals which kinds of gaze, which forms of truth and which linguistic practices are actually privileged. In practice, we argue that value positions associated with the expression of sensibilities – empathy, discernment, ‘finer feelings’ – are particularly rewarded by teacher-examiners. However open the invited response may be, the preferred response is that associated with

Table 8.1

Models of school English in contemporary schools Functional Language Studies

New Literacy Studies

Analytical Categories

Basic Skills

Cultural Heritage

Personal Growth

Objects of study

Discrete skills for basic reading and a little writing (alphabet, spelling, parts of speech, grammatical analysis, paraphrasing).

Accessing sensibility and culture through engagement with ‘the canon’. Learning to read literary texts as embodiments of great and universal truths.

Processes of talking, listening, reading and writing, stimulated by texts and experiences. Learning to read and write as journeys of personal discovery.

Knowledge about spoken and written language; analysis and production of texts in a range of social contexts (through applications of functional grammar).

Multiple readings of different texts (spoken, written, literary, popular, multimodal). Critical analysis of texts for dominant discourses and socio-cultural effects.

Literacy events and their associated literacy practices (both spoken and written). Learning to read and write as ‘situated’ activities.

Subject positions or ‘gazes’

A well-regulated citizenry, inculcated with the basics; developing inner discipline, Christian values and personal responsibility.

A culturally valued sensibility – the preserve of élites who can understand (and interpret) great works in the received tradition.

A sensitive and self-conscious individuality – in dialogue with texts and with others who share their personal responses to texts and experience.

A linguistically informed analysis, production and critique of texts and contexts powerful in mainstream culture.

Interrogation of the subjective positions (ideologies) embodied in texts; questioning, critiquing, even subverting texts and their social conditions.

Investigation of social contexts in which literacy is used and of the ideological positionings and meanings involved.

Cultural Studies

Table 8.1

(continued)

Analytical Categories

Basic Skills

Cultural Heritage

Personal Growth

Linguistic practices

Use of language that builds discrete competences. * Texts such as spelling and grammar books, dictionaries, poems, composition books and biblical passages. (Basic literacy competences)

Use of language that interprets and creates fictive worlds and literary sensibilities. * Texts such as novel, poetry and drama; analytical texts such as interpretive essays. (Cultural literacy)

Use of language that is close to ‘self’. * Texts such as journals, dialogues, personal letters, imaginative reconstruction of literary texts, emails to friends. (Personal literacy)

Functional Language Studies Use of language that is analytical (metalinguistic) and well structured. * Texts such as narratives, expositions, multiliterate text, analysis and text interpretations. (Powerful meta literacies)

Cultural Studies Use of language that identifies and challenges ideologies enshrined in texts. * Texts such as deconstructive essays, evaluations, satires, spoofs. (Multimodal texts and multiliterate practices)

New Literacy Studies Use of language that is localized and particular to social events. * A variety of texts, valued for their relevance to immediate setting. (Situated literacy practices)

170

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Cultural Heritage English. The tendency to value texts which demonstrate control of the rhetoric of ‘finer feelings’ is clear, despite the apparent privileging of ‘Cultural Studies’ in contemporary curricula, such as the New South Wales senior syllabus for English, already referred to. In order to illustrate some of this, we turn to a text written by a school student in England as part of the national testing regime in that country. Text 8.1 was used by the Qualifications and Curriculum Authority in preparing advice to markers for the English Test for Key Stage 3, which is the last of the Key Stages identified under the terms of the National Curriculum in England. The Mark Scheme for Paper 1 for Key Stage 3 (Qualifications and Curriculum Authority 2002) gave a set of performance criteria which were illustrated using exemplars drawn from students’ written work produced in a pilot phase. Text 8.1 was recognized as ‘in the range above level 7’ (the highest level used), while it was given the highest mark available, namely a score of 33. (The student’s spelling and punctuation are preserved.) Text 8.1 As she eased her way past the leafy tendrils of the branches that beckoned to her; a sense of excited anticipation mixed with sad reflection filled her mind, accentuated by the tingling drops of rain that fell on to her head. Crystal drops. Each one glistened, as if crafted by a happy memory. The rocks, weather worn and covered in ancient moss, were silouetted (sic) by the smiling sun and they towered above Marie as she timidly stepped onto the path that led its winding way to the heart of the gorge. Closing her eyes now, she felt giddy with happiness; this was the place – her place; untouched the rocks and moss knew her story; the gorge had hidden depths; stories of times long forgotten. Marie didn’t know exactly where she was going, although she was certain of where she needed to be. Her thoughts became voices which echoed from the rocks, recalling the conversations that had filled her childhood. She glanced down at her hands and, feeling the wrinkles there, looked up at the rocky ways of the gorge in faint disgust. Gripping the box tightly, she continued through the living arch of trees. A tear rolled down her cheek; surprised to find it there, she brushed it aside. As the birds of the air agreed with her, their song resonating with joy, she found it. ‘It’ was a pool, quicksilver amongst the green; a treasure Marie’s mother had shown her, when she was a child. And there, beside the pool, was the rock her mother had always sat on. The air around it whispering and laughter ringing off the rocks. A glimmer of a smile caressed her cheeks. She laid the box down. A fragile flower graced the bank of the pool. Marie picked it and laid it by the box. ‘Goodbye’, she murmured. She left the box there, a treasure amongst treasures, and walked back into the gathering gloom.

VERTICALITY IN SUBJECT ENGLISH

•

CHRISTIE & MACKEN-HORARIK

171

The task was relatively open-ended. Students were to write about ‘someone going for the first time into a secret place – for example, a cave or an abandoned building’. The experience could be ‘real or imagined’ and students ‘could describe the reactions of the person as he or she goes into and explores this place’; they should ‘build up an atmosphere of tension and suspense . . . (and) include an effective ending’ (English Test 1 Paper 1. Key Stage 3 (Levels 4–7), 2002, QCA, page 9.) Text 8.1 was judged ‘a confidently written response which engages the interest of the reader through its selection of effective detail and control of language and structure to create a sense of place and mood’. The writing, it was said, ‘moves smoothly through the description of the scene and events of childhood, reflecting on the effects of the passage of time, both in terms of the changes in scenery and in the writer’s own responses’. The student is a skilled writer and uses language resources to build a sense of time past and of a nostalgic regret for a happy childhood. Let us look closely at the semantic choices made in this text, and the linguistic realizations of these. Semantically, the text creates a world of atmosphere and regret through dense nominal groups which provide physical descriptions of leaves and rain and rocks and elaborate on these in greater detail in embedded (or rank-shifted) clauses. These embedded clauses ‘plump out’ the nominal group through detailed elaborations of the information contained in its preceding head noun. Nominal groups are underlined below, head nouns in bold, and embedded clauses indicated through double square brackets, in the following examples: Example 1: as she eased past the leafy tendrils of the branches [[ that beckoned her]] Example 2: accentuated by the tingling drops of rain [[ that fell on her head]] Example 3: and there, beside the pool, was the rock [[ her mother had always sat on]]. A related feature of this student’s sentence structure is the use of enclosed clauses. Here the additional information is not part of what it is modifying (i.e. not embedded) but inserted into the middle of it. Halliday (1994: 227) calls these ‘enclosed clauses’ and represents them by means of angled brackets, as in: Example 4: the rocks << weather worn and covered in moss >> were silhouetted by the smiling sun Two other resources related to this student’s sentence structure are in evidence here. They include the use of non-finite clauses, often in thematic positions, and the occasional use of elliptical clauses. Perera (1984: 235–6) observed that non-finite clauses emerged in children’s writing at about age 12, though not all children will have mastered this by that age. When placed first (i.e. in theme position), non-finite clauses foreground the particulars of psychological and sensory experience, framing and deferring the message encoded in the main clause. Examples of non-finite clauses which are either thematic or enclosed include:

172

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Example 5: Closing her eyes now, she felt giddy with happiness (thematic) Example 6: Gripping the box tightly she continued through the living arch of the trees (thematic) Example 7: She glanced down at her hands and, << feeling the wrinkles there >> looked up at the rocky walls of the gorge in faint disgust (enclosed) As for the elliptical expressions, these provide particular emphasis in part because they are not full sentences. They serve to distil significance by leaving out elements of the clause that can be taken as given (and supplied by the active reader). Example 8: Crystal drops Example 9: Stories of time long forgotten An additional feature of the literate language of Text 8.1 is what Halliday calls elaboration. In elaboration one clause elaborates on the meaning of another by further specifying or describing it. The secondary clause does not introduce a new element into the picture but rather provides a further characterization of one that is already there, restating it, clarifying it, refining it, or adding a descriptive attribute or comment. The thing that is elaborated may be the primary clause as a whole, or it may be just some part of it – one or more of its constituents. (Halliday 1994: 225)

Using elaboration, the student is able to provide specific details about the natural phenomena which her character, Marie, observes in her journey back to her childhood ‘mecca’. Example 10: The rocks, weather worn and covered in ancient moss, were . . . Example 11: This was the place, her place, untouched . . . Elaboration and its related feature, apposition, is a defining feature of successful narratives and of highly valued interpretations of narrative (see MackenHorarik, in press for 2006, for extended discussion of this resource in a wider range of texts). We have briefly touched on three linguistic features – atmospheric nominal groups, thematic non-finite clauses, and elaboration – which are crucial to the control of literate language. These features are developmentally a feature of late childhood to adolescence. (Christie 2002; Derewianka 2003). They enable writers to distil the meaning of experience and to expand on this in the ways typical of ‘highly literate’ language. These are related to the genesis of another linguistic feature – that of metaphor (lexical and grammatical). Metaphor is the

VERTICALITY IN SUBJECT ENGLISH

•

CHRISTIE & MACKEN-HORARIK

173

gateway to alternative realms of experience and is crucial to success in English – in both reading and writing. In lexical metaphor, rhetorical transference is created through the creative doubling of one word or image with another. The reader is invited to construe experience both literally and figuratively. This transcoding of experience, construing the sun as ‘smiling’, the drops of rain as ‘crafted by a happy memory’, and bird song as ‘agreement’ with the character’s happiness, enables a writer to construe experience symbolically, even if the following examples are jarring in their self-conscious artifice: Example 12: The rocks were silouetted (sic) by the smiling sun Example 13: Each one (i.e. crystal drops) glistened, as if crafted by a happy memory Example 14: As the birds of the air agreed with her, their song resonating with joy . . . In the literary idiom adopted by this student, lexical metaphors serve not just an ideational function (through their transcoding of experience); they also serve an interpersonal function – provoking attitude in the reader through powerful, connotative realizations of attitude (see Martin and White 2005: 61–8, for discussion of the role of lexical metaphor within evaluation). The other kind of metaphor important in English (and other disciplines) is what Halliday has called ‘grammatical metaphor’ (Halliday 1994). This involves not a doubling of meanings (literal and figurative) but of grammatical forms (congruent and non-congruent). Here processes such as ‘to anticipate’ or ‘to reflect’ (congruently realized in verb forms) are encoded in nominal form, as in: Example 15: A sense of excited anticipation [[mixed with sad reflection]] filled her mind. The effect of the nominalization in this latter example is to create an abstraction – a sense of excited anticipation mixed with sad reflection. These non-congruent forms are good examples of grammatical metaphor – a resource which ‘unties texts from situations and allows writers to re-construe activities as things and thus break the iconic connections between linguistic and material activity’ (Martin and Rose 2003: 245). Using this resource enables writers to interpret experience from a ‘meta’ point of view, to abstract away from material activity through linguistic activity. What can we say about the gaze that is adopted in this text? Semantically, the student demonstrates an adult sensibility in his or her nostalgic regret for times past. The narrator adopts a subject position something like this: happy times occur in childhood in company with one’s mother, and old age is about looking back. The narrator’s voice is one of empathic engagement with the past, and this is projected through thoughts and feelings.

174

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Example 16: Marie didn’t know exactly where she was going, although she was certain of where she needed to be. Her thoughts became voices which echoed from the rocks, recalling conversations that had filled her childhood. Subject positions involving empathy and discernment are important if students are to adopt and demonstrate the appropriate gaze in English (Macken-Horarik 2003). A key linguistic resource for producing this kind of sensibility is appraisal, which is concerned with ‘the kinds of attitudes that are negotiated in a text, the strength of feelings involved and the ways in which values are sourced and readers aligned’ (Martin and Rose 2003: 22). In this text, as in others like it, the student encodes a particular reading of experience through the combination of inscribed affect (via attitudinal lexis) and invoked appraisal (via connotative renderings of experience). In this text, invoked affect is either flagged via non-core vocabulary to describe the character’s physical behaviour, or provoked by means of lexical metaphor (see Martin and White 2005, for a fuller treatment of this important domain of interpersonal meaning). The following examples display the combination of invoked affect through underlining and inscribed affect through bold face. Example 17: Each one glistened, as if crafted by a happy memory. Example 18: Closing her eyes now, she felt giddy with happiness. Example 19: Gripping the box tightly, she continued . . . The sensibility embodied in this narrative and valued by the examiners was not explicitly called for in the wording of the question. Furthermore, no fourteenyear-old might be expected to have had the necessary experience to produce Text 8.1. The sensibility rewarded here is far more likely to come out of exposure to fiction than out of lived experience of old age, loss and memory. The difficulty for many students (and their teachers), however, is that the QCA comments, while laudatory, do not identify the language resources used to produce the sensibility which is so highly valued. However, if indeed such texts and their values are to be rewarded in subject English, we argue that we should teach children the language skills necessary to access and produce the vertical discourse involved. From the point of view of Functional Language Studies, this means that we should make the forms of knowledge – the gaze and the linguistic resources that are valued – explicitly available to students, both by modelling and by analysis of examples (see Gray 1998, about a programme to do this for remote Indigenous children in Australia). In terms of the gaze rewarded in this situation, careful scrutiny reveals that what is required is an expression of the linguistic demands of Cultural Heritage English. Having examined an instance of a ‘literary’ narrative produced in examination conditions, we now turn our attention to the ubiquitous challenge, in both Australia and England, of the unseen text and the open question in response to it, to be answered in examination conditions.

VERTICALITY IN SUBJECT ENGLISH

•

CHRISTIE & MACKEN-HORARIK

175

Interpretation and the linguistic underpinnings of the symbolic reading of a literary text Like their English counterparts, Australian educators have adopted teaching and testing policies which reward the creation of alternative worlds and of ‘finer feelings’, though again within terms which obscure the language proficiency actually needed. As we shall see, the linguistic features of vertical discourse shift only slightly when we move from narrative to interpretation of narrative. Text 8.2 is taken from the NSW School Certificate Reference Test in English, 19955 – an examination which all sixteen-year-old students face at the end of Year 10. In this test, students were presented with a text adapted from a story called The Red-Back Spider by Peter Skryznecki. The story concerns a young migrant boy and his mother going to work clearing weeds in the garden of an Australian woman, Mrs Burnett, who appears to be a racist. Mrs Burnett shows no consideration for the pair and refuses to let the young boy (through whose eyes events are seen) play with toy soldiers which he finds while sheltering from the hot sun under her house. She also ignores the plight of the boy’s mother when she discovers a nest of dangerous red-back spiders in a tin while weeding. The story focuses on the young boy’s growing awareness of Mrs Burnett’s persecutory attitude to them and of his own mother’s emotional response to her rejection. It is a challenging narrative, made more difficult by the response task students were given. Although the story is called The Red-Back Spider, it ends with the words, ‘. . . I knew it was nothing to do with the spider’. What do you think the story is about? How does the writer make it an effective story?

It is not hard to imagine the difficulty of this question for any students who assumed (understandably) that the title of this story would be related to what it is about (i.e. about a spider). However, the wording of this question directs students to reconstrue the question as ‘What do you think the story is (really) about?’ This makes the whole process even more mysterious than it appears, or needs to be, for students (see Macken-Horarik 2006, for more extensive discussion). Text 8.2 – like others in the higher range of achievement – ignores the red herring of the question and produces a response in line with a Cultural Heritage agenda and the privileging of the symbolic reading (Macken-Horarik 2006). The text is presented below. Text 8. 2 ‘The Red Back Spider’ is a story about an old woman who finds it necessary to build a wall between herself and her worker, about a woman who is so selfish that she will not allow a young child to play with toys which have remained unused and unwanted for some time. The old woman lives alone, and feels like she needs to separate herself from the gardener and her child, ‘she closed the door slowly, deliberately, with a metallic click’. The story’s ending revolves around the burning of the spider and the sharp words of the old woman, ‘Please return all the toys and do not play with them’. However the sadness

176

LANGUAGE, KNOWLEDGE AND PEDAGOGY

of the child’s mother runs deeper than her fear of the spider. She is upset that this old woman, like the spider, hides her precious things away, and runs to protect them at any sign of danger. The fact that her child finds enjoyment in playing under the house and with blocks of wood in the dirt and weeds seems to show that they don’t have too many possessions. The mother is distressed at the scenario that the only enjoyment her son is getting in his school holidays has been cut off by the old woman. The separation between the two lifestyles, that of the lonely, selfish old woman, and the hardworking mother is the essence of the story. An old woman who is clinging to a past of child’s toys and a house, with no feelings for the people living in the present, as opposed to a tough, self-sufficient wife who struggles to provide for her family and misses her husband. The writer uses descriptive metaphorical language, ‘whose brown-as-leather skin hung over her frame like a synthetic material’, to enable to reader to picture the exact scene in their head. This brings the reader closer to the characters and the backyard of the house. The questions the child asks to themselves make the reader consider ‘was there something evil in it’s nature that it had to hide?’ This question seems as if it is being asked also about the old woman. The author shifts the focus of the story from the description of the house, to the history of the family, to the physical characteristics, and onto the child’s games and the discovery of the spider. This movement of the story at a reasonably quick speed keeps the reader interested. The story is effective because it brings the reader into the story, allows them to think about it, and moves at a reasonable pace. This interpretation of The Red-Back Spider was awarded an A+ (15 marks out of 20). It is typical of those in the higher range of achievement. Semantically, it embodies several choices which typify ‘verticality’ in the discourse of subject English. Foremost is control of symbolic abstraction. In this linguistic task, sensory experience (e.g. what happens in a story) is construed in non-sensible terms (the narrative significance of characters and events). In the highly valued literary interpretation, the student draws on a key lexical metaphor in the stimulus narrative to distil the abstract meaning of events. This symbolic reading is typically represented in dense nominal groups containing generalizations about the overarching significance of the text. Note the combination of dense nominal groups (underlined) and the sequence of three embedded clauses (in double square brackets) in the following interpretation of the character of the old woman in the story: Example 20: ‘The Red Back Spider’ is a story about an old woman [[ who finds it necessary to build a wall between herself and her worker, about a woman who is so selfish [[ that she will not allow a young child to play with toys [[which have remained unused and unwanted for some time]]]]]. Nominalizations which distil the organizing oppositions of the narrative are powerful resources in the service of abstraction, particularly when combined

VERTICALITY IN SUBJECT ENGLISH

•

CHRISTIE & MACKEN-HORARIK

177

with elaboration (indicated with an equals = sign). In the following example, the opposition between two lifestyles is identified and elaborated: Example 21: The separation between the two lifestyles (=) that of the lonely, selfish old woman, and the hardworking mother is the essence of the story. In this task, the student needs to discern key metaphors (such as the spider) and use these to organize the response (contra the confusing wording of the examination question). The semantic strategy underpinning this symbolic reading involves making an extended analogy between the spider and some other aspect of the narrative. In this interpretation, the student relates the spider to the old woman and interprets her actions through the symbolic mesh of this analogy: Example 22: She (the mother) is upset that this old woman, like the spider, hides her precious things away, and runs to protect them at any sign of danger. All A-range responses in this corpus (approximately 10) focused sharply on the spider motif and created connections between this and other parts of the narrative. In short, they ignored the confusing lead given in the question (see Macken-Horarik 2006, for discussion of some of these other responses). What kind of gaze is created here? As in Text 8.1, the sensibility projected is one of empathy (feeling with the main character), realized primarily by expressions of affect, as in the following two examples: Example 23: The mother is distressed that the only enjoyment her son is getting has been cut off by the old woman. Example 24: However the sadness of the child’s mother runs deeper than her fear of the spider. However, in this text there is also a strong dose of judgement, especially negative judgement of the old woman, Mrs Burnett. As in the earlier narrative text, a literary sensibility is created through a combination of direct and indirect expressions of attitude. Tokens of judgement (indirectly invoked through non-core lexis) are underlined while explicit (directly inscribed through attitudinal lexis) inscriptions of judgment are in bold. Example 25: . . . about a woman who is so selfish . . . Example 26: . . . an old woman who is clinging to a past of child’s toys and a house, with no feelings for the people living in the present, as opposed to a tough, self-sufficient wife who struggles to provide for her family. In other words, empathy is combined with discernment in this response, tuning into the way the story narrativizes prejudice. Being able to read and re-enact both emotional sensitivity to characters’ feelings and ethical judgement of their behaviour is crucial to the gaze in English, and often rewarded within Cultural

178

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Heritage models of the subject (both in the disciplinary and the personal sense of the word ‘subject’). Table 8.2 presents an overview of the most salient linguistic features of the two texts we have considered, along with relevant examples.

Table 8.2 Text 1

Semantic features of Texts 1 and 2 with examples

Narrative

Text 2

Literary interpretation

Abstraction (via dense nominal groups and embedding) • The leafy tendrils of the branches [[that beckoned her]] • the tingling drops of rain [[that fell on her head]] • the rock [[her mother had always sat on]]

• a woman [[who finds it necessary to build a wall between herself and her worker]] • the separation between the two lifestyles, that of the lonely old woman and the hardworking mother is . . .

Metaphor (lexical and grammatical) • The rocks were silhouetted by the smiling sun • Each one glistened, as if crafted by a happy memory • The birds of the air agreed with her. • Her thoughts became voices which echoed from the rocks

• However the sadness of the child’s mother runs deeper than her fear of the spider • She is upset that his old woman, like the spider, hides her precious things away and runs to protect them at any sign of danger.

Elaboration and apposition • The rocks, weather worn and covered in ancient moss • This was the place, her place, untouched ... • the gorge had hidden depths, stories of times long forgotten . . .

• This is a story about a woman [[who finds it necessary to build a wall . . .]] about a woman who is . . . • the separation between the two lifestyles, that of the lonely old woman and the hardworking mother is . . .

Appraisal (via various resources that build affect or judgement) • Each one glistened as if crafted by a happy memory • Closing her eyes now, she felt giddy with happiness • Gripping the book tightly she continued ...

• She is upset that this old woman hides her precious things away • The mother is distressed • About a woman who is selfish • An old woman with no feelings for the people living . . .

VERTICALITY IN SUBJECT ENGLISH

•

CHRISTIE & MACKEN-HORARIK

179

Conclusion: building a metalanguage for enhancing verticality in subject English Bernstein (2000: 155–74) was clear that horizontal knowledge structures, marked by ‘weak grammars’ as he suggested, tend to produce discourses that are segmentally organized, and that ‘there is no necessary relation between what is learned in the different segments’ (ibid.: 159). Our discussion has revealed a range of different knowledge structures (models) and their associated gazes which have emerged over the history of subject English, such that their various languages are often incommensurate, even contradictory. Paradoxically, and partly as a result of the dilemmas caused by a history that has been so segmented, as school subject English has evolved, its tendency has been to diminish the status of overt teaching of knowledge about language, while promoting various valued subject positions, though the tools for their expression have become increasingly invisible. This is especially clear in examinations of students’ written responses to open questions and unseen texts in the later years of junior secondary English (age 16 and up). We have argued that the metalanguage of SFL has proved useful in exploring the linguistic character of texts valued in many contexts of instruction and evaluation in English, and we have sought to demonstrate this by reference to two valued texts from England and Australia. In making the features of these texts visible through the SF metalanguage we are able to make subject English more visible. English teachers have access to a metalanguage through which to articulate requirements, and hence to make it more accessible to a greater number of students. In its work within Functional Language Studies, functional linguistics has created links across ‘segments’ of communication, and this has helped create mechanisms for identifying and integrating the diverse specialized languages of various subjects with horizontal knowledge structures (see Martin, this volume). As we have noted, SFL gives us tools for connecting meanings to wordings. The wording choices of successful students in English give evidence of a particular orientation to meaning (gaze) which is summarized in Table 8.2. The orientations which are highly valued include preferences for abstraction (realized in use of dense nominal groups, nominalization and embedded clauses), for restatement or reformulation (realized in elaboration and apposition), for lexical and grammatical metaphor, and for empathy (realized in direct and indirect expressions of affect) and ethical discernment (realized in direct and indirect expressions of judgement). The metalanguage of SFL can be used to highlight commonalities and differences across the specialized languages and methodologies of English, and we suggest it can be used to outline a scope and sequence for progressing learning in English at both primary and secondary levels. The work of doing this has already begun (e.g. Christie and Derewianka, in progress; Christie, MackenHorarik and Maton, in preparation), though more needs to be done to develop a coherent, theoretically rigorous account of the English curriculum from Kindergarten to Year 12. At this stage, students are often forced to intuit requirements as they progress up the years, without access to mechanisms for deciding which responses to which texts are going to be valued. SFG provides us

180

LANGUAGE, KNOWLEDGE AND PEDAGOGY

with tools for translating across, and comparing, texts and contexts, showing what they share linguistically and how their covert fashions of meaning are related to one another. In this sense, functional linguistics offers a new set of lenses for creating a new gaze, for making the subject at least partly visible, and hence for strengthening the discipline.

References Arnold, M. (1869), Culture and Anarchy. An Essay in Political and Social Criticism. London, Edinburgh and NY: Thomas Nelson and Sons. ‘Australia’s Language. The Australian Language and Literacy Policy’ (1991), Canberra: Australian Government Publishing Service. Barton, D. (1994), Literacy. An Introduction to the Ecology of Written Language. Oxford UK and Cambridge USA: Blackwell. Belsey, Catherine (2002), Critical Practice (second edn). London and NY: RoutledgeFalmer. Bernstein, B. (1990), The Structuring of Pedagogic Discourse. Class, Codes and Control, Vol. IV. London and NY: Routledge. Bernstein, B. (2000), Pedagogy, Symbolic Control and Identity. Theory, Research, Critique (rev. edn). Lanham, Boulder, NY and Oxford: Rowman and Littlefield. Board of Studies (1999), English Stage 6 Syllabus. Sydney: New South Wales. http://www. boardofstudies.nsw.edu.au (accessed on 30 December 2005) Britton, J. and Squires, J. (1975), ‘Foreword’, in Dixon, J., 1975, pp. vii–xviii. ‘Bullock Report, The’ (1975), A Language for Life. Report of the Committee of Inquiry appointed by the Secretary of State for Education and Science under the Chairmanship of Sir Alan Bullock. London: HMSO. Cathcart Borer, M. (1976), Willingly to School. A History of Women’s Education. Guildford and London: Lutterworth Press. Christie, F. (1976), ‘The teaching of English in elementary schools in New South Wales 1848–1900: an inquiry into social conditions and pedagogical theories determining the teaching of English’. Unpub. MEd Thesis, University of Sydney. Christie, F.G. (2002), ‘The development of abstraction in adolescence in subject English’, in M.J. Schleppegrell and M.C. Colombi (eds), Developing Advanced Literacy in First and Second Languages. Meaning with Power. Mahwah, NJ and London: Erlbaum, pp. 45–66. Christie, F. (1993), ‘The “received tradition” of English teaching: the decline of rhetoric and the corruption of grammar’, in B. Green (ed.), The Insistence of the Letter. Literacy Studies and Curriculum Theorizing (Critical Perspectives on Literacy and Education). London and Washington: Falmer Press, pp. 75–106. Christie, F. (2003), ‘English in Australia’, in RELC Journal 34, 1, pp. 100–19. Christie, F. (2004a), ‘The study of language and subject English’, in Australian Review of Applied Linguistics 27, 1, pp. 15–29. Christie, F. (2004b), ‘Revisiting some old themes: the role of grammar in the teaching of English’, in J.A. Foley (ed.), Language, Education and Discourse. Functional Approaches. London and NY: Continuum, pp. 145–73. Christie, F. (2004c), ‘Authority and its role in the pedagogic relationship of teaching’, in L. Young and C. Harrison (eds), Systemic Functional Linguistics and Critical Discourse Analysis. Studies in Social Change. London and NY: Continuum, pp. 173–201. Christie, F. (2005), Language Education in the Primary Years. Sydney: University of New South Wales Press. Christie, F., Gray, P., Gray, B., Macken, M., Martin, J. and Rothery, J. (1990), Exploring Procedures: Books 1–4 and Teachers Book. Sydney: Harcourt Brace Jovanovich.

VERTICALITY IN SUBJECT ENGLISH

•

CHRISTIE & MACKEN-HORARIK

181

Christie, F., Gray, P., Gray, B., Macken, M., Martin, J. and Rothery, J. (1992a), Exploring Explanations: Books 1–4 and Teachers Book. Sydney: Harcourt Brace Jovanovich. Christie, F., Gray, P., Gray, B., Macken, M., Martin, J. and Rothery, J. (1992b), Exploring Reports: Books 1–4 and Teachers Book. Sydney: Harcourt Brace Jovanovich. Christie, F. and Soosai, A. (2000), Language and Meaning 1. Melbourne: Macmillan Education. Christie, F. and Soosai, A. (2001), Language and Meaning 2. Melbourne: Macmillan Education. Christie, F. and Unsworth, L. (2005), ‘Developing dimensions of an educational linguistics’, in R. Hasan, C. Matthiessen and J.J. Webster (eds), Continuing Discourse on Language. A Functional Perspective, Vol.1. London and Oakville: Equinox, pp. 217–50. Christie, F. (in press), ‘Literacy and current debates over reading’, in R.Whittaker, M. O’Donnell and A. McCabe (eds), Language and Literacy: Functional Perspectives. London and NY: Continuum. Cope, B. and Kalantzis, M. (eds) (2000), Multiliteracies. Literacy Learning and the Design of Social Futures. South Yarra, Victoria: Macmillan. Cox National Curriculum English Working Group, The (1989), English for Ages 5 to 16: Proposals of the Secretary of State for Education and Science and the Secretary of State for Wales. London: Department of Education and Science and the Welsh Office. Davies, F. (1973), Teaching Reading in Early England. London: Pitman Publishing. Derewianka, B. (1990), Exploring How Texts Work. Sydney: PETA. Derewianka. B. (1998), A Grammar Companion for Primary Teachers. Sydney: PETA. Derewianka, B. (2003), ‘Grammatical metaphor in the transition to adolescence’, in A.M. Simon-Vandenbergen, M. Taverniers and L. Ravelli (eds), Grammatical Metaphor. Amsterdam and Philadelphia: Benjamins, pp. 185–220. Dixon, J. (1967), Growth through English. London: National Association for the Teaching of English and Oxford University Press. Dixon, J. (1975), Growth through English (third edn). London: National Association for the Teaching of English and Oxford University Press. Droga, L. and Humphrey, S. (2002), Getting Started with Functional Grammar. Berry, NSW: Target Texts. Droga, L. and Humphrey, S. (2003), Grammar and Meaning. An Introduction for Primary Teachers. Berry, NSW: Target Texts. ‘English: a Statement for Australian Schools’. A Joint Project of the States, Territories and the Commonwealth of Australia initiated by the Australian Education Council (1994). Carlton, Victoria: Curriculum Corporation. ‘English Curriculum and Standards Framework’, Victoria. (http://www.vcaa.vic.edu.au/ prep10/csf/ accessed 30 December 2005) Fitch, J. (1880), Lectures on Teaching. Cambridge: CambridgeUniversity Press. Gee, J.P. (1996), Social Linguistics and Literacies: Ideology in Discourses (second edn). London: Taylor and Francis. Goldstrom, J.M. (1972), Education: Elementary Education 1780–1900. (David and Charles Sources for Social and Economic History). Devon: David and Charles. Gosden, P.H.J.H. (1969), How They were Taught. An Anthology of Contemporary Accounts of Learning and Teaching in England 1800–1950. Oxford: Blackwell. Gray, B. (1998), ‘Accessing the Discourses of Schooling: English Language and Literacy Development with Aboriginal Children in Mainstream Schools’. Unpublished PhD thesis, University of Melbourne. Halliday, M.A.K. (1994), An Introduction to Functional Grammar (second edn). London: Edward Arnold. Halliday, M.A.K., McIntosh, A. and Strevens, P. (1964), The Linguistic Sciences and Language Teaching. London: Longmans. Halliday, M.A.K. and Hasan, R. (in press), ‘Retrospective on systemic functional linguis-

182

LANGUAGE, KNOWLEDGE AND PEDAGOGY

tics and literacy’, in Whittaker, R., O’Donnell, M. and McCabe, A. (eds), Language & Literacy: Functional Approaches. London and Washington: Continuum. Hasan, R. and Martin, J.R. (1989), Language Development: Learning Language, Learning Culture. Meaning and Choice in Language: Studies for Michael Halliday. (Advances in Discourse Processes Series, Vol. XXVII). Norwood, NJ: Ablex. Hasan, R., Matthiessen, C. and Webster, J.J. (2005), Continuing Discourse on Language. A Functional Perspective, Vol. 1. London and Oakville: Equinox. Hawashir, G.E. and Selfe, C.LK. (2000), Global Literacies and the World Wide Web (Literacies Series). London and NY: Routledge. Heath, S.B. (1983), Ways with Words. Cambridge: Cambridge University Press. Hunter, I. (1994), Rethinking the School. Subjectivity Bureaucracy Criticism. St Leonards, NSW: Allen and Unwin. Kachru, B.J. (1990), ‘World Englishes and applied linguistics’, in M.A.K. Halliday, J. Gibbons and H. Nicholas (eds), Learning Keeping and Using Language, Vol. 11. Amsterdam and Philadelphia: Benjamins, pp. 203–29. ‘Kingman Report, The’ (1988), Report of the Committee of Inquiry into the Teaching of English Language. Committee chaired by Sir John Kingman. London: HMSO. Kress, G. (2000), ‘Multimodality’, in Cope and Kalantzis, pp.182–202. Kress, G. (2003), Literacy in the New Media Age (Literacies Series). London and NY: Routledge. Leavis, F.R. (1933), For Continuity. Cambridge: The Minority Press. Leavis, F.R. (1934), ‘Introduction’, in Determinations. Critical Essays. London: Chattus and Windus, pp. 1–9. Leavis, F.R. (1965), The Common Pursuit. London: Chatto and Windus. Leavis, F.R. and Thompson, D. (1933), Culture and Environment. The Training of Critical Awareness (this edition 1964). London: Chatto and Windus. ‘Literacy for all, the Challenge for Australian Schools’ (1998). Canberra: Commonwealth Department of Employment, Education and Training. Macken-Horarik, M. (1998), ‘Exploring the requirements of critical school literacy: a view from two classrooms’, in F. Christie and R. Misson (eds), Literacy and Schooling. London and NY: Routledge, pp. 74–103. Macken-Horarik, M. (2002), ‘“Something to shoot for”: a systemic functional approach to teaching genre in secondary school science’, in A.M. Johns (ed.), Genre in the Classroom. Multiple Perspectives. Mahwah, NJ: Erlbaum, pp. 17–42. Macken-Horarik, M. (2003), ‘Appraisal and the Special Instructiveness of Narrative’, in Macken-Horarik, M. and Martin, J. R. (eds), Text 23, 2, pp. 285–312. Macken-Horarik, M. (2004), ‘Interacting with the Multimodal Text: reflections on image and verbiage in ArtExpress’, in Visual Communication 3 (1), pp. 5–26. Macken-Horarik, M. (2006), ‘Recognizing and realizing “what counts” in examination English: Perspectives from systemic functional linguistics and code theory’, in Functions of Language 13, 1; pp. 1–35. Martin, J.R. and Rose, D. (2003), Working with Discourse. Meaning beyond the Clause. London and NY: Continuum. Martin, J.R. and Rose, D. (2006), Genre Relations. London and Oakville: Equinox. Martin, J. R. and White, P.R.R. (2005), The Language of Evaluation: Appraisal in English. Hampshire: Palgrave Macmillan. Maybin, J. (2000), ‘The canon. Historical construction and contemporary challenges’, in J. Davison and J. Moss (eds), Issues in English Teaching. (Issues in Subject Teaching Series). London and NY: Routledge, pp. 180–95. Muller, J. (2000), Reclaiming Knowledge. Social Theory, Curriculum and Education Policy. London: RoutledgeFalmer. Muller, J. (2001), ‘Intimations of boundlessness’, in A. Morais, I. Neves, B. Davies and H. Daniels (eds), Towards a Sociology of Pedagogy. The Contribution of Basil Bernstein to Research. (History of Schools and Schooling Series, Vol. 23). NY: Peter Lang, pp. 129–52.

VERTICALITY IN SUBJECT ENGLISH

•

CHRISTIE & MACKEN-HORARIK

183

Muller, J. (2004), ‘On the shoulders of giants: a digression on knowledge, curriculum, and finally, the teacher’. A plenary paper given at the Conference on Reclaiming Knowledge: Registers of Discourse in the Community and School, held at the University of Sydney, 13–15 December 2004. Murray, L. (1795, 1968), English Grammar. Facsimile Reprint No.106. Menston: Scolar Press. ‘National English Curriculum, The’ (1995), London: Department for Education and Science and HMSO. ‘Newbolt Report, The’ (1921), Report on the Teaching of English in England: being the Report of the Departmental Committee appointed by the President of the Board of Education to inquire into the Position of English in the Educational System of England. London: HMSO. New London Group (1996), ‘A pedagogy of multiliteracies: designing social futures’, in Harvard Educational Review 66, pp. 60–92. New South Wales Board of Studies Senior English Syllabus. http://www.education.tas. gov.au/english/sylrev.htm, accessed 7 March 2006. Peel, R., Patterson, A. and Gerlach, J. (2000), Questions of English. Ethics, Aesthetics, Rhetoric, and the Formation of the Subject in England, Australia and the United States. London and NY: RoutledgeFalmer. ‘Plowden Report, The’ (1967), Children and their Primary Schools. London: HMSO. Stenhouse, L. (1975), An Introduction to Curriculum Research and Development. London: Heinemann Educational Books. Street, B. (1984), Literacy in Theory and Practice. Cambridge: Cambridge University Press. Street, B. (ed.) (1993), Cross-cultural Approaches to Literacy. Cambridge, NY and Oakleigh, Victoria: Cambridge University Press. Street, B. (1997), ‘The implications of the “New Literacy Studies” for literacy education’, in English in Education, Vol. 321, No. 3, pp. 45–59. Street, B. (ed.) (2001), Literacy and Development. London and NY: Routledge. Sutherland, G. (ed.) (1973), Arnold on Education. London and Melbourne: Penguin Education. Unsworth, L. (ed.) (2000), Researching Language in Schools and Communities. Functional Linguistic Perspectives. London and Washington: Cassell. Unsworth, L. (2001), Teaching Mutiliteracies across the Curriculum. Changing Contexts and Images in Classroom Practice. Buckingham and Philadelphia: Open University Press. Unsworth, L., Thomas, A., Simpson, A. and Asha, J. (2005), Children’s Literature and Computer Based Teaching. Maidenhead: Open University Press.

Notes 1

2

3

4

5

The authors acknowledge the assistance of Joan Rothery who contributed to an early draft of this paper which was delivered at a conference on ‘Reclaiming Knowledge’ at the University of Sydney in December 2004. English is a major international language, spoken by millions. Significant policy statements attach to the teaching of English in the many countries in which it is not the mother tongue, though they are not our concern. This discussion concerns the ‘Inner Circle’ countries (Kachru, 1990: pp. 203–29) of English speakers, though it draws most on the experience of England and Australia. Maybin, 2000, p. 184, says the Newbolt Report was the culmination of the work of the English Association, formed in 1907, though this does not in itself diminish the significance of the First World War as a contributing factor. As this book went to press, an Australian National Inquiry into the Teaching of Literacy, had produced a Report on Teaching Reading, released in January 2006. http://www.dest.gov.au/nitl/report.htm (accessed 8 March 2006) The nature of this test has changed since 1995, though similar tasks are still used.

9

Vertical and horizontal discourse and the social sciences Peter Wignell

Introduction Bernstein (1999) distinguishes between what he refers to as vertical and horizontal discourses. He defines a vertical discourse as follows: ‘a vertical discourse takes the form of a coherent, explicit, and systematically principled structure, hierarchically organised, as in the sciences, or it takes the form of a series of specialised modes of interrogation and specialised criteria for the production and circulation of texts, as in the social sciences and humanities’ (Bernstein 1999: 159). In contrast: ‘a horizontal discourse entails a set of strategies which are local, segmentally organised, context specific and dependent, for maximising encounters with persons and habits’ (Bernstein 1999: 159). In addition to vertical and horizontal discourses, Bernstein also characterizes vertical discourses as having either hierarchical knowledge structures (the physical sciences) or horizontal knowledge structures (the humanities and social sciences) (Bernstein 1999: 162). Bernstein represents a hierarchical knowledge structure as a triangle:

And he states: ‘This form of knowledge attempts to create very general propositions and theories, which integrate knowledge at lower levels, and in this way shows underlying uniformities across an expanding of apparently different phenomena. Hierarchical knowledge structures appear, by their users, to be motivated towards greater and greater integrating propositions, operating at more and more abstract levels. Thus, it could be said that hierarchical knowledge structures are produced by an “integrating” code’ (Bernstein 1999: 162). In contrast Bernstein (1999: 162) describes horizontal knowledge structures as consisting of ‘a series of specialised languages with specialised modes of interrogation and criteria for the construction of specialised texts. Thus, any one of the specialised disciplines within the form of a horizontal knowledge structure found within the humanities can be visually portrayed as: L1 L2 L3 L4 L5 L6 L7 . . . Ln’ (Bernstein 1999: 162). To continue the quote:

VERTICAL AND HORIZONTAL DISCOURSE

•

WIGNELL

185

Thus, in the case of English literature, the languages would be the specialised languages of criticism; in Philosophy, the various languages of this mode of inquiry and in Sociology . . . the languages refer, for example, to functionalism, post-structuralism, post-modernism, Marxism, etc. (Bernstein 1999: 162)

Bernstein’s classification of the physical sciences as hierarchical knowledge structures and the humanities as horizontal knowledge structures can be calibrated quite readily with systemic functional work on the discourses of the physical sciences and humanities. Hierarchical knowledge structures more or less correlate with discourses which establish an elaborate technical framework as a means of categorizing and interpreting the world. The physical sciences are a prime example. Horizontal knowledge structures more or less correlate with discourses which use abstraction (see Martin, 1997) as their major resource for interpretation, such as the humanities. The case of the social sciences, however, is not so clear. Wignell (1997) proposes that the discourse of social science is different in kind from both the physical sciences and the humanities but, because of its origins and evolution, shares some features of both. In a very brief summary of the conclusions of a substantial body of research, the physical sciences use technicality as a primary resource for naming, categorizing, analysing and interpreting the world, whereas the humanities use abstraction as their principal resource. The social sciences are both abstract and technical, having evolved over several hundred years as a discourse which had its origins in the abstract humanities discourse of moral philosophy (how people should behave), but over time that abstract discourse evolved into one which is both technical and abstract, resulting in the social sciences (how humans do behave). Wignell (1990, 1992, 1997, 1998) argues that the language of what came to be the social sciences evolved as a hybrid of the languages of the humanities and of the physical sciences, with, over time, the language patterns of the physical sciences taking a more and more prominent role. This suggests that the social sciences are, to varying degrees, likely to be hierarchical rather than horizontal knowledge structures. Wignell’s (1997) work certainly implies this. This work analysed the orthodoxy of three social science disciplines – economics, sociology and political science – as presented to undergraduate students. In analysing the orthodox this work did not examine what happens to unorthodox voices within the disciplines. It is in focusing on the orthodox that a conclusion that the social sciences are hierarchical knowledge structures might be favoured. In an effort to resolve this issue, what I propose to do in this chapter is to present a diachronic perspective on the social sciences and then refer back to Bernstein’s analysis. What I hope to demonstrate by using a number of samples from several canonical texts from the history of social science is the building over time of a hierarchical knowledge structure which then ruptured into a number of smaller knowledge structures. That is, what are now known as the social sciences began as and evolved as a single hierarchical knowledge structure for over two hundred years, and then experienced a Tower of Babel syndrome towards the end of the nineteenth century. What I hope to discover from my analysis and discussion is what kind of knowledge structures resulted: did what

186

LANGUAGE, KNOWLEDGE AND PEDAGOGY

became the various social sciences then evolve into horizontal knowledge structures, or continue as smaller-scale hierarchical knowledge structures? I will be basing my discussion on examples from some key texts in the history of social science. The texts are Thomas Hobbes’ Leviathan (1651), Adam Smith’s The Wealth of Nations (1776), David Ricardo’s Principles of Political Economy and Taxation (1817) and Karl Marx’s Capital, Volume One (1867).

The evolution of the social sciences What became the social sciences was originally called civil philosophy by Thomas Hobbes in the mid seventeenth century. Then, from the late eighteenth century to around the end of the nineteenth century, the discipline that evolved into the social sciences was called political economy. What I hope to demonstrate is that political economy grew as a hierarchical knowledge structure until Marx. After Marx a process of fragmentation set in. Part of the discipline of political economy continued on under another name as Marxism. The rest of it either broke into or spawned other major social science disciplines such as sociology, economics and political science. What is recognizable as social science, or the application of the discourse of science to discussion of the social, appears to have begun with Thomas Hobbes. The extract below is from Hobbes’ Leviathan (1651). It is an example of a discourse of the same kind as the example from Newton’s Treatise on Opticks (published in 1704; written 1675–87). Halliday (1988) treats Newton’s Treatise on Opticks as ‘registering the birth of scientific English’. Hobbes was writing in a similar style a quarter of a century before Newton’s work was written. The two examples are far from identical. The example from Newton is a description of a set of experiments. Hobbes’ example is an explanation, written in a ‘this is how things are’ manner without experiments as verification. This is a difference in method rather than a difference in kind. I argue that the two examples are close enough in kind to be considered as belonging to the same discourse. Extract One: ‘The causes of sense’ – Hobbes (1651: 85) The cause of sense, is the Externall Body, or Object, which presseth the organ proper to each sense, either immediately, as in the Tast and Touch, or mediately, as in Seeing, Hearing, and Smelling which pressure, by the mediation of Nerves, and other strings, and membranes of the body, continued inwards to the Brain, and Heart, causeth there a resistance, or counter pressure, or endeavour of the heart, to deliver it self: which endeavour because Outward, seemeth to be some matter without. And this seeming, or fancy, is that which men call sense; and consisteth, as to the Eye, in a light, or Colour figured; To the Eare, in a sound; To the Nostrill, in an Odour; To the Tongue and Palat, in a Savour; And to the rest of the body, in Heat, Cold, Hardnesse, Softnesse, and such other qualities, as we discern by feeling. All of which qualities called Sensible, are in the object that causeth them, but so many several motions of the matter, by which it presseth the organs diversly.

VERTICAL AND HORIZONTAL DISCOURSE

•

WIGNELL

187

Extract Two: Newton 1704 (from Halliday 1988) I found moreover, that when Light goes out of air through several contingous refracting Mediums as through Water and Glass, and thence goes out again into Air, whether the refracting Superfices be parallel or inclin’d to one another, that Light as often as by contrary Refractions ’tis so corrected, that it emergeth in Lines parallel to those in which it was incident, continues ever after to be white. But if the emergent Rays be inclined to the incident, the Whiteness of the emerging Light will be by degrees in passing on from the Place of Emergence, becoming tinged in its Edges with Colours. This I try’d by refracting Light with Prisms of Glass placed within a Prismatick Vessel of Water. Now those Colours argue a diverging and separation of the heterogeneous Rays from one another by means of their unequal Refractions, as in what follows will more fully appear. And, on the contrary, the permanent argues, that in like Incidences of the Rays, and by consequence no inequality of their whole Refractions. Whence I seem to gather the two following theorems.

What Hobbes did was apply the emerging discourse of the physical sciences to discussions of the social, which, up until then, had been the domain of the humanities, notably moral philosophy. Hobbes did not adopt the discourse of science and apply it to the social for its own sake: Hobbes was not writing a treatise on sense. His goal was similar to that of moral philosophers: to instruct people on how they should behave. Hobbes’ main task in the first half of Leviathan is to present an irrefutable argument for the need for sovereign government. The example above comes from Chapter 1 of Leviathan, which establishes one of the fundamental premises of Hobbes’ argument: that humans are analogous to bodies in motion. In trying to arrive at first principles Hobbes needs to establish the argument that the primary cause of motion in humans is sensory input. In paraphrase the argument runs along the following lines: sensory input creates appetites and aversions in individuals. These are unlimited and, in a state of nature, each individual has complete liberty to act on them. The inevitable result of this is a ‘warre of all against all’, or total anarchy. This state is intolerable, therefore some form of sovereign government is needed to keep people away from each other’s throats. Incidentally, references to a ‘state of nature’ are a recurring theme in Enlightenment discourse. This hypothetical state is meant to be the state humans lived in between the Flood and when they became ‘civilised’. Each writer who refers to it tends to construct it for his own purposes: for Thomas Hobbes it is a hell on earth, for John Locke it is a liberal property owner’s paradise, for Adam Smith it is an entrepreneur’s Eden, and for Karl Marx it is a prototypical communist utopia. The ‘science’ in Hobbes is there to further the argument. But why? We need to look at this in the context of the times. Hobbes was writing at a time of social turmoil in England. He was trying to convince the people who could do something about it to do something about it. I will not speculate on whether or not Hobbes chose to use the discourse of the physical sciences simply because it

188

LANGUAGE, KNOWLEDGE AND PEDAGOGY

would further his cause best, or because he believed in it. Either way it was a voice that people of influence were likely to listen to. The methods of the physical sciences were, in Hobbes’ time, in the process of completely overturning prevailing notions about the nature of the universe. The combination of observation, experimentation and calculation had provided (or almost provided in Hobbes’ time) proof of a uniform, solarcentred cosmos governed by universal, mathematically derived physical laws. The emerging discourse of the physical sciences was, even then, a very powerful discourse, not yet quite ‘naturalised’ or hegemonic (although it was shortly after, post Newton), but it was a voice with status that was listened to, especially by the emerging power brokers, the bourgeoisie. The physical sciences had status and were gaining more: science was a discourse on the ascent. It was ascending not only because it could prove things but also, and just as importantly, because it could, through its application to technology, produce things. Therefore, in using a discourse most likely to convince, Hobbes made the logical choice. In economic terms he adopted the discourse with the most purchasing power. Hobbes lived in the early (or late-early) stages of what is often referred to as the scientific revolution. This revolution did not happen in isolation, but was part of a broader sweep of social change. In terms of class, the bourgeoisie was ascending in economic and political power. In conjunction with this revolution, liberal, individualistic and utilitarian ideas were ascending in philosophy and capitalism was in its early stages. Both material and semiotic relations were shifting in favour of the bourgeoisie. The social class that derived its wealth and prestige from manufacture and trade had more to gain from the application of science to technology than the class that relied on ownership of land as its wealth and power base. Science, however, was not only in the interests of the bourgeoisie. Especially in its applications, science had direct and indirect benefits for all of those who were dominant in society at the time: the aristocracy, the church, the military, the emerging capitalists, even intellectuals and scientists themselves. These interests can be summed up in six words: discover, conquer, plunder, convert, civilize and trade. Discover and conquer happen sequentially. The rest happen concurrently. Improvements in shipbuilding, navigation and ballistics enabled relatively large numbers of well-armed men to be moved around the world to more places, and more quickly, than ever before. What emerged was a very powerful discursive alliance between the established and the emerging because there was something in it for all of them. Hobbes did not change completely the way the social was written about. Philosophers continued, after Hobbes, to write like philosophers, not like physical scientists. Hobbes did, however, introduce a possibility, a new choice. He negotiated the potential for a new discourse to evolve. This choice was taken up 125 years after Leviathan by Adam Smith. Adam Smith’s The Wealth of Nations (1776) marks the second beginning of the social sciences with the discipline of political economy. By the time Adam Smith was writing in the late eighteenth century, the discourse of physical science was well developed and virtually the only ‘recognised’ paradigm for explaining the

VERTICAL AND HORIZONTAL DISCOURSE

•

WIGNELL

189

physical world. In the broader social context, the bourgeoisie had attained political dominance, the philosophical climate (in England) was liberal, individualistic and utilitarian, and early capitalism (pre-large-scale industrialism) was the dominant economic practice. Hobbes, therefore, cannot be given sole credit for influencing Smith. Had the physical sciences not been the dominant paradigm for interpreting the world there would have been little point in drawing on Hobbes. Likewise, had self-interested individualism and utilitarianism not been the dominant themes of British philosophy, they would not necessarily have been taken up by Smith as assumptions about human behaviour. There are, however, identifiable legacies of Hobbes in Smith. As well as direct citation, the use of the ‘state of nature’ and mechanistic, self-interested assumptions about human nature, if not originating with Hobbes, were given explicit voice by him. Smith’s reasoning in the early chapters of The Wealth of Nations is very similar to Hobbes’ reasoning in Leviathan until Smith begins to build field. In building field, Smith uses the same set of resources as are used in the physical sciences. Adam Smith began the building of the hierarchical knowledge structure that evolved into the social sciences of today, and he did it while trying to work out one of humanity’s fundamental questions: what are things really worth? In doing this he used Hobbes as his point of departure through the following quote: The Value, or Worth of a man, is as in all other things, his Price; that is to say, so much as would be given for the use of his power: and therefore is not absolute; but a thing dependant on the need and judgement of another. Wealth, as Mr Hobbes says, is power. (Hobbes, T. 1651, cited in Smith, A. 1776: 26)

Smith then immediately continues by linking the citation to Hobbes to his own work: But the person who either acquires, or succeeds to a great fortune, does not necessarily acquire or succeed to any political power, either civil or military . . . The power which that possession immediately and directly conveys to him, is the power of purchasing: a certain command over all the labour, or over all the produce of labour, which is in the market . . . (Smith, A. 1776: 26)

I will take up this question of value and look at how it is treated in examples from the works of three political economists: Adam Smith, David Ricardo and Karl Marx. These examples will demonstrate the emergence of social science as a small but growing hierarchical knowledge structure. What happened, in brief, was that Smith established a technical, theoretical framework, using much the same resources as those used by the physical sciences. This framework was challenged, critiqued, modified and added to by Ricardo. Both Smith and Ricardo were then challenged, critiqued, modified and added to by Marx. What is important in this critique and modification is that they all used the same language. The critique is from within the field and is used to alter and build the field.

190

LANGUAGE, KNOWLEDGE AND PEDAGOGY

In the first four chapters of The Wealth of Nations a number of extended lexical motifs are developed. Three of these are crucial. There is a motif of trade: barter, buy, sell, exchange, trade, money, price, market, commerce, goods, commodities. There is a motif of labour: toil, work, labour, division of labour. There is a motif of quantification: amount, number, quantity, value. These motifs are realized both congruently and metaphorically but, as they continue, the metaphorical forms are the preferred realization and can thus be regarded as abstractions. At the end of Chapter 4 Smith makes a major shift. Some of the abstractions from these lexical motifs are distilled into technicality. This pivotal point marks both the beginning of Western economic theory as a technical discipline, and the new beginning of what evolved into the social sciences. Extract Three: Adam Smith 1776: 24 It is in this manner that money has become in all civilised countries the universal instrument of commerce, by the intervention of which goods of all kinds are bought and sold, or exchanged for one another. What are the rules which men naturally observe in exchanging them either for money or for one another, I shall now proceed to examine. These rules determine what may be called the relative or exchangeable value of goods. The word value, it is to be observed, has two different meanings, and sometimes expresses the utility of some particular object, and sometimes the power of purchasing other goods which the possession of that object conveys. The one may be called ‘value in use’; the other, ‘value in exchange’. The things which have the greatest value in use have frequently little or no value in exchange; and, on the contrary, those which have the greatest value in exchange have frequently little or no value in use. In order to investigate the principles which regulate the exchangeable value of commodities, I shall endeavour to show, First, what is the real measure of this exchangeable value; or wherein consists the real price of all commodities. Second, what are the different parts of which this real price is composed or made up. And, lastly, what are the different circumstances which sometimes raise some or all of these different parts of price above, and sometimes sink them below their natural or ordinary rate; or, what are the causes which sometimes hinder the market price, that is, the actual price of commodities, from coinciding exactly with what may be called their natural price.

In this extract Smith for the first time explicitly does what the physical sciences do. He defines technical terms, begins to arrange them taxonomically, and says that he is about to explain how things work. This emerging technical framework is to be deployed in the search for semiotic abstractions: principles, and institutional abstractions: rules. Or, in Bernstein’s terms, ‘very general propositions and theories’ (Bernstein 1999: 162). The following taxonomy of value can be drawn up from the passage above:

VERTICAL AND HORIZONTAL DISCOURSE

•

WIGNELL

191

value

value in use

value in exchange (exchangeable value)

real price (natural price)

Figure 9.1

market price (actual price)

Taxonomy of value

It is this initial technicality which is used for analysis in the rest of the chapters in Book One of The Wealth of Nations. The next chapter, Chapter 5, integrates the lexical motif of labour into this new technical framework by arguing that labour is the true measure of value:

Extract Four: Smith 1776: 26–9 First, is a justification for why labour is used as the measure of value. ‘The value of any commodity, therefore, to the person who possesses it, and who means not to use or consume it himself, but to exchange it for other commodities, is equal to the quantity of labour which it enables him to purchase or command. Labour, therefore, is the real measure of the exchangeable value of all commodities.’ ‘What is bought with money or with goods is purchased by labour as much as what we acquire by the toil of our own body. That money or those goods indeed save us this toil. They contain the value of a certain quantity of labour which we exchange for what is supposed at the time to contain the value of an equal quantity. Labour was the first price, the original purchase-money that was paid for all things.’ Second, the difficulty of seeing labour as the underlying measure of value. ‘But though labour be the real measure of the exchangeable value of commodities, it is not that by which their value is commonly estimated.’ Next, money as the conventional measure of value. ‘Every commodity, besides, is more frequently exchanged for, and thereby compared with, other commodities than with labour. It is more natural, therefore, to estimate its exchangeable value by the quantity of some other commodity than by that of the labour which it can purchase.’ ‘But when barter ceases, and money has become the common instrument of commerce, every particular commodity is more frequently exchanged for money than for any other commodity.’

192

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Then, money itself is variable in value, therefore it is not an accurate measure of value. ‘Gold and silver, however, like every other commodity, vary in their value, are sometimes cheaper and sometimes dearer, sometimes of easier and sometimes of more difficult purchase.’ ‘Labour alone, therefore, never varying in its own value, is alone the ultimate and real standard by which the value of all commodities can at all times and places be estimated and compared. It is their real price; money is but their nominal price only.’ Apparent variation in the value of labour is an illusion. ‘But though equal quantities of labour are always of equal value to the labourer, yet to the person who employs him they appear sometimes to be of greater and sometimes of smaller value. He purchases them sometimes with a greater and sometimes with a smaller quantity of goods, and to him the price of labour seems to vary like that of all other things. In reality, however, it is the goods which are cheap in the one case, and dear in the other. In this popular sense, therefore, labour, like commodities, may be said to have a real and a nominal price. Its real price may be said to consist in the quantity of the necessaries and conveniences of life which are given for it; its nominal price, in the quantity of money. The labourer is rich or poor, is well or ill rewarded, in proportion to the real, not to the nominal price of his labour.’

It should be noted that much of this argument is carried out using the technical framework that has been established previously. For example, exchangeable value, real price and nominal price (synonymous with market price) are all used in the argument. This technical framework is interwoven with the motif of measuring through the use of nominal group structure. The actual price [at which any commodity is commonly sold for] Value

is called

its market price. Token (Smith, A. 1776: 49)

How market price is determined is then explained. At the same time new technical terms (natural price, effectual demand and absolute demand) are introduced. Extract Five: Building technicality (from Smith, A. 1776: 49) The market price of every particular commodity is regulated by the proportion between the quantity which is actually brought to the market, and the demand of those who are willing to pay the natural price of the commodity, or the whole value of the rent, labour, and profit, which must be paid in order to bring it thither. Such people may be called the effectual demanders, and their demand the effectual demand; since it may be sufficient to effectuate the bringing of the commodity to market. It is different from the absolute demand.

VERTICAL AND HORIZONTAL DISCOURSE

•

WIGNELL

193

Compare the extract above with an example of science discourse from about the same time. Extract Six: Comparison extract. From Priestley’s The History and Present State of Electricity, with Original Experiments (1757) (Taken from Halliday 1988) When the equilibrium of this fluid in any body is not disturbed; that is, when there is in any body neither more nor less of it than its natural share, or than that quantity which it is capable of retaining by its own attraction, it does not discover itself to our senses by any effect. The action of the rubber upon an electric disturbs this equilibrium, occasioning a deficiency in one place, and a redundancy of it in another.

Smith uses much the same grammatical resources as those used by Priestley. The extract from Smith parallels the discourse of the physical sciences of its own time. I will now skip forward 41 years to David Ricardo’s Principles of Political Economy and Taxation (1817). By Ricardo’s time industrial capitalism had become the dominant mode of production, distribution and exchange in England. Enlightenment, liberal and utilitarian philosophical trends were fully established and the bourgeoisie were becoming dominant in politics and economics. Ricardo takes Smith to task on what should be the measure of value. What Smith introduced at the end of a chapter and then developed, Ricardo now picks up as his point of departure. It has been observed by Adam Smith that ‘the word Value has two different meanings, and sometimes expresses the utility (of some particular object), and sometimes the power of purchasing other goods which the possession of that object conveys. The one may be called value in use; the other value in exchange. The things’, he continues, ‘which have the greatest value in use, have frequently little or no value in exchange; and, on the contrary, those which have the greatest value in exchange, have little or no value in use.’(Ricardo, D. 1817: 5)

Ricardo then moves into dispute with Smith’s analysis. Ricardo uses an extensive quote from Smith as his point of departure. The quote outlines Smith’s position that the value of something is determined by labour: ‘The real price of everything,’ says Adam Smith, ‘what everything really costs to the man who wants to acquire it, is the toil and trouble of acquiring it. What everything is really worth to the man who has acquired it, and who wants to dispose; of it, or exchange it for something else, is the toil and trouble, which it can save to himself, and which it can impose upon other people.’ ‘Labour was the first price – the original purchase-money that was paid for all things.’ (Ricardo, D. 1817: 6–7)

Here Ricardo contends that Smith erred in equating ‘the quantity of labour bestowed on the production of any object’ and ‘the quantity which it can

194

LANGUAGE, KNOWLEDGE AND PEDAGOGY

command in the market’. Ricardo argues these two are different and that, in terms of what it can ‘command in the market’, labour varies in value. From this point Ricardo presents his alternative analysis of exchangeable value. This is principally done by supporting the argument by alternative explanations. For example: In the same country double the quantity of labour may be required to produce a given quantity of food and necessaries at one time than may be necessary at another and a distant time; yet the labourer’s reward may possibly be very little diminished. If the labourer’s wages at the former period were a certain quantity of food and necessaries, he probably could not have subsisted if that quantity had been reduced. Food and necessaries in this case will have risen 100 per cent, if estimated by the quantity of labour necessary to their production, while they will scarcely have increased in value if measured by the quantity of labour for which they will exchange. ... It cannot then be correct to say with Adam Smith, ‘that as labour may sometimes purchase a greater and sometimes a smaller quantity of goods, it is their value which varies, not that of the labour which purchases them;’ and therefore, ‘that labour, alone never varying in its own value, is alone the ultimate and real standard by which the value of all commodities can at all times and places be estimated and compared;’ — but it is correct to say, as Adam Smith had previously said, ‘that the proportion between the quantities of labour necessary for acquiring different objects seems to be the only circumstance which can afford any rule for exchanging them for one another;’ or in other words that it is the comparative quantity of commodities which labour will produce that determines their present or past relative value, and not the comparative quantities of commodities which are given to the labourer in exchange for his labour. (Ricardo, D. 1817: 8–9)

The second paragraph is Ricardo’s conclusion to his refutation of Smith’s analysis. In his refutation Ricardo argues and concludes by using much the same resources as were used by Smith. In arguing with Smith, Ricardo uses the same language as Smith. What he is doing is fine-tuning Smith’s taxonomies and explanations. A combination of technicality and abstraction carries the experiential weight of the argument. Value is not in dispute, but how it is measured is. Ricardo concludes that money should be used to measure value. Suppose money to be that commodity. If a salmon were worth $1 and a deer $2, one deer would be worth two salmon. But a deer might become of the value of three salmon, for more labour might be required to obtain the deer, or less to get the salmon, or both these causes might operate at the same time. If we had this invariable standard, we might easily ascertain in what degree either of these causes operated. If salmon continued to sell for $1 whilst deer rose to $3, we might conclude that more labour was required to obtain the deer. If deer continued at the same price of $2 and salmon sold for 13s 4d, we might then be sure that less labour was required to obtain the salmon; and if deer rose to $2 10s and

VERTICAL AND HORIZONTAL DISCOURSE

•

WIGNELL

195

salmon fell to 16s 8d, we should be convinced that both causes had operated in producing the alteration of the relative value of these commodities. (Ricardo, D. 1817: 16)

We now skip forward another 50 years to Karl Marx. When Marx was writing industrial capitalism was in full swing as the dominant mode of production, distribution and exchange in England. In Capital, Volume One (1867), right at the beginning, in Chapter One, Marx challenges Ricardo and Smith. Marx both uses and builds on the technicality of the discipline. He begins by making explicitly technical a term used as a generic abstraction but never fixed by definition by Smith and Ricardo: a commodity. Six pages of Chapter 1 are taken up by argumentation which arrives at a final definition of a commodity. The argumentation derives from the field as already established, and in its course reorganizes the taxonomic relations of part of the field. Marx first defines a commodity as: A commodity is, in the first place, an object outside us, a thing that by its properties satisfies human wants of some sort or another. (Marx, K. 1867: 43)

Marx then uses the theory of value to assign attributes to a commodity: The utility of a thing makes it a use value. But this utility is not a thing of air. Being limited by the physical properties of the commodity, it has no existence outside that commodity. A commodity, such as iron or corn, or a diamond, is therefore, so far as it is a material thing, a use-value, something useful.(Marx, K. 1867: 44)

Marx then introduces quantity and quality and uses these to rework the existing theory of value to arrive at a different definition of a commodity. Marx reworks the taxonomy established by Smith and assumed by Ricardo: value

use value

Figure 9.2

exchange value

Taxonomy of value

The arguments Marx presents in reworking this taxonomy are somewhat complex. In paraphrase, using the notions of quality and quantity, Marx reverses the proposition, so that at first sight it appears that use value is quantitative and that exchange value, also quantitative, is derived from the exchange of different amounts of use value, that ‘exchange-value, at first sight, presents itself as a quantitative relation, as the proportion in which values in use of one sort are exchanged for those of another sort’. Marx proceeds to eliminate use value as a measure of value. He does this by arguing that, in order to be exchanged for one another, commodities must have

196

LANGUAGE, KNOWLEDGE AND PEDAGOGY

something in common which can be used as a measure. He argues that different uses represent different qualities, not quantities, therefore use cannot measure value. Still, in order to be exchanged, they must have something in common. Since commodities all have different uses, are all different, but are exchanged for relative quantities of each other, they must have something in common besides their use value. That ‘something’ is external to the qualities of commodities themselves (use value). This external must be ‘a total abstraction from their use-value’. This ‘something’ is their exchange value. Use value is seen as a qualitative relation, while exchange value is seen as a quantitative relation which does ‘not contain one atom of use-value’. Use value is thus excluded from any estimation of how much of one commodity will exchange for another. What all commodities have in common is that they are the ‘products of labour’. This is not any particular kind of labour but ‘human labour in the abstract’. The taxonomy in Figure 9.2 above is rearranged so that value is no longer in a relation of hyponymy to the co-hyponyms use value and exchange value. Use value is removed, and value and exchange value become synonymous. The following quote illustrates this: . . . there is nothing left but what is common to them (commodities) all; all are reduced to one and the same sort of labour, human labour in the abstract. Let us now consider the residue of each of these products; it consists of the same insubstantial reality in each, a mere congelation of homogenous human labour, of labour power expended without regard to its mode of expenditure. All that these things tell us is, that human labour power has been expended in their production, that human labour is embodied in them. When looked at as crystals of this social substance, common to all of them, they are – Values. (Marx, K. 1867: 46)

Here Marx also ‘solves’ a previous problem: that of different kinds of labour and different intensities of labour over time being ‘worth’ more or less than each other. By excluding use value as a qualitative relation Marx also excludes the different qualities of labour. As labour is averaged, so commodities are averaged: ‘each individual commodity, in this connexion, is to be considered as an average sample of its class’ (Marx 1867: 47). Thus: as values,

all commodities

Circumstantial role

Token

are

only definite masses of congealed labour-time. Value (Marx, K. 1867: 47)

This involves a distillation of several preceding pages into a redefinition of commodities. A distinction between products and commodities is then introduced. Things produced for one’s own use are not regarded as commodities because these things do not enter into any relationship with other things: they

VERTICAL AND HORIZONTAL DISCOURSE

•

WIGNELL

197

are not exchanged. Taking this addition into account, a final definition of commodities concludes this section of the chapter. This definition is accomplished by enhancement in a clause complex: x

β To become a commodity

α a product must be transferred to another, who it will serve as a use-value by means of an exchange

Token

Value (Marx, K. 1867: 48)

This final definition of a commodity is then used as the basis for Marx’s analysis of labour and of exchange value in the next two sections of the chapter. The term labour was a continuing motif in the works of Smith and Ricardo. From Smith can be inferred a distinction between labour as the most superordinate abstraction and work as a more subordinate abstraction through his different lexical choices when referring to labour and work. This distinction is not formalized: the words are simply used in different contexts, not defined. Neither does Ricardo make a formal distinction. Marx undertakes a formal analysis of labour using the technical framework he has reordered in the previous section. That is, Marx builds the field by making technical things that were latent earlier in the discourse (commodities and labour). In analysing labour Marx uses the same method as he used to analyse commodities. He takes up a point introduced in the previous section, human labour in the abstract, labour power. He distinguishes concrete (specific) labour from labour in the abstract and extracts qualitative differences in labour from calculations of exchange value (value) just as he extracted use value from the determination of value. Concrete labour, it is argued is, like use-value, qualitative. Marx (49) distinguishes concrete/useful labour from abstract labour by defining it: the labour, [[whose utility is thus represented by the value in use of its product]] or [[which manifests itself by making its product a use-value]]

we

Value

Assigner

call

useful labour.

Token

This qualitative difference of labour allows exchange to happen. Things are not exchanged for things that are qualitatively the same: ‘coats are not exchanged for coats, one use-value is not exchanged for another of the same kind’ (Marx, K. 1867: 49). The above is restated more technically two paragraphs later:

198

LANGUAGE, KNOWLEDGE AND PEDAGOGY

In the use-value of each commodity there is contained useful labour, ie productive activity of a definite kind and exercised with a definite aim. Use values cannot confront each other as commodities unless the useful labour embodied in them is qualitatively different in each of them. (Marx, K. 1867: 49)

Useful labour is then excluded: Productive activity, if we leave out of sight its special form, viz., the useful character of labour, is nothing but the expenditure of human labour-power. Tailoring and weaving, though qualitatively different productive activities, are in each a productive expenditure of human brains, nerves, and muscles, and in this sense are human labour. They are but two different modes of expending human labourpower . . . It is the expenditure of simple labour-power, ie., of the labour power which, on an average, apart from any special development, exists in the organism of every ordinary individual . . . Skilled labour counts only as simple labour intensified, or rather, as multiplied simple labour, a given quantity of skilled being considered equal to a greater quantity of simple labour . . . A commodity may be the product of the most skilled labour, but its value, by equating it to the product of simple unskilled labour, represents a definite quantity of the latter labour alone. (Marx, K. 1867: 51)

Marx has theorized labour so that it can be used to measure value (which he has also redefined): Coats and linen, however, are not merely values, but values of a definite magnitude, and according to our assumption, the coat is worth twice as much as the ten yards of linen. Whence this difference in values? It is owing to the fact that the linen contains only half as much labour as the coat, and consequently, that in the production of the latter, labour-power must have been expended during twice the time necessary for the production of the former. While, therefore, with reference to use-value, the labour contained in a commodity counts only qualitatively, with reference to value it counts only quantitatively, and must first be reduced to human labour pure and simple. (Marx, K. 1867: 52)

So far Marx has specified the field specific meanings of two terms, commodities and labour, previously regularly used but undefined and untheorized, and has both redefined value and reordered the taxonomy of kinds of value. Having defined a commodity and proposed a measure for how to measure its value, Marx now uses the new developed technical framework to analyse the relativities of exchanging one commodity for another. To borrow an analogy from Systemic Functional Linguistics, the relationship is akin to the relationship between Token and Value in a relational identifying clause: one thing represents another. This is embodied in the following:

VERTICAL AND HORIZONTAL DISCOURSE

•

WIGNELL

199

A. Elementary or Accidental form of value x commodity A = y commodity B, or x commodity A is worth y commodity B. 20 yards of linen = 1 coat, or 20 yards of linen are worth 1 coat. Here two different kinds of commodities (in our example the linen and the coat), evidently play two different parts. The linen expresses its value in the coat; the coat serves as the material in which that value is expressed. The former plays an active, the latter a passive, part. The value of the linen is represented as relative value, or appears in relative form. The coat officiates as equivalent, or appears in equivalent form. (Marx, K. 1867: 55)

coat = Token, linen = Value This relationship between two commodities is the basis of exchange-value. The next form expands this to include all commodities: B. Total or Expanded form of value z Com A = U Com B or v Com C or W Com D or x Com E or &c (20 yards of linen = 1 coat or = 10 lbs tea or = 40 lbs coffee or = 1 quarter corn or = 2 ounces gold or = 1/2 ton iron or = &c The value of a single commodity, linen, for example, is now expressed in terms of numberless other elements of the world of commodities. Every other commodity now becomes a mirror for linen’s value. It is thus, that for the first time, this value shows itself in its true light as a congelation of undifferentiated human labour. (Marx, K. 1867: 68)

Marx argues that the first form of value could result from accident but that this expanded form could not. It is put up as proof of labour as the measure of value. From this Marx reaches a conclusion which reverses the conclusions of Smith and Ricardo: It becomes plain, that it is not the exchange of commodities which regulates the magnitude of their value; but, on the contrary, that it is the magnitude of their value which controls their exchange proportions. (Marx, K. 1867: 69)

Marx points out practical problems with this form of value; potentially every commodity stands in relation to every other commodity. Everything is a potential Token and a potential Value. This leads to the next form:

200

LANGUAGE, KNOWLEDGE AND PEDAGOGY

C. The General form of value 1 coat 10 lbs of tea 40 lbs of coffee 1 quarter of iron = 20 yards of linen 2 ounces of gold 1/2 a ton of iron x com A, &c 1.The altered character of the form of value All commodities now express their value (1) in an elementary form, because in a single commodity; (2) with unity, because in one and the same commodity. (Marx, K. 1867: 70)

Here the roles of Token and Value are reversed, and linen becomes the Token through which the values are realized. From here it is a small step to the final form of value, the money-form: The Money-form 20 yards of linen 1 coat 10 lbs of tea 40 lbs of coffee = 2 ounces of gold 1 qr of corn 1/2 ton of iron x commodity A In passing from Form A to Form B, and from the latter to Form C, the changes are fundamental. On the other hand, there is no difference between forms C and D except that, in the latter gold has assumed the equivalent form in the place of linen. Gold is in Form D, what linen was in Form C – the universal equivalent. The progress consists in this alone, that the character of direct and universal exchangeability – in other words, that the universal equivalent form – has now, by social custom, become finally identified with the substance, gold. (Marx, K. 1867: 75)

Gold has become the universal Token. While Smith and Ricardo both argued that money itself had value, Marx treats any value money might have as irrelevant. Money now signifies, or represents, value: it is the Token that represents value.

VERTICAL AND HORIZONTAL DISCOURSE

•

WIGNELL

201

Summary and conclusions Having laboured through Hobbes, Smith, Ricardo and Marx, what have we found out? What is illustrated is the slow building over a period of 200 years of a hierarchical knowledge structure. Hobbes plants the seed, Smith begins the theory, Ricardo and Marx refine it, rework it and add to it. In doing so they use the language of the field that they are building. What they have done appears to satisfy Bernstein’s criteria for a hierarchical knowledge structure: This form of knowledge attempts to create very general propositions and theories, which integrate knowledge at lower levels, and in this way shows underlying uniformities across an expanding of apparently different phenomena. Hierarchical knowledge structures appear, by their users, to be motivated towards greater and greater integrating propositions, operating at more and more abstract levels. Thus, it could be said that hierarchical knowledge structures are produced by an ‘integrating’ code. (Bernstein 1999: 162)

Smith outlines the theory, arriving at very general propositions: propositions that are intended to apply over all cases and all time. Ricardo works through Smith’s theory and finds what he thinks are a couple of propositions that don’t quite work. He doesn’t throw the theory out or use a different language to talk about it. He uses the same kind of language and argument that Smith used, and uses it to refine the theory: he tries to make its general propositions work better. Marx then takes the level of generality and abstraction further than either Smith or Ricardo. Marx used his economic analysis to build an integrated social theory. However, with Marx the discipline of political economy came to a fairly abrupt halt. After Marx, arguments about and analysis of value just about disappear from capitalist economics: the focus is on price. If you look in just about any standard Western economics textbook you will find mention, often along with pictures, of canonical figures such as Smith, Ricardo, and even Keynes, but no mention, and certainly no picture, of Marx. What happened after Marx? What I speculate happened with Marx was that his economic analysis was so flawless and his political/social analysis so threatening to those wielding power at the time that the best way for them to deal with him was simultaneously to ignore him and turn him into a demon. Ignore his economics, and demonize his politics. The discipline called political economy ended with Marx, but the social sciences didn’t. Political economy spawned a large family. Political economy itself didn’t hit a dead end. It underwent a name change to Marxism and continued as an integrated theory of the economic and the social. Marxism wasn’t only a theory. Attempts at the practical application of Marxism changed the world in ways very few social theories have ever done. Leaving Marxism aside, the social science disciplines of economics, political science and sociology all emerged in the West from the splintering of political economy. Orthodox capitalist economics left Marx out of its pantheon and went back to Smith as the founder of the discipline, taking on the economy part of

202

LANGUAGE, KNOWLEDGE AND PEDAGOGY

political economy. Political science picked up on the political part, and sociology drew on both parts and added some. Sociology, through Durkheim, has a direct link back to Adam Smith. Durkheim’s canonical work The Division of Labour in Society (1893) picks up on, theorizes and makes technical the term the division of labour, which is discussed at length in Smith but not incorporated into his technical/theoretical framework. Durkheim’s mechanical solidarity also has echoes of notions of a state of nature and thus can be linked back to Hobbes, Smith, Ricardo and Marx as well as to philosophers like Locke. The social sciences had a lineage that was an emerging hierarchical knowledge structure. Following the rupture of political economy, did the various social science disciplines simply become smaller hierarchical knowledge structures or did they, as Bernstein argues, become horizontal knowledge structures? My answer is that both options were possible and that, depending on circumstances, both paths were taken to varying degrees. Despite these different paths, in their underlying grammatical and discourse semantic patterns the various disciplines of the social sciences do represent the science of the social. Wignell (1997) shows that the social science disciplines of economics, sociology and political science use the same set of linguistic resources to identify and define technical terms, arrange those terms into taxonomic frameworks, use those terms to explain phenomena, and from there build further technicality. These patterns are both similar to and different from the physical sciences and can be traced back to being derived from a hybrid of the language of the physical sciences and the humanities. As it evolved, social science became a discourse which is both technical and abstract. However, the balance between the abstract and the technical differs across the social sciences. That is, the balance between the influence of the language of the physical sciences and the language of the humanities alters. Of the three social sciences analysed by Wignell (1997), the language of orthodox Western economics was the most technical and the most like the language of the physical sciences. The language of sociology was also highly technical but relied less on quantification than economics. Political science, although still technical, was much less so than either economics or sociology. Since the language of the social sciences evolved as a hybrid of the language of the physical sciences and the language of the humanities, there is always a kind of dynamic tension between the science and the social in the discourse. What happened after the fragmentation of the discipline of political economy is complex and, I speculate again, stems from the dynamic tension between the science and the social and from the balance of power between the science and the social among the practitioners of each of the social sciences. Following from Wignell’s (1997) discussion, orthodox Western economics is clearly an example of a hierarchical knowledge structure in the social sciences. It has continued to evolve from the seeds planted by Smith and Ricardo. The discipline of economics has an elaborate technical framework which appears to satisfy Bernstein’s criteria for a hierarchical knowledge structure. The case of a discipline such as sociology is less clear. Wignell (1997) examined sociological orthodoxy as laid out in undergraduate textbooks. These

VERTICAL AND HORIZONTAL DISCOURSE

•

WIGNELL

203

textbooks showed the discipline to be an example of a hierarchical knowledge structure according to Bernstein’s criteria. However, a textbook presenting orthodoxy to undergraduates cannot afford to spend a lot of time on dissenting voices. I believe that the key to whether a social science discipline evolves as a hierarchical knowledge structure, a horizontal knowledge structure, or some blend of the two, depends on what happens with dissenting voices. I hypothesize that the stronger the boundaries around a discipline and the stronger the concord within that discipline, then the easier it will be for that discipline to prevent outside voices from intruding and for it to expel or re-educate internal dissenting voices. These circumstances would foster the evolution of a hierarchical knowledge structure within that discipline since less time and effort would be used up in dealing with internal bickering and attacks from outside. With the risks of revolution and invasion reduced, and with relative concord among insiders, there is more energy available for hierarchical knowledgebuilding. This would be the case with a discipline like orthodox Western economics. Conversely, the weaker the boundaries around a discipline (the more it overlaps with other disciplines), the more it listens to other voices, and the more dissenting voices there are within that discipline, then the more likely it is that it will evolve as either a horizontal knowledge structure or as a blend of hierarchical and horizontal. I would place sociology as a blend of hierarchical and horizontal. In its orthodoxy it is a hierarchical knowledge structure, but it is a discipline with both internal dissent and open to many influences. For example, as Bernstein (1999: 162) points out, there are Marxist sociologists, post-structuralist sociologists, and a variety of other sociologists. The more contending voices there are, then the more disagreement there is likely to be even on fundamentals. In such cases it is likely that one strategy to combat orthodoxy will be each faction inventing a new language. This would result in the knowledge structure of the discipline over time becoming less hierarchical and more horizontal.

References Bernstein, B. (1999), ‘Vertical and Horizontal Discourse: an essay’, in British Journal of Sociology of Education, Vol. 20, No. 2. London: Taylor and Francis. Durkheim, E. (1893), The Division of Labour in Society (De la division du travail social) (1964 reprinting). New York: The Free Press. Halliday, M.A.K. (1988), ‘On the language of Physical Science’, in Halliday, M.A.K. and Martin, J.R. (eds) (1993), Writing Science: literacy and discursive power. London: Falmer. Halliday, M.A.K. (1987), ‘Language and the Order of Nature’, in Halliday, M.A.K. and Martin, J.R. (eds) (1993), Writing Science: language and discursive power. London: Falmer. Hobbes, T. (1651), Leviathan (1986 reprinting). Harmondsworth: Penguin. Locke, J. (1690), ‘On Property’, in Macpherson, C.B. (ed.) (1980), John Locke: Second Treatise of Government. Indianapolis: Hackett. Martin, J. R. (1997), ‘Analysing Genre: functional parameters’, to appear in Christie, F. and Martin, J.R. (eds), Genres and Institutions: social processes in the workplace and school, pp. 3–39. London: Cassell.

204

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Marx, K. (1867), Capital, Volume One (1974 reprinting). Moscow: Progress Publishers. Ricardo, D. (1817), Principles of Political Economy and Taxation (1973 reprinting). London: Dent. Smith, A. (1776), The Wealth of Nations (1964 reprinting). London: Dent. Wignell, P. (1998), ‘Technicality and Abstraction in Social Science’, in Martin, J.R. and Veel, R. (eds), Reading Science. London: Routledge. Wignell, P. (1997), ‘Making the Abstract Technical: on the evolution of the discourse of social science’. PhD thesis, Department of Linguistics, University of Sydney. Wignell, P. (1992), ‘Abstraction and technicality in Social Science’. Paper presented to the International Systemic Functional Conference, July 1992, Macquarie University, Sydney. Wignell, P. (1990), ‘Flowers in the Sentence: a look at the discourse of the sociology of medicine’. MA (Applied Linguistics) Special Topic. Department of English and Linguistics, Macquarie University, Sydney.

10 Mathematical and scientific forms of knowledge: a systemic functional multimodal grammatical approach Kay O’Halloran

Introduction Bernstein (2000: 155) calls for a shift of focus in educational research beyond the pedagogic transmission and acquisition process (e.g. Bernstein 1990; Christie 1999, 2002; Christie and Martin 1997) to ‘the forms of the discourse’ being transmitted; ‘that is the internal principles of their construction and their social base’. The nature of subject-specific discourses has been investigated from the systemic functional (SF) perspective for some time, however (e.g. Coffin 1997; Halliday and Martin 1993; Martin and Veel 1998; Unsworth 2000; Veel 1999), and increasingly, this approach is adopting a multimodal discourse analysis (SF-MDA) perspective which extends the scope of study beyond language use to consider the contributions of other semiotic resources (e.g. Guo 2004; Iedema 2003; Kress 2003; Kress et al. 2001; Lemke 1998, 2003; O’Halloran 1999, 2003a, 2005a; Thibault 2001; Unsworth 2001). As part of this emerging tradition, this chapter explores how an SF-MDA approach to mathematics and science furthers our understanding of Bernstein’s (2000) formulations of knowledge, knowledge structure and grammaticality. In particular, Bernstein’s view of the grammaticality of mathematics and science (see Muller, this volume) is enhanced through the concept of multimodal grammaticality, where linguistic, visual and symbolic grammatical systems are seen to function integratively to construct mathematical and scientific knowledge. Lastly, pedagogical implications of the SF-MDA multimodal grammatical approach are briefly outlined. Bernstein’s (2000) formulations provide a basis for exploring and comparing mathematical and scientific knowledge. Mathematics is seen to provide the tools for the control and manipulation of the material world through the scientific description of the relations of time, space and matter for the purposes of prediction and the establishment of causality. Habermas (1974: 8) explains, ‘[t]here is a systematic relationship between the logical structure of a science [i.e. mathematics] and the pragmatic structure of the possible applications of the information generated within its framework [i.e. science]’. The ensuing discussion reveals how mathematics and science form different knowledge structures which complement each other. Through this collaboration, the scientific project is successful in the material realm through technological innovation within everincreasing terms of reference (e.g. physics, chemistry, biology, nuclear physics and genetics), and increasingly the approach shapes our understanding of what it means to be human (e.g. through the social sciences, psychiatry, medicine and so forth). An SF-MDA multimodal grammatical approach to understanding

206

LANGUAGE, KNOWLEDGE AND PEDAGOGY

mathematics and science as forms of knowledge enhances Bernstein’s (2000) theorizations of knowledge and knowledge structures.

Bernstein’s forms of knowledge and knowledge structures Bernstein (2000) proposes horizontal discourse and vertical discourse as a first step in conceptualizing forms of knowledge. Horizontal discourse exists within the realm of the everyday knowledge and ‘it is likely to be oral, local, context dependent and specific, tacit, multi-layered and contradictory across but not within contexts’ (Bernstein 2000: 157). Bernstein (2000: 157) explains, ‘the crucial feature is that it is segmentally organized’ across different situational contexts. On the other hand, vertical discourse takes ‘the form of a coherent, explicit and systematically principled structure, hierarchically organized as in the sciences, or it takes the form of a series of specialized languages . . . in the social sciences and humanities’ (Bernstein 2000: 157). Bernstein explains that vertical discourse is distributed through processes of recontextualization, which may be contrasted with the discrete segmented forms of horizontal discourse. Following Bernstein (2000), mathematics and science are viewed as vertical discourses which function through the principle of recontextualization. Bernstein (2000) describes vertical discourse as consisting of hierarchical and horizontal knowledge structures. In the first instance, hierarchical knowledge structures are conceptualized as integrating discourses of increasing abstraction, such as those found in physics and biology. In the second instance, horizontal knowledge structures ‘consist of a series of specialized languages with specialized modes of interrogation and criteria for the construction and circulation of texts’ (Bernstein 2000: 161). Horizontal knowledge structures are distinct, and the result is an accumulation of non-translatable languages. ‘[T]he speakers of each language become as specialized and as excluding as the language’ (Bernstein 2000: 162). Bernstein considers mathematics to contain horizontal knowledge structures ‘as it consists of a set of discrete languages for particular problems’ (Bernstein 2000: 163). These formulations raise some issues, which are explored below. The classification of hierarchical and horizontal knowledge structures in science and mathematics respectively may oversimplify matters, where the reality is closer to a hybrid mix of the two. Muller (this volume) explains that hierarchical knowledge develops through integration ‘towards a unitary convergent shape’ which Bernstein (2000: 161) conceptualizes as a triangle (see Wignell, this volume), while progress in horizontal knowledge structures occurs through the introduction of new languages by new speakers to investigate new sets of problems. However, as Muller (this volume) points out, while mathematics consists of horizontal knowledge structures with discrete fields of study (e.g. fields of study within pure mathematics, applied mathematics and statistics), forms of verticality associated with hierarchical knowledge structures exist within these different fields. From this perspective, mathematics would look something like a series of triangles which, following Wignell (this volume), have the potential to be in conflict with each other.

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE Mathematics

•

O’HALLORAN 207

Science (e.g. physics)

The same may be said of science, where specialization (e.g. nuclear physics) involves the hierarchical knowledge structure of physics and the horizontal knowledge structure specific to the field, in this case, quantum mechanics. The result would be triangles within triangles which together form an integrated hierarchy of knowledge, as shown in the diagram. The same may be said of mathematics, which integrates smaller hierarchical knowledge structures within the larger horizontal knowledge structures of mathematics. One major issue which remains unexplained is the reasons why mathematics expands through the proliferation of new languages which become separate fields of study (i.e. horizontal knowledge structures), while science progresses through the integration of different fields of study within a unified structure, i.e. a hierarchical knowledge structure (see Muller, this volume, on science and Wignell, this volume, on the social sciences). This question is explored in relation to Bernstein’s formulation of grammaticality. Before moving into this discussion, however, some preliminary remarks regarding the ideological effects of simple classifications of knowledge and knowledge structures are made (see Martin, this volume, on taxonomies). Dichotomies such as vertical/horizontal knowledge and hierarchical/horizontal knowledge structures have the potential to simplify the discussion, with one consequence being the maintenance of particular ideologies about the nature of knowledge (Muller 2004). The tendency for simplification through dichotomization is one effect of using linguistic categories to describe what in essence are complex phenomena. For example, ‘to know’ and ‘to structure’ become ‘knowledge structures’ which are differentiated into two types, hierarchical and horizontal. The semantic shift arising from grammatical metaphors such as ‘knowledge structures’ (e.g. Halliday 2004a; Halliday and Martin 1993; Halliday and Matthiessen 1999; Martin, this volume; Simon-Vandenbergen, Taverniers and Ravelli 2003) means that the discussion centres around the description of metaphorical entities, which gain a sense of permanence with attached values. While Bernstein’s (2000) theorizations are useful from a global perspective, for example, the concepts of vertical and horizontal discourse are usefully employed for a variety of purposes (e.g. this volume), including the investigation of school discourse (Bourne 2003; Muller 2000), it appears some caution is required with respect to the ideological side-effects of such classifications. The ideological effects of dichotomization become evident with ‘resemioticization’, which involves a shift in semiotic construal (Iedema 2001, 2003); in this case, the visualization of horizontal and vertical discourses. Two conceivable visualizations are the well-organized city landscape with flat planes for the

208

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Figure 10.1a

Bollmann map of New York (Nr. 40 – New York 1962)

Vertical discourse (recontextual)

Horizontal discourse (contextual)

Figure 10.1b

Stable city landscape

segmentally organized horizontal discourse and tower blocks for vertical discourse, as seen in Figure 10.1a, a three-dimensional Bollmann map of New York, and Figure 10.1b, a virtual city landscape. In these two city scenes, everyday horizontal discourses take place in the streets while vertical discourses take place in the institutional tower blocks. Alternatively, the spatial metaphors horizontal and vertical are visualized as lines in Figure 10.1c. The visual mappings in Figures 10.1a–c (and the triangles pictured above) each promote a wellordered, stable view of knowledge with different values accorded to horizontal

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

•

O’HALLORAN 209

Vertical discourse (recontextual)

Horizontal discourse (contextual)

Figure 10.1c

Horizontal and vertical lines

and vertical discourses if height is considered culturally representative of power, status and seclusion. While keeping in mind the possible ideological effects of dichotomization, Bernstein’s (2000) formulation of grammaticality provides a platform for exploring the reasons why mathematics expands through the proliferation of new languages (i.e. horizontal knowledge structures) and science progresses through the integration of new knowledge structures (i.e. hierarchical knowledge structures). From this point, multimodal grammaticality provides an alternative view of the grammaticality of mathematics and science to provide an account of how mathematics and science differ from other forms of knowledge.

Bernstein’s grammaticality and knowledge structures Bernstein (2000) formulates grammaticality as a form of knowledge variation which applies to the internal and external aspects of language description. Bernstein (2000: 163) suggests there are ‘strong/weak’ grammars within the different languages of horizontal knowledge structures, where the distinction is based on the relative ability of the grammar to model empirical relations in the world. That is, a strong grammar has the capacity to describe and model external relations in the world (e.g. economics, linguistics and parts of psychology), while the power of a weak grammar to model such relations is much lower (e.g. sociology, social anthropology and cultural studies). The strong grammars of horizontal knowledge structures ‘often gain their power by rigorous restrictions on the empirical phenomena which they address’ (Bernstein 2000: 163). On the other hand, the strong grammar of hierarchical knowledge structures is a feature which remains constant (e.g. physics). Bernstein (2000) admits, however, that his formulations of the relations between grammaticality and knowledge structures remain unclear. For instance, mathematics creates a problem because it is seen to possess the strongest grammar of all horizontal knowledge structures, yet it does not have the empirical fit with the external world which is the definition of this form of grammar.

210

LANGUAGE, KNOWLEDGE AND PEDAGOGY

This contradiction warrants further investigation, for it helps us to explore the reasons why knowledge develops in different forms in mathematics and science. That is, mathematical knowledge expands through proliferation (i.e. horizontal knowledge structures) while science progresses through integration of new knowledge structures (i.e. hierarchical knowledge structures). Bernstein’s view of strong/weak grammars as a means for investigating and differentiating between mathematical and scientific knowledge is undertaken in two steps. First, mathematics and science are discussed with respect to their relationship to each other, and their relative ability to model empirical relations in the external world. Second, comparisons between mathematics and science provide a key to understanding the relationship between knowledge structures and the ways in which the two disciplines expand and progress through the principles of proliferation and integration respectively. From here, the nature of mathematical and science knowledge is developed from the perspective of SF approach to multimodal grammaticality in the next section: ‘SF Approach to Multimodal Grammaticality’. This discussion is aimed at exploring the relationship between mathematics and science, and how they differ from other forms of knowledge. The relationship between the disciplines of mathematics and science is intimate. Bell (1987), for example, sees mathematics as the ‘queen and servant of science’. While different in fundamental ways, mathematics is none the less called ‘the queen of the sciences’ and is ranked first amongst the sciences: It is customary to call mathematics a science and, after the positivist philosopher I.A.M.F.X. Comte (1798–1857), to place it first in the classification of the sciences. So as long as we remember the radical difference between mathematics and the physical or biologic sciences, no harm is done in calling mathematics a science. Something of the distinction of the mathematical method and the strictly scientific, however, must be seen before we attempt to uncover the mysteries of mathematical prophecy in scientific discovery. The matter is extremely simple, but none the less profound. (Bell 1987: 259)

The ‘radical difference’ between mathematics and science lies in the relationship with the external world. ‘A science has “real” content [with respect to the world], or claims to have’ (Bell 1987: 259), and it proceeds through experimentation and correlation with empirical data. On the other hand, ‘[t]he essence of mathematics is deductive reasoning from explicitly stated assumptions called postulates’ (Bell 1987, p. 261), which may have nothing to do with the external world. Electromagnetism, for example, is the organized body of information that has been acquired concerning electricity and magnetism as they ‘actually’ appear in human experience. Likewise astronomy, unless scrutinized too critically, is a systematized accumulation of ‘facts’ about heavenly bodies, these celestial objects being assumed to have an ‘existence’ outside our ‘sense perceptions’. With mathematics, it is entirely different . . . So far as the external world is concerned, mathematics is as empty as a game imagined in a dream and forgotten on waking. Nothing whatever is in mathematics except the rules of the game, and these rules are prescribed at will by the player. (Bell 1987: 259)

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

•

O’HALLORAN 211

Mathematics is seen to be the ‘servant’ or ‘handmaiden’ of science because it is called upon to provide the means for scientific descriptions of empirical relations in the external world. Furthermore, the scientific experimental method depends on scientific laws which are mathematically expressed. Emeritus Professor Philip Davis (Department of Applied Mathematics, Brown University) explains how mathematics is applied to areas beyond the immediate concerns of mathematics itself: Mathematics is the science, art and a language of quantity, space, and pattern. Its materials are organized into logically deductive and very often computational structures. Its ideas are abstracted, generalized, and applied to concerns other than mathematics itself. When mathematics mixes with ‘outside concerns’, the mixture is called applied mathematics, and for reasons that are by no means clear and may even remain a perpetual mystery, mathematics has been found to be of utility and an indispensable aid to the physical sciences. It has often been called the ‘Handmaiden of the Sciences’. The expositions of theoretical physics are completely mathematical in character. (Davis, personal correspondence, 17 March 2005)

There are several points to be made here. First, mathematics and science share the same grammatical basis, a point which is developed in the next section: ‘SF Approach to Multimodal Grammaticality’. Second, mathematics has weak empirical relations with the world only within certain fields of mathematics (e.g. ‘pure mathematics’). Fairly obviously, applied mathematics is concerned with modelling real, empirical data. If we put aside applied mathematics, and accept pure mathematics as the benchmark of mathematical activity, mathematics appears to be at odds with Bernstein’s view of strong/weak grammars because it has the strongest grammar of horizontal knowledge structures, yet it is not concerned with empirical relations within the world. On the other hand, science has the strongest grammar of hierarchical knowledge structures and the strongest empirical relations with regard to the external world. Why is this the case? Why the discrepancy between the strong grammaticalities of mathematics and science, and their respective relations to the external world? And how does this discrepancy relate to the expansion of knowledge in mathematics and the progression of knowledge in science? The answer to the question concerning the strong grammaticality of mathematics and science becomes self evident; mathematics and science both have strong grammars because they both have a grammar based on mathematics which has the potential to describe empirical relations in the world (i.e. the case of applied mathematics). Furthermore, the methodologies of mathematics and science both involve reasoning based on mathematical postulates and scientific laws which are mathematically expressed. For reasons which philosophers continue to debate (see Wigner 1960), mathematics is effective in describing empirical relations in the world. In conclusion, the grammar looks inwards (internally) in the case of mathematics, and the grammar looks outwards (externally) in the case of science (and applied mathematics). The strong grammaticality of mathematics

212

LANGUAGE, KNOWLEDGE AND PEDAGOGY

and science is at one with each other: the difference is a matter of application. One could reasonably ask, why is this the case? How does the close partnership between mathematics and science work, and what are the mutual benefits? And why does mathematics expand through proliferation of new horizontal knowledge structures, and science through the integration of new knowledge structures to form a unified hierarchical knowledge structure? The question of expansion and progression through proliferation and integration is not quite so clear-cut. Muller (this volume) sees progress in science as the product of the integrative nature of hierarchical knowledge structure of science which combines with a strong grammar to give science a stronger explanatory power than other forms of knowledge. One speaks of progress, and one thinks of science. In other words, grammaticality determines the capacity of a theory or language to progress through worldly collaboration; verticality determines the capacity of a theory or language to progress integratively through explanatory sophistication. Together, we may say that these two criteria determine the capacity of a particular knowledge structure to progress. [my emphasis]

Muller’s (this volume) link between verticality and grammaticality is a significant step in explaining the progress of science in terms of an increasing explanatory power. The question which remains is the discrepancy between expansion through proliferation in mathematics and progress through integration in science. Given the strong relationship between mathematics and science, why do the two disciplines differ in this respect? The answer to this question has several dimensions which are developed below. First, one possible reason for the integrative nature of scientific knowledge is that by seeking to explain the external world, science possesses a strongly unified agenda. Science seeks to explain external phenomena, and thus scientific endeavours are orientated towards achieving that common goal, in one way or another. Second, science depends upon support from outside for scientific knowledge to progress. That is, the mapping of experience (and hence knowledge) is tied to political, economic and military interests (see Wignell, this volume) where associated bodies fund and support the development of science. Scientific progression, in particular, is linked to the accumulation of power and capital through technological advancement. One could almost say that science serves that goal. From this perspective, the progression of scientific knowledge is determined by factors outside the intellectual arena. Scientific knowledge progresses according to the interests of stakeholders, and most typically new knowledge is integrated within existing structures because if isolated outside existing fields of research, the new knowledge will disappear unless institutional, financial and intellectual support is obtained to develop that knowledge. If such support is obtained, the new knowledge becomes integrated within the existing structures of the science which seeks to fulfill the same goal. The other alternative is a major rupture (e.g. see Koestler, 1959; Kuhn, 1970), which shifts the axis of verticality of the scientific knowledge structure from one position to another. But this rupture doesn’t bring down the whole hierarchical knowledge

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

•

O’HALLORAN 213

structure of a science, because – and this is the third major point – science possesses a strong mathematical infrastructure which can withstand alternative descriptions of phenomena. That is, a range of mathematical tools can be integrated for the common pursuit of describing external phenomena. This brings us to a fourth (inter-related) point. If a knowledge structure is concerned with specific areas of experience – i.e. if it has strong relations with the external world – for progress to take place, the new knowledge structure has to offer benefits above and beyond other competing approaches which explain that same phenomenon. Knowledge progression becomes a political/economic/ military/intellectual exercise where alternatives co-exist in the same hierarchical knowledge structure if they provide rewards to those maintaining the structure. The perceived value of a mathematical and scientific theory relates to the objectives which it may achieve. Muller’s (this volume) view of the ‘worldly collaboration’ of science takes on a new meaning in this light. Mathematics, on the other hand, does not compete with regard to explaining the phenomena in the real world. Unless called into action to support a cause in the external world, mathematics is generally focused inwards (with the exception of applied mathematics, as explained above). New knowledge can, up to a certain point, proliferate unproblematically. Following Wignell (this volume), if conflict does occur within a horizontal mathematical knowledge structure, the hierarchy is maintained through internal struggle, or alternatively, another horizontal discourse is established. Wignell (this volume) suggests that the outcome of struggle depends on the nature of the boundaries which are established; strong boundaries provide protection for practitioners within horizontal knowledge structures. Given that mathematics has strong boundaries, dissenting voices in mathematics are either quietened or leave to form new horizontal knowledge structures. This is not to say that mathematics is free to expand with unbounded proliferation; there are limits on what type of mathematics is developed at different periods of time. Some constraints relate to the close relationship of mathematics to science, which selectively calls upon mathematics to solve particular problems. The ability of mathematics to expand through proliferation is, however, one key factor which explains the successful progression of science. This point is developed below. The significant point arising from this discussion is that science progresses through accessing the strong grammar of mathematics which is free to proliferate to provide alternative approaches and descriptive tools. These approaches are integrated with the hierarchical knowledge structure of science to serve particular agendas relating to the description of material reality. The relationship of mathematics to science is a two-way affair, what Bell (1987: 2–3) calls the ‘right-hand, left-hand aspect of mathematics’: [T]he pure serves the applied, the applied pays for the service with abundance of new problems that may occupy the pure for generations. The debt may then be reversed, when art for art’s sake [in mathematics] pays off in the solution of difficult problems in science and technology.

214

LANGUAGE, KNOWLEDGE AND PEDAGOGY

As O’Halloran (2005) and Wignell (this volume) explain, science is a discourse which rapidly expanded because so many stood to gain from the power it unleashed: the bourgeoisie, aristocracy, military, merchants, capitalists, intellectuals and scientists themselves. Being the queen and handmaiden for the sciences, mathematics underscores scientific progression and the development of new technology. Bell (1987), for example, surmises that war outranks economics with regard to being influential in the development of mathematics, a trend which must have increased in the past few decades. More mathematics was developed in some departments ‘in less than a decade of [the second world] war than might have been found in half a century of peace’ (Bell 1987: 3). The mathematics was used to develop military technology and armaments. The close relationship between mathematical and scientific forms of knowledge provides a platform to understand how horizontal and hierarchical knowledge structures function productively together as a related phenomenon which fuels scientific progression. That is, the verticality and strong grammars of the sciences which progressively integrate knowledge depend upon a strong grammar which proliferates in the horizontal knowledge structures of mathematics. In what follows, the basis for Bernstein’s formulation of the strong grammars for mathematics and science (which are one) is explored through the concept of multimodal grammaticality, where ‘the languages’ of mathematics and science are re-conceptualized as the product of the integration of three semiotic resources: language, mathematical symbolism and visual imagery. The aim of the ensuing discussion is to explore how mathematical and scientific knowledge provides the means for rewriting the external world.

SF approach to multimodal grammaticality Multimodal grammaticality is proposed as an approach which furthers our understanding of the grammaticality of mathematical and scientific knowledge beyond that proposed by Bernstein (2000). The approach departs from Bernstein’s formulation of the strong and weak grammar distinction between mathematics and science with the aim of investigating the underlying similarities between the two forms of knowledge which permit the ‘right-hand, left-hand aspect of mathematics’ described by Bell (1987: 2–3). For this purpose, mathematics and science are conceptualized as forms of knowledge constructed through the integrated use of grammatical choices from language, visual images and symbolism. In what follows, multimodal grammaticality is theorized from the Systemic Functional (SF) social semiotic perspective of language (Halliday 1978, 2004a; Halliday and Matthiessen 1999; Martin 1992; Martin and Rose 2003), which has been extended to visual images and other semiotic resources (e.g. Kress and van Leeuwen 1996; Martinec 2000, 2001; O’Toole, 1994; van Leeuwen 1999, 2005). More recently, SF multimodal discourse analysis (SF-MDA) research efforts are directed towards investigating the ways in which semiotic choices interact to create meaning (e.g. Baldry 2000; Baldry and Thibault 2006; Iedema 2003; Lemke 1998, 2002; Martin 2002; Martinec 2004; O’Halloran 2004, 2005; Royce and Bowcher, in press; Thibault 2000; Ventola 2004).

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

•

O’HALLORAN 215

An SF multimodal grammatical approach to mathematics and science is concerned with the individual and integrated functionality of language, mathematical symbolism and visual images in the form of diagrams, and abstract and statistical graphs (see Lemke 1998, 2003; O’Halloran 2005). In the first case, the three semiotic resources are seen to be uniquely functional in the construction of mathematical and scientific knowledge. That is, each resource makes an individual intra-semiotic contribution to meaning through the meaning potential of that resource. Second, the three resources function inter-semiotically to make meaning. That is, mathematical and scientific discourse functions through the grammatical integration of linguistic, symbolic and visual elements. Furthermore, shifts from one semiotic resource to another (e.g. linguistic to visual to symbolic) in larger discourse moves permit metaphorical expansions of meaning to take place, while simultaneously enabling the meaning potential of the new semiotic resource to be accessed. Using pre-established results, laws and theorems, the construction of knowledge in mathematics and science takes place through: • • •

the meaning potential of the three semiotic resources; the integration of grammatical choices from the three semiotic resources; the expansions of meaning which take place as the discourse shifts from one resource to another.

One aim of multimodal grammaticality is to theorize the three semiotic resources as an integrated phenomenon to explore the nature of mathematics and scientific knowledge. One result is formulation of the ‘strong grammatical interconnectivity’ of the three semiotic resources within a restricted semantic field (see section (2) ‘Strong grammatical interconnectivity: the visual image and the symbolism’). The SF-MDA approach to multimodal grammaticality is developed with respect to the concept of visibility (see Miller 1984; Virilio 1994) where ‘to make visible’ is to capture semiotically; in this case, using visual imagery, mathematical symbolism and language. Visibility relates to the concept of semiotic reality (Martin 1992) where the discourse shifts from ordering the material realm to ordering the semiotic realm. The scientific project is the result of the expanded visibility made possible through semiosis: that is, semiosis involving visual imagery and symbolism, the two central semiotic resources upon which modern mathematics and science are built. Language is seen to play a significant supportive role in this enterprise. However, the functions of language are only discussed in general terms in this paper (see Halliday 2004b; Halliday and Martin 1993; Martin and Veel 1998) because the focus is directed towards the visual imagery and the symbolism, the distinguishing trademarks of mathematical and scientific knowledge. In what follows, the expanded visibility derived from the use of mathematical symbolism and visual imagery is explored.

216

LANGUAGE, KNOWLEDGE AND PEDAGOGY

The mathematical and scientific project The mathematical and scientific project is unashamedly aimed at the control and manipulation of the material realm. In the beginnings of modern science, Descartes (1596–1650) sought to establish a ‘methodology for a universal science which is as concerned with increasing human power as well as understanding’ (Barry, 1996: 35). Barry (1996: 35) explains that ‘Descartes’ programme is technological, because it seeks to penetrate the secrets of the material realm in order to manipulate them’. For these ends, Descartes dismisses that which is perceived through senses and accepts only that which is conceivable through the mind (see Descartes 1952). This decision is based on Descartes’ view of the unreliability of sense data and on his success in using mathematical symbolism to construct geometrical curves. Descartes has the mathematical symbolism as a tool for reasoning and the geometrical images as a tool for seeing in what becomes a limited and well-defined semantic field (see Descartes, 1683, 1954). Descartes moves to constructing a new realm of semiotic reality, and in doing so, the everyday language of the sensory world is dismissed and the mathematical symbolism and geometrical imagery is developed. The result is a mind/body duality with a privileging of two forms of semiosis. That is, the mathematicized world constructed using geometrical images and mathematical symbolism is admitted as true knowledge, while sensory data constructed through everyday language is dismissed as secondary knowledge (see Descartes 1985 and Muller, this volume). Newton (1642–1727) develops Descartes’ symbolic and visual semiotic tools to reformulate the material sensory realm in what becomes an empirical-based science which uses scientific instrumentation in experimental laboratories. Newton constructs his mathematical principles of the natural world based on a model of linearity, where complex wholes are broken into constituent parts in order to understand the system. Although the model of linearity has been replaced by non-linear dynamical systems theory, Newton nonetheless makes visible the non-visible through the symbolic and visual formulation of entities such as force, gravity, velocity and acceleration (Barry 1996), a scientific tradition of visualization which continues today within computerized environments. In what follows, visibility is discussed in relation to the functions and grammar of mathematical and scientific visual imagery. From this point, the strong grammatical interconnectivity between the visual imagery and the symbolism is explored, before the grammar of mathematical symbolism is discussed. (1) Mathematical and scientific visual imagery The impact of visual imagery is so significant that scientific, political, religious and artistic pictorial images, icons and models are sites of construction, destruction and defacement (Latour and Weibel 2002). However, the major theoretical approach to visual imagery is often restricted to the cognitive domain for investigating cognitive psychological processing of visual images, the physiology of perception, and the dynamics of creative thinking. This emphasis on the cognitive may be explained by the available funding for research in the ‘science and technology of instrumented aiding of optically-guided spatial behaviour’ (Koenderink 1993: viii) for military purposes, space exploration, medical diagnostics and other fields where the human operator is vital but often

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

•

O’HALLORAN 217

performs as ‘the weakest in the chain’ (Koenderink 1993: ix). Koenderink (1993: ix) explains that ‘[a] fundamental understanding of human abilities and idiosyncrasies is necessary, but only fragments exist’. A second major approach to visual imagery is the graphic design perspective for developing effective displays of information for advertising, signs, maps and so forth (e.g. Lester 2000; Wildbur 1989; Wildbur and Burke 1998). The intended use of the visual image thus plays an important role in determining the theoretical approach to visual semiosis. The social semiotic approach to the semantic realm which mathematics and science make visible remains a relatively unexplored field of study. This approach to visual imagery in mathematics and science is developed below. God plays a central role in the beginnings of modern science (Funkenstein 1986). The need for God, however, disappears with the ongoing success of science in formulating time, space, matter relations for purposes of prediction and establishing causality. With an increased focus on geometrical shapes, lines and curves, human figures in material settings gradually disappear from mathematical and scientific drawings (O’Halloran 2003b, 2005a). Human bodies are replaced by body parts, such as hands to depict material actions and eyes for acts of perception. Scientific procedures, for example, involve disembodied hands as displayed in Figure 10.2. Military objectives such as hitting targets are no longer represented. Eventually the context of the problem disappears altogether and only the geometrical lines and curves remain. The ongoing success of science meant that God became irrelevant and man became the invisible agent in the programme for controlling the material world. Today mathematical results rarely directly display the objectives for which they were developed, and the reader sees non-violent applications of the theory after the mathematical results are established in a de-contextualized environment. An examination of the beginnings of modern science is productive, however, for understanding the semantics and grammar of modern mathematical and scientific visual images. In what follows, Newton’s early work is closely examined.

Figure 10.2 Scientific experiment with frogs (Plate 1, Galvani 1792) reproduced in Stafford (1996: 91)

218

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Figure 10.3 is the page opposite the title page of Newton’s (1736) The Method of Fluxions and Infinite Series. The problem of men shooting flying birds is visualized, complete with hunting dogs to collect the victims. The visual image is interesting for several reasons. First, it becomes possible to see how mathematical and scientific images make visible the invisible. For example, the paths of the two birds are semiotically constructed over time using parallel lines AB and CD. Instances of time in the flight path of two birds are displayed spatially as points E, F and G and H, I and K. The conflation of matter (in this case, the birds) with respect to particular instances of time in the depicted scene are mapped as points G and K. Three lines of fire are captured visually as dotted lines. The potential collision of the gunshot with the birds is depicted as points E, F and G and H, I and K, with the intended hit pictured at G and K. Thus the visual image makes visible the flight path, the lines of fire, and the potential interaction between material matter (the gunshot and the birds) at different instance of time. That is, the drawing is concerned with the relations between space, time and matter for the purposes of prediction.

Figure 10.3

Space, time, matter and prediction (Newton 1736: Frontispiece)

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

•

O’HALLORAN 219

Second, Figure 10.3 demonstrates how spatiality is linked with temporality in ways which are unique to mathematics and science. The experience of visual space is explained by Saint-Martin (1990: xi–xii): ‘. . . spatiality can be defined as the apprehension of a simultaneous coexistence of multiple elements in an autonomous form of organization, which is considerably different from that of the temporal order of these elements’. From Newton’s (1736) drawing, however, it can be seen that mathematical and scientific diagrams capture spatiality with respect to a continuous temporal order, which simultaneously incorporates several specific instances of time. We find the conflation of space with time in a way that combines matter with predictive ability, a feat which does not seem to occur in other genres of visualization. This point warrants further discussion. A photograph captures a single frozen instance of time; an instance where someone may be crying, laughing or running. The sculptor Auguste Rodin explains that sculpture and other art forms, however, have the potential to give the impression of movement over time, rather than capturing the single instance of the photograph. Rodin claims that in doing so, ‘[i]t is art that tells the truth and photography that lies. For in reality time does not stand still, and if the artist manages to give the impression that a gesture is being executed over several seconds, their work is certainly much less conventional than the scientific image in which time is abruptly suspended . . . ’ [excerpt from Paul Gsell interview with Rodin cited in Virilio (1994: 2)]. Rodin appears to be simultaneously right and wrong. Time is abruptly suspended in the mathematical and scientific image, but it may be suspended for any number of instances within a continuous timeframe. That is, the mathematical and scientific image contains repeated photographic snapshots, where the potential exists for any moment to be seized indefinitely, as seen in Newton’s drawing in Figure 10.3. The peculiar capturing of unlimited photographic-type instances within a continuous framework in mathematical and scientific visual imagery is explored in relation to Figure 10.3. There is a sense of movement and action in Newton’s drawing, with the men shooting the birds and the dogs sniffing and watching. However, the mathematical lines and points capture a time-sequence which includes and extends the action sequence of the men and the dogs. That is, the time-frame realized by the flight paths of the birds includes a time before and a time after the hunting scene which is drawn. The hunting scene serves as a visual backdrop to contextualize the spatial relations of matter over time for the purpose of prediction; i.e. to hit the bird at times G and K, or more generally, to hit a moving target at some pre-determined time. However, the mathematical semiotic construction of the problem, the geometric lines and points, captures change over time in a manner which extends beyond other genres of visual representation such as drawings, photographs and sculptures. The lines and curves in mathematical graphs display an uninterrupted continuity in terms of spatio-temporal relations, where the temporal-frame of mathematical visual images extends to the infinite. The mathematical visual curve typically depicts a relationship between entities; it makes the relationship visible, it makes the relationship concrete and it makes the relationship continuous. From this point, the relationship is described using mathematical symbolism.

220

Figure 10.4

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Revisiting the moment mathematically (Newton 1736: 67)

If we consider Newton’s (1736: 67) mathematical writings in The Method of Fluxions and Infinite Series displayed in Figure 10.4, relations of time, space and matter appear as geometrical figures in the form of triangles, curves, lines and points, and numbered paragraphs consisting of linguistic and symbolic components. The de-contextualized nature of the visual images means that the experiential and logical meanings are foregrounded. There are no visual distractions in mathematical and scientific visual imagery. From this point, having achieved visibility in the form of geometrical images, the strong grammatical relations between these images and their symbolic description (e.g. the numbered paragraphs in Figure 10.4) provide one path for understanding the functionality of the mathematical symbolism.

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

•

O’HALLORAN 221

(2) Strong grammatical interconnectivity: the visual image and the symbolism Descartes’ method of using mathematical symbolism to describe curves means that the space, time and matter relations displayed in the visual image are relocated within a new semantic map; that is, the visual patterns of co-variation are exactly described symbolically, a feat which is impossible using language (Lemke 1998, 2003). The meaning potential of language simply does not permit curves to be described in such an exact manner. Mathematical symbolism evolved from language, however, to develop new grammatical systems of meaning so that symbolic description and manipulation of continuous patterns of relations became possible. Before the grammatical strategies developed in mathematical symbolism to achieve these functions are investigated (see section (3) ‘The grammaticality of the mathematical symbolism’), the nature of the grammatical relations between the visual image and the symbolism is discussed. A point in the mathematical visual image represents an instance in the state of the system which captures the spatio-temporal relations between entities (e.g. the birds and the bullet in Figure 10.3). The instance, and every instance, is visualized in relation to other states of the system (e.g. the line segments for the flight path of the birds and the line of fire), just as the hand for dissecting the frogs in Figure 10.2 is viewed in relation to the activity sequence of the experiment which is portrayed. The ‘temporal’ frame of the mathematical and scientific image is not necessarily related to clock-time, rather the framing relates to unfolding patterns of relations. For example, the tonnage of nuclear bombs and the levels of radioactivity may be visualized as an unfolding system along a scale which is not a temporal-clock version of time. Rather, the system is a pattern of co-varying relations where specific instances may be located and suspended. Significantly, the semantic framework of the visual image has a strong relationship with the semantics of the symbolism, so the relations of the entities visually displayed can be described exactly using mathematical symbolism, with the potential for describing the state of the system at specific instances. This means that the visual image is re-contextualized within the semiotic landscape of the symbolism, and vice versa. The re-contextualization process is bi-directionally semiotic; in other words, one form of semiotic construction (the visual image) is re-contextualized with respect to a second form of semiosis (the mathematical symbolism), and vice versa. The linguistic text is also involved in the re-contextualization process (e.g. Figure 10.4) making mathematics and science multi-directionally semiotic. There is room for such ‘bi-directional’ and ‘multi-directional’ re-contextualization processes within other semiotic landscapes; for example, the photograph in the newspaper is re-contextualized with respect to the linguistic text, and vice versa. However, the directedness of the re-contextualization process in mathematics leads to what could be called a strong grammatical interconnectivity. That is, there is little room for ambiguity in the re-contextualization process, and so the ‘semantic effervescence’ (Cheong 2004) arising from the interaction between the visual image, the mathematical symbolism and the linguistic text is minimal. In fact, mathematics developed as a written semiotic discourse where specific types of grammatical interconnectivity were established in order to solve problems through the mathematical symbolism. The strategies for achieving the

222

LANGUAGE, KNOWLEDGE AND PEDAGOGY

strong grammatical interconnectivity between the visual image and the symbolism are discussed below.

Figure 10.5

Suspending the instant (Newton, 1736: 10)

Space becomes a metaphor for the temporal in the form of the unfolding relations in the mathematical and scientific visual image, and the temporal becomes located within space, a visual space which is describable using mathematical symbolism. For example, Newton’s (1736: 10) grid in Figure 10.5 demonstrates how an instant * can be spatially suspended in relation to an unfolding sequence of relations, in this case represented by the dotted line. The capturing, suspension and mapping of temporal-spatial relations in twoand three-dimensional symbolic grids may be seen in Figures 10.6a Newton (1968: 31), 10.6b Prigent et al. (2003: 112) and 10.6c Golovin et al. (2003: 199). According to the symbolic organization or mapping of space, (i) a point (i.e. an instant in the relationship between entities), (ii) the line, the curve and three-dimensional shapes (i.e. the actual relationship between entities over some temporal frame) and (iii) space (i.e. the potentiality of the relationship) have symbolical formulations. That is, the strong grammatical interconnectivity between the visual image and the symbolism occurs because the visual entities (the point, the curve, shape or object, and space) have symbolic descriptions. These symbolic descriptions are manipulated to describe other relationships and to predict the behaviour of systems. As we see from Figure 10.6c, new forms of visualization are appearing through the use of computer graphics (see section ‘Visibility in the computer age’). Mathematical and scientific visual images are concerned with the spatialtemporal display of relations which are suspended within a symbolic grid. The visual entities have symbolic identities (e.g. x, y and z) and the symbolic configuration of the relationship (the line, the curve and three-dimensional shapes)

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

Figure 10.6a

•

O’HALLORAN 223

Suspending the visual image: Newton (1968: 31)

takes place through the formulation of the patterns of variation using mathematical operative processes (O’Halloran 1999, 2003a, 2005) (see section (3) ‘The grammaticality of the mathematical symbolism’). The ongoing success of the symbolic descriptions means that visualization, however, is no longer a necessary requirement for mathematical and scientific theorizations. For example, geometry has moved beyond three-dimensional Cartesian space to higher nonvisible dimensions, and calculations in nuclear physics predict the behaviour of electrons which cannot be visualized. That is, the verticality of mathematical and scientific discourse has moved beyond the boundaries of perceptual spacetime-matter reality through the semiotic potential of the symbolism. Being

224

Figure 10.6b

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Suspending the visual image: Prigent et al. (2003: 112)

caught in a visual grid and/or symbolic forms of reasoning, however, means that an effort is required to escape the confines of the semantic field which is subsequently established. The increasing potential to manipulate visual imagery through computer generated imagery (CGI) is providing semiotic tools to move beyond traditional approaches to mathematics and science (see section ‘Visibility in the computer age’). The grammaticality of the symbolism is investigated below.

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

Figure 10.6c

•

O’HALLORAN 225

Suspending the visual image: Golovin et al. (2003: 199)

(3) The grammaticality of the mathematical symbolism Virilio (1994: 5) explains how war contributed to condensing and making efficient written characters: ‘[t]he tendency to make reading time as intensive as speaking time . . . stemmed from the tactical necessities of military conquest and more particularly of the battlefield . . .’. However, despite its written efficiency, language is limited for capturing precise degrees of variation in relationships (Lemke 2003), for example between time, space, and matter. While visual images capture variation in the form of unfolding relations, traditionally visual

226

LANGUAGE, KNOWLEDGE AND PEDAGOGY

displays are manipulated and re-reassembled for predictive purposes with some difficulty, although this situation has changed considerably with computer technology (see section ‘Visibility in the computer age’). The semiotic resource through which relationships are encoded and manipulated is the mathematical symbolism because, as those following Descartes discovered, the symbolism incorporates exact descriptions of visualized relations through the strong grammatical interconnectivity described above. Particular forms of grammaticality of the mathematical symbolism (which differ from language) were developed so that it could be used as a tool for reasoning about the visualized relations. However, despite the significance of mathematical symbolism, it remains largely an untouched area of research in the humanities, as discussed below. Elkins (1999: 134) explains that symbolic mathematical formulations are largely under-theorized domains of writing in the humanities: ‘[t]raditionally, mathematical expressions, formulae, diagrams, and equations are excluded from publications in the humanities. They are virtually absent from texts that explore deconstruction, and it is even possible to write about philosophic logic without using formalisms.’ Following Rotman (1995), it is difficult to justify the exclusion of mathematical notation for the study and deconstruction of written texts. One reason for the omission is that mathematical symbolic expressions are typically not accessible to a wide audience. The symbolism requires theorization, however, and the current situation highlights the need for transdisciplinary studies to bridge mathematics, the sciences, the humanities and education. A step is made in this direction in this paper. Mathematical symbolism is a ‘symbolic’ sign system according to Peirce’s (1991: 251) classification of sign systems, where ‘every sign is determined by its object’. The Icon possesses characteristics of the object it refers to, the Index has connections to the object, but the meaning of the Symbol must be acquired through ‘habit’, i.e. the culturally assigned meanings and values must be learnt. Language and mathematical symbolism are such symbolic sign systems. Mathematical and scientific visual images typically bear some relationship to perceptual reality, although this changed with quantum mechanics and, more recently, visualizations made possible through computer graphics. The majority of traditional visual images tend to be iconic (e.g. a three-dimensional cube to represent a box) and indexical (e.g. the graph where increase/decrease appear as rising/falling curves). However, despite these relations to perceptual reality, the meanings of mathematical visual imagery need to be taught and learnt, and the difficulties for students learning mathematics increase with the symbolic status of language and mathematical symbolism. The difficulties associated with scientific language and mathematical symbolism are explored below. In the first case, scientific language departs from the congruence of everyday language (Halliday 2004b; Halliday and Martin 1993; Halliday and Matthiessen 1999; Martin and Veel 1998). For example, Halliday (1993: 69–85) lists the difficulties of scientific and mathematical language as including interlocking definitions, technical taxonomies, special expressions, lexical density, syntactical ambiguity, grammatical metaphor and semantic discontinuity. The metaphorical nature of scientific language is conceptualized through grammatical metaphor and other linguistic constructions which aid the flow of reasoning (e.g. Halliday

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

•

O’HALLORAN 227

and Martin 1993; Martin and Veel 1998; Simon-Vandenbergen et al., 2003). In the second case, it appears that the major problem in teaching and learning mathematics and science relates to mathematical symbolism, the semiotic through which theory is constructed and problems are solved. Mathematical symbolism is most clearly a Peircian symbolic sign which must be acquired through learning, the problem being that most people fail to learn the ‘habit’. The question which plagues many mathematics educationalists is why this is the case. In what follows I give some possible reasons. Mathematical symbolism is the major tool for reasoning because it developed as a form of semiosis capable of (i) encoding the continuity of relations displayed by visual images, (ii) suspending and capturing those relations at any point, and (iii) rearranging those relations to solve problems. In other words, the symbolism captures exactly the relations between space, time and matter to establish states of being for predictive purposes. But how is mathematical symbolism organized so that these functions may be fulfilled? This question is explored below in terms of semantics, grammar and grammatical strategies for encoding meaning in mathematical symbolism. Mathematical and scientific symbolic descriptions construct a different version of reality compared to linguistic constructions. Firstly, the semantic field is radically reduced to time, space, and matter relations (i.e. experiential meaning) for purposes of prediction and establishing causality (i.e. logical meaning). With respect to experiential meaning, Relational Identifying processes establish states of being (e.g. =, ≤ and ∝) which are described using Operative processes (e.g. ×, +, ÷ and ±) between generalized mathematical participants (e.g. x, y and z) (O’Halloran 1999, 2005a). The mathematical participants, in the form of numbers and variables, have direct referents within the context of the problem, or otherwise the referent may remain unspecified. The nature of Operative mathematical processes radically departs from the process types found in language. Operative processes are symbolic processes of dividing, adding, subtracting and multiplying and so forth which capture the dynamic patterns of variation as a complex of interactions within and between process/participant configurations. Operative processes developed from materially based processes (sharing, increasing and decreasing material goods, for example) to describe degrees of increase, decrease and relativity with virtually unlimited boundaries of delicacy. However, Operative processes are semantically different from linguistic processes. Operative process/participant configurations, for example, do not necessarily contain one major participant (the Medium) with the potential for agency (the Agent), which is the hallmark of linguistic constructions (see Halliday 2004a). The potential exists within the mathematical symbolism to configure an unlimited number of Operative processes and participants; for example, x 1 × x 2 × x 3 . . . x n where x i and n take any value, and × can be replaced with other forms of Operative processes. Given the narrow semantic field of the mathematical symbolic descriptions, the choices for circumstance are limited. The nature of symbolic processes, participants and circumstance requires further research in order to investigate the ways in which mathematics and science construct the world (see O’Halloran 2005, Chapter 5).

228

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Operative process/participant symbolic configurations are maintained so they can be manipulated and rearranged to solve problems. This may be compared to the encoding of meaning through grammatical metaphor in language where dynamic processes are lost, for example, in nominalized word groups. The means for maintaining the relationships between participants is embedding where process/participant structures are rankshifted to function within other process/participant structures. Mathematical symbolism has the potential to embed process/participant configurations to an unlimited degree. To make the embedding and consequent potential for rearrangement of relations a viable option, an economy of expression is blended with conciseness in the symbolism. This took place through the development of new grammatical systems. That is, modern mathematical symbolic notation developed as a written system with the aim of solving particular types of problems, and therefore specific grammatical systems for encoding meaning were developed. The new grammatical systems include, for example, generalized symbolic participants (x, y and z), special symbols for processes (e.g. Σ, √ and ±, ×), spatial notion (e.g. xy indicate division, powers to indicate multiplication x2 and subscript/superscripts x 12), use of brackets and other forms of spatial notation (e.g. in matrix notation). In addition, the Rule of Order for the unfolding of the processes (e.g. brackets, multiplication/division and addition/subtraction) means that the Operative processes do not unfold sequentially in a left to right format (see O’Halloran, 2005a, Chapter 5, for further discussion). The new systems for encoding meaning are apparent if software applications for mathematical notation are examined. The grammatical systems in mathematical symbolic notation mean that the description of the relations can be rearranged for the solution to problems. For example, the equation in Figure 10.4 (Newton 1736: 67) may be rearranged to obtain an expression for y. x3 = by2 3

y2 = x b 3 y= x b

√

Embedded process/participant configurations are indicated below using squared brackets [[ . . . ]]: [[x × x × x]] = [[b × y × y]] [[y × y]] = [[ y = [[

[[x × x × x]] ]] b

√[[ [[x × bx × x]]]] ]]

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

•

O’HALLORAN 229

The symbolism deals with a limited experiential field primarily consisting of Relational and Operative processes and variable participants. The aim of encoding spatio-temporal relations of matter for predictive purposes means the range of circumstances which need to be symbolized is similarly narrow. Interpersonal meaning is steady and repetitive; giving information, and a restricted range of commands with a consistent level of modality with a high truth value. The net gain is an expansion of experiential meaning within a given domain accompanied by an increased proficiency for logical meaning made possible through the potential of the symbolism to reorganize the relationships via a grammar dedicated to this purpose. The basis for the logical relations is preestablished mathematical axioms, laws, theorems and results. Patterns of variation are captured in mathematics and science, but only within the semantic domains of Operative processes, participants and circumstance. This is the Descartes’ price, the capture of dynamic relations at the expense of the rich semantic realm of the everyday. Natural language has a larger repertoire of meanings, but the symbolism extends the semantics afforded by language in terms of capturing the patterns of relations within the confines of systems which are de-contextualized from the complexity of material reality. The grammar and semantics of mathematical symbolic notation require further investigation, including exploration of how this form of semiosis combines with visual imagery and language to construct a scientific view of the world. However, at this stage we can see that the price for the semantic expansions afforded by mathematics and science is limitations with respect to the field of the description and the complexity of the situational context. The payoff is the discipline of mathematics which expands through proliferation, thus offering new tools for scientific description, explanation and prediction of empirical data. The strong grammar of mathematics looks inwards, but the semiotic tools are employed in the service of science which looks outwards to the external world. The progress of science, the building of a complex hierarchical knowledge structure with increasing powers of description and prediction, is the result of the ‘left hand right hand’ relationship of mathematics to science. It is hoped that the SF-MDA approach contributes to an understanding of that relationship. Finally, SF-MDA social semiotic perspective is promising with regard to understanding the relations between mathematics, science and technology, and for developing the practices through which mathematics and science can be more effectively taught in schools. These two issues are discussed in turn below.

Visibility in the computer age The impact of technology cannot be downplayed with respect to the development of mathematics and science. The printing press in the mid-fifteenth century led to the development of mathematical symbolic notation because, for the first time, algorithms and symbolic notation could be extensively studied, developed and standardized (Eisenstein 1979; O’Halloran 2005a; Swetz 1987).

230

LANGUAGE, KNOWLEDGE AND PEDAGOGY

While the printing press aided the development of the grammar of mathematical symbolism and the spread of the Hindu Arabic numerical system, the printed visual images were initially difficult to reproduce. In the ongoing development of modern science, the emphasis was directed towards the symbolic notation and the written word (Davies et al. 1990). The symbolic was thus favoured over the visual image. In other fields the race between the ‘transtextual’ and the ‘transvisual’ continued until the advent of the audio-visual mix (Virilio 1994). Davies et al. (1990), for example, discuss the ‘picture explosion’ from the 1950s onwards arising from television, video, the digital computer, the radio telescope, remote sensing technology and the Internet. The ‘picture explosion’ is a ‘meaning potential explosion’ afforded by manipulable dynamic visual imagery. Development in semiotic forms of representation, and thus knowledge, are linked to technology. Virilio (1994: 4) explains that instruments have a long tradition in the quest for visibility: ‘Al-Hasan ibn al-Haitam aka Alhazen’s camera obscura in the tenth century, Roger Bacon’s instruments in the thirteenth, the increasing number of visual prostheses, lenses, astronomic telescopes and so on from the Renaissance on’ resulted in the ‘imperative to re-represent oneself, the imaging of the imagination which was such a great help in mathematics according to Descartes . . .’. Visual images and scientific instruments were accompanied by threedimensional geometrical models which were popular in the 1800–1860s in Europe, and models made from ‘wood and plaster, wire and paper, and glass and brass’ appeared in the late 1800s in Germany (Galison 2002: 304). Today, scientific visualizations are produced in a variety of forms through computergenerated imagery. The use of computers is changing the face of mathematics and science. For example, mathematical results are increasingly visualized using computerized numerical routines. These visualizations are slowly replacing the traditional symbolic solutions to problems favoured by Descartes and Newton. ‘Conjuring with fluid patterns was one way to cope with dynamic phenomena eluding the reductionist thinking that had dominated the mathematical sciences since Descartes and Newton’, claims Waldrop (1992: 13), author of Complexity: The Emerging Science at the Edge of Order and Chaos, cited in Stafford (1996: 94). Similarly, mathematical results are increasingly visualized using computerized numerical routines which have replaced analytical solutions. Virilio (1994: 76) expects that ‘synthetic imagery’ (as a mode of statistical representation) should soon be contributing to the development of the digitalized visual image as a form of reasoning. A systems-based multimodal grammatical approach to mathematics and science is sensitive to such changes in orientation, which are afforded through computer technology. Virilio (1994: 69–70) speaks of the current drive from visibility to invisibility in the military-industrial complex, where ‘. . . what is perceived, is already finished . . .’. ‘Repeating weapons’ have been replaced with ‘repeating images’ in the form of computer graphics for the automatic recognition of shapes; the ability to remotely scan, detect and track objects means that once an object is located, it may be destroyed (Virilio 1994: 70). The investment in decoy research is aimed at hiding real targets through deceptive techniques.

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

•

O’HALLORAN 231

It is a war of images and sounds, rather than objects and things, in which winning is simply a matter of not losing sight of the opposition. The will to see all, to know all, at every moment, everywhere, the will to universalised illumination; a scientific permutation on the eye of God, which would forever rule out the surprise, the accident, the irruption of the unforeseen. (Virilio 1994: 70)

While there has been a ‘shift of interest from the thing to its image, and especially from space to time, to the instant . . . [from the] real-figurative dichotomy to the more relative actual-virtual’ (Virilio 1994: 70), tension still remains over the roles of the visual image versus the symbolism in mathematics and science (e.g. Davies et al. 1990; Davis 1974; Galison 2002; Miller 1984; O’Halloran 2005a). Despite the recent innovations in computer graphics, solutions are usually presented using text-based analytical means rather than visual imagery. The traditional prejudice against the use of visual images for establishing proofs (Davis, 1974) is gradually being eroded through the meaning potential made available though computer-generated imagery and the ongoing expansion of new fields of mathematics, such as chaos theory and non-linear dynamical systems theory. Multimodal grammaticality and the systems-based SF-MDA approach foreground mathematical and scientific knowledge as the product of interacting systems involving technology and political, business and military interests. Complexity is acknowledged, for example, in the shift from conceptualizing language as an isolated phenomenon to understanding language as a form of semiosis which developed – and makes meaning – in combination with other semiotic resources. The systems-based multimodal grammatical approach promises new insights across a wide range of fields, including mathematics and science pedagogy.

Pedagogical implications of an SF multimodal grammatical approach Some students successfully learn the discourse of mathematics and science, while many are relegated to failure. Patterns of achievement can be mapped to schools which differ in terms of school type (e.g. private versus state), social class, funding, resources and physical location. Students in private schools and state schools which draw from affluent socio-economic populations enjoy greater success than those students from low socio-economic backgrounds. Increasingly, the political trend is to support these differences through funding and tax breaks, creating sectors within the education system which range from the private/academic to the state/semi-quasi-academic-vocational. Within this framework, cognitive ability is an acceptable guise for making invisible the political practices which enable wealth and status to be one major factor in success at school. Muller (2004) and others (e.g. Bourne 2003) believe that successful teaching/learning rests with the teacher. Muller (2004) states ‘teacher competence is by far the most important factor in learner attainment; the second is that in-service teacher training has had almost no effect’. The differences in

232

LANGUAGE, KNOWLEDGE AND PEDAGOGY

teaching/learning practices in mathematics classrooms differentiated on the basis of socio-economic status have been documented (O’Halloran 2004, forthcoming). As the divide grows between different types of schools, so do the experience, qualifications and salary of the teaching staff. These trends point to the urgent need to develop theoretical and practical approaches for developing effective teaching strategies, particularly for those teachers working with disadvantaged students. One answer lies with the systems-based SF-MDA approach, where recontextualizing practices in the classroom would include a discussion of the nature of mathematical and scientific knowledge as seen intra- and inter-semiotically through the grammar of visual images, mathematical symbolism and language. This would include an explanation of the meaning potential of mathematics and science, leading to an appreciation of how mathematics and science extend our understanding of the world, but only within limited domains. Descartes’ price needs to be explained. Teachers can be provided with such knowledge so that students understand what mathematics and science can do, and the means through which this is achieved. In addition, the systems-based approach takes into account how politics, business, the military and technology impact on the development of mathematics and science. Mathematics and science education becomes a discourse of critique, for an understanding of the mathematical and scientific view of the world means an appreciation of realms to which it can be reasonably applied. The scientific reductiveness of the social sciences, for example, masks the complexity of the everyday human realm and preserves the inequalities generated through current political, economic, military and business-based practices. Finally, the SF-MDA approach reveals the metaphorical nature of pedagogical discourse, where language serves as the metalanguage for discussing mathematical and scientific knowledge (O’Halloran 2000, 2004, forthcoming).

Conclusion Mathematics and science impact on our life every day. ‘The three great authorities over life and death’, ‘the law, the army and medicine’, with ‘the power to detain, destroy and objectify’ (Virilio 1994: 43) embrace technology and a scientific view of reality. Enlightenment did not deliver the promised results (Horkheimer 1972; Horkheimer and Adorno 1972). Our material reality is the product of science, and we ourselves are increasingly the scientific subjects of physical, psychological, social and cultural innovation (Foucault 1991). Narratives of space, time and causality impact on lives in western consumer-orientated societies in the form of day-to-day schedules and routines (see Tufte 1990). Our lives are literally the product of mathematics and science, discourses we hardly understand given the limited access to the semiotic tools of their construction: mathematical visual images, symbolism and scientific language. One place to address this imbalance is the education system, in order to make new generations aware of alternatives to the path we ourselves have chosen with regard to the development and uses of mathematics, science and technology.

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

•

O’HALLORAN 233

Acknowledgements Figure 10.1a is reproduced with permission from Bollmann-Bildkarten-Verlag GmbH & Co. KG, Lilienthalplatz 1, 38108 Braunschweig, Germany. Figures 10.3–10.5 are reproduced with permission of the Bodleian Library, University of Oxford (Reference: Newton (1736), Call Number: SAVILE JJ 11). Figure 10.6a is reproduced with permission of Dawson Books, Rushden, UK. Figures 10.6b–c reproduced with permission from Elsevier Ltd, Oxford, UK.

References Baldry, A.P. (ed.) (2000), Multimodality and Multimediality in the Distance Learning Age. Campobasso, Italy: Palladino Editore. Baldry, A.P. and Thibault, P.J. (2006), Multimodal Transcription and Text Analysis. London: Equinox. Barry, J. (1996), Measures of Science: Theological and Technical Impulses in Early Modern Thought. Illinois: Northwestern University Press. Bell, E.T. (1987), Mathematics: Queen and Servant of Science (first published 1951). Redmond, WA: Tempus Books of Microsoft Press. Bernstein, B. (1990), Class, Codes and Control, Volume IV: The Structuring of Pedagogic Discourse. London: Routledge. Bernstein, B. (2000), Pedagogy, Symbolic Control and Identity: Theory, Research and Critique. Lanham: Rowman & Littlefield. Bourne, J. (2003), ‘Vertical discourse: the role of the teacher in the transmission and acquisition of decontextualized language’, in European Educational Research Journal 2(4), pp. 496–521. Cheong, Y.Y. (2004), ‘The construal of ideational meaning in print advertisements’, in O’Halloran, K.L. (ed.), Multimodal Discourse Analysis. London and New York: Continuum, pp. 163–95. Christie, F. (ed.) (1999), Pedagogy and the Shaping of Consciousness: Linguistic and Social Processes. London and New York: Continuum. Christie, F. (2002), Classroom Discourse Analysis: A Functional Perspective. London and New York: Continuum. Christie, F. and Martin, J.R. (eds) (1997), Genre and Institutions: Social Processes in the Workplace and School. London and New York: Continuum. Coffin, C. (1997), ‘Constructing and giving value to the past: an investigation into second school history’, in Christie, F. and Martin, J.R. (eds), Genre and Institutions – Social Processes in the Workplace and School. London: Continuum, pp. 196–230. Davies, D., Bathurst, D. and Bathurst, R. (1990), The Telling Images: The Changing Balance between Pictures and Words in a Technological Age. Oxford: Clarendon Press. Davis, P.J. (1974), ‘Visual geometry, computer graphics and theorems of perceived type’, in Proceedings of Symposia in Applied Mathematics 20, pp. 113–27. Descartes, R. (1683), Geometria, anno 1673 Gallice Edita / Postea Cum Notis Fl. de Beaune in Latinam Linguam Versa / et Commentariis Illustrata Opera Francisci a Schooten: Volume 1. Amsterdam. Descartes, R. (1952), Descartes’ Philosophical Writings (ed. and tr. Norman Kemp Smith). London: Macmillan. Descartes, R. (1954), The Geometry of Rene Descartes. New York: Dover. Descartes, R. (1985), The Philosophical Writings of Descartes: Volume 1. Cambridge: Cambridge University Press. Eisenstein, E.L. (1979), The Printing Press as an Agent of Change. Cambridge: Cambridge University Press.

234

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Elkins, J. (1999), The Domain of Images. Ithaca, New York: Cornell University Press. Foucault, M. (1991), Discipline and Punish. London: Penguin. Funkenstein, A. (1986), Theology and the Scientific Imagination from the Middle Ages to the Seventeenth Century. New Jersey: Princeton University Press. Galison, P. (2002), ‘Images scatter into data, data gathers into images’, in Latour, B. and Weibel, P. (eds), Iconoclash: Beyond the Image Wars in Science, Religion, and Art. Cambridge, MA: The MIT Press, pp. 300–23. Galvani, L. (1792), ‘Electrophysiology: experiments with frogs’, in De Viribus Electricitatis. Plate 1. Golovin, A.A., Matkowsky, B.J. and Nepomnyashchy, A.A. (2003), ‘A complex SwiftHohenberg equation coupled to the Goldstone Mode in the non-linear dynamics of flames’, in Physica D 179, pp. 183–210. Guo, L. (2004), ‘Multimodality in a Biology textbook’, in O’Halloran, K. L. (ed.), Multimodal Discourse Analysis. London: Continuum, pp. 196–219. Habermas, J. (1974), Theory and Practice. Boston: Beacon Press. Halliday, M.A.K. (1978), Language as Social Semiotic: The Social Interpretation of Language and Meaning. London: Edward Arnold. Halliday, M.A.K. (1993), ‘Some grammatical problems in scientific english’, in Halliday, M.A.K. and Martin, J.R. (eds), Writing Science: Literacy and Discursive Power. London: Falmer Press, pp. 69–85. Halliday, M.A.K. (2004a), Introduction to Functional Grammar. (third edn revised by C.M.I.M. Matthiessen; first edn 1985, second edn 1994). London: Arnold. Halliday, M.A.K. (2004b), ‘The Language of Science’, in Collected Works of M.A.K. Halliday, edited by J. Webster. London and New York: Continuum. Halliday, M.A.K. and Martin, J.R. (1993), Writing Science: Literacy and Discursive Power. London: Falmer. Halliday, M.A.K. and Matthiessen, C.M.I.M. (1999), Construing Experience through Meaning: A Language Based Approach to Cognition. London: Cassell. Horkheimer, M. (1972), Critical Theory. New York: The Seabury Press. Horkheimer, M. and Adorno, T.W. (1972), Dialectic of Enlightenment. New York: The Seabury Press. Iedema, R. (2001), ‘Resemioticization’, in Semiotica 137, 1/4, pp. 23–39. Iedema, R. (2003), ‘Multimodality, resemioticization: extending the analysis of discourse as a multisemiotic practice’, in Visual Communication 2(1), pp. 29–57. Koenderink, J. (1993), ‘Foreword’, in Ellis, S.R. (ed.), Pictorial Communication in Virtual and Real Environments. London: Taylor & Francis, pp. viii–ix. Koestler, A. (1959), The Sleepwalkers: A History of Man’s Changing Vision of the Universe. Harmondsworth: Penguin. Kress, G. (2003), Literacy in the New Media Age. London: Routledge. Kress, G., Jewitt, C., Ogborn, J. and Tsatsarelis, C. (2001), Multimodal Teaching and Learning: The Rhetorics of the Science Classroom. London: Continuum. Kress, G. and van Leeuwen, T. (1996), Reading Images: The Grammar of Visual Design. London: Routledge. Kuhn, T.S. (1970), The Structure of Scientific Revolutions. Chicago: University of Chicago Press. Latour, B. and Weibel, P. (eds) (2002), Iconoclash: Beyond the Image Wars in Science, Religion, and Art. Karlsruhe: ZKM. Lemke, J.L. (1998), ‘Multiplying meaning: visual and verbal semiotics in scientific text’, in Martin, J.R. and Veel, R. (eds), Reading Science: Critical and Functional Perspectives on Discourses of Science. London: Routledge, pp. 87–113. Lemke, J.L. (2002), ‘Travels in Hypermodality’, in Visual Communication 1(3), pp. 299–325. Lemke, J.L. (2003), ‘Mathematics in the middle: measure, picture, gesture, sign, and word’, in Anderson, M., Sáenz-Ludlow, A., Zellweger, S. and Cifarelli, V.V. (eds),

MATHEMATICAL & SCIENTIFIC FORMS OF KNOWLEDGE

•

O’HALLORAN 235

Educational Perspectives on Mathematics as Semiosis: From Thinking to Interpreting to Knowing. Ottawa: Legas, pp. 215–34. Lester, P.M. (2000), Visual Communication: Images with Messages. Belmont, CA: Wadsworth/Thompson Learning. Martin, J.R. (1992), English Text: System and Structure. Amsterdam: John Benjamins. Martin, J.R. (2002), ‘Fair trade: negotiating meaning in multimodal texts’, in Coppock, P. (ed.), The Semiotics of Writing: Transdisciplinary Perspectives on the Technology of Writing. Turnhout, Belgium: Brepols, pp. 311–38. Martin, J.R. and Rose, D. (2003), Working with Discourse: Meaning Beyond the Clause. London and New York: Continuum. Martin, J.R. and Veel, R. (eds) (1998), Reading Science: Critical and Functional Perspectives on Discourses of Science. London: Routledge. Martinec, R. (2000), ‘Types of processes in action’, in Semiotica 130–3/4, pp. 243–68. Martinec, R. (2001), ‘Interpersonal resources in action’, in Semiotica 135–1/4, pp. 117–45. Martinec, R. (2004), ‘Gestures that co-concur with speech as a systematic resource: the realization of experiential meanings in indexes’, in Social Semiotics 14 (2), pp. 193–213. Miller, A.I. (1984), Imagery in Scientific Thought: Creating Twentieth-Century Physics. Boston: Birkhäuser. Muller, J. (2000), Reclaiming Knowledge: Social Theory, Curriculum and Educational Policy. London: Routledge. Muller, J. (2004), ‘On the shoulders of giants: a digression on knowledge, curriculum, and finally, the teacher’. Plenary paper at Reclaiming Knowledge Conference. Faculty of Education, University of Sydney, 13–15 December 2004. Newton, I. (1736), The Method of Fluxions and Infinite Series; with its Application to the Geometry of Curve-Lines; With Comments by John Colson. London: printed by Henry Woodfall and sold by John Nourse. Newton, I. (1968), Mathematical Principles of Natural Philosophy. London: Dawson. O’Halloran, K.L. (1999), ‘Towards a systemic functional analysis of multisemiotic mathematics texts’, in Semiotica 124–1/2, pp. 1–29. O’Halloran, K.L. (2000), ‘Classroom discourse in mathematics: a multisemiotic analysis’, in Linguistics and Education 10(3), pp. 359–88. O’Halloran, K.L. (2003a), ‘Educational implications of mathematics as a multisemiotic discourse’, in Anderson, M., Sáenz-Ludlow, A., Zellweger, S. and Cifarelli, V. V. (eds), Educational Perspectives on Mathematics as Semiosis: From Thinking to Interpreting to Knowing. Ottawa: Legas, pp. 185–214. O’Halloran, K.L. (2003b), ‘Intersemiosis in mathematics and science: grammatical metaphor and semiotic metaphor’, in Simon-Vandenbergen, A.-M., Taverniers, M. and Ravelli, L. (eds), Grammatical Metaphor: Views from Systemic Functional Linguistics. Amsterdam: John Benjamins, pp. 337–65. O’Halloran, K.L. (2004), ‘Discourses in secondary school mathematics classrooms according to social class and gender’, in Foley, J. A. (ed.), Language, Education and Discourse: Functional Approaches. London and New York: Continuum, pp. 191–225. O’Halloran, K.L. (2005), Mathematical Discourse: Language, Symbolism and Visual Images. London and New York: Continuum. O’Halloran, K.L. (forthcoming), A Multimodal Approach to Classroom Discourse. London: Equinox. O’Toole, M. (1994), The Language of Displayed Art. London: Leicester University Press. Peirce, C.S. (1991), ‘Prolegomena to an apology for pragmaticism’, in Hoopes, J. (ed.), Peirce on Signs: Writings on Semiotic by Charles Sanders Peirce. Chapell Hill: University of North Carolina Press, pp. 249–52. Prigent, A., Gregoroire, G., Chate, H. and Dauchot, O. (2003), ‘Long-wavelength modulation of turbulent shear flows’, in Physica D 174, pp. 100–13. Rotman, B. (1995), ‘Thinking dia-grams: mathematics, writing and virtual reality’, in South Atlantic Quarterly 94(2), pp. 389–415.

236

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Royce, T. and Bowcher, W. (eds) (in press), New Directions in the Analysis of Multimodal Discourse. New Jersey: Lawrence Erlbaum Associates. Saint-Martin, F. (1990), Semiotics of Visual Language. Bloomington: Indiana University Press. Simon-Vandenbergen, A.-M., Taverniers, M. and Ravelli, L. (eds) (2003), Grammatical Metaphor: Views from Systemic Functional Linguistics. Amsterdam: John Benjamins. Stafford, B. M. (1989), ‘The new immaterialism: Kenneth Snelson imagines the atom’, in Kenneth Snelson: The Nature of Structure (Exhibition Catalogue). New York: New York Academy of Sciences, pp. 51–7. Stafford, B.M. (1996), Good Looking: Essays on the Virtue of Images. Cambridge, MA: MIT Press. Swetz, F.J. (1987), Capitalism and Arithmetic. La Salle, IL: Open Court. Thibault, P. (2000), ‘The multimodal transcription of a television advertisement: theory and practice’, in Baldry, A. P. (ed.), Multimodality and Multimediality in the Distance Learning Age. Campobasso, Italy: Palladino Editore, pp. 311–85. Thibault, P. (2001), ‘Multimodality and the school science textbook’, in Carol Taylor Torsello, G.B., and Nicoletta Penello (eds), Corpora Testuali per Ricerca, Traduzione e Apprendimento Linguistico. Padua: Unipress, pp. 293–335. Tufte, E.R. (1990), Envisioning Information. Cheshire: Graphics Press. Unsworth, L. (2000), ‘Investigating subject-specific literacies in school learning’, in Unsworth, L. (ed.), Researching Language in Schools and Communities. London and New York: Cassell, pp. 245–74. Unsworth, L. (2001), Teaching Multiliteracies across the Curriculum: Changing Contexts of Text and Image in Classroom Practice. Buckingham, UK: Open University Press. van Leeuwen, T. (1999), Speech, Music, Sound. London: Macmillan. van Leeuwen, T. (2005), Introducing Social Semiotics. London: Routledge. Veel, R. (1999), ‘Language, knowledge and authority in school mathematics’, in Christie, F. (ed.), Pedagogy and the Shaping of Consciousness: Linguistic and Social Processes. London: Cassell, pp. 185–216. Ventola, E., Charles, C. and Kaltenbacher, M. (eds) (2004), Perspectives on Multimodality. Amsterdam: John Benjamins. Virilio, P. (1994), The Vision Machine. Bloomington: Indiana University Press. Waldrop, M.M. (1992), Complexity: The Emerging Science at the Edge of Order and Chaos. New York: Simon & Schuster. Wigner, E.P. (1960), ‘The unreasonable effectiveness of mathematics in the natural sciences’, in Communications on Pure and Applied Mathematics 13, pp. 1–14 (reprinted in Mickens, R. E. (1990), Mathematics and Science. Singapore: World Scientific, pp. 291–306). Wildbur, P. (1989), Information Graphics: A Survey of Typographic, Diagrammatic and Cartographic Communication. New York: Van Nostrand Reinhold. Wildbur, P. and Burke, M. (1998), Information Graphics: Innovative Solutions in Contemporary Design. London: Thames and Hudson.

Section 3

Research Prospects – Exploring Uncommon Sense

This page intentionally left blank

11 Taking stock: future directions in research in knowledge structure Frances Christie, J.R. Martin, Karl Maton and Johan Muller

We have organized this closing stage of our conversation here as a kind of interview, based on an agenda of issues which we hope will inspire future research. The contributions are from the editors, Fran Christie and Jim Martin, along with Karl Maton and Joe Muller, who converse on a first name basis below.

1. Why talk? Jim: As we signal early in this book, we’ve tried here to push along the negotiation of ideas initiated by Bernstein, Halliday and Hasan into a second generation of research. What is it that enables this conversation to continue, across disciplines which usually have very little to say to one another? This is my own third intense engagement with Bernstein’s thinking. The first stemmed from my concern with educational failure, and its relation to the coding orientations privileged or not by institutionalized learning; the second had to do with our struggle to dislodge the hegemonic position of progressive education as far as literacy teaching was concerned; this time round we’re focusing on hierarchy and knowledge structure, and its implications for learner pathways in school. It seems to me that while my functional linguistics colleagues and I are good at seeing how meaning is realized in texts, we’re very much weaker at understanding how meaning is distributed in society. The ways Bernstein and his colleagues talk about this unfolds for me as a kind of revelation, and an essential underpinning for any kind of intervention we try to make with social justice in mind. It is hard to read across our respective knowledge structures; but without this reading we can’t engineer socially significant changes to educational practice – we tend to waste our time. Fran: My path into the issues here was different from Jim’s. As a young school teacher I developed an interest in language in two senses: firstly, I found that traditional school grammar was not relevant for most of the children I taught; secondly, I became interested in the ways teachers and children used language in schools, and I observed that children performed in differential ways which needed explanation in some systematic way. It was through these interests that I turned to Halliday’s SFL. It offered both a possible pedagogic grammar and a tool for the analysis of classroom talk. The interest in classroom talk led me to Bernstein’s work on pedagogy, and thence to notions of knowledge structure.

240

LANGUAGE, KNOWLEDGE AND PEDAGOGY

Bernstein’s unfinished work on a theory of knowledge structure provided a powerful response to the spread of progressivist and constructivist theories of curriculum, which had taken grip on much Western schooling by the end of the twentieth century. What enables us to keep talking is that the two disciplines provide a theoretical frame for addressing the nature of knowledge structure and of curriculum design in such a way that each informs the other. Joe: The work of Halliday and SFL has until recently been a part of my consciousness only in a general sort of way. It is only since my colleagues and I have been working on rendering the knowledge structure of curricula visible, using Bernstein’s theoretical tools, that I began to read the work of Fran Christie, Jim Martin and their colleagues seriously. I found to my surprise that they had made considerable progress in making visible the metalinguistic structure of disciplines and curricula. This enterprise, I found, is not only directly parallel to our work in curriculum but shares many of the same epistemological and political convictions. And this work makes clear not only that knowledge lives in texts, but how. The advance is considerable. Karl: The question of what enables our conversation to continue is crucial. Contemporary social science is replete with calls for ‘multidisciplinary’, ‘interdisciplinary’, even ‘post-disciplinary’ work, but all too often these are realized as non-disciplinary monologues rather than fruitful cross-disciplinary dialogue. For me, three attributes of SFL and Bernstein sociology stand out here. First, both share what Bernstein described as an allegiance to a problem rather than to an approach. Intellectuals in both traditions are willing to look beyond the confines of their knowledge structures for conceptual tools that enable them to explain better that part of the world they focus on. Second, they both attempt to generate strong external languages of description, concepts that get to grips with problems in empirical research. This gives them the possibility of a shared purchase on the world, enabling dialogue. For a sociologist, SFL is extremely impressive for its technical armoury – its conceptual framework is elaborate, detailed and aims for clarity. Lastly, both theories downplay the significance of who is speaking and instead focus on the explanatory power of their approaches in addressing defined problems. The degree of detail SFL can analyse in texts is eye-opening to me, and working with linguists has made me more aware of the significance of language to my own interests. Whether we borrow from each other’s conceptual toolkits or manage to integrate the two theories, or perhaps evolve from cross-fertilization towards integration, is something to be seen – it’s an exciting prospect. Politically, what impresses me about both theories is that they start with questions rather than with solutions – what Isaiah Berlin called a culture of consequence, rather than one of commitment, recognizing it is not enough to be well intentioned, one also needs epistemologically powerful knowledge.

FUTURE DIRECTIONS IN RESEARCH

241

2. Horizontal and vertical discourse 2.1 The borders of horizontal and vertical discourse Jim: One thing I remain puzzled about, not addressed in this book, is why Bernstein places crafts in vertical discourse. Since he characterizes these as having weak grammars and tacit transmission (showing and modelling proceeding doing), I expect he would include trades, hobbies, sport and recreation here too. For various reasons, I’d prefer to treat these as horizontal discourses of a specialized kind. They’re learned the same way as other horizontal discourses (via participation and ostensive definition), don’t traditionally involve institutionalized learning and don’t depend on grammatical metaphor to construct their expertise. What are the grounds for considering discourses of this kind ‘vertical’? This is a very important educational issue since radical progressive and constructivist educators seem to believe that all vertical discourses have to be reconstructed as crafts in order for students, especially younger or less successful students, to really learn them. Fran: Well, however we address the last issue that Jim raises here, I don’t think it will be by revisiting what Bernstein said of the crafts in vertical discourse. The fundamental issue that faces us for pedagogy is: how do we make it possible for people to move from horizontal to vertical discourse? As a first step we need better teacher education, which, among other things, must be more upfront about addressing the pedagogic discourse of schooling, and more politically sensitive in the best sense. Thus, for example, it will pay more attention to difference (though not the rather vacuous notions of ‘individual difference’ with which educational psychologists have long had a field day!). It will acknowledge the different meaning codes that children act with when they come to school, and it will not, like much twentieth-century educational theory, subscribe to a rather naive notion of the idealized subject who is ‘self-regulating’, and who develops benignly, with little assistance from formal instruction (Bernstein 2000: 43). It will also, for reasons explored by both Joe and Karl (this volume), not disguise the essentially hierarchical nature of much school knowledge, making the different knower codes and knowledge codes more explicitly available to students. Here is a nice instance of the ways the two disciplines assist each other: the sociology of Bernstein, Muller and Maton provides a theoretical frame to identify the problem, while the SFL theory provides a language for addressing it. Joe: Through a 2004 PhD study of a master-apprentice relation in cabinet making (a trade close to the carpentry-woodwork examples Bernstein often uses), Gamble illuminates the nature of a tacit knowledge base in craft. Her work shows that while the pedagogic outcome is an external performance that must meet explicit and exact criteria, this performance rests crucially on an internalized competence – or a capacity to visualize the relationship between parts and whole, in both space and time. Visualization stands in place of a nonarticulable or ineffable ordering principle that acts as a ‘glue’ to make all the segments hang together. It is this, what Polanyi calls an ineffable relationship

242

LANGUAGE, KNOWLEDGE AND PEDAGOGY

between particulars jointly forming a whole, even though all the particulars are explicitly specifiable, that makes modelling the only possible transmissionacquisition practice. Visualization thus compensates for the lack of a clear syntax by allowing a formal principle of arrangement to be grasped in embodied form. This goes for the generation of novelty too: as Bernstein explained in an email, ‘There will be a strong visualizing of the new. That may be why it is difficult to explain in words. You “see” the design rather than verbalize it.’ This makes craft homologous to knowledge forms, which although far more strongly developed in terms of vertical and codified chains of abstraction, share a common feature in that they make the jump from ‘token’ to ‘type’ and thus generate meaning at a remove from the everyday. Bernstein acknowledged in an email the difficulty of talking about craft as having an ‘internal grammar’, and proposed the term ‘condensed recontextualization’ for the tacit technical principles of craft operation. Constructivist interpretations of apprenticeship subsume the nature of the knowledge transmitted and the form of its transmission into the notion of ‘shared or situated practice’ that fits with post-modern versions of heterogeneous local knowledges not transferable to context-independent knowledge. In this version the notion of apprenticeship becomes so pervasive that there is nothing that is not apprenticeship. The term loses its meaning. Karl: Fran’s question is the key issue here: what enables the move into vertical discourse? Is it helped by recontextualizing existing disciplines to become more homologous to horizontal discourse? As Fran suggests, it requires making the codes underlying curriculum and knowledge structures more explicit – we need to work out ways of giving different kinds of pupils the keys to the code. Not all pupils arrive with the means to recognize or realize the code required to learn successfully. We need to work at creating forms of curriculum and pedagogy that provide pupils with what we possess, as quickly as possible. Calls to change education to be more like everyday knowledge fail to understand the difference between horizontal and vertical discourses; in effect, they call for an end to education – they want to make the sacred profane. In the field of information technology education research, for example, a widely made argument currently is that children are ‘digital natives’ who arrive at school with immense amounts of knowledge. The argument that they already know how to use mobile phones, computers, ipods, etc. is then used to justify moves to peergroup learning and a reconstruction of the curriculum around everyday uses of technology. Such arguments can only be maintained so long as one does not have a theory of education or knowledge. It relies on negating differences between horizontal discourse and its forms of circulation, and vertical discourse and its specific forms of pedagogy. What research is beginning to show in this example is that such moves to ‘authentic learning contexts’ and peer-teaching tend to leave pupils where they began: with only context-dependent practical mastery rather than context-independent symbolic mastery. This is why recognizing the differences between horizontal and vertical discourse is crucial for enabling pupils to succeed in education. Here SFL and sociology have a potentially crucial role to play in enabling everyone to learn.

FUTURE DIRECTIONS IN RESEARCH

243

2.2 Grammatical metaphor and vertical discourse Jim: From the perspective of functional linguistics, grammatical metaphor is the key resource used to construct the uncommon sense knowledge of vertical discourses and is characterized in functional linguistics as involving stratal tension, with grammar symbolizing alongside realizing semantics, and the meaning of the metaphor involving its grammatical reading, its semantic reading and the import of the tension between the two. The technicality of the functional linguistic account makes the concept a challenging one to explain to outsiders. Does anyone have any thoughts on strategies we might use to make this crucial understanding accessible to academics in other disciplines, and equally importantly for teachers and students of vertical discourses in secondary school? Fran: This is a hard issue. We already know, from curriculum experience in Australia at least, that school teachers in discipline areas like the sciences are more disposed than their counterparts in the humanities and social sciences to engage with technical discourse, and this reflects their training in the university disciplines. There is, in other words, a relationship between the discourse of the university disciplines and their recontextualized forms in schools, although as Jim notes, under the impact of constructivist models of curriculum, some accounts of the secondary science curriculum have shed their commitment to a vertical discourse. The dilemma for those who espouse constructivist models of curriculum in all areas of schooling is that they actually look for and reward vertical discourse in their teaching and their assessment. This is probably the way into persuading academics and school teachers in the various areas of knowledge to have another look at what they are doing: identify instances of text types that teachers reward in their students’ work, bring some of their linguistic features to consciousness, and turn this understanding into teaching practice. Karl: Excellent question, Jim. The truth is no guarantee of belief. Two interrelated issues spring to mind here: appealing to the people we wish to help, and making our technical language accessible. When presenting papers or writing in sociology, education or cultural studies, I sometimes hear reactions to the relatively simple conceptual framework I employ, suggesting the world is more complex than the concepts suggest or implying the analysis is somehow cold and needs more warm bodies of knowers. My own strategy is to appeal to the rational, to the fact that in higher education we are still ostensibly concerned with understanding and explaining the world. I simply ask whether the analysis seems to make sense, could the concepts be of use in explaining issues of concern, and if not let’s come up with something better. Similarly, as Fran says, one can appeal to the desire of teachers to help pupils learn. It’s an obvious point but, in terms of making a technical language more accessible, we need different ways of talking to different audiences. In speaking across the languages of the horizontal knowledge structure of social science as a whole, one has to be a little multilingual, or at least appeal to the codes they operate with. For an audience of academics unused to the kind of language we use in the Bernstein tradition, I employ concepts such as ‘knowledge code’ or ‘knower code’ in a less explicitly defined way (such as ‘knower approach’ or simply ‘emphasizes the

244

LANGUAGE, KNOWLEDGE AND PEDAGOGY

knower’), without using classification and framing or epistemic relation and social relation. However, this may be easier to do than for systemic linguists, because our language is less technical, or at least less extensive. I’m not saying we should work against the technicalization of our discourse by dumbing down or not using technical language at all, but simply highlighting that academics and teachers may be operating with different codes from our own, which may require different modes of expression.

3. Horizontal and hierarchical knowledge structures 3.1 Verticality and grammaticality Jim: Joe (this volume) develops Bernstein’s notion of hierarchality with respect to what he calls grammaticality and verticality: ‘grammaticality determines the capacity of a theory or a language to progress through worldly corroboration; verticality determines the capacity of a theory or language to progress integratively through explanatory sophistication. Together, we may say that these two criteria determine the capacity of a particular knowledge structure to progress.’ Following on from the discussion of grammatical metaphor above, I’m wondering about the role of technicality in determining degrees of grammaticality and verticality. Technicality involves deploying grammatical metaphor to define terms and place them in precise relationships to one another. In my chapter (this volume) I gave examples from Bernstein (e.g. ‘a horizontal discourse entails a set of strategies which are local, segmentally organised, context specific and dependent, for maximising encounters with persons and habitats’). Technical terms engendered by the process of definition subsume the knowledge created by the grammatical metaphors, but are not themselves metaphorical. In effect, technicality lightens up the discourse, making room for further grammatical metaphors to move the development of hierarchical knowledge along – and thus contribute to verticality. At the same time technical terms contribute to the development of an explicit network of precise concepts which can be tested against data – and thus contribute to grammaticality (i.e. the possibility of empirical disconfirmation). So, from a linguistic perspective, if grammatical metaphor is the watershed demarcating horizontal and vertical discourse, is distilling grammatical metaphor as technicality the key to establishing degrees of verticality and grammaticality? Fran: Well yes, I think it is, though the whole issue needs a great deal of unpacking; this is one of the issues that Mary Macken-Horarik, Karl and I would like to pursue in the future with respect to subject English – surely one of the most difficult of all subjects in the curriculum in which to propose to teachers that there is a ‘verticality’ for them to explore! The difficulty about subject English is that the language is both the instrument of teaching and learning and the object of study. If English is to be ever understood as having ‘verticality’, this will be built around what we know of grammatical metaphor. Its emergence is a development of late childhood to adolescence (though many children with literacy difficulties don’t master it), and its emergence occurs when students enter the

FUTURE DIRECTIONS IN RESEARCH

245

secondary school, with its different curriculum (though the nature of that curriculum is often subverted these days by constructivist notions). Successful control of literacy for the secondary school, in which grammatical metaphor plays a crucial role, facilitates entry to ‘uncommonsense’ experience, because it allows the writer to be distanced from the immediate experience, achieving a degree of detachment from the event. It is precisely this capacity to draw back from experience and build abstraction – be that achieved in writing a valued story, in reviewing a novel or film, or in writing an expository text on some social issue – that subject English actually rewards, though for the most part, English teachers are not aware of this. In the first instance, English teachers need to develop an understanding of the technical terminology involved in understanding how the rewarded language is constructed, and then turn this into a metalanguage they can use in teaching and working with their students. Thus, they can make rather more visible than presently applies what it is that people do with language in order to build the kinds of texts and their knowledge that subject English rewards. An infinitely trickier issue, however, will involve taking the next step: how we take an understanding of the kinds of capacities in using written language laid bare in the first step just described, and develop a coherent and integrated account of the English curriculum, such that a degree of sequence and progression across the years of schooling is achieved. This is an issue that cries out for empirical research! Joe: The issue of what would constitute technicality in horizontal knowledge structures has not been addressed. I would suspect that in disciplines that proliferate parallel languages, hence with weak subsumption, technicality is too infused with lexical metaphor to be stable, with the result that neophytes often don’t know whether they are using the perhaps imperfectly technicalized terms correctly or not. This is why students often learn usage of such discourses not by learning the principles but by mimicking the discursive style of their teachers, and in so doing acquire the disciplinary ‘gaze’, a rather weak form of consciousness specialization. Karl: This is one of the questions that I think this book really presses us on as the next stage for our thinking. To be speculative, if grammatical metaphor is a key to moving from horizontal to vertical discourse, then perhaps it is the form taken by the process this comprises that shapes differences between hierarchical and horizontal knowledge structures. As Joe is saying, in horizontal knowledge structures technicality is prolific but of a different order. Sociologists, for example, often ‘pack up’ (to be crude) a whole set of meanings, allusions, political attitudes, aptitudes and beliefs into particular terms, such as ‘ideology’, ‘hegemony’ or ‘rhizomes’. There are extensive literatures on their definitions, but in common use such terms often signal the author’s stance, with a high normative and political loading. One aspect of the difference may be that underlying grammatical metaphor in hierarchical knowledge structures is a knowledge code (see my paper, this volume), where its strongly bounded and controlled epistemic relation to the object and other knowledge helps regulate the form it takes, enabling subsumption and a stronger grammar; while, in

246

LANGUAGE, KNOWLEDGE AND PEDAGOGY

horizontal knowledge structures, technicalization may be shaped by inter alia a knower code, where this epistemic relation is weaker and more emphasis is placed on the social relation to dispositions, aptitudes, attitudes, etc. Though this gives the resulting language a weaker external grammar, the process (and the basis of legitimate insight) is still regulated, affecting the way it gets taught and the kinds of consciousness specialized. 3.2 SFL and verticality Jim: Joe (this volume) characterizes verticality as having to do with ‘how theory develops. In hierarchical knowledge structures, it develops through integration, towards ever more integrative or general propositions, the trajectory of development of which lends hierarchical knowledge structures a unitary convergent shape.’ Bernstein uses the imagic metaphor of a triangle to symbolize this convergent apicality. Triangularity is a useful metaphor for considering the progress in canonical hierarchical knowledge structures such as physics and biology, and perhaps within some of the specialized languages of certain horizontal knowledge structures (those with proliferation inhibiting verticality in Joe’s terms). But I wonder how appropriate the triangle metaphor is for a polysystemic knowledge structure such as systemic functional linguistics, and whether the nature of its object of inquiry, language (and semiosis in general) has engendered its multiperspectival path of evolution (so we end up with a prism, not a pyramid). Alongside hierarchy, in other words, I’d like to introduce here the notion of complementarity – the idea that we can’t always find ‘ever more integrative and general propositions’ but have to learn to live with (and enjoy) complementaries which taken together exhaust our understanding of the field but which cannot be reduced to a single apical insight. A familiar example of such a complementarity would be light having to be interpreted as both particle and field in physics (consider also Joe’s complementarity of verticality and grammaticality as co-determinants of knowledge structure progression). Is the image of ‘yin and yang’ in other words a possible reading or extension of ‘ever more integrative or general propositions’, alongside the image of the triangle? Joe: I would regard complementaries as belonging to the same apical node, so to speak. It’s the supplementaries that are the competitors, and that logically can’t co-exist on the same apical level. That a proposition can be broken down into independent propositional clusters doesn’t mean that the clusters are conceptually incompatible. (Recall that it is of the essence of Derrida’s approach to deconstruction that there must always be a supplement that can’t be subsumed; for Bernstein, the more hierarchical the knowledge structure, the more intolerant of supplements it has to be.) From my point of view, complementarity and hierarchy are, well, complementary. On the other hand, Bernstein was at least a closet Hegelian, developing his thoughts by antitheses followed by syntheses. Perhaps this gets us closer to yin and yang? Karl: I agree with Joe – complementarity and triangularity are not mutually exclusive. A question: are the complements (the systems) brought together at a

FUTURE DIRECTIONS IN RESEARCH

247

higher level of abstraction? If not, can they be? Why stop and declare the complements as the height of the ambition to climb? 3.3. Science, social science and humanities Jim: Joe (this volume) proposes degrees of verticality as a criterion for consigning disciplines to hierarchical or horizontal knowledge structures, and within horizontal knowledge structures to constrained proliferation or unconstrained proliferation types. Degrees of grammaticality complement this with respect to the possibilities for worldly corroboration. Does this give us adequate resources for interpreting Bernstein’s hierarchical and horizontal opposition as a cline, and satisfactorily positioning the knowledge structures of social science along it? I remain puzzled by my own field of linguistics, for example, which presents itself as the science of language and as such strives for as much verticality and grammaticality as possible. This would seem to predict that the proliferation of new theories of language would be highly constrained and that decades of research would lead to integration. It might be argued however that proliferation of incommensurable theories is in fact rampant, and that what we in fact experience is a language proliferating horizontal structure par excellence. This may simply be because linguists can’t agree on what language is, and without consensus on data verticality charges off in different directions, and grammaticality ‘tests’ these divergent models against radically different phenomena. This certainly inhibits co-operation within the discipline, and can cause confusion in the community as far as interventions and applications are concerned. Fran: Most of Jim’s last observations apply equally to English, as that is conceived in university departments of that name. All this suggests to me that we have a way to go in straightening out the degrees of verticality that are found/are possible in the hierarchical and horizontal knowledge structures in the social sciences and humanities. About this matter, some of Halliday’s observations are worth noting. Thus, he is given to saying that the semiotic sciences in the twentieth century, including linguistics, stood in relation to the objects of their study close to where the natural sciences stood to their objects of study in the sixteenth century. The twenty-first century, he has said, is the century in which the semiotic sciences will come of age, and he is of the view that linguistics will be a pivotal discipline in the endeavour. The paradigms found in his linguistics will have relevance for the other social science disciplines, though that does not preclude adoption of other paradigms. He also has said that it is no accident that the natural sciences got their acts together rather earlier than the social sciences: the phenomena of the natural world are certainly hard to study, but those of the social world – ourselves in fact – are infinitely harder (Halliday 1993). Hence he would agree with a point Joe makes (this volume) that we still are clumsy about the ways we talk about the things we do talk about – we are still shaping the intellectual tools to build our sense of the social world. Joe: I wonder whether SFL and theoretical linguistics, say, are simply two alternative languages of linguistics or whether, as is the case for sociology and anthropology, and human geography, these are disciplines which, though

248

LANGUAGE, KNOWLEDGE AND PEDAGOGY

taking the same broad class of objects, are more fairly treated as different though related disciplines? I would regard the range of approaches within sociolinguistics as different languages, for example; and the rest of us social scientists would dearly love to see a synthesis within that grouping. But I would not have thought that it could be easily achieved across all the ‘hyphen-linguistics’ in the textbook. But perhaps I mistake the integrative ambitions of SFL? Karl: For me this highlights a couple of issues. First, a discipline is not merely its knowledge structure. A strong grammar and high degree of integration and subsumption are perhaps necessary but not sufficient conditions for avoiding proliferation of languages within the knowledge structure. One has also to contend with the tradition of the intellectual field as a social field of practice, where actors are invested in different approaches. This brings us back to the earlier question of speaking across segments and the issue of persuasion. Second, perhaps what is required is a metatheory, a theory which can determine which of the languages of linguists are more ontologically sound than others. This is perhaps one reason why SFL is looking to approaches such as that of Bernstein – to provide the basis of a means for understanding what is going on in linguistics as a whole, and to help construct a ruler for determining what approaches are more or less empirically adequate and for what phenonema. Bernstein did not focus very much on issues of ontology and epistemology, though there are approaches compatible with his and, I believe, with SFL that may offer further tools for such a task. In the sociology of education, for example, some of those working with Bernstein’s approach are drawn towards what is called social realism or ‘critical realism’, a metatheory that underlabours other theories by providing a means of discussing and criteria for judging between theories on the basis of which are more or less empirically adequate. It is an ‘underlabourer’ because it describes what a theory would have to do, but not what that theory would say. This has had perhaps its biggest impact in economics, where mainstream economics has been strongly critiqued for being divorced from the real world and lacking fruitful empirical application, and the nature of theories required to make sense of the economic world are being outlined, underpinning the rise of what is being called ‘post-autistic economics’. 3.4 SFL as metalanguage Jim: What does it mean to model discourse through the metalinguistic and metasemiotic prisms of SFL? As far as the professional discourse of knowledge production is concerned, this remains of course a linguistics exercise – and linguists need to keep in mind that this modelling involves recontextualization from one discipline to another, and that the amount of verticality a discipline deploys relative to linguistics will impact strongly on the nature of the recontextualization. For pedagogic discourse, the impact of such modelling is even more challenging. For hierarchical knowledge structures like science, language educators seem to be suggesting that an additional language be introduced, in effect horizontalizing the school discipline. For horizontal knowledge structures like English or history, language educators have advocated introducing an addi-

FUTURE DIRECTIONS IN RESEARCH

249

tional language with far more verticality than the host discipline, in effect verticalizing it. I remain convinced that intervening across disciplines along these lines is the only way to make the knowledge structures schools were invented to transmit democratically available, across the range of coding orientations students bring with them. Is there any way of making these marriages of convenience more comfortable ones for the host and intervening disciplines? Joe: What Fran and Mary have shown (this volume) is how the school subject English, by moving in Australia from what Bernstein called a performance pedagogy to a competence pedagogy, has had its knowledge structure rendered invisible. A central feature of competence modes is that they focus on the acquirer and her sensibilities, and it is the latter which are to be tutored and refined towards an end which is invisible to the acquirer. At the same time, hierarchy – that is, progression and progress – is rendered invisible, and the acquirers are stranded in the present tense of their own productions. The paper also shows that the only recourse for concerned pedagogues (like Mary and Fran) who wish to render verticality visible is to reconstruct a hierarchy in the metalanguage. For ideological reasons they are prevented from making verticality visible by retrieving a canon (making the knowledge structure visible), but, as they show, they can at least make it visible metalinguistically. The ‘knowledge’ option is blocked, but not the ‘metalinguistic’ option. What this makes plain to me is that it is virtually impossible to speak about the knowledge structures of subjects in a competence mode because they are, in principle, invisible. To make them visible is to change the mode, which one can’t do unilaterally. Therein lies the difficulty in talking about English as a knowledge structure: under prevailing conditions, one literally can’t. Does that mean there is no verticality? No, it does not. Jim shows (this volume) that certain metalinguistic features are entailed by vertical discourse. Thus, by showing verticality in the metalanguage Fran and Mary show, by implication, that the knowledge structure must be vertical; it’s an inference, but a pretty well-founded one because, as Jim says, ‘if no grammatical metaphor then no verticality’, and vice versa. In this way, SFL allows one to circumvent the prevailing ideology, and becomes a powerful tool for intervention. This clarifies for me the power of the term ‘verticality’; it applies both to the knowledge structure and to the metalanguage, and allows one to make inferences from one to the other. It may be that all knowledges are virtual, and come to light only by conscious effort on the part of pedagogic recontextualizers. SFL researchers have done this in two ways: Jim Martin, Veel and others have reconstructed the metalinguistic structure of history and science, inter alia, from textbooks – that is, from already recontextualized texts, which they take for representative tokens of the knowledge structure. Fran and Mary have less to work with, and have reconstructed the metalinguistic structure of English almost from scratch. Their next step, I think, is to show that it holds up when measured by some or other equivalent representative token of the knowledge structure. This can’t be done through the current curriculum, because it is an invisible/competence pedagogy, hence radically under-stipulated. They might then try with some

250

LANGUAGE, KNOWLEDGE AND PEDAGOGY

older textbooks that do reference representative text types. Or from tokens from another national system; or even more powerfully, through a selection of tokens from a number of national curricular systems. All of this begs the larger question: what is the necessary and sufficient global archive of knowledge systems? This brings to light the fact that knowledges do not exist naturally in a finished representational form. The object we call ‘a knowledge’ is virtual. The objects we have that we take for tokens of the knowledge are the product of various kinds of representational reconstruction. The SFL community has begun to construct metalinguistic representations, as we have seen. In logic, maths, some of the sciences, one finds propositional reconstructions – the order that one can see retrospectively (i.e. when doing pedagogic recontextualization) in terms of the necessary sequence of logical subsumption. There are probably other kinds of representational reconstruction too. None of these can sufficiently capture the entirety of the knowledge. Metalinguistic reconstruction; propositional reconstruction; pedagogical sequence reconstruction; there are probably other kinds as well. This digression sheds light on why the question of knowledge arises in the light of the need for its pedagogic recontextualization. We can agree with Bernstein, and with Karl, that the knowledge structure is not curriculum structure, but it is only the question of curriculum structure that brings to light that there might be a knowledge structure worth bringing to light. For scholars at the cutting edge, the knowledge structure is old news not worth reporting on – until they put on their pedagogical caps and try to figure out how to teach the retrospectively accumulated store, as in: ‘I know I stand on the shoulders of giants, but what exactly am I standing on?’ To date, it is the SFL community that have given the most systematic answer. Fran: Joe is right when he says Mary and I (this volume) have shown that in the unfolding of the English curriculum, there has been a decisive shift from performance to competence pedagogy. I am grateful to him for sharpening the discussion here, for we can at least say that the road is cleared to think about where we can go next (see my earlier comments about subject English above). I am excited by Joe’s observation that ‘all knowledges are virtual’, and that they ‘come to light only by conscious effort of the pedagogic recontextualizers’. It is indeed in coming to terms with issues of knowledge and curriculum structure in schools that we face not only the fact of recontextualization, but also – what is more difficult – what it is that is recontextualized anyway. If we confront the issue, ‘what is the knowledge we seek to recontextualize for the purposes of school learning?’, we are driven – as this book reveals – to revisit quite a lot of history (Muller, Moore, O’Halloran, Wignell, Christie and Macken-Horarik). Thus, echoing some of Joe’s remarks to me by email, in revisiting history we ‘pay regard to the retrospect and the prospect: where the knowledge has come from and where it is going’. This, Joe further states, ‘illuminates the communal learning path traversed which in turn is the guarantor for the individual learning path that neophytes must traverse’. Does this – I ask myself – mean that the various models of English that Mary and I identify as all to some extent

FUTURE DIRECTIONS IN RESEARCH

251

jostling for position in the English curriculum – mean we have to (re)accommodate them all in whatever model we adopt anew? I think not, but following what Karl and Joe (joint chapter, this volume) argue, we should be able to establish which past theories of subject English (if any) can be shown to share some commensurable elements, making possible the articulation of a ‘more vertical’ model than presently applies. Karl: I’m glad Fran and Mary’s chapter has come up, as it’s a real step forward to take an aspect of Bernstein’s discussion of knowledge structures (the way in which they progress through integration and subsumption of past knowledge or through segmental addition) and homologously transfer that to the study of curriculum and learning. This forms the basis of an insightful exploration of the different principles that have been underlying the recontextualization of knowledge into educational knowledge (to become subject English), and the consequences of these different principles for the form taken by transmission/acquisition in the subject area and the educational experiences of learners. As Joe rightly points out, the competence mode of subject English is associated with rendering invisible the principles underlying the form taken by the curriculum structure, if not the educational knowledge itself. Instead of a visible structure of educational knowledge we get a series of ideal knowers as the basis of recontextualization. In short, the educational knowledge structure is made less visible and the educational knower structure becomes the key. This knower code renders integrative and progressive learning of knowledge problematic. As Fran and Joe discuss above, the key question is: what is to be done? Two things stand out for me. First the basis of the hierarchy in the curriculum as it currently stands must be rendered visible. There are, after all, means of judging success or failure in the subject and thus a basis of hierarchy. To not highlight this is to leave the knower code tacit, making it an extremely difficult position to get to grips with or change – it is all the more powerful for not being brought out of the shadows and into the light. Second, the alternative basis of a knowledge code can be articulated and make visible. This is what I think Fran and Mary are doing with SFL. As Joe puts it, they’re not doing it directly – the subject’s powerful knower code makes this difficult to do in terms of the knowledge structure. It’s an interesting move to do this via the metalanguage. A question raised for me by all this is how the nature of English as a knowledge structure (not subject English but rather its field of knowledge production) shapes the form taken by the humanities and social sciences. If, as Kay O’Halloran shows so well in this volume, mathematics helps enable the integrative and subsumptive nature of science, does English help enable the segmental, proliferative nature of humanities and social science? Is English the key subject area for enabling more epistemologically powerful knowledge in other subjects that use it as the principal medium of expression? If so, then SFL has an extremely powerful role to play as an underlabourer, determining what is and is not the best means of articulating knowledge claims.

252

LANGUAGE, KNOWLEDGE AND PEDAGOGY

3.5 Values Jim: From an SFL point of view the discussion of knowledge structure in Bernstein’s work seems very ideational. But we know from work on appraisal in SFL that the way in which concepts are valued by members of a discipline are every bit as important as what these concepts are and how they are arranged – including consideration of how research is appreciated, how researchers are judged and how researchers and their apprentices are positioned emotionally. Karl’s work on knower structures and Rob Moore’s on judgements (this volume) is part of this picture – the kind of identity researchers and learners construct for themselves is an essential feature of disciplinarity. How might our discussion of knowledge structure be expanded to bring a concern with the disciplinarity of feeling into the picture? Fran: I think the very notion of pedagogic subject positioning following Bernstein has consequences for a view of knower codes and knowledge codes as carrying values. It is in my view a relatively easy extension of both theories here to view all areas of the curriculum as involving the adoption of certain value positions. The study I did a few years ago (2002) of classroom discourse did not use appraisal theory, though with hindsight I see that it could have done. Joe: Bernstein didn’t speak directly about disciplinary identities, but in the last volume he does discuss both pedagogic identities and local (cultural) identities. As one might expect, for him, identities are the product of symbolic formation and they relate to the ‘social base’ within which the identity is embedded. It is not clear to me how this might dovetail with appraisal theory or ideation more generally, since for Bernstein identities of either variety are not self-constructed. Karl: The notion of knower structures brings into the picture something that was tacit but already there in Bernstein’s account of knowledge structures. I should emphasize that it’s not the sole basis or equivalent of talking about identities. Basil showed how identity is shaped by different forms of knowledge structures and curriculum structures in various ways. So identity is already a part of the picture. But whenever talking about those structures of knowledge or curriculum that render knowledge less visible, then it was not clear what the basis of identity is. ‘Knower structures’ simply brings out what was in the shadows. There is always both a knowledge structure and a knower structure, where the latter simply refers to the ways in which knowing subjects are arranged by the discipline. For example, with physics one could say that alongside its ‘explicit, coherent, systematically principled and hierarchical organisation of knowledge’ (Bernstein 1996: 172) – its hierarchical knowledge structure – it is also characterized by a series of strongly bounded knowers, each with their own specialized modes of being and acting. That is to say, physicists represent a series of segmented knowers in terms of their (non-scientific) ‘gaze’ – it matters less who you are in terms of your social background, for example; they can be very different people, so long as they use the procedures of physics. Conversely, disciplines with horizontal knowledge or curriculum structures, such as that of English explored by Fran and Mary, may be characterized by a hierarchical

FUTURE DIRECTIONS IN RESEARCH

253

knower structure – the basis is who you are, your aptitudes, dispositions, attitudes, etc. based on your biological, social or cultural background. If we don’t get to grips with the knower structure, and especially the role it can play in humanities and social science subjects, then we’ll find it difficult to understand why the existing state of affairs exerts such a powerful grip over many people. If they’re not focused primarily on the knowledge structure or they render knowledge less visible, then what is it that they’re focusing on? This is important not only for being able to understand what it is that different disciplines do, but also for reaching out to other disciplines, being able to see what is underlying their ways of working and so be able to speak across languages. It may help explain why it is difficult to get some academics and teachers to engage in discussion of ideas as ideas. The hierarchical knower structure of English, for example, has been the basis of criticisms of canons (see Moore, this volume) of the ‘patriarchal’, ‘bourgeois’, etc. basis of knowledge claims. It’s also what makes many approaches so appealing – they seem to speak to the experiences of people, and they’re viewed as warmer, more caring and sharing, more obviously human than cold, inhuman technicality. I have been recently working with Sue Hood on using appraisal theory alongside these concepts and it’s becoming clear to me just how well SFL can get to grips with these issues. 3.6 SFL and hierarchical knowledge Joe: I have at times, perhaps mistakenly, discerned something of an apologetic tone in SFL discussions of ‘abstraction’ and ‘technicalization’, as if these terms were somehow politically incorrect. Yet everything in the theory tells us that discursive power is made available only by this route. The first part of the question is I suppose this: does this apologetic-ness stem from the neo-Marxist belief that abstraction and power are intrinsically tools of domination? Or is it rather allied to a more post-modern spirit of de-differentiation, as in the post-modern David Bohm’s objection to an analytic that ‘divides things into separate entities’ thus distorting ‘undivided wholeness’, and his wish to denominalize scientific language (‘which gives the basic role to the verb rather than the noun’)? All of this I take to be quoted approvingly by Halliday (p. 108) in Halliday and Martin (1993). It is true that there are signs of a lingering progressivism in early Bernstein too (my students fight every year as to whether he is actually for integrated codes and against collection codes). But this is by and large gone by mid-period Bernstein, and by late Bernstein his colours are clear. Is this embarrassment about dealing with the necessary instruments of power something we should talk about? Fran: I don’t see that Joe’s interpretation is justified. This discussion is very Hallidayan in character, in that, as often, he is interested in the relationship between language and the social construction of experience and the claims sometimes made that language can’t always cope with the pressures of change, and of the meanings people want to make. It all points to another issue: is there another reality that language somehow only partly captures? I don’t find anything particularly ‘approving’ about the ways Halliday quotes Bohm. In fact, he writes of Bohm and others that they may have overlooked one thing: namely

254

LANGUAGE, KNOWLEDGE AND PEDAGOGY

that ‘you do not need to keep engineering a language in order to change it; it will change anyway – because that is the only way it can persist’ (Halliday, ibid.). Jim: I take Halliday’s point as being that whereas language offers both more nominal (attic) and more verbal (doric) construals of experience, science has by and large adopted the attic mode, and has become, now and again, selfconscious about the limitations of this bias. Halliday is suggesting, contra Bohm, that it’s not language which is at fault, but the language of science, and that language has the resources to serve the doric mode if scientists wish to take them up. This of course begs the question of whether science in the doric mode would recognize itself as science. I suspect not, since it is not at all clear how a hierarchical knowledge structure can be developed verbally, without the resources of grammatical metaphor developed in the attic mode. I hope I have never expressed a distaste for abstraction and technicality in my own writing, since I have always seen myself as mounting a defence of such in the face of progressivism, constructivism, the plain English movement and so on. As Maton points out (this volume) hierarchical knowledge structures can be a great deal more democratic that horizontal ones, since what you know is privileged over who you are. I have always found the non-technical discourses of the various languages in horizontal knowledge structures far more elitist and insidiously dominant than discourses featuring verticality. Joe: A follow-on to Jim: Halliday in the chapter just quoted goes on to speculate that Bohm might be heralding ‘a new type of order’ where ‘communication’ becomes of the essence – and perhaps linguistics can then become the new queen of the sciences (‘Their coat-tailing days are over’). There is a similar kind of aspiration expressed at the end of Jim’s chapter. Do your ambitions for linguistics rest on a similar foundation to Halliday’s, Jim? That is, do you think that the future will be a de-differentiating one? Or do you think SFL will have to stiffen its spine somewhat (become a hierarchical knowledge structure) before it can challenge for pole position? Fran: SFL theory is often heard as being very imperialistic in that it is said to want to overtake everything else in its claims to offer ultimate explanations and/or ultimate accounts of experience. Well, that is a risk I suppose, but it misses an important point. Halliday and Bernstein were bold and ambitious in their scholarly visions, for both wanted to develop tools with which to explain the nature of social life. Both might well be accused of being arrogant because of the boldness of their visions – they did want to do nothing less than explain social experience! I think the very adventurousness of what they both aspired to do is part of their appeal, but also part of the intellectual power they both unleashed in the world. Theirs was a noble endeavour, and one that remains relevant in the twenty-first century. Jim: The basic propaganda here as far as SFL is concerned is to argue that linguistics should not be modelling itself as the science of language, since science builds knowledge structures responsible for describing physical and

FUTURE DIRECTIONS IN RESEARCH

255

biological systems – and social semiotic systems are more complicated than these, having evolved out of them (their emergent complexity is of a different order). This means that the theories we build for social semoisis will have to be different in kind from those we’ve built before, and we can expect them to have more than enough power to interpret metalinguistically and metasemiotically what science has been saying about physical and biological systems all along. If this metasemioic purchase makes linguistics the queen of sciences, so be it, long may she reign. But let me stress that this purchase will not make her a scientist, because her object of inquiry is discourse, not material reality per se. I don’t think linguistics has to stiffen her spine and become more of a hierarchical knowledge structure to achieve this, because becoming more like science as we know it is not going to do the trick. As I said above, we need prisms not pyramids, and so we need a theory that embraces complementarity alongside hierarchy. If complementarity as I presented it above can be construed as a dimension of verticality, as Joe suggests, then, OK, yes, linguistics does need to stiffen its spine. Currently, the main technological problem standing in the way is that we can’t process data fast enough and deeply enough to foster the grammaticality we need to move on. In short, our computers can’t do discourse analysis; they can barely do the kind of functional grammatical analysis we need in fact. So we can’t see very well what we are looking at. This should change in another generation or so and then we’ll begin to see which languages of linguistics have the right stuff. Fran: I have not much to add here, other than to say that we are getting better at seeing what we are looking at. The next generation will no doubt do better still, but we are on the way. Karl: One thing I’d like to add is to echo Fran’s admiration for the boldness of the enterprise of SFL and Bernstein’s sociology. In the contemporary climate of social science such boldness is easily denigrated as arrogance or imperialism. F.R. Leavis (someone often accused of all kinds of arrogance, not always mistakenly) once said that the only way to escape misrepresentation is never to commit oneself to any critical judgement that makes an impact – that is, never say anything. What began as a much needed dose of modesty became a form of intellectual cowardice. To always caveat everything as only being from our own knower perspective, as if we should only share our own individual story, leaves us where we began in terms of understanding the world. This has become an all too common strategy in at least my own disciplines of Education and Sociology. The image that always springs to my mind is of a tower of knowledge. The aspirations of SFL, it seems to me, are akin to a grand undertaking such as building a cathedral, involving many thousands of workers, taking many decades, where the original architect and most of those who laboured will not live to see its completion. In contrast, many other approaches are akin to the relatively rapid creation of suburban housing in a new development, each building being constructed quickly, requiring near-constant maintenance, with new members of the workforce moving rapidly to begin building the next house. Instead of a cathedral or tower, building upwards and aiming to last, this represents the

256

LANGUAGE, KNOWLEDGE AND PEDAGOGY

spread horizontally across the intellectual landscape of comparatively low-level, largely identical buildings. Hierarchical knowledge structures are intellectual creation as the building of towers; horizontal knowledge structures are intellectual creation as suburban sprawl. As John Ruskin put it, when we build let us think that we build for ever. One concern I would have though, concerning understanding knowledge and curriculum structures, is the possibility of linguistic reductionism, that is seeing these objects solely as comprising their linguistic features. Knowledge is more than the sum of its linguistic parts. Exactly how, and what form the interface between our approaches will take, are exciting issues to explore. 3.7 The frontiers of knowledge Jim: One final question I’d like to pose has to do with discourses that don’t get mentioned in our discussions of knowledge structure. In our early 1990s work on uncommon sense, one such discourse we tried to tackle was the discourse of administration. This is full of grammatical metaphor, but in the service of proposals (directives, rules and regulations) rather than propositions (statements about the world). Are we suggesting that a procedural discourse of this kind is not a knowledge structure? Another discourse we considered was the discourse of technology, which bridges between science and control of our material world. This turned out to be more like horizontal discourse than I expected, although knowledge in science industry clearly bridges across the horizontal and vertical discourse divide. What do we do about bridging discourses of this kind? Perhaps my general point here has to do with what we do about discourses evolving in the service of other discourses. Kay O’Halloran (this volume) shows the crucial role played by mathematics for the grammaticality of science. This interplay in turn engenders the technology that manages material resources. Reasoning along similar lines, we might see the language of administration as a technology that manages people, as informed by humanities and social science. In short then, what qualifies a discourse as our object of inquiry? Why are everyday conversation and traditional disciplinary knowledge in, and things like the discourse of technology and bureaucracy out? Fran: The answer to this is surely social: it is a matter of what gains recognition and respect. I suspect that many of the discourses that evolve in the service of other discourses are often the most invisible (and like women’s work, often least respected!). Their very invisibility is a source of their strength. In Bernstein’s sense of the word ‘pedagogy’, the discourses involved in the service of science or of administration are profoundly important as sites for pedagogic practice. SFL theorists might well re-open this whole matter in the future. Joe: Reading through Jim’s comment above, it struck me that nowhere in the book do we discuss a key distinction Bernstein makes between knowledge singulars and knowledge regions. Our entire discussion has been assuming that all knowledges take the form of singulars, when they evidently don’t. A region is an ensemble of singulars combined sometimes with segments of everyday or

FUTURE DIRECTIONS IN RESEARCH

257

procedural knowledge. Architecture is one example of a region. The various technical or technological discourses would be the same. Public administration, as a subject taught at university is also a region. What then is the discourse of administration? Bernstein’s answer would be to ask: is there integration at the level of meaning – in which case it is a vertical discourse – or is integration segmental? I would say that it is probably segmental, which makes it horizontal discourse. However, it borrows a number of features from its recontextualized parent public administration, so that it exhibits features of a region. And so on. In other words, either we can deal with administration, and all types of discourse for that matter, with Bernstein’s tools, or the theory is lacking. There can’t be a question of leaving out this or that class of discourse. What the question highlights for me is that although we may say: if no grammatical metaphor then no verticality, we can’t necessarily say: if grammatical metaphor then it must be a vertical discourse. Karl: As Fran and Joe suggest, I don’t think the focus of our discussions has been to negate other discourses, such as administration and technology, as not being worthy of study. But Jim is right to highlight that these must be brought into the scope of the discussion. As the concepts are applied to such discourses they will undoubtedly be pushed to their limits, necessitating theoretical development, and raising further questions for us all to consider. The dialogue between the theoretical and the empirical will continue, as I hope will that between systemics and sociology. This, we could say, is not the beginning of the end, but simply the end of a new beginning.

References Bernstein, B. (2000), Pedagogy, Symbolic Control and Identity (rev. edn). Lanham and Oxford: Rowman & Littlefield. Gamble, J. (2004), ‘Tacit Knowledge in Craft Pedagogy: A Sociological Analysis’. Unpublished PhD thesis, University of Cape Town, Cape Town. Halliday, M.A.K. (1993), ‘Language in a Changing World’. Occasional Paper 13, Applied Linguistics Association of Australia. Halliday, M.A.K. and Martin, J.R. (1993), Writing Science: Literacy and Discursive Power (Critical Perspectives on Literacy and Education). London: Falmer.

This page intentionally left blank

Index

Page numbers in italics refer to figures and tables. absolutism 115, 119, 120 abstractions embedded clauses in nominal groups 171, 176–7, 178 (proto-)grammatical metaphor and 145–8 and technicality 190–200 passim, 202 activity sequences 34–6, 37, 41, 43–5 Agent and Medium 54–5 All-or-Nothing Fallacy 114, 122 appraisal 178 arts 112–13 high and low 114–15 music 98–104 sculpture and photography 219 see also humanities; judgement(s) Basic Skills, English teaching model 159–60, 161, 162, 168–9 Bell, E.T. 210, 213, 214 Berlin, I. 73–4 Bernstein, B. 3–8passi, 14–29, 36, 38, 47–80 passim, 87–96 passim, 104, 109, 110, 112, 126–7, 131–9 passim, 157–67 passim, 184–5, 201–14 passim, 239–44 passim, 252 Bloom, H. 125, 134 Bourdieu, P. 91, 111 Brookes, B.C. 95

canons, critical approaches 110–12 Carey, J. 112–21, 122, 123, 124–6 Cartesian rationalism (Descartes) 73–4, 113–14, 216, 226, 229 ‘cause in the clause’ 45–7, 48 childhood see language learning in early childhood Christie, F. 17, 56, 159, 160, 162, 239–57 passim and Martin, J. 34 class see social class Classics 91 classification 17, 36, 38–40 see also taxonomy of value coalitions of the mind 121–4 cochlear ear implant, example of composition 40 code(s) 6–7, 16–18 collection and integrated 96–7 elaborate 6, 16, 22, 131–2 élite 98–9, 100, 102–4 integrating 8 knowledge–knower 27, 96–8 legitimation, of specialization 93–5, 96–8, 99, 100, 101 orientation to meaning 16, 18, 20 relativist 98 restricted 6, 16, 22 codification of disciplines 69–70 Collins, R. 65, 68, 109, 112, 121–6, 127 common and uncommon sense fields 37, 38, 66, 68–9 composition 36, 37, 40, 42

260

INDEX

computer technology 224, 225–6, 229–31 context 5 decontextualization 122, 123 language learning in early childhood 139, 140–3 recontextualizing rules 7, 19, 22, 28–9, 162–3 social 35 Cultural Heritage, English teaching model 160–1, 162, 165, 166, 167–70, 174, 177–8 Cultural Studies, English teaching model 164–5, 167, 168–9 curriculum 8 England English 156, 170–4 music 98–104 NSW, Australia, English 156, 165, 175–8 planning 72 South Africa 79–82 structures and knowledge structures 28–9, 80–2 Davis, P. 211, 231 et al. 230 decontextualization 122, 123 Descartes, R. (Cartesian rationalism) 73–4, 113–14, 216, 226, 229 dialects, social 6 dialogue development 5–9 interlocutory interpolation 121 language learning in early childhood 132–5 ‘situation-transcending’ 125 discursive saturation 69–70 distributed social goods, knowledge as 66 distributive rules 7, 19, 162–3

division of labour, mental and manual 19 Donne, John 74–5, 117 Durkheim, E. 15–16, 19, 22, 28, 67–8, 75, 109, 112, 117–18, 127, 202 economics see political economy; value (economic theory) educational knowledge–knower structures 96–7 elaborate codes 6, 16, 22, 131–2 elaboration 172, 176–7, 178 élite code 98–9, 100, 102–4 Elkins, J. 226 elliptical clauses 171, 172 embedding 56–7, 171, 178, 228 empathy 167–74 passim, 175–8 passim English interpretation/symbolic reading of literary text 175–8 ‘literary’ narrative 167–74, 178 metalanguage of systemic functional linguistics (SFL) 179–80 and science, pupil/university entrant perceptions 102 teaching models 158–67, 168–9 see also literature epistemic relation (ER) to knowledge structures 93–4, 96–8, 101 ethical judgement 177–8 ethical theory and aesthetics 120–1 evaluative rules 7, 19, 162–3 events and objects, joint attention in early childhood 133 external language of description 71 field(s) 34–8, 55–6, 111 common and uncommon sense 37, 38, 66, 68–9

INDEX

production 88–96 reproduction 96–104 framing 17 Functional Language Studies, English teaching models 163–4, 166, 168–9, 174, 179 functional linguistics see systemic functional linguistics (SFL) Gellner, E. 91, 95 genealogical organization of living things 38, 39 generic reference, language learning in childhood 142–3 genre 9, 55–6, 57 Golovin, A.A. et al. 222, 225 grammar(s) 5, 26–7 matching 52–3 modelling 136 weak and strong 71, 78, 209–10, 211, 221–4 grammatical metaphor 49–55, 58–9, 60–1, 173 (proto-)grammatical metaphor and abstraction 145–8 and vertical discourse 53–61 passim, 243–4 grammaticality and knowledge structures 209–14 of mathematical symbolism 225–9 multimodal approach 214–15, 216–29 and verticality 71–2, 244–6 habitus 120, 126 Hacking, I. 68 Haire, M. et al. 38, 39, 40 Halliday, M.A.K. 3, 4, 5–6, 9, 17, 47, 51, 53, 54, 60, 61, 132, 133, 145, 147, 163, 164, 172, 173, 186, 187, 193, 205–9, 214, 226, 239, 240, 247, 253–4

261

and Hasan, R. 9 McIntosh, A. and Strevens, P. 9, 163 and Martin, J. 226–7, 253 and Matthiessen, C. 214, 226 Hasan, R. 3, 6–7, 17, 66 Halliday, M.A.K. and 9 hierarchical knower structures 91, 92–3 hierarchical knowledge structures 8, 23–4, 70–1 political economy 186, 189, 190–200 science 38–41, 90 social sciences 189, 201–3 sociology 202–3 and systemic functional linguistics (SFL) 253–6 triangle metaphor 60, 90, 91, 92, 184, 206–7 vs horizontal knowledge structures 25–8, 109–10 hierarchy, in discourse and society 65–70 high and low art 114–15 history Australian Aboriginal people 45–7 Australian population and immigration 42–5, 55–6 genre 57, 58, 59 horizontal knowledge structure 42–9 Philippines, Spanish colonization 47–9 of science 73–8, 186, 187–9, 193 of social sciences 186–203 Hobbes, Thomas 186, 187–8, 189, 201 Hoggart, R. 111–12, 115 horizontal discourse definitions of 7–8, 22–3, 54, 66–7 language learning in childhood

262

INDEX

131, 132, 135–9 science and humanities 95 see also vertical and horizontal discourse horizontal judgements 112–17 horizontal knower structures 92 horizontal knowledge structures 8, 23–4, 25–6, 70–1 English 167 history 42–9 humanities 90 sociology 202–3 vs hierarchical knowledge structures 25–8, 109–10 humanities hierarchical knower structure 91 horizontal knowledge structure 90 knower–knowledge structure 93–4 and science and social science 247–8 ‘two cultures’ debate 88–95, 111–12, 119 idealization 67–8 ideational metafunctions of language 5 identity see knower code; knower structures imaginative nexus 125–6 information exchange, in early childhood 136–7 intellectuals 122–3 internal language of description 71, 72, 78 interpersonal metafunctions of language 5 judgement(s) critical approaches 110–12 ethical 177–8 horizontal 112–17

literary 116–17, 118–21, 123, 124–6 preferences and 114–17, 120–1, 126–7 sociality of 112, 124–6 vertical 117–24 Kant, E. 113, 118 Kepler, Johannes 75, 76 knower code 27, 96–8 knower structures 91–3 see also knowledge–knower structures knowledge 14–16, 29–30 and dilemma of progress 70–2 ethical theory and aesthetics 120–1 making visible 21–2, 24–5 in school 34 knowledge about language (KAL) 57, 58–9, 163 knowledge frontiers 256–7 knowledge progression 73–7, 212–14 knowledge structures and curriculum structures 28–9, 80–2 differences between 25–8 discourses and 22–5 future research directions 239–57 linguistics as 56–60 and pedagogic structures 79–83 science and humanities cultures 89–90 knowledge theory development 14–16 knowledge–knower structures educational 96–7 in production fields 88–96 in reproduction fields 96–104 science and humanities 93–6 knowledge code 27, 96–8

INDEX

language learning in early childhood after three years old 140–53 centrality of dialogue 132–5 first two and a half years 135–9 horizontal discourse 131, 132, 135–9 lavatories, grammatical metaphor 49–51 Leavis, F.R. 94–5, 160–1 legitimation codes of specialization 93–5, 96–8, 99, 100, 101 lexical metaphor 173, 176 lexical motifs and technicality 190–3 ‘literary’ narrative in English 167–74, 178 literature judgement 116–17, 118–21, 123, 124–6 teaching models 160–1, 165 logical positivism 71, 77, 82 McEwan, Ian 125 Macken-Horarik, M. 164, 167, 174, 175, 244, 250–1 magnification 60 Mares, P. 43, 55–6 Martin, J.R. 9, 34, 37, 38, 55, 66–7, 68, 70, 164, 214, 239–57 passim Christie, F. and 34 Halliday, M.A.K. and 226–7, 253 and Rose, D. 34, 164, 173, 174, 214 and Veel, R. 226–7 and White, P.R. 9, 55 Marx, Karl (Marxism) evolution of social sciences 186, 187, 189, 195–200, 201–2 reductionism 111 vs modernist history 47–8, 49 mathematical symbolism 215, 216, 219–20, 221–4, 229–30 grammaticality of 225–9

263

mathematics knowledge structure 80–2 and science 210–14 computer technology 214–15, 216–29, 231–2 knowledge and knowledge structures 206–9 knowledge structures and grammaticality 209–14 SF-MDA (multimodal grammatical) approach 214–15, 216–29, 231–2 Maton, K. 8, 88, 91, 110, 239–57 passim Moore, R. and 4, 30, 74, 88, 93, 110, 122 meaning knowledge as 66 orientation to 16, 18, 20 Medium, Agent and 54–5 Merton, R. 30, 69, 70, 75–6 metafunctions of language/SFL 5, 35, 55 metalanguage development in early childhood 142 systemic functional linguistics (SFL) as 248–51 English teaching 179–80 metaphor 172–3, 178 triangle 60, 90, 91, 92, 184, 206–7 modelling grammar, early childhood 136 Moore, R. 87, 118, 122 and Maton, K. 4, 30, 74, 88, 93, 110, 122 and Muller, J. 8, 77–8, 87 Morais, A. et al. 65 Morris, M. 89 Muller, J. 30, 55, 82, 158, 167, 207, 231, 239–57 passim Moore, R. and 8, 77–8, 87

264

INDEX

multiliteracies, English teaching model 166, 167 multinocular vision 60 Murray, L. 132–3 music 98–104 New Critics, English teaching model 161 ‘New Cynics’ 77–8, 82 New Literacy Studies (NLS), English teaching model 166–7, 168–9 Newton, Isaac 75–6, 186, 187, 216–22 passim, 223, 228 nominalization/nominal groups 52–3, 171, 172, 176–7, 178, 192 non-finite clauses 171–2 objects and events, joint attention in early childhood 133 observational understandings of text 143–5 O’Halloran, K.L. 214, 217, 222–3, 227, 229–30, 231–2 operative processes/participant configurations 227–9 oral and written transmissions 68–9 orientation to meaning 16, 18, 20 Painter, C. 67, 68, 133 parochialism 123 participant configurations/operative processes 227–9 pedagogic device 7, 18, 19–21, 22, 160 pedagogic discourse 7, 21, 22, 56–7, 162–3 pedagogic and knowledge structures 79–83 pedagogies to knowledges 14, 22 Personal Growth, English teaching model 161–3, 164, 165, 167, 168–9

philosophy 122–3 photography, sculpture and 219 political economy 186, 187, 189, 190–200 Popper, Karl 24, 26–7, 77 pre-linguistic communication 132–3 preferences and judgements 114–17, 120–1, 126–7 Prigent, A. et al. 222, 224 production fields, knowledge–knower structures in 88–96 profane and sacred worlds 67–8 progression, knowledge 73–7, 212–14 projection relations in pedagogic discourse 56–7 (proto-)grammatical metaphor and abstraction 145–8 ‘proto-language’ phase of language development 133–4 realism see science recontextualizing rules 7, 19, 22, 28–9, 162–3 reductionism 111 register (variation) theory 5, 9 relationalism 111 relations between social groups 115 relativism 111, 112, 114–15, 116, 121–2 local 123–4 subjective 118 relativist code 98 reproduction fields, knowledge–knower structures in 96–104 research administration 72 code theory 17–18 future directions 239–57

INDEX

‘resemioticization’ 207–9 restricted codes 6, 16, 22 Ricardo, David 189, 193–5, 199, 200, 201 Rodin, Auguste 219 Rose, D. 58–9, 68, 83 Martin, J.R. and 34, 164, 173, 174, 214 sacred and profane worlds 67–8 science biology 38–40 codification 69 and English, pupil/university entrant perceptions 102 geology 34–6, 37 hierarchical knowledge structure 38–41 history of 73–8, 186, 187–9, 193 horizontal discourse 95 and humanities and social science 247–8 ‘two cultures’ debate 88–95, 111–12, 119 knowledge progression 73–7, 212–14 and mathematics see mathematics, and science meterology 41, 42, 47 post-modernist account 113 sculpture and photography 219 ‘segmental organization’ of horizontal discourse 138–9 self-contextualisation, language learning in childhood 141–3 semiotic(s) ‘resemioticization’ 207–9 resources, mathematical and science 215 social 56–7, 60 Shapin, S. 74, 75

265

skills instruction, early childhood 135–6 Smith, Adam 187, 188–9, 190–2, 193, 199, 200, 201–2 Snow, C.P. 88–9, 90, 91–2, 93, 94, 95 ‘social base’ 15–16 social class 3, 6–7, 16, 17–18 and science 91–2, 188, 231–2 social context as register and genre 35 social dialects 6 social phenomenon of language 5–6 social relation (SR) to knower structures 93–4, 96–8, 101 social sciences historic evolution of 186–203 relativism in 114–15 science and humanities 247–8 social semiotics 56–7, 60 sociality of judgement 112, 124–6 of knowledge 111, 118, 120, 122 sociology of education, Bernsteinian perspectives 3–8 evolution of 201–3 spatial/temporal relationship 217–29 passim specialization 96–8 legitimation codes of 93–5, 96–8, 99, 100, 101 specialized lexis 41 subjective relativism 118 subjectivism/inter-subjectivism 115, 116–17, 121, 124–5 Sydney School 34, 38, 56 ‘symbolic control’ 7, 15–16 symbolism, mathematical see mathematical symbolism systematic knowledge building, language learning in childhood 148–53

266

INDEX

systemic functional linguistics (SFL) 3, 8–9 dialogue development 5–9 English 179–80 future research directions 239–57 and hierarchical knowledge 253–6 and knowledge construction 34–64 as metalanguage 248–51 English teaching 179–80 scales and complementarities 60 SF-MDA (multimodal grammatical) approach, mathematics and science 214–15, 216–29, 231–2 and verticality 246–7 systemic grammar 5 tacit acquisition of values 157 tacit learning in childhood 136–8, 139 tastes see preferences and judgements taxonomy of value 190–5 technical lexis 41 technicality and abstraction 190–200 passim, 202 history 42, 43, 45, 48–9 linguistics 58 see also science text interpretation/symbolic reading 175–8 observational understandings 143–5 and oral transmissions 68–9 vs lived experience 143–4 textual metafunctions of language 5 thematic non-finite clauses 171–2 ‘thingification’ 44, 45, 53 time

activity sequences 34–6, 37, 41, 43–5 spatial/temporal relationship 217–29 passim Token and Value 53–4, 199–200 triangle metaphor 60, 90, 91, 92, 184, 206–7 truth 118, 119–20, 121–2 ‘two cultures’ debate, science and humanities 88–95, 111–12, 119 value (economic theory) commodities and 194–200 and labour 191–2, 193–4, 196–8, 199 taxonomy of 190–5 Value, Token and 53–4, 199–200 values aesthetic 111 SFL 252–3 tacit acquisition of 157 vertical discourse/verticality definitions of 7–8, 22–3, 53–4, 70–2 in English 157 and grammatical metaphor 53–61 passim, 243–4 and grammaticality 71–2, 244–6 judgements 117–24 kinds of 77–8 and systemic functional linguistics (SFL) 246–7 vertical and horizontal discourse borders of 241–2 city landscape visualization 207–9 definitions of 7–8, 22–3, 53–4, 66–7 and knowledge structures 22–5 maths and science 206 see also horizontal discourse; vertical discourse/verticality Virilio, P. 219, 225, 230–1, 232

INDEX

visibility concept of 215 of knowledge 21–2, 24–5 see also mathematical symbolism; visual imagery visual imagery city landscape 207–9 computer technology 224, 225–6, 229–31

267

maths and science 216–20 Vivo, Giambattista 73–4 weak and strong grammars 71, 78, 209–10, 211, 221–4 weak and strong progression 73–4 Wignell, P. 185, 202