Literacy, Narrative and Culture [1 ed.] 9781136858031, 9780700714971

Citation preview

LITERACY, NARRATIVE AND CULTURE

WORLD OF WRITING Series Editor Roy Harris Editorial Board Jaques Anis, University de Paris X Jesus Cameraro, Universidad del Pais Vasco Francesco d'Errico, Institut de Prehistoire, Talence Harmut Gunther, Max Planck Institute, Nijmegan Per Linnell, Linkoping Universtity Giovanni Lussu, Milan Polytechnic Peter Miilhiiusler, University of Adelaide James McCawley, University of Chicago Christian Puech, Universite de Paris III Robert Richardson, De Montfort University Colette Sirat, Ecole Pratique des Hautes Etudes, Paris This series is prompted by a number of recent developments which are set to reshape the world of writing. The traditional Western view of writing being merely a form of 'visible speech' no longer commands the unquestioned acceptance it once did. New technologies are replacing books and manuscripts as the principal means of storing written records. Computer-based texts tend more and more to blur the distinction between verbal and non-verbal symbols. We are entering an era in which the study of writing can no longer be taken for granted as a appendage to linguistics, sociology or psychology, but demands recognition as a field in its own right. This series is the first to recognise these complex requirements. Its aim is to serve as a forum for articulating and analysing the ideas that emerge for the world of writing in the 21 st century. The series editor and the publishers welcome proposals for books in this series.

LITERACY, NARRATIVE AND CULTURE

Edited by

Jens Brockmeier, Min Wang and David R. Olson

~ 1 Routledge ~~

Taylor & Francis Group

LONDON AND NEW YORK

First Published in 2002 by Curzon Press Published 2013 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OXI4 4RN 711 Third Avenue, New York, NY, 10017, USA

Routledge is an imprint of the Taylor & Francis Group, an informa business

Editorial Matter

©

2002 Jens Brockmeier, Min Wang and David R. Olson .

Typeset in Times by LaserScript Ltd, Mitcham, Surrey All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers.

British Library Cataloguing in Publication Data A catalogue record of this book is available from the British Library Library of Congress Cataloguing in Publication Data A catalogue record for this book has been requested ISBN 13 978-0-700-71497-1 (hbk)

Contents

Introduction: What is a culture of literacy? Jens Brockmeier and David R. Olson Part 1: Written culture 2

The literacy episteme: The rise and fall of a cultural discourse Jens Brockmeier

17

3

Literacy and the future of writing: An integrational perspective Roy Harris

35

4

The construction of mind and self in an interpretive community Carol Fleisher Feldman

52

5

Hunting, tracking and reading J Edward Chamberlin

67

6

Narrative distancing: A foundation of literacy Jerome Bruner

86

Part 2: The shaping of modern written culture 7

Letters and pictures in seventeenth-century education John H. Astington

8

Painters and literacy William Blissett

110

9

"Dumb significants" and Early Modern English definition Ian Lancashire

131

10

The spread of culture: Subscription libraries in France in the nineteenth century Graham Falconer v

97

155

Contents

11

The essay as a literary and academic form: Closed gate or open door? Margaret Procter

170

Part 3: Literacy as cultural learning 12

Writing as a form of quotation David R. Olson and Deepthi Kamawar

13

Children's conceptions of name: A study on metalinguistic awareness in Italian children Ilaria Grazzani Gavazzi and Veronica Ornaghi

187

199

14

The distinction between graphic system and orthographic system and their pertinence for understanding the acquisition of orthography 215 Emilia Ferreiro

15

Children's analysis of oral and written words Sofia A. Vernon

229

16

Young children's "clever misunderstandings" about print Janette Pelletier

245

17

Literacy and metalinguistic thought: Development through knowledge construction and cultural mediation Bruce D. Homer

266

Making new or making do: Epistemological, normative and pragmatic aspects of reading a text Linda M. Phillips

283

18

Contributors

301

Subject Index Name Index

303 309

vi

1 Introduction: What is a culture of literacy? Jens Brockmeier and David R. Olson

The "discovery" of writing as a particular form of language has produced a revolution in the human sciences. While a few decades ago the subject of writing and literacy was largely ignored, if not explicitly denied, it has become an important field of research in many disciplines. With it has come increasing differentiation and specialization of the research. Terms like "writing" and "reading," and "literacy practices" are used and understood quite differently by historical anthropologists exploring the origins of writing among the Piro of Eastern Peru, than by historians counting signatures in early Rome. Similarly, "semi-literacy" means something quite different to a historian describing Medieval Europe than to an educator worried about performance on a standardized test or an international Non-Governmental Organization concerned with the consequences of universal schooling in Africa. Even within the same discipline, literacy means something quite different to a reading researcher focusing on knowledge than to a cultural or discursive psychologist focusing on social interactions. There is a widening gap between those studying the "processing" of differently shaped letters, and those who focus on the pragmatics of "language in use". Diverse disciplinary perspectives can, nonetheless, contribute to a common understanding. The contributions to this volume include perspectives on literacy from developmental psychology, linguistics, literary theory, history and sociology of literature, philosophy, anthropology, and history of art and culture. But they share some essential presuppositions. First, even iftheir authors, among them leading figures in their fields, set out specialized areas of research, they do so with a minimum of technical jargon. They address not only their academic peers but also a multidisciplinary forum. This is a result of the fact that all the papers were first presented to audienccs of two multidisciplinary conferences on literacy which took place in 1999 at the University of Toronto, the first as the 21st annual University College Symposium, the second as a workshop sponsored by the Social Sciences and Humanities Research Council of Canada, the Connaught Research Fund, and the Ontario Institute for Studies in Education. 1

Introduction

Second, the papers in this volume share an understanding of literacy as socially and historically embedded activities and cultural practices. They are concerned primarily with inscriptions and documents as cultural artifacts, how they are produced and distributed, how they are consulted and interpreted, and how these activities contribute to intellectual and cultural life. This culturalhistorical view is elaborated in the first part (Written culture) and the second part (The shaping of modern written culture) in papers ranging from the pre-historical beginnings of "reading signs" in the tracking activities of hunters and gatherers (1. Edward Chamberlin), to the emergence of modem literate traditions in the seventeenth to nineteenth century in Europe (John H. Astington, Ian Lancashire, Graham Falconer), and on to the implications of electronically mediated writing in times of the post-Gutenberg galaxy (Roy Harris). How literacy creates as well as reflects culture is examined in the formation of particular interpretive "textual" communities (Carol Feldman) and of genres of writing such as the academic and literary essay (Margaret Procter). Likewise, it is argued that the emergence of writing as an "epistemic subject," that is, an object of thought and intellectual reflection, is linked to the rise of electronic media (Jens Brockmeier). All studies of the book make the point that there is no theory and history of writing that does not presuppose a theory of culture. At the same time, the papers also demonstrate that every theory and history of culture must unavoidably entail a theory and history of writing and written culture. Just what is involved in learning to live in a culture of literacy is the concern of several chapters in the third part of the book (Literacy as cultural learning) that represent recent advances in developmental and educational psychology (David Olson and Deepthi Kamawar, Haria Grazzani and Veronica Ornaghi, Emilia Ferreiro, Sofia Vernon, Janette Pelletier, Bruce Homer, Linda Phillips). A third theme underlies a number of papers in this book: the issue of narrative. The study of narrative discourse in its multifaceted forms claims a central place in what has been called the "narrative turn" in psychology, anthropology, sociology, philosophy, and other human sciences. Narrative, it is argued, is a unique linguistic and psychological form, oral and written, that integrates human experiences and social practices with the canonical registers of a culture. In several papers of the present book, the linguistic and cognitive functions of narrative are examined. In narrating experiences, memories, or intentions, we create a mode of cognitive and emotional distancing (Jerome Bruner), a mode of distancing that can also be seen as characteristic of writing. Yet it is not only writing, but also the process of reading, the deciphering and understanding of signs, that can be described in terms of narrative and narrative interpretation (1. Edward Chamberlin). In narrative we create order and coherence, and this applies not only to linguistic discourse in the narrow sense but also to visual-iconic forms, such as pictures and paintings (John H. Astington, William Blissett) as well as children's drawings and scribbles (Janette Pelletier). In making possible sophisticated use of indirect speech and quotation 2

Jens Brockmeier and David R. Olson

(David Olson and Deepthi Kamawar), writing further develops the possibilities of narrative construction. Both writing and narrative are linguistic and metalinguistic practices. Intentionally or not, in writing and narrating we do something in and on language, we are engaged into an activity that "stages" language itself. What becomes visible when language is "staged" in this way is the question to which the chapters of the third part suggest answers. It is undisputed that the use made of language in the manifold contexts of

writing, reading, and narrating depends on the cultural matrix that defines the functions of language. As well, there is wide agreement on the claim that the historical development of narrative as well as of writing and reading has had far-reaching cultural consequences. However, what exactly the term "culture" or "cultural" means in these contexts remains obscure. Although there has been an inflationary spread of the cultural vocabulary in the human sciences, this spread has been at the expense of precision. In many contexts, the term "culture" could easily be replaced by such terms as "society," "history," "social representations," or "politics" without any loss or gain in meaning. The elusiveness of the meaning of "culture" is not unique to discussions on writing and literacy, nor is it a recent phenomena. In their famous survey, carried out half a century ago, Kroeber and Kluckhohn (1952) discussed 161 different meanings of "culture" in the human sciences. This may be an underestimate given the rise and academic institutionalization of new disciplines such as "Cultural Studies," "Communication," and "Media Studies". Consider the emergence of "Cultural Psychology". In the Editorial of the first issue of the then newly launched journal Culture & Psychology, the editor (Val siner, 1995) made the plea for cultural psychologists to explicate what they mean by "culture". Five years later, the conclusion was that "the contributions received [by the journal] rarely show signs that efforts have been made to fulfill this expectation. Often 'culture' is used as a term to define a perspective through its opposition to some existing labeled perspectives (e.g. 'cognitive,' 'nativist,' etc.), rather than serving as a general intellectual heuristic for a new understanding of complex issues". (Crawford and Valsiner 1999: 262). How, then, do we make sense of the many "cultural" approaches to literacy? And what is the idea of a "culture of literacy" that we wish to offer? Perhaps we can address the question by taking a closer look at the different ways these concepts have been used in a classical debate about the relationship between literacy, culture, and the mind. Although the problem of language, mind, and culture has been a subject of scholarly discussion at least since Enlightenment (Jahoda 1993), the issue of written language and written culture has only in recent decades become the focus of debate. The origins of this debate on literacy and culture can be traced back to a number of disciplines, notably sociology, anthropology, psychology, history, classics, and history of literature. One line of this debate found its point of departure in a psychological context in the 1930s with the work of Lev

3

Introduction

Vygotsky, today regarded to be one of the founding fathers of the culturalhistorical view in psychology. With his colleague Alexander Luria, he set out to examine how sociohistorical conditions affect human consciousness. A series of psychological field studies in Central Asia (Luria 1976) focused on the effects of modernization, including literacy, education, and collectivization, on forms of perception and thought. Historically, this was a timely issue. The socialist government was carrying out literacy campaigns in the new Asian republics, and the idea of historical materialism that the social and economic conditions of life determined human consciousness was a principle of official educational and cultural politics. Vygotsky's and Luria's working hypothesis was that the effects of societal literacy would manifest themselves in such cognitive abilities as logical reasoning and systematic classification. And in fact, the findings of the Central Asia expedition seemed to confirm the view that changing cultural circumstances had an impact on individuals' mind. Among other findings, the testing results showed that literate subjects were more able to think in a formal, abstract, and self-reflexive manner than their non-literate neighbors. Moreover, Vygotsky and Luria suggested that the cultural institution of writing not only allowed people to think in a new, decontextualized way, but also drew attention to writing and language as the central "tools" of consciousness. As a consequence, the relationship between language, thought, and reality could itself become an object of rational reflection and, as a consequence, impinge on consciousness. In Vygotsky's Marxist view, the notion of consciousness and abstract rationality was a crucial parameter not only of cognitive and linguistic development, but also of cultural change, historical progress, and political emancipation (Wertsch 1996). Literacy played a prominent role in this vision. As a developmental psychologist, Vygotsky (1978, 1987) reported that in learning to write and to read, children become able to understand the logic of general and abstract concepts, "scientific concepts," as he dubbed them. "Scientific concepts" also include concepts that refer to language and its structural properties, such as words, expressions, and sentences, what today would be called metalinguistic knowledge. Viewing or producing a written form makes children aware of language in its own right and as an object of thought thereby bringing thought and language under conscious and deliberate control. Language, as Vygotsky put it, offers "cultural tools" of communication and representation. Growing up in a culture of writing and other '·literate tools," allows language users to become conscious of these particular tools. For Vygotsky and Luria, then, culture was the material and symbolic ensemble of communicational and representational systems transmitted across generations through the institutions of education and the practices of literacy. Vygotsky's and Luria's approach to the dialectics of literacy, thought, and culture fOlmd new support in the 1960s and 1970s. In a series of cross-cultural studies on, among others, the Vai people of Liberia, Michael Cole, Silvia Scribner, and their colleagues compared the cognitive effects of illiteracy,

4

Jens Brockmeier and David R. Olson

schooling and literacy (Cole, Gay, Glick, and Sharp 1971, Scribner 1975, Scribner and Cole 1981). Literacy was investigated in a variety of forms: one part of the Vai population were literate in an indigenous syllabic script system, another one in Arabic, and a third one in English. While the use of Vai script was transmitted informally, English and Arabic were taught in schools. For the researchers, these condition provided a natural laboratory to test the various relations between literacy, thought, and culture. Again, focusing on subjects' logical reasoning, problem solving, and metalinguistic awareness, the studies were intended to follow in the footsteps of Luria's investigations in Central Asia (Cole 1979: 215). Yet the conclusions drawn from the findings were importantly different from those reported by Luria. While they confirmed the influence of schooling on some forms of formal thinking, they contradicted the hypothesis that thought and, in particular, logical reasoning was a direct effect of learning to read and write. Rather, Scribner and Cole (1981) suggested that the specific profile of individuals' intellectual abilities was dependent on the social and practical context in which these reading and writing were used and taught. In this way, they argued for a more contextual and situation-specific approach to the literacy-culture relationship, linking, for example, the cognitive performance of schooled subjects to a particular mode of discourse associated with the institution of formal schooling, rather than the direct consequence of learning to read and write. In his later writings, Cole (1996) further developed this view of culture as a set of contextualized practices and artifacts, which also had become the organizing idea of other sociocultural studies (e.g., Heath 1983, Street 1984, Gee 1990, Wagner 1993, John-Steiner, Panofsky, and Smith 1994, Hamilton, Barton, and Ivanic 1994). In this literature, the cognitive effects of literacy depend first of all on its particular social embeddedness (in terms of class, race, education, religious, national, and local traditions), and on the specific situation in which writing and reading is used to fulfil a concrete function. But how can we tell different contexts and effects? Because there are numerous cognitive effects due to numerous socialcultural contexts in which writing can be embedded, these contextual effects are "difficult to disentangle from those of the ability to read and to write," write Nicolopoulou and Cole (1999: 81). Ultimately, then, the "contextual approach," while renouncing to isolate literacy as a distinctive, mode of communication and representation, suggests a fusion of the concept of literacy with that of culture, conceived of as an array of socially situated practices. Influenced by the works of Clifford Geertz (1973, 1983) and other anthropologists and sociolinguists, this understanding of culture as a multilayered fabric (or "text") of situated practices, discourses, artifacts, and belief systems became wide-spread in the human sciences in the 1980s and 1990s. However, it sometimes tended to cut the explicit links to writing and literacy altogether; while the importance of culture generally was widely recognized, the role of writing as a distinctive historical invention and a cultural practice was minimized, if not denied. 5

Introduction

To be sure, this was not the case in another line of argument about the relation between literacy and culture which has had a strong impact on many discussions in this area. This line leads back to the early 1960s when a number of publications appeared that introduced what became known as the "literacy hypothesis" in sociology, history, classics, and media studies. Especially, in the writings of Eric Havelock (1963, 1982, 1991), Marshall McLuhan (1962) and Jack Goody and Ian Watt (1963, Goody 1987), and, some years later, in those of Walter Ong (1982) a theory of literacy was outlined that made strong claims for the cultural and cognitive implications of writing. It was argued that alphabetic literacy is an unique technology of representation and communication which has been of fundamental importance for the development of Western culture. According to this theory, oral language and written language are intellectual technologies which are causally responsible of two different types of culture, cultures of orality and of literacy. Some critics of the "literacy hypothesis" thus spoke of a "great-divide theory" (Finnegan 1988). The watershed, to stick to the metaphor, between speech and writing, oral and literate culture was the invention (or, once it was invented, the introduction) of the alphabet. As Havelock (1991 : 25) summarizes: At first the alphabet was used to record oral language as previously composed for memorization in Greek epics, lyrics, and drama. The conceptual revolution began when it was realized that the full register of linguistic sound could be placed in a new kind of storage no longer dependent on the rhythms used in oral memory recall. It could become a document, a permanent set of visible shapes, no longer fleeting vibration in the air but shapes that could be laid aside until rescanned for some purposes and indeed forgotten. The mechanisms of the oral memory could then be slowly superseded in favor of documented prose, the first histories, the first philosophies, the first bodies of prosaic law, the first bodies or prosaic rhetoric. Still more, the narrative requirement, the activist syntax, and the living agents required for all oral speech held in the memory could also be laid aside, replaced by a reflexive syntax of definition, description, and analysis. Such was the prose of Plato and all his successors, whether philosophic, scientific, historical, descriptive, legal, or moral. European culture slowly moved over into the ambience of analytic, reflective, interpretative, conceptual prose discourse. Patently, the domain of culture upon which literacy was expected to have its impact was exceedingly broad. Literacy was claimed to impinge upon the entire gamut of cultural phenomena from the intellectual to the aesthetic and political, including the production of science, philosophy, history, literature, art, and religion, as well as the institutions of education, documented law, and democratic forms of social organization. Further, literacy was seen as having an impact on the individualism of modern Western thought along with forms of mentality (rational and logical), cognition (conceptual and analytical), memory (objective

6

Jens Brockmeier and David R. Olson

and accumulative), as well as forms of communication (decontextualized and emotionally distanced) and grammar (reflective and prescriptive). Here, the vision of culture that unfolded with literacy, printing, and the alphabet, merged with the idea of civilization in general. At the same time, however, the idea of writing suggested by the proponents of the "literacy hypothesis" was extremely narrow and reductive in two fundamental ways. It reduced all writing to one form of writing, the alphabet, and it offered a monocausal explanation for vast social changes rather than acknowledging the multiple causes involved in social change. Havelock (1963), for example, examined the differences between the linguistic forms of Homer (oral) and Hesiod (literate) and concluded that the differences could be traced to Hesiod being among the first generation of alphabetic writers. Lloyd (1979; 1990), without denying the relevance of writing, countered that other institutional arrangements, such as the disputative nature of the Greeks and the existence of a public forum were as important and that no one cause could explain radical social change. To understand the essence of alphabetic writing (or, as many would say, to misunderstand it) as being just a translation of sounds into graphic signs is not only characteristic of Havelock's works but wide-spread in linguistics and reading research. In contrast, Harris (1986) and Olson (1994) argued that writing is poorly understood exactly because of this assumption, namely, to treat written signs as transparent to the oral form and (mental contents) it represents. Derrida (1974) goes further to argue that both spoken and written signs tend to be ignored by the tendency to "Iogocentrism," common to all Western thought about language: that is, the view that all signs are only hints to and loose expressions of deeper meanings and truths which ultimately belong to the realm of the logos, the "pure thought" (Brockmeier 1992). Confronted with this realm, the sign always appears to be derivative, profane, transitory. For Harris (1986: 37), to view all writing through the alphabet is the logical outcome of the "ethnocentric bias of an European approach to non-European languages". It fails to take into account the fact that the practical utility of having separate signs for vowels and consonants, the "true alphabet," varies according to the phonological structure of the language concerned. As comparative and historical linguistics has amply illustrated, among the 10,000 or so languages on earth there are fundamental phonological and syntactic differences. A writing system viable for one language is not necessarily viable for another. Harris (1986: 36) sees the irony implicit in the prevailing teleological explanations of the development of the alphabet, "to find writing systems classified and evaluated as if they should have been designed not to meet the practical needs of particular linguistic communities, but rather to serve the universal descriptive purposes of an Abstract Phonology," - as if the historical evolution of writing systems "could be seen to have been gradually working towards the creation of an 'ideal' alphabet as its long-term goal" (Harris, 1986: 37). 7

Introduction

The "literacy hypothesis" has been subject to even more criticism. Many authors have questioned what they regarded as its teleological and ethnocentric bias, cultural reductionism, phonocentrism, technological determinism, and the claim that written texts are qualitatively different from oral discourse because of their ability to fix decontextualized "autonomous" meanings. For a superficial observer, much of this debate on the relationship between literacy and culture might have looked like, at some point, as if there were only two contrasting positions in favor and against the idea of the great divide between the spoken and written language. However, this certainly is a too static picture of a debate in which the positions have been evolving, not least as a result of the debate itself which has continuously been fueled by new evidence from empirical and historical research. Ageliki Nicolopoulou and Michael Cole (1999: 85) have commented that "much of the most interesting work of the last few decades has involved efforts to formulate the problem in a nondichotomous fashion". Speaking from what they call the "contextualist" vantage point, Nicolopoulou and Cole (1999: 85) observe that the literacy hypothesis regarding the cultural and cognitive implications of literacy has become a "moving target". Most of the originally strong claims have been in various ways further developed and redefined, and many initial arguments have been qualified and differentiated (e.g., Good 1987, Olson 1994, Olson and Torrance, 2001) - again both in the light of new research and in response to "contextualist" and "sociocultural" criticism. At the same time, basic assumptions of the contextualist line of argument, as represented, for example, by Scribner and Cole (1981), also have become a subject of critique. Goody (1987, 1999) argued that Scribner and Cole's (1981) crosscultural research applied a concept of literacy that was, first, socioculturally too narrow and, second, based on a merely mentalistic view of linguistic and intellectual operations. To understand what a culture of literacy means, Goody pointed out, it is not enough, if not misleading, to narrow the focus of analysis to the effects that writing has on an individual. Writing is a cultural resource whose implications only unfold in historical time, not in the here and now of individual behavior. Examining the influence of writing on the development of written law, Goody (1987, chapter 4) demonstrated how the listing and classification of laws gradually made legal procedures more comparable and "objective," that is, less dependent on SUbjective decisions based on reign, power, and local custom. On the other hand, once introduced, written law tended to become increasingly extensive and complicated. It needed to be systematized, expounded, interpreted, setting free a new cultural dynamics, again closely connected to specific literacy practices. None of these cultural dynamics, as Goody made clear, would come into light if one reduced the focus of analysis to how individuals were taught to write and to read a law code. To understand how a student learns to solve a mathematical equation does not explain why mathematics is necessary for the construction of railroad bridges, nor why the history of mathematics of the last two hundred years has been closely connected to the evolution of notations and

8

Jens Brockmeier and David R. Olson

the history of education. Thus, from this sociohistorical point of view, Scribner and Cole's (1981) research program missed an essential point of every culture of literacy: namely, that it cannot be determined by testing the direct impact on the cognitive abilities of literate or illiterate individuals. In fact, Cole (1996) has criticized his own earlier research on Vai culture with similar arguments. Again, one can argue that Cole's revisions might have been as much an effect of the dynamics of this debate as of further research and the development of the new Cultural Psychology in the 1980s and 1990s (Bruner 1990, 1996, Shweder 1991, Shweder and LeVine 1984, Val siner 1987, Wertsch 1985). In the wake of this development, for the first time, sociohistorical and cultural perspectives on language, literacy, and thought entered centre stage in psychological discussions (Lee and Smagorinsky 1999). Thus, we would argue that writing has two linguistic and psychological dimensions; both are cultural. One is sociohistorical, the other is individual. This twofold nature is reflected in two different, but interwoven, conceptual components of the term "literacy," a term which can and, in fact, has been used in both contexts of meaning. In front of these two sides of literacy, we distinguish two cross-disciplinary families of theory ofliteracy. The members of one family have focussed their explanatory and investigative efforts on writing as a social, societal, and historical phenomenon, while regarding the individual process (including the process of learning to write and to read) as a derivative issue. In this view, the chief issue is to explain the social functions of literacy in its historical political, economic, and ideological contexts. Once a societal system, a culture of literacy, is established, the individual seems to simply "take over" a set of given material and symbolic resources. Terms such as "acquisition," "appropriation," "socialization," "interiorization," and "internalization" of culture indicate some of the conceptual options offered. In abstracting, in various degrees, from the individual's active potentials of agency and construction, these options often have turned out to get caught in the pitfalls of social or cultural determinism. In contrast, the other family of theories have primarily been concerned with the individual. The chief issue here is the cognitive skills and concepts acquired in learning to read and write and participate in what is vaguely referred to as "the culture". On this view, the historical and evolving set of cultural artifacts and institutions is taken for granted if not ignored. Literacy, then, is only mental skills, not institutional systems. We believe that one lesson to be learned from the debate we have outlined is that neither of the two approaches can capture what defines a culture of literacy in a satisfactory manner without giving an answer to the question with which the other approach is primarily concerned. Moreover, a theory of the relationship between literacy and culture that is conscious of the arguments developed in our debate must embrace not only the two dimensions of writing as historical system and as individual process, but also explain the interface and, that is, the mutual transitions between these two dimensions.

9

Introduction

A pivotal function in these mutual transitions is played by the "graphic body" which is constituted by the spectrum of signs used in a literate culture: systems of writing, scripts, notations, maps, formulas, pictures, texts, and paratexts. Given the central role of written signs in modern intellectual life, it is surprising, as Harris (1995: 6) emphasizes, the extent to which psychology, in particular, has treated the sign "as something externally given, an object already provided by society for the learner to 'acquire' and utilize". Harris argues for making the analysis of the written sign itself the key to the theory of writing. For what he calls "semiological" reasons, he rejects treating writing and reading as activities made possible by the prior existence of written signs; instead, he explains the written sign by reference to the "contextualized integration of the activities of writing and reading" (1995: 7). That is to say, in this view, the written sign is not just a means (given by culture) that is used in the activities of writing and reading; rather, it is the product of these activities, their "integration". Learning to write and to read, to master the "graphic body" is at best an introduction to the world of literacy. In learning to understand and to use written signs, children must learn to make distinctions which the culture as a whole took several millennia to produce. While it is true that children become literate not simply by being taught how to write and to read but by being raised in the universe of discourse of a literate culture, it is nonetheless the case that in becoming literate they are actively engaged in constructing their own theory of writing. For example, they entertain a series of hypothesis about how writing systems work, what words and names mean, and how they connect to the material and social world they live in. These hypotheses about words and names, as has been demonstrated in a number of studies (e.g., Homer and Olson 1999, Homer, Brockmeier, Kamawar, and Olson 2000, Ferreiro 1997), are just those which have been exploited in culture-specific "literacy practices" .. That is to say, the alphabetic writing that children learn and, in the process, re-construct in their way, provides them with a model for understanding what words and names are. It offers them a model for thinking about speech and, that is, for bringing language into consciousness. Writing, in this sense, is a cultural tool of exploration of the unknown, not merely a system of signs for recording the known. Becoming literate implies the active mastery of an ensemble of material, discursive, cognitive, and institutional practice of writing and reading - practices that are in a simplifying fashion called "writing and reading". But they only open the entrance to the "symbolic space" of literacy (Brockmeier 2000), a space of writing that embraces the accumulated cultural resources not only of the history of writing, but also of the writing of history: that is, of storing and transmitting knowledge, of constructing and interpreting archival traditions. It would be naive to assume that an individual reader, just by learning to write and read, will immediately have access to all resources developed in a literate tradition. Literacy is neither only a societal structure, and neither can it be reduced to a basic set of mental skills isolated from everything else; nor, as we argue, can it be captured by the exclusively linguistic-semiological logic of the written signs

10

Jens Brockmeier and David R. Olson

as such. Rather, it is a concept that embraces the cultural resources of a literate tradition - including the writing system(s) of this tradition - and the ensemble of the abilities necessary to exploit these cultural resources. Goody (1987: 222) suggested the idea of "intervening variables" between a culture of literacy and corresponding forms of thought and cognition. One of these variables are cities. Only urban life has produced social and cultural activities and institutions that need and engender writing: law, literature, science, religion, philosophy, and other forms of public discourse. It is in this sense that "cognitively as well as sociologically, writing underpins 'civilization,' the culture of cities" (Goody 1987: 300). Literacy, then, is not simply learning to cope with a script nor to exploit the resources of a literate culture stored in books and other print-media. Literacy is a form of cultural organization itself, what we may call "societal literacy" (Olson and Torrance 2001). This concept has found a place in the early writings of Max Weber (1947) on bureaucracy and documentation as well as in the more recent writings of Michel Foucault (1979), Dorothy Smith (1990), and Georg Elwert (2001), all of whom emphasize the role of written documents in the construction of the modern, bureaucratic social order. Our sciences, arts, literature, economy, and government are all institutional practices based on accessible documentation. The loss of written records would bring a modern society to its end. Evidently, to participate in such institutions requires not only basic literacy but years of professional schooling. This schooling consists in large part in learning to handle the documents specialized to the domain of work - for example, how to read and interpret a manual or a text. Furthermore, the notion of societal literacy undercuts the somewhat romantic view that literacy is merely a convenience. To become a state already assumes documentation, the archives to store them, and the competencies to add to, consult, and make decisions based on them. We already mentioned that the very idea of "the rule of law" assumes either a written form, as in the tradition of Roman Law, or an indexed archive of written cases which can be consulted and compared as in the tradition of the Common Law. Such laws must be sufficiently explicit that judgments may be transparently based upon them. In his careful analysis of the modernization of a Muslim state, Yemen, Brinkley Messick (1993) traces the shifting role of "wisdom" and documentation in legal judgment. Criticism of traditional Shari 'a or Muslim law was based on its reliance on the opaque processes involved in the way religious judges, muftis, reached their decisions. Consequently, it was replaced by codified and legislated forms of law. Messick writes: As the simply organized patrimonial imamate gave way to proliferating bureaucratic segments and to the beginnings of representational government, and as a face-to-face society of witnesses and known reputations yielded to a citizenry of equivalent strangers, so individualized licenses for 11

Introduction the transmission of specific texts were replaced by state diplomas, the unitary opinion of the judge by the collective voice of ( ... ) the bench, and the stand-alone authority of the notary's hand by official registration. In the process, the social basis of the polity is shifting from reckoning by status and kinship ( ... ) to the imagined homogeneity of national citizenry

(253-254). Developing countries aspiring to nationhood consequently devote a substantial portion of their resources to both building the necessary bureaucratic institutional forms and training persons to participate in them. A culture of literacy presupposes a large spectrum of social and psychological conditions. Without taking into account these conditions, their mutual interplay, and their concrete sociohistorical context, there will be neither a satisfactory notion of literacy nor of culture.

References Brockmeier, J. (1992) "Reines Denken ". Zur Kritik der teleologischen Denkform ["Pure Thought": A Critique of the Teleological Form of Thinking], Amsterdam & Philadelphia: B. R. Gruner - John Benjamins. Brockmeier, J. (2000) Literacy as symbolic space. In J. W. Astington (Ed.), Minds in the making (pp. 43-61), Oxford: Blackwell. Bruner, J. S. (1990) Acts of meaning, Cambridge, MA: Harvard University Press. Bruner, J. S. (1996) The culture of education, Cambridge, MA: Harvard University Press. Cole, M. (1979) Epilogue: A portrait of Luria. In A. R. Luria (Ed.), The making of mind (pp. 189-225), Cambridge, MA: Harvard University Press. Cole, M. (1996) Cultural psychology: A once and future discipline, Cambridge, MA: Harvard University Press. Cole, M., Gray, J., Glick, J. & Sharp, D. (1971) The cultural context of learning and thinking, New York: Basics Books. Crawford, V. M., and Val siner, J. (1999) Editorial: Varieties of discursive experience in psychology: Culture understood through the language used. Culture & Psychology, 5, 259-269. Derrida, J. (1974) Ofgrammatology, Baltimore, MD: Johns Hopkins University Press. Elwert, G. (2001) Social literacy: Writing culture and development. In D. R. Olson & N. Torrance (Eds.), The making of literate societies, (pp. 54-67), Oxford: Blackwell. Ferreiro, E. (1997) The word out of (conceptual) context. In C. Pontecorvo (Ed.), Writing development: An interdisciplinary view (pp. 47-59), Amsterdam & Philadelphia: Benjamins. Finnegan, R. (1988) Literacy and orality: Studies in the technology of communication, Oxford: Blackwell. Foucault, M. (1979) Discipline and punish: The birth of prison, New York: Vintage. Gee, J. (1990) Social linguistics and literacies: Ideology in discourses, Brighton: Falmer Press. Geertz, C. (1973) The interpretation of cultures, New York: Basic Books. Geertz, C. (1983) Local knowledge: Further essays in interpretive anthropology, New York: Basic Books.

12

Jens Brockmeier and David R. Olson

Goody, J., & Watt, 1. (1963) The consequences of literacy. Comparative Studies in Society and History. 5. 304-345. Goody, ]. (1987) The interface between the written and the oral. Cambridge: Cambridge University Press. Goody, ]. (1999) The implications of literacy. In D. A. Wagner, R. L. Venezky, & B.Y. Street (Eds.) Literacy: An international handbook, (pp. 29-33), Boulder, co: Westview Press. Hamilton, M., Barton, D., & Ivanic, R. (Eds.) (1994) Words of literacy, Toronto: Ontario Institute for Studies in EducationlClevedon, Philadelphia & Adelaide: Multilinguals Matters. Harris, R. (1986) The origin of writing, London: Duckworth. Harris, R. (1995) Signs of writing, London & New York: Routledge. Havelock, E. A. (1963) Preface to Plato, Cambridge, MA & London: Belknap Press of Harvard University Press. Havelock, E. A. (1982) The Literate revolution in Greece and its cultural consequences, Princeton, NJ: Princeton University Press. Havelock, E. A. (1991) The oral-literate equation: A formula for the modern mind. In D. R. Olson & N. Torrance (1991). Literacy and orality (pp. 11-27), Cambridge: Cambridge University Press. Heath, S. B. (1983) Ways with words: Language, life and work in communities and classrooms, Cambridge: Cambridge University Press. Homer, B. D., & Olson, D. R. (1999) Literacy and children's conception of language. Language and Literacy, 2, 113-140. Homer, B. D., Brockmeier, ]., Kamawar, D., and Olson, D. R. (2001) Between nominalism and realism: Learning to think about words and names. Genetic, Social and General Psychology Monographs, 127, 5-26. Jahoda, G. (1993) Crossroads between culture and mind: Continuities and change in theories of human nature, Cambridge, MA: Harvard University Press. John-Steiner, V., Panofsky, C. & Smith, L. (Eds.) (1994) Sociocultural approaches to language and literacy, New York: Cambridge University Press. Kroeber, A. L., & Kluckhohn, C. (1952) Culture: A critical review of concepts and definitions, Cambridge, MA: Peabody Museum, Vol. 47, No. 1. Lee, C. D., & Smagorinsky, P. (Eds.) (1999) Vygotskian perspectives on literacy research, Cambridge: Cambridge University Press. Lloyd, G. E. R. (1979) Magic, reason, and experience, Cambridge: Cambridge University Press. Lloyd, G. E. R. (1990) Demystifying mentalities, Cambridge: Cambridge University Press. Luria, A. R. (1976) Cognitive development: Its cultural and social foundations, Cambridge, MA: Harvard University Press. McLuhan, M. (1962) The Gutenberg galaxy: The making of typographic man, Toronto: University of Toronto Press. Messick, B. M. (1993) The calligraphic state: Textual domination and history in a Muslim society, Berkeley, CA: University of California Press. Nicolopoulou, A., & Cole, M. (1999) Literacy and cognition. In D. A. Wagner, R. L. Venezky, & B. V. Street (Eds.), Literacy: An international handbook (pp. 81-86), Boulder, CO: Westview. Olson, D. R., & Torrance, N. (Eds.) (1991) Literacy and orality, Cambridge: Cambridge University Press. Olson, D. R. & Torrance, N. (2001) Conceptualizing literacy as a personal skill and as a social practice. In D. R. Olson & N. Torrance (Eds.), The making of literate societies (pp. 3-18), Oxford: Blackwell.

13

Introduction

Olson, D. R. (1994) The world on paper: The conceptual and cognitive implications of writing and reading, Cambridge: Cambridge University Press. Ong, W. (1982) Orality and literacy, London: Methuen. Scholes, R. ]., & Willis, B. J. (1991) Linguists, literacy, and the intensionality of Marshall McLuhan's Western man. In D. R. Olson & N. Torrance (Eds.), Literacy and orality (pp. 215-235), Cambridge: Cambridge University Press. Scribner, S. (1975) Recall of classical syllogisms: A cross-cultural investigation of error on logical problems. In R. Falmagne (Ed.), Reasoning: Representation and process, Hillsdale, NJ: Erlbaum. Scribner, S., & Cole, M. (1981) The psychology of literacy, Cambridge, MA: Harvard University Press. Shweder, R., & LeVine, R. (Eds.) (1984) Culture theory: Essays on mind, self and emotion, Cambridge: Cambridge University. Shweder, R. A. (1991) Thinking through cultures: Expeditions in cultural psychology, Cambridge, MA: Harvard University Press. Smith, D. E. (1990) The conceptual practices of power: A feminist sociology of knowledge, Toronto: University of Toronto Press. Street, B. V. (1984) Literacy in theory and practice, Cambridge: Cambridge University Press. Val siner, ]. (1995) Editorial. Culture & Psychology, 1, 5-10. Val siner, ]. (1987) Culture and the development of children s action: A culturalhistorical theory of developmental psychology, New York: Chichester & Wiley. Vygotsky, L. S. (1978) Mind in society: The development of higher psychological processes, Ed. M. Cole, V. John-Steiner, S. Scribner, & E. Souberman. Cambridge, MA: Harvard University. Vygotsky, L. S. (1987) Thinking and speech, Ed. N. Minick. New York: Plenum. Wagner, D. A. (1993) Literacy, culture, and development: Becoming literate in Marocco, New York: Cambridge University Press. Weber, M. (1947) The theory of social and economic organization, New York: The Free Press. Wertsch, J. V. (1985) Vygotsky and the social formation of mind, Cambridge, MA: Harvard University Press. Wertsch,]. V. (1996) The role of abstract rationality in Vygotsky's image of mind. In A. Tryphon &]. Voneche (Eds.), Piaget - Vygotsky: The social genesis of thought (pp. 25-44), East Sussex: Psychology Press.

14

Part 1 Written culture

This page intentionally left blank

2 The literacy episteme: The rise and fall of a cultural discourse Jens Brockmeier

The literacy episteme: Or how writing was discovered

With the invention of the alphabet began a development of far-reaching importance for the entire Western culture. Although it may be hazardous to summarize ongoing debates about how best to understand the cultural and historical significance of alphabetic literacy, one can say that most scholars today agree on a picture with a Greek origin of the alphabet, its Latin dissemination through Roman antiquity and the Middle Ages, the invention of the printing press, the missionary expansion of alphabetic literacy through the Renaissance and Enlightenment, Christianity and colonialism, the introduction of universal schooling and institutions of letters like public libraries and, finally, the electronic media revolution. At present, an increasing number of researchers are adding countless details to this picture, both in historical and systematic terms. We are witnessing an unprecedented interest in issues of writing, a general cultural concern with literacy and literacies that surely goes far beyond an intellectual fashion wave or the emergence of a new academic topic. How powerful this discourse is, becomes evident by the export of its central categories. Twenty years ago, Clifford Geertz (1983) observed that the "text analogue" was to become a wide-spread model in the human sciences. In the meantime, "writing" and "reading" also have turned into travelling concepts, in intellectual metaphors that effortlessly cross borders between disciplines and discourses. Not only do we read books but also pictures, landscapes, and the "texts" of cultures, lives, and minds. The issue of literacy has become inextricably intermingled with social and societal issues (see, e.g., Olson and Torrance 2001). A recent handbook, which claims to present a compendium of research on literacy at the beginning ofthe new Millennium, outlines its field self-consciously: "The study ofliteracy combines all the social science disciplines, from psychology and linguistics to history, anthropology, sociology, and demographics, but the field itself broadens beyond research to both policy and practice, form childhood though adulthood" (Wagner 17

The literacy episteme

1999: 1). In fact, literacy is discussed as part of the human rights agenda and as an educational right; it has become a rationale of national and international strategies for improved education, human development, and well-being (Triebel 2001). Patently, the new concern with literacy is embedded in a variety of institutional orders which include the economies of knowledge and education, communication and entertainment, public and private administration, politics and the military. Never before have so many interests been engaged in debates about literacy, fostering and funding a variety of theoretical and applied research programs. I have mentioned these cultural and institutional economies because I think that the present theoretical interest in writing and literacy is neither sufficiently characterized in terms of the development of a new object of research nor as the rise of a new paradigm. Since its introduction into the history of science by Thomas S. Kuhn (1962), the term "paradigm" means a perspectival framework of assumptions and beliefs that organizes the concepts, models, theories, and research methodologies of scientific knowledge. The new interest in literacy, however, is not simply tied to the rise of a new paradigm in the human sciences. The idea I wish to offer is that a development of interests and theories like the present one in literacy can only be understood if located within an overarching cultural discourse, a discourse that is laid out not only by academic but also societal and technological coordinates. I shall call this discourse the literacy episteme. I suggest that it is only within such a cultural-historical trajectory that we can capture the rise and fall of writing as a new intellectual and academic field of inquiry. Only within this order, a number of phenomena associated with writing and literacy take on an epistemic form, that is, they become subjects of thought and theoretical curiosity. Only within the literacy episteme can the psychological, intellectual, and cultural implications of writing become epistemic objects - objects that present themselves as intelligible, that is, they can be investigated and their investigation fulfils societal demands and cultural interests. To explain this argument, I shall begin by specifying the notion of episteme, locating it against the backdrop of two epistemological traditions in philosophy and the human sciences. Both traditions have explored the general conditions that make human knowledge possible, the a priori of empirical knowledge. I then shall tum to the more specific reasons and circumstances that allow us to apply this notion to the cultural discourse of literacy. Finally, I wish to address the, at first view, strange absence of linguistics, the discipline of language par excellence, from this picture. I will argue that the discovery of writing as an epistemic object is not an outcome of the traditional academic discourse on language but a result or, more precisely, a side-effect of the cultural transformations that result from twentieth-century media revolution. The episteme as historical a priori An episteme, I have suggested, is a concept that helps to understand what organizes the theoretical order of a culture. Differently put, an episteme is the

18

Jens Brockmeier

cultural order of ideas and concepts that determine, at a given moment in history, what knowledge is and how we gain it - defining, for example, those material and symbolic practices (materialized, e.g., in an experimental laboratory) that make certain things visible and others invisible. Only when I have a notion of, say, "drives" and "the unconscious," can I look at human behavior as being driven by psychological forces that are beyond our conscious control. As the concepts of "drives" and "the unconscious" do not exist in most cultures, nor did they exist during some millennia in Western culture, the difficult question arises whether the same idea of drives and the unconscious that Sigmund Freud used to understand his fellow Viennese around the tum of the last century can be attributed to the self-understanding of Chinese or ancient Greeks. For anthropological, linguistic, cognitive, and psychoanalytical universalists, the answer to this question would be yes. Yet if one holds the view that the meaning of a concept depends on its "historical semantics" - that is, as Reinhard Koselleck (1985) pointed out, the cultural-historical context that gave linguistic concepts and expressions their specific meanings - then the answer is no. It is no because the modem psychoanalytical idea of human nature does not cohere with Asian and ancient Greek thinking about the human condition. The meanings of terms like "drives," "consciousness," and "the unconscious" are embedded in the specific cultural grammar of Western discourse that developed in the nineteenth century (Hacking 1995). Clearly, this understanding depends on how we understand "meaning" itself. The concept of meaning that underlies this historical and contextualized view comes close to Wittgenstein's (1958) idea of meaning as the use of a term. That is to say, meaning is defined by the way a word or expression is used; and this use can change from one episteme to another in much the same way as do the social rules of its use change from one culture to another. Similarly, Charles Taylor (1989) argued that the "moral ontology" of a culture might change from one historical episteme to another. A "moral ontology" is constituted by the set of concepts and assumptions that layout what we believe is good and bad, right and wrong, and ethically appropriate and inappropriate. Closely associated with the moral ontology there is an epistemic ontology, the system of beliefs and rules that set out what is considered to be knowledge and worth knowing - including the relationship between the knowing subject and the possible object of knowledge. The epistemic ontology is the order of being of those things, events, and ideas that, within an episteme, come into the range of our intellectual vision and imagination. All these elements constitute an episteme. Considered as a whole of semantic, ontological, and epistemological ideas, an episteme outlines the horizon of possible knowledge in a particular intellectual and institutional domain of an epoch. In this sense, it is an "historical a priori," as Michel Foucault (1972) suggested. Foucault's famous example is the "episteme of Man," the amalgamation of ideas, forms of knowledge, social and cultural practices that emerged in the eighteenth century and made possible modem thinking about the individual. For Foucault, the "episteme of Man" is the

19

The literacy episteme

cultural system of representation that shaped the modem idea of human being and thinking: be it by introducing categories of the individual subject, self, and personal identity, be it by categories of the individual mind, conscience, and consciousness (including the unconscious). All these categories, Foucault (1972: 127) argued, depend on a priori conditions which attribute "reality" to them, that is, "conditions of emergence of statements, the law of their coexistence with others, the specific form of their mode of being, the principles according to which they survive, become transformed, and disappear". Accordingly, a statement like "writing makes us understand what language is" only has "reality" if there is a "law of coexistence" with other statements that also can claim "reality," such as the statement "writing is a form of language on its own". Speaking of the cultural system of literacy in terms of an historical a priori, we should bear in mind that this implies a rupture with the Kantian tradition in which the notion of the a priori has long played a crucial role. Kant suggested the term a priori to highlight what he viewed as the transcendental dimension of human knowledge. This idea is based on the thought that human thinking always depends on concepts that cannot derive from experience but are necessary preconditions of experience. For Kant, the categories of space and time were such intellectual "conditions of the possibility" of all experience, which he regarded as necessarily given to the individual mind in an a priori fashion. For example, in making possible the localization of concrete experience, time and space constitute such a transcendental framework of categories that give structure and coherence to all possible experience. In the Kantian scheme, this framework is entirely mental and universal, a "diamond net of categories" that is not subject to historical change but itself a precondition of any idea of history. In contrast, for an historical discourse analysis of the a priori conditions of individual knowledge the focus is on the concrete cultural context in which ideas, institutions, and practices emerge and concepts are being used. Accordingly, categories are understood as historical "forms of life" rather than transcendental (or logical, linguistic, or cognitive) universals. It is in this sense that I suggest the term episteme to describe the cultural discourse of literacy. The idea of an episteme organizes what to my mind is the cultural-historical version of Kant's attempt to investigate the intellectual conditions of the possibility of a system of knowledge. My argument, then, is to understand the literacy episteme as an historical a priori. It defines the terms under which we conceptualize the "symbolic space" of literacy that characterizes our culture (Brockmeier 2000). Most notably, the literacy episteme has brought to our consciousness two ideas that, a few decades ago, were absent from academic and public discourse on language and culture. One is that writing is an independent and peculiar form of language, and not just a secondary representation of speech; the other one is that the traditional form of written language, as a dominant feature of our civilization, is a particular historical form which has, as all historical phenomena, a beginning and an end. And the end has just come into sight. 20

Jens Brockmeier

Pioneering years

One of the most amazing aspects of the literacy episteme is that it has taken form only very recently. This is all the more astonishing given the long history of the alphabet and the even longer history of writing in general. The phenomenon of writing confronts us with an interesting time gap between invention and discovery. An invention creates something new in life, a discovery uncovers something already present but unknown. Whereas the invention of the first writing systems dates back to prehistoric times, and the invention of the alphabet to about three thousand years, the discovery of writing has taken place only over the last few decades. I should emphasize that by "discovery" I refer to what I have described as the emergence of an epistemic object. Of course, the phenomenon of writing had already been dealt with by earlier authors in philosophical, religious, social, and political contexts - from Luther, Erasmus, and the early Protestants to the philosophers of the Enlightenment such as Rousseau, Vico, Turgot, and Astle (see Brockmeier 1998; Olson 1994). However, the "discovery" of writing and literacy which I am addressing here has a different epistemic quality. Amazingly enough, we can pinpoint the events in question quite precisely. It was at the beginning of the 1960s, when writing, its significance, origin, and future became an epistemic object. All of a sudden, literacy seemed to have turned into a prestigious subject, worthy of intellectual attention and systematic academic investigation. Within a few years, a number of books set the stage for a new epistemic ontology of language. One of these books was Preface to Plato by the classicist and theorist of culture Eric Havelock. Published in 1963, it was to become a milestone in the short history of the discovery of literacy. Preface to Plato set out to present Havelock's view of a "literate revolution" that happened upon the introduction of the alphabet into the oral culture of ancient Greece. In hindsight, Havelock (1986) mentioned three more works that came out at the same time as his study. Recalling Havelock's suggestions, I would like to add some further publications which might broaden the frame of those events sketched by Havelock. The discovery of the fact that both orality and literacy were not naturally given ways of language use but rather distinct semiotic and cultural productions was from the very beginning accompanied by a new valuation of traditional, nonliterate or oral cultures - a shift, we might say, in the moral ontology. A turning point in this reassessment is Claude Levi-Strauss's structural anthropology. When Levi-Strauss published in 1962 his La Pensee sauvage, it already indicated in its title a new approach to a kind of thinking which most anthropologists and philosophers had until then considered to be the "primitive" (Malinowski), "pre-logical" (Levy-Bruhl), and "mythic" (Cassirer) thought of native people. In this book, Levi-Strauss describes the structures of a "savage thinking" whose cognitive and communicative basics he attributes to "natural" orality which is in sharp contrast with the "artificial" structures of literacy of Western culture. Rightly, Havelock (1991: 21) wrote, what Levi-Strauss was

21

The literacy episteme

investigating was not la pensee sauvage but what he hoped to identify as la pensee oraliste. A few years earlier, Albert Lord (1960) had made a similar point about the "oral mind" and its sophisticated mnemonic techniques. It is not least this evocation of a deep gap between oral and literate cultures that caught the imagination of an entire generation of intellectuals. La Pensee Sauvage was published one year before Havelock's Preface to Plato. Also in 1963, the sociologists Jack Goody and Ian Watt presented a first version of their cross-cultural studies The Consequences of Literacy, and the biologist Ernst Mayr published his book Species and Evolution which, among others, highlighted the pivotal role of language in the evolution of the human species (because language, as Mayr showed, allows for cultural development to be superimposed upon the biological). In the same year appeared the classicist Herman Koller's study on the relation between ancient orality and literacy Dichtung und Musik im friihen Griechenland [Poetry and Music in Early Greece], and the volume L'ecriture et la psychologie des peuples [Writing and Folk Psychology], edited by Marcel Cohen. And of course, there was, in 1962, the publication of Marshall McLuhan's world-wide classic The Gutenberg Galaxy which formulated the spirit the new orientation just taking shape in a most spectacular manner. Media revolutions, McLuhan claimed, are revolutions of consciousness. They are not just about the form of communication and cognition, but about their very content. A work that belongs to still a different intellectual tradition but ought to be seen in this context is La geste e la parole [Gesture and Speech] (volume 1: Technique et langage and volume 2: La memorire e les rythmes) by prehistorian and paleontolgist Andre Leroi-Gourhan. Published in 1964/65 (in English 1993), it drew an evolutionary and cultural-historical line from the first traces of prehistoric writing practices to the computer age - the line of a continuous "externalization of memory". In 1967, another influential study on the "technologies" of orality and literacy appeared, Walter Ong's The Presence of the Word. In the same year, Jacques Derrida published his two books De la grammatologie [Of Grammatology, in English 1974] and L 'ecriture et la difference [Writing and difference] that would refine the philosophical discourse of modernity in the light of what Derrida called ecriture. Derrida outlined a philosophical notion of writing that suggested that we can understand the history of Western metaphysics as based on a kind of repression of the significance of the written and material sign. As a consequence, Derrida argued, philosophy and linguistics focussed on the idea that words can communicate fixed meanings without any direct link to the material sign or mediation, which implies conceiving of signs as completely transparent. Derrida called this idea the "metaphysics of presence". Finally, let me add here another important work that in 1962 was published for the first time in the West, Lev S. Vygotsky's Thought and Language. Certainly, the semiotic tradition of psychological research on (oral and written) language and thought that has developed in the wake of Vygotsky and his

22

Jens Brockmeier

colleague Aleksandr R. Luria did not start to flourish until the 1970s and 1980s. But in retrospect, it appears to be more than just an accident that 1962 also was the starting point of this tradition that has not only highlighted the importance of language for thought and culture but also emphasized the particular role of writing and, what's more, the impact of growing up in a culture of literacy for cognitive and social development (see the Introduction to this volume). Looking at these publications in the 1960s with the distance of some decades, they present an astounding coincidence, all the more, because before then no noteworthy communication had taken place among the various authors. However, we can suspect that the sudden and almost simultaneous appearance of these investigations - which soon would be followed by a great number of further studies, conferences, research projects, university programs, and the like - must have been the result of an ongoing multi-centered discussion, a new awareness and sensitivity that must have begun to surface some time before. An episteme, r have argued, is not just about books and bookishness. But books can be indicators. They can be seismographs, and the publications r have mentioned certainly were. Their almost concurrent appearance marked a turning point, a "watershed," to borrow Havelock's picture (1991: 12), "that had been reached, or perhaps more accurately they point to a dam starting to burst," releasing a flood of cultural interest and intellectual activity devoted both to the evocation of the importance ofliteracy and to understanding it. Yet before I raise again the question what brought these and many more authors to direct their attention to issues of writing and reading, let me address another question posed by the rise of the literacy episteme. When we look at the protagonists of the intellectual scene just outlined, we find classicists, anthropologists, philosophers, theologians, theorists of media and culture, sociologists, psychologists, biologists, and paleontologists. However, what we do not find are outstanding representatives of that discipline which, one should assume, is meant to playa major role at the forefront of the study of language, oral and written, namely: linguists. How do we account for this surprising absence? It is worthwhile to discuss this question in more detail, for this may help us see more clearly why such a seemingly manifest phenomenon like writing had to be discovered at all. Why did writing become an epistemic object only thousands of years after its invention? Linguistics without literacy

Most linguists agree on dividing their field into the four areas of phonology, morphology, syntax, and semantics, with grammar traditionally being considered the central axis of the subject. It is grammar that the most scientific attention has been dedicated to, where the highest intellectual claims have been formulated, and where the greatest academic prestige could be gained. The kind of linguistics that together with this conception of language has claimed a similar prominence is often called "theoretical synchronic microlinguistics". It is 23

The literacy episteme

"theoretical" linguistics as opposed to applied (which would include the study of "practical" contexts in which language is used, for example, where it is learned, taught, translated, written and read); it is "synchronic" linguistics as opposed to diachronic (which would include the historical development of language and, in this way, the manifold changes that language undergoes as part of a cultural, and that is temporal, life world); and it is "micro linguistic" as opposed to macrolinguistic (which would include the communicative, social and psychological functions of language, that is, its character as a cultural system of forms of life, to use Wittgenstein's terms). In this sense, Saussure and Chomsky, the two scholars who most shaped the academic study of language in the twentieth century, are both quintessential theoretical and synchronic microlinguists (or, in short, theoretical linguists). As has often been pointed out, for both scholars the social and discursive functions of language were equally irrelevant as the role language plays in a culture and as a form of culture. It is not difficult to see that the idea of linguistics as the science of universal laws or rules that underlie all languages necessarily led to the exclusion of attention to the concrete forms of use in which language (and languages) exists. As a consequence of the conviction that the nature of languages is to be determined by "fixed codes," which is the central assumption of what Roy Harris (1981) has called the "language myth" of Western linguistic thought, the Saussurian-Chomskyian mainstream of general linguistics never conceptualized "language in use". Expelled from the core definition oflanguage, the pragmatic and discursive dimension of language was instead delegated to a kind of independent sub-discipline called sociolinguistics. And even this has happened only over the last decades. More importantly in the context of my argument, the systematic exclusion of "language in use" from the theoretical and empirical center of linguistics has, not least, resulted in another epistemic exclusion: that of literacy. Historically, this exclusion can be traced back to Aristotle's (1963) view that spoken words are the symbols of mental experience and written words are the symbols of spoken words - in short, that writing is speech "put down". The Aristotelian paradigm, it has been argued (Derrida 1974, Harris 1986, Olson 1994), has even dominated twentieth-century linguistics, philosophy, and psychology of language; significantly enough, the three most influential linguistic theories about the origin and development of writing, those of Gelb (1963), Cohen (1958), and Diringer (1968) suggested a picture of writing systems teleologically evolving towards the phonocentric ideal of the alphabet which postulates the precise representation or "perfect transcription" of sound patterns by visible signs. However, there are a few exceptions. One of them is Jacques Derrida. His philosophical approach to writing is particularly interesting because, as already mentioned, it also aims to explain why Western thought about language persistently avoided thematizing the materiality of the linguistic sign. For Derrida, philosophy's and linguistics' difficulties with recognizing the semiotic 24

Jens Brockmeier

independence of writing is part of a bigger intellectual project that he called phonocentrism. Examining the works of Rousseau, Hegel, Husserl, de Saussure, and other philosophers and theorists oflanguage, Derrida (1974, 1993) pointed out that phonocentrism must be conceived of as an intrinsic part of Western history of ideas. The focus on the voice (phone), on oral speech, is but a consequence of the rejection of writing as a mere secondary representation, an appendage that only has function and meaning in as far as it is a transparent symbol of the spoken word. Why, then, could this idea become such a pervasive conviction? Derrida argues that phonocentrism and the "metaphysics of immediacy" is not just a linguistic project but the symptom of a more fundamental tendency of exclusion and inclusion that relates phonocentrism to logocentrism. Logocentrism is the belief that both spoken and written signs are only hints at, and external expressions of, deeper meanings and truths that lie either in the thoughts of men or the minds of Gods. As a consequence, the first and last things are "ethereal" entities such as Logos and the Divine Word, soul and spirit, Hegel's Geist (mind! spirit) and Kant's transcendental subject. Viewed in this way, phonocentrism and logocentrism appear to be bound into a moral ontology that privileges the mind over the body, and the spiritual and intelligible over the sensible and material. In sum, the priority of the immaterial voice over writing is interwoven with a complex fabric of philosophical and religious motives (Brockmeier 1990). Against this backdrop, I suspect that the linguistic expulsion of the use- and context-bound dimension of language - and written language, I have argued, is a part of this dimension - is a side effect of this underlying fabric of phonocentrism and logocentrism. Let me highlight another reason why the Aristotelian paradigm of phonocentrism has remained unrivaled in modem theoretical linguistics. This is the assumption of universalism, the claim that there is such a thing as one universal model of language, be it theological, transcendental, logical, or biological. For each category, there are longstanding theoretical traditions that make the case of universalism. All of them use arguments based on the distinction between natural sciences and cultural-historical science, brought up in the Enlightenment and, philosophically, dating back to Plato (Berlin 1981, Shweder 1991). In consequence, it has been regarded as the principal task of theoretical (that is, "scientific") linguistics to find and describe these universal laws of language. I also subsume under this assumption the weaker variant that there are such things as diverse languages, each of them, however, with a determinable structure, a "fixed code". Harris (1981) remarked that already the singular noun form of "language," the taken-for-granted basic category of theoretical synchronic linguistics, carries a number of precarious implications; as if the fact that nature had provided the human species with a unique biological equipment, the ability to speak, would suggest that from this genetic universal somehow flowed the whole range of diverse practices acknowledged as "linguistic". The philosophical, linguistic, and narratological search for an

25

The literacy episteme

"universal grammar" (Herman 1995) and, in a broader cultural picture, the search for a "perfect language" (Eco 1995) are probably the most advanced and, at the same time, most curious and bizarre forms in which the idea of a singular ur-Ianguage has pervaded European culture. Needless to say, both "universal grammar" and the "perfect language" have exclusively been conceived of as gestalts of oral language. Arguably, the universalist assumption plays a pivotal role in this connection. Imagine for a moment it did not have this hegemonic status in linguistics and cognitive sciences. The alternative approach would be to investigate the variety of concrete ways language is used in different cultures, including different historical periods - an approach that certainly would make it much more difficult, if not impossible, simply to ignore that writing is a specific linguistic mode on its own, a communicational and representational medium sui generis. No linguistics without literacy

With the abstraction of grammar and semantics from pragmatics and the universalist assumption I have singled out two principal aspects of linguistic phonocentrism. I would like to add a third point that will shed a more dialectical light on the relation between writing and literacy on the one hand, and the phonocentrism of linguistics and philosophy of language and mind on the other hand. Although it is true that the written word has been entirely in the shadow of spoken language, it has, at the same time, been present in all linguistic discourse. In a sense, it has even been a basic condition of it. In contrast with the far-reaching claims of the "literacy hypothesis" put forward in the 1960s and 1970s by authors like Havelock, Goody and Watts, and Ong, today, the specific intellectual or cognitive functions attributed to alphabetic writing are regarded as more modest. Writing, it has been argued, restructures modes of thought, including thought about language (e.g., Harris 1989); in "quoting" language, it provides a model for thinking about language (e.g., Olson 1994, Olson and Kamawar, in this volume), for example, in defining linguistic properties such as words, clauses, and sentences (e.g., Ferreiro 1997, Saenger 1997); and it makes one aware of language as a particular semiotic and communicative medium on its own (e.g., Brockmeier 1998). Considering the gist of these arguments, one can be inclined to believe that linguistics, the chief academic endeavor of thinking about language, has been possible primarily because of the existence of writing and literacy. One might even suspect that the kind of decontextualization of language which is realized in the abstraction of grammar from pragmatics and in the universalist assumption, presupposes writing and "literate thinking" as an intellectual and semiotic condition of its very possibility. A similar case has been made for the traditional linguistic view of orality. Scholes and Willis (1991) discussed the implications of Whorf's principle of linguistic relativity for linguistics itself. Whorf's claim was that a culture's

26

Jens Brockmeier

grammar serves as a filter through which the members of this culture view the world. Applied to linguistics, this suggests that the description of a highly context-bound and use-oriented oral language by a formally sophisticated linguist is more likely to reflect the specific linguistic and cognitive competence of the (literate) linguist than that of the (oral) language user. Scholes and Willis (1991: 229-30) therefore conclude that "in the context of literate linguists characterizing illiterate speaker-hearer ( ... ), Whorf's hypothesis is surely true". In language-based social sciences such as anthropology and ethnography, the problem of the orality-literacy transformation, to which Scholes and Willis refer, has led to a crisis in the traditional concept of "objective representation." But not so in theoretical linguistics. In light of the arguments outlined, this does not come as a surprise. However, one can argue that Scholes and Willis' point is also true for the systematic understanding of language in terms of theoretical linguistics and, especially, for what I am here concerned with, namely, the exclusion of literacy from linguistics - and that is, in turn, the exclusion of theoretical linguistics from the literacy episteme. Considering the concrete activities of professional linguists, it undoubtedly seems strange, as Goody (1987: 261) writes, that "a group of human beings who probably spend more time reading and writing than they do speaking and listening, have been so oblivious to the social and psychological implications of their craft". One should not forget that languages are all but the primary subjects of the linguist's inquiry. They are second-order products of a scientific community whose actual linguistic experience (based on a set of highly formalized literacy practices) is what first has to be made sense of - both in order to reconstruct how linguistic categories are produced and to understand what "languages" are (Harris 1997). The construction of "linguistic facts" is in this respect not basically different from the construction of "facts" or "epistemic things" in the natural sciences (e.g., Latour and Woolgar 1979, Rheinberger 1997) and social sciences (e.g., Smith 1990). Yet the critical focus on the intellectual construction of concepts and theories is only half the story. A scientific community is not only constituted by shared conceptual and theoretical assumptions but also by institutional realities. More than merely being collective mental or linguistic representations, these realities include financial and administrative constraints, as well as the actual supervisory activities of authorities and powerful organizations. In his recent work, Olson (in press, Olson and Torrance 2001) has emphasized that the linguistics of spoken and written language consists not only of ideas and theories above and beyond what speakers and writers of a language know and share. These theories and ideas are always bound into institutional structures that come into play when teachers, academics, politicians, and other rule-monitoring institutions come onto the scene, institutions which have the power to enforce norms of linguistic practice. Such norms range from standards of academic writing to elementary rules of language practices at school. "When teachers tell children how to

27

The literacy episteme

correctly spell 'weight' and when those teachers have the power to enforce, through testing, the children's spelling [and thereby literacy standards of orthography], we encounter the significance of institutional knowledge and power" (Olson, in press). Similarly, the institutional norms and disciplinary standards of "academic literacies" (Lea and Street 1999) aim to shape scientific discourse, including discourse on spoken and written language. Recent studies have demonstrated in much detail that academic discourse is not the neutral linguistic expression of new findings and insights but that body of style and genre conventions which the institutions of science have legitimized as in conformity with an established epistemic code (eg., Bazerman 1988, Geissler 1994, Berkenkotter and Huckin 1995). Margaret Proctor (in this volume) argues that a particular genre of academic literacy, the essay, is essential in establishing the hegemony of this epistemic code in today's North American university. Students' obligatory mastering of the essay is not simply a gate-keeping ritual, but also plays a crucial role in imposing the very cultural grammar of this epistemic code. The irony thus is that although linguistics as a discipline not only has emerged as a by-product of writing and, in fact, depends in almost every aspect of its existence on practices of writing and reading, it has from its beginning ruled out the subject matter of literacy from its institutional domain. It is not least for this reason that the rise of the literacy episteme took place almost entirely beyond the boundaries of mainstream linguistics. To be sure, over the last few decades there has been a remarkable increase in the number of linguists "willing to consider writing as a form of language in its own right," as Harris (1995: 3) observes; and this is particularly true if confronted with the situation at the time that I. J. Gelb (1952) published his A Study of Writing, for a long time the standard work on the issue. However, these changes patently only began to take place after the dam had already burst under the pressure of technological and cultural innovation and the flood of a novel discourse had given shape to the new literacy episteme - fueled by ongoing debates in many fields, from anthropology and classics to poststructuralist theories of text, intertextuality, and culture. Notably enough, the first systematic linguistic theory of writing that has been proposed in the light of the new episteme is Signs of Writing by Roy Harris (1995), an outspoken outsider and critic of the linguistic mainstream.

The medium turn Today, there can be little doubt that the emergence of this novel orientation cannot be accounted for within academic constellations. Rather, it can only be understood within the trajectories of a far-reaching cultural transformation, a transformation that originated in a technological revolution. To be more precise, what we face here is a number of, by now, well-known revolutions in the technology of media and communication. 28

Jens Brockmeier

Not many realized the significance of twentieth-century electronic media revolutions for the shift in the architecture of our knowledge about orality and literacy as early as McLuhan (1962; 1964). And he set the tone. Insisting on a mutual connection between technological development and cultural change in the Western means of communication, reflection, and imagination, he probably was the most timely media-eloquent prophet of the new episteme. However, as we saw, McLuhan was not the only one who began to reflect upon the fact that television and its predecessors, radio and telephone, were transforming the reach of the spoken word (that is, of orality), while at the same time establishing a spectrum of new literacies. Certainly, McLuhan did not give an answer (as he himself saw it), but raised a question or, more exactly, a series of questions. What do we mean by "orality" and "literacy," and what is the difference between their "old" and "new" (that is, electronically mediated) forms? And, on a more fundamental plane, how does oral discourse relate to non-oral, literate discourse, and how do both affect thinking? These questions were about to take shape in various cultural milieus in North America and Western Europe. However, as Havelock (1991) remarked years later, it is misleading to view the appearance of this new sensitivity as the conscious and coordinated emergence of a new awareness of a problem or even a new vision. Rather, it was as a reaction to something else: "Was this grouping as it occurred a pure accident or did it reflect a common and widespread response, even if an unconscious one, in France, England, the United States, and Canada, to a shared experience of a technological revolution in the means of human communication?" (Havelock 1991: 12) If the shared experience of the technological transformation of public discourse, which seemed to have culminated in the idea of the "global village" of 1960s' television culture, can be seen as the backdrop of what I have called the rise of the literacy episteme, then it fully blossomed with the electronic revolutions in 1970s, 1980s, and 1990s - that is, the emergence of the personal computer, e-mail, and the Internet. Yet it was not the new electronic technology alone that led to a new sensitivity towards the cultural significance of literacy. Another development overlapped with the diffusion of the new technological possibilities to the point that it became difficult to distinguish between the two. This was the development of traditional literacy which was gaining an unprecedented momentum, in fact, a new quality. Not least as a consequence of the new technological possibilities, the last decades have witnessed an exponential increase in all kinds of printed documents - a flood of writing unimaginable to previous generations. In the second half of the twentieth century more books, newspapers, journals, and other written materials have been printed and published than in all the centuries since Gutenberg. There is a wide-spread perception that writing has spun out of control. No library can contain it all, and even Jorge Luis Borges' (1988) fantasy of megalomanic bookishness, the "Library of Babel" which contains all possible and impossible books, would fall short of embracing the countless virtual libraries that come every day into existence in cyberspace. The most disciplined 29

The literacy episteme

reader can read any but a minute fraction of them and often does not even know of their existence. Nevertheless, I believe that Borges' famous narrative captures well the diffuse anxieties emerging from this human condition. These are anxieties, to borrow another almost Borgeian image from Roy Harris, that perceive writing like a fungus that spreads everywhere and clogs everything; there is no pesticide that will halt it. Being a human invention, it now threatens to displace its inventor. As the total corpus of written knowledge expands beyond all imagination, so the share of knowledge that the individual can hope to attain in a lifetime diminishes proportionately. The increase of human knowledge as a whole comes along with the increased perception of individual powerlessness. Today, this anxiety feels even stronger, as it seems that writing itself will more and more pass out of control. In a sense, it is already taken over by machines (see also Harris essay on Literacy and the future of writing in this volume). If there is a place where the poststructuralist's vision of a decentered and continuously expanding intertextual universe takes on material reality, then it is the symbolic space of the internet: its textual discourse is without limit, center, and control. Back in the 1960s, McLuhan was, again, not the only one who spelled out the vision that media revolutions redraw the maps for culture and the mind. But he gave the most articulated voice to this idea that would have been developed, in the decades to come, in countless variations and applied to almost every facet of the "digital culture" (for a survey see Levinson 1999). Entire new disciplines - such as Communication, Cultural Studies, and Media Studies - were built upon this thesis. McLuhan was both, a sharp theorist and the Andy Warhol of the fledgling media and communication research. Earlier than most, he brought the new spirit up to the point. Warhol was the first artist whose message was intrinsically connected to mass media and television, in particular. His art was part of the rise of a public discourse that was not only mediated but also staged by electronic media. Warhol became the icon and, to no small extent, the instrument of a change in public perception, a change in which publicity and media attention fused. And he got both in abundance, because "the Sixties in New York reshuffled and stacked the social deck: the press and television, in their pervasiveness, constructed a kind of parallel universe in which the hierarchical orders of American society - vestiges, it was thought, but strong ones, and based on inherited wealth - were replaced by the new tyranny of the 'interesting' ," as Robert Hughes (1982: 212) wrote. The "interesting" attracted the media, this was the old formula. The media define, and what's more, they create the "interesting," this was the new formula, the formula of Warhol's aesthetics. If Warhol found an expression - actually, he was the expression - for the emerging culture of media, art, and money, McLuhan did so for the emerging culture of media, communication, and globalization. What he articulated was the 30

Jens Brockmeier

insight that there is no such thing as a neutral medium. The semiotic media of communication, thinking, and imagination have a decisive impact on the very content of communication, thinking, and imagination. Ultimately, every representation of a meaning leads to its recreation, every mediation of a meaning constitutes a new meaning. Viewed again in retrospect, this new focus on the semiotic mediation of culture and thought - with literacy being just one form or mode of such mediation - appears to be an advanced version of the linguistic tum in philosophy and the human sciences, that is, the tum to the study of the manifold functions oflanguage in human life. It seems as if the new medium turn, we may also say semiotic turn, which has come along with the rise of the literacy episteme, continues, extends, and radicalizes a central motif of an old-age tradition of thinking about language and culture. This tradition ranges from Humboldt's notion of language as energeia, as activity, and not just as ergon, as tool or instrument (see Taylor 1985), to Sapir-Whorf's relativity theory of language, Vygotsky's idea of language as cultural sign system, and Wittgenstein's view of language as form of life. In this vein, historians of culture have suggested a "relativity theory of media" (Assmann 1990). Media revolutions are revolutions of consciousness and human senses. They not only establish new ways to mediate communication and thought, but also create new realities. Arguably, the technological and social transformation of modem communication has already had a culturally much more far-reaching impact on our thoughts and imagination than all linguistic thinking. It says a great deal about the closed and self-referential discourse of Western academic culture that the question of the particular nature of writing could be posed clearly for the first time only after the traditional phonocentric dogmas about the relationship between speech and writing had been questioned from the outside, subjected both "to the brash counterpropaganda of a McLuhan and to the inquisitorial scepticism of a Derrida," as Harris (1986: 158) put it. Harris goes on to state that it says even more "that the question could not be posed clearly until writing itself had dwindled to microchip dimensions". It is another ironic point that the registers of the literacy episteme whose rise, as I have argued in this essay, seems to have only been possible in the wake of the technological media revolution, at the same time witness how the once universal cultural hegemony of writing is itself coming to an end. To be clear about this, it is not the practice of writing that is coming to an end but the intellectual and cultural attitude, originated in the modernist episteme, that has regarded the traditional form of writing and its institutions as the epitome and warranty of civilization - not only of Western civilization but of human Civilization as such. Ultimately, then, the condition of the possibility of the rise of the literacy episteme is the fall of the traditional literacy culture. In this sense, what I have described as the "discovery" of this culture, the configuration of writing as an epistemic object, is the unmistakable sign of its end. It seems thus, as if once

31

The literacy episteme

again the owl of Minerva, the heraldic animal of philosophical knowledge and reflection, has started its flight only when dusk begins to fall and the events of the day are over. Reference Aristotle (1963) Categories and De Interpretatione (transl. by J. L. Ackrill). Oxford: Clarendon Press. Assmann, A. &. J. (1990) Schrift, Kognition, Evolution. Eric A. Havelock und die Technologie kultureller Kommunikation [Writing, cognition, evolution: Eric A. Havelock and the technology of cultural communication.] Introduction to E. A. Havelock, Schriftlichkeit. Das griechische Alphabet als kulturelle Revolution (pp. 1-35). Weinheim: VCH Acta Humaniora. Bazerman, C. (1988) Shaping written knowledge: The genre and activity of the experimental article in science. Madison, WI: University of Madison Press. Berkenkotter, c., & Huckin, T. (1995) Genre knowledge in disciplinary communication: Cognitioniculture/power. Mahwah, NJ: Lawrence Erlbaum. Berlin, 1. (1981) Against the current: Essays in the history of ideas. Oxford: Oxford University Press. Borges, J. L. (1988) Labyrinths: Selected stories and other writings, Ed. by D. A. Yates & J. E. Irby. New York: New Directions Brockmeier, J. (1990) Language, thought, and writing: Hegel after deconstruction and linguistic turn. Bulletin of the Hegel Society of Great Britain, 20121, 20-54. Brockmeier, J. (1998) Literales BewujJtsein. Schriftlichkeit und das Verhiiltnis von Sprache und Kultur [The literate mind: Literacy and the relationship between language and culture]' Munich: Fink. Brockmeier, J. (2000) Literacy as symbolic space. In J. W. Astington (Ed.), Minds in the making (pp. 43-61), Oxford: Blackwell. Cohen, M. (1958) La grande invention de l'ecriture et son evolution. Paris: Klincksieck. Cohen, M. et al. (1963) L'ecriture et la psychologie des peuples. Centre International de synthese. Paris: Colin. Derrida, J. (1967) De la grammatologie. Paris: Editions de Minuit. Derrida, J. (1967) L 'ecriture et la difference. Paris: Editions du Seuil. Derrida, J. (1974) Ofgrammatology. Baltimore, MD: Johns Hopkins University Press. Derrida, J. (1978) Writing and difference. Chicago, IL: University of Chicago Press. Derrida, J. (1993) Speech and phenomena. Evanston, IL: Northwestern University Press. Diringer, D. (1968) The alphabet: A key to the history of mankind (1st Ed. 1948), London: Hutchinson. Eco, U. (1995) The search for the perfect language. Cambridge, MA: Blackwell. Ferreiro, E. (1997) The word out of (conceptual) context. In C. Pontecorvo (Ed.), Writing development: An interdisciplinary view (pp. 47-59), Amsterdam & Philadelphia: John Benjamins. Foucault, M. (1972) The order of things, New York: Pantheon Books. Geertz, C. (1983) Local knowledge: Further essays in interpretive anthropology, New York: Basic Books. Geisler, C. (1994) Academic literacy and the nature of expertise: Reading, writing, and knowing in academic philosophy, Hillsdale, NJ: Lawrence Erlbaum. Gelb, 1. J. (1963) A study of writing (1st Ed. 1952), Chicago, IL: University of Chicago Press.

32

Jens Brockmeier

Goody, J., & Watt, 1. (1963) The consequences of literacy. Comparative Studies in Society and History. 5. 304-345. Goody, J. (1987) The interrace between the written and the oral, Cambridge: Cambridge University Press. Hacking, 1. (1995) Rewriting the soul: Multiple personality and the sciences ofmemorv.. Princeton, NJ: Princeton University Press. Harris, R. (1981) The language myth. London: Duckworth. Harris, R. (1986) The origin of writing, London: Duckworth. Harris, R. (1989) How does writing restructure thought? Language and Communication, 9, 99-106. Harris, R. (1995) Signs of writing. London & New York: Routledge. Harris, R. (1997) From an integrational point of view. In G. Wolf & N. Love (Eds.), Linguistics inside out: Roy Harris and his critics (pp. 229-310), Amsterdam & Philadelphia: John Benjamins. Havelock, E. A. (1963) Preface to Plato, Cambridge, MA & London: Belknap Press of Harvard University Press. Havelock, E. A. (1986) The muse learns to write: Reflections on orality and literacy from Antiquity to the present, New Haven, CT: Yale University Press. Havelock, E. A. (1991) The oral-literature equation: A formula for the modern mind. In D. R. Olson & N. Torrance (Eds.), Literacy and orality (pp. 11-27), Cambridge: Cambridge University Press. Herman, D. (1995) Universal grammar and narrative{orm, Durham, NC and London: Duke University Press. Hughes, R. (1982) The rise of Andy Warhol. The New York Review ofBooks, February 18. Quoted from Anthology: Selected Essaysji"()m thejirst 30 years ofThe New York Review o{Books (pp. 205-224), New York (1993). Koller, H. (1963) Dichtllng und Musik im Fiihen Griechenland, Bern & Munich: Francke. Koselleck, R. (1985) Futures past: On the semantics of historical time, Cambridge, MA & London: MIT Press. Kuhn, T. S. (1962) The structure o{scientijic revolutions, Chicago, IL: The University of Chicago Press. Latour, B., & Woolgar, B. (1979) Laboratory lire: The social construction o{scientijic jacts, Beverly Hills, CA: Sage. Lea, M., & Street, B. (1999) Writing as academic literacies: Understanding textual practices in higher education. In C. Candlin & K. Hyland (Eds.), Writing: Texts, processes and practices (pp. 62-81), London: Longman. Leroi-Gourhan, A. (1964, 196.5) La geste e la parole (Vol. 1 Techique et langage. Vol. 2 La memorire e les rythmes), Paris: A. Michel. Leroi-Gourhan, A. (1993) Gesture and speech. Cambridge, MA: MIT Press. Levinson, P. (1999) Digital McLuhan: A guide to the injormation Millennium, London and New York: Rourledge. Levi-Strauss. C. (1962) La pensee sauvage, Paris: Pion. Lord, A. B. (1960) The singer of'tales, Camhridge, MA: Harvard University Press. Mayr, E. (1963) Animal specie\' and el'olurioll, Camhridge, MA: Belknap Press of Harvard University Press. McLuhan, M. (1962) The Gutenberg galaxy: The making o/tvpographie man, Toronto: Univ('fsity of Toronto Press. McLuhan, 1\1. (1964) Understanding media: The extensions of man, New York: McGraw-Hill. Olson, D. R. (1994) 771e world on paper: The cognitive and conceptual implication of writing alld reading, Cambridge: Cambridge University Press.

33

The literacy episteme

Olson, D. R. (in press) The cognitive consequences of literacy. In P. Bryant & T. Nunes (Eds.), Handbook of literacy. Dordrecht: Kluwer. Olson, D. R., & Torrance, N. (2001) Conceptualizing literacy as a personal skill and as a social practice. In D. R. Olson & N. Torrance (Eds.), The making of literate societies, (pp. 3-18), Oxford: Blackwell. Ong, W. J. (1967) The presence of the word, New Haven, CT: Yale University Press. Ong, W. J. (1982) Orality and literacy: The technologizing of the word, London: Methuen. Rheinberger, H.-J. (1997) Toward a history of epistemic things: Synthesizing proteins in the test tube, Stanford, CA: Stanford University Press. Saenger, P. (1997) Space between words: The origins of silent reading, Stanford, CA: Stanford University Press. Scholes, R. J., & Willis, B. J. (1991) Linguists, literacy, and the intensionality of Marshall McLuhan's Western man. In D. R. Olson & N. Torrance (Eds.), Literacy and orality (pp. 215-235). Cambridge: Cambridge University Press. Shweder, R. A. (1991) Thinking through cultures: Expeditions in cultural psychology, Cambridge, MA: Harvard University Press. Smith, D. (1990) The conceptual practices of power, Toronto: University of Toronto Press. Taylor, C. (1985) Theories of meaning. In: Human agency and language. Philosophical papers I. Cambridge: Cambridge University Press. Taylor, C. (1989) Sources of the self: The making of the modern identity, Cambridge: Cambridge University Press. Triebel, A. (2001) Literacy and development: The roles of literacy practices in the activities and institutions of developing societies. In D. R. Olson & N. Torrance (Eds.), The making of literate societies (pp. 19-53), Oxford: Blackwell. Vygotsky, L. S. (1962) Thought and language, Cambridge, MA: MIT Press. Wagner, D. A. (1999) Rationales, debates, and new directions: An introduction. In D. A. Wagner, R. L. Venezky, & B. V. Street (Eds.), Literacy: An international handbook (pp. 1-8), Boulder, CO: Westview. Wittgenstein, L. (1958) Philosophical investigations (2nd Ed.), Oxford: Blackwell.

34

3 Literacy and the future of writing: An integrational perspective Roy Harris

According to a commonly accepted definition of history, the history of literacy is coterminous with history itself. For history begins with written records, and whatever antedates the emergence of writing is by definition pre-history. Does it follow from the same definition that when writing becomes obsolete, as it may well do at some future time, history will come to a stop? If our descendants survive the extinction of literacy, will the human race have entered the posthistorical era? We do not have to accept this narrow definition of history in order to appreciate the importance of the connection between literacy and the concept of record-keeping, which has come to playa central role in what we understand by the term civilization. This connection emerges in all kinds of ways. The expressions oral history and oral literature, although nowadays current in academic circles and no longer regarded as self-contradictory, nevertheless are linguistically marked expressions on a par with free admission and decaffeinated coffee. In other words, the form of the expressions themselves reflects the assumptions that normally one might expect admission not to be free and coffee to have caffeine. A similar expectation hovers over the expression pre-literate culture; and the study of pre-literate cultures by anthropologists and others often seems to have focused on the means by which, in the absence of writing, such cultures have nevertheless managed to maintain a sense of historical identity and continuity. Again the implicit norm seems to be a state of affairs in which those are the functions of literacy. So if we are to anticipate in the future the emergence of post-literate cultures, rather than general cultural collapse, we are inevitably led to consider the possibility that writing might be replaced by other forms of communication which take over the cultural role we have grown accustomed to associating with writing. Indeed, according to some prophets, the obsolescence of writing and that process of replacement have already begun. In this paper I wish to argue that this kind of prognostication is the product of a widely held but deeply mistaken view of communication. It is, unfortunately,

35

Literacy and the future of writing

the view that has predominated in modern linguistics for most of the twentieth century. The only serious alternative view, both of language and of communication, is one that has emerged - comparatively recently - under the label "integrational" (Davis and Taylor 1990, Harris 1998, Harris and Wolf 1998, Wolf 1999). The integrational perspective is one that has very particular implications for written communication and the way writing should be studied (Harris 1993, 1995). The essential difference between the orthodox and the integrational perspectives is that the former treats a system of communication as an object of knowledge existing independently of any context in which it might or might not be used; whereas the latter, on the contrary, denies not only the autonomy of systems of communication but their logical and ontological priority over the act of communication itself. This denial, it need hardly be added, is little short of scandalous not only in the eyes of orthodox professional linguists but to the great majority of professional philosophers as well. It also strikes many honest citizens as an affront to common sense. But this is because what passes for common sense - in this case as in so many others - is a watered-down version of what people have been taught to believe at school. Now writing constitutes an interesting test case, precisely because writing systems are traditionally regarded as secondary systems based on - but nevertheless quite separate from - the spoken languages they subserve. For many theorists, this status is established by the simple argument that writing has to be learnt separately from speech - as witnessed by the fact that, whereas every known community (other than those comprising persons physiologically deprived of speech or hearing) has a spoken language, until relatively recently very few communities were literate. This argument is circular. It is based on a concept of literacy which is not merely static and backward-looking, but already incorporates the assumption that systems of communication are autonomous. By contrast, an integrational perspective holds the key to a quite different concept of literacy and a quite different prognostication for the future of writing. Orthodox linguistics is totally incapable of dealing with questions of literacy, because it is committed to the view that writing is basically a representation of speech. This phonocentricity stands the truth on its head. The truth is that in every civilization that we know of, writing constantly recreates speech in its own image. As Olson puts it in his neat reformulation of Whorf, "we introspect our language along lines laid down by our scripts" (Olson 1994: 90). In recent years it has become increasingly common to hear two kinds of gypsies' warning on the subject of literacy. One is that in the so-called advanced industrial societies literacy is in decline, and that if care is not taken communities which now think they arc literate will lapse into illiteracy. The other is that new techniques of communication are now becoming available on a scale that will devalue literacy itself. 36

Roy Harris

These two prophecies, although perhaps superficially contradictory, are head and tail of the same coin. What is being complained about is that certain forms of technological innovation threaten to undermine the educational virtues that literacy formerly stood for. This undermining is not peripheral, but apparently threatens the expertise of professional writers themselves. Two recent protests are symptomatic. Michael Heim has objected that because of the computer, news reporters are no longer learning how to rewrite their copy; while Gore Vidal complains: "The word processor is erasing literature". The first point an integrationist would want to make here is that although in the major civilizations of both East and West literacy is traditionally associated with the practice of writing, what that involves may differ very widely from one cultural context to another. Anyone who has watched a skilled Chinese artist making a seal will probably have witnessed a process of about thirty or forty minutes duration, beginning with the blank surface of a polished piece of soapstone held in a vice, and concluding with the waxed impression on a sheet of paper. No Chinese watching such a demonstration will doubt for a moment that the artist is highly literate. It will typically involve not only drawing Chinese characters in ink on the surface of the soapstone, but being able to draw them in reverse. As the West measures literacy, few if any of its literate members would be able to do anything comparable, given the same material implements: the inks, the stone and the various cutting instruments. This is clearly not because in the West no one is capable of learning to paint in reverse with a brush, but because there are no privileged cultural practices which require this skill. In China, on the other hand, the use of seals has a long-established place in the tradition of literacy. Sometimes it is suggested that the history of literacy consists essentially in the replacement of one technology by another. On this view we might anticipate a future in which our descendants will be counted "illiterate" if they cannot operate a word processor. But what has operating a word processor to do with being able to carve a seal, or with knowing how to use a ballpoint, for that matter? If we reflect on the contrast between these operations, we are led to address the question of where exactly literacy begins and ends. It appears to require the systematic use of specialized tools, in other words to be a technology, or at least to involve mastery of a technology. But does it? In his book entitled On Literacy Robert Pattison argues eloquently against what he regards as the misleading equation of literacy with the practical skills of reading and writing, and against what he sees as a consequential series of educational fallacies based on that equation. For Pattison, it is obvious that Homer was a paragon of literacy, and yet Homer could neither read nor write. Therefore those who define literacy in terms of the ability to read and write, and thus by implication relegate Homer to the ranks of the world's illiterate masses, merely demonstrate their own cultural myopia and a complete failure to understand what it is that makes literacy worth having.

37

Literacy and the future of writing

It might perhaps be objected to Pattison that Homer would never have become a paragon of literacy had his poems not been fixed in written form by a later generation of Greeks; and consequently Pattison's prime historical example proves the very opposite of what he takes it to demonstrate. Nevertheless Pattison's argument, even if we do not agree with it, is sufficiently provocative to compel attention. In effect, what he is saying is that it makes no more sense to dismiss Homer as illiterate than to judge musicianship by whether a player can play by sight from a musical score. If that were the test, then there could be no blind musicians. The mistake in both cases is to erect what is no more than a secondary, ancillary technique into the primary criterion of a complex ability which that technique happens to serve. Just as it makes no sense to suppose that playing from a score is the only kind of musical performance, likewise it makes no sense to suppose that reading and writing are the only kinds of activity that manifest literacy. It is interesting to consider how Pattison's argument fares when applied to the specific form of literacy which is sometimes called numeracy. The modern concept of numeracy treats it as essentially involving the manipulation of quantification symbols according to principles and patterns which have to be studied and learnt. That is why when we come across apparent examples of early civilizations capable of undertaking the building of large temples and palaces and engaging in extensive enterprises of irrigation or town planning we are reluctant to accept that they could have done all this on the basis of a purely oral culture, even if they have left behind no record of their writing system. We prefer to think that they must have developed a written mathematics and a geometry now lost, because we cannot otherwise imagine how they could have managed the complex calculations apparently required by their feats of planning and construction. But let us hypothesize for a moment that somehow they did manage precisely that. Would we then dismiss them as having failed to achieve numeracy, on the same ground that some people would describe Homer as illiterate? Or would we not rather be forced to conclude that they must have been so highly numerate that they did not need to commit their calculations to papyrus or to clay tablets? In his book The Domestication of the Savage Mind, Jack Goody recalls that he found he could not count cowrie shells as quickly or as accurately as boys of the LoDagaa in Northern Ghana who had had no schooling at all. And this was because, in the manner of Western literates, he counted the shells one by one. The native boys, on the other hand, counted them according to a traditional method in successive groups of three and two, which was both faster and easier to check. Furthermore, they had different methods of counting different objects. Counting cowrie shells had its own special technique. When it came to multiplication, however, Goody found that he could easily manage calculations that were beyond the expertise of the local cowrie counters. And this superiority Goody attributes to the fact that multiplication, as distinct from addition, is essentially a literate operation. The native boys, he says, "had no

38

Roy Harris

ready-made table in their minds" which they could use for purposes of calculation. What might be questioned about Goody's explanation, however, is precisely what the connection between multiplication and literacy is. In some cultures an illiterate can use an abacus to calculate at a speed which will match any literate mathematician's pencil and paper. Perhaps Goody would reply that using an abacus involves an operational technology which is in all respects equivalent to that of manipulating figures on paper. But that simply brings us back to Pattison's argument about what literacy is. For however we may describe using an abacus, we certainly do not call it "writing". A conclusion similar to Pattison's might perhaps be reached by a different argumentative strategy. Suppose the first explorers from Earth to arrive on Mars reported the existence of a curious reversal of the familiar terrestrial relationship between speech and writing. In other words, let us suppose it was discovered that Martians communicated primarily for everyday purposes by means of making visible marks on surfaces, and were biologically equipped to do this because their fingers constantly exuded a coloured liquid which they used in much the same manner as we use ink. With this coloured liquid Martians from their earliest years were used to tracing graphic symbols on any convenient surface that came to hand. But only few Martians ever learnt to make sounds corresponding to these graphic symbols, because this involved learning to use a special piece of equipment invented for the purpose, which looked like a small box and was worn strapped round the throat. This box was known in Martian as the "vocal apparatus," but very few Martians could afford to buy one, and in any case this apparatus could be used effectively only after years of special training in the correspondences between Martian graphic symbols and the sounds that the box could produce. Perhaps the first question that might occur to an anthropologist on the mission from Earth is: "Why do these Martians bother with this clumsy vocal apparatus at all?'' Taking this science-fiction story one stage further, let us suppose that the Martian answer to this question turned out to be that whereas any fool on Mars could write, using the vocal box required a special form of intelligence, and furthermore conferred certain advantages on speakers over writers. For instance, speakers could communicate to one another by means of sound even in the absence of a writing surface. Furthermore, speakers could communicate even when doing something else with their hands and eyes, whereas writers could not. Third, speakers could communicate with one another audibly in ways that those who could only write were quite unable to understand. Fourth, vocalization had the inestimable practicality of leaving no trace, so that it was impossible for anyone subsequently to prove what had been said. It thus required mental alertness to engage in vocal communication, and quick reactions of an order quite beyond the average slow-witted writer. In short, speaking was a privileged form of communication shared by an elite, but beyond the grasp of the masses. Any sensible Martian, therefore, could see that it was well worth buying a vocal apparatus and learning to use it, because being

39

Literacy and the future of writing

able to use a vocal apparatus brought with it all kinds of communicational superiority. I have, of course, constructed this allegory so as to reverse, on every crucial point, the usual terrestrial assumptions about the relationship between speech and writing. The lessons to be drawn from it do not need to be spelled out in laborious detail. Whether writing is judged to be a better form of communication than speech depends on one's point of view, and that point of view will be shaped by certain biological and cultural presuppositions. But in one sense it does not matter at what point and in what precise form technology enters the picture. As the allegory illustrates, it is possible to imagine a culture in which speech depends on the availability of certain tools, just as in our more familiar case writing depends on the availability of certain tools. What matters in both cases is the use made of those tools. And the utility of the tools is always measured against what could be done without them. This, at least as one reader understands it, is precisely Pattison's point about Homer. In other words, the identification of literacy with mastery of the technology of writing commits a double error. The first mistake is to have confused a merely contingent use of artifacts with the cognitive consequences of their use. The second mistake is a failure to see that the technology would be pointless unless subserving pre-existing goals, which are not set by the technology itself. Both of these are essential points to grasp if we hope to discuss the future of human communication without falling into the trap of nonsense. Unfortunately, this is an area of speculation where prestigious nonsense abounds. Much of it derives from two sources. One of these sources is what has been called scriptism: that is to say, the tendency prominent throughout the Western tradition to analyze spoken language as if it were written language avant fa lettre (Harris 1980: 6ff.). How deeply scriptism is entrenched in Western thinking about language emerges in various paradoxes. Perhaps the most striking is that linguistic theorists who subscribe to the doctrine of the primacy of speech as the characteristic form of human communication (Saussure, for example) nevertheless feel constrained to analyze speech in such a way that its units correspond in a quasi-miraculous fashion to the units of writing, even while proclaiming that writing practices are quite extrinsic to language. This may even carryover to metalinguistic terminology. A quite remarkable case is the paper published by Bloomfield in 1927, in which he applied the distinction between "literate" and "illiterate" speech to a pre-literate speech community, the Menomini of Wisconsin. The so-called "International Phonetic Alphabet" is another classic example of scriptism disguised as linguistic analysis, based as it is on one culturally localized system of writing. Scriptism in all its forms encourages the view that literacy involves some kind of transference of linguistic skills from a natural medium (namely, that of speech) into an artificial medium (namely, writing). The other principal source of nonsense about literacy can be traced to a body of pronouncements from what may be termed the anti-scriptist school. The

40

Roy Harris members of this school include Ong, Havelock and McLuhan. Their major thesis is that writing, far from being merely speech made visible in the guise of inscriptions, constitutes a radically different cognitive enterprise. In their account, writing is not just a convenient way of recording speech, but involves a restructuring of thought. Anti-scriptism in its various forms encourages the view that literacy is a profoundly different mental condition from that of pre-literate humanity. Adherents to the anti-scriptist school often emphasize this radical change by speaking of the "literacy revolution". As the terms scriptist and anti-scriptist suggest, there is certainly a deep divergence of views here. While accepting this, one may nevertheless wish to argue that the conflict between these views is often presented as a conflict over the wrong issue. The important difference between literacy and pre-literacy is not essentially a difference between typical ways of thinking about the world, of classifying and ordering, of overcoming memory limitations, or of strategies for acquiring knowledge, although all these differences doubtless correlate with the advent of writing. But they are all manifestations of something more fundamental; and this something more fundamental is a shift in conceptions of language itself. The change comes about for two reasons. One is that the advent of writing destroys, once and for all, the former equation between language and speech. In a preliterate culture the world of language is the world of sound. Writing changes all that. With writing, language invades the world of visual communication. It enters into competition - and partnership - with pictorial images of all kinds. Thus writing ushers in the concept of language as something that is mediumtransferable: words can be spoken, transferred into a biomechanically different form where they are visible but no longer audible, and then transferred back again into speech. This process sometimes appears to preliterate communities, on their initial acquaintance with writing, as a form of "magic"; but it does so only because their preJiterate conception of language cannot immediately cope with the notion of medium-transferability. The second reason is no less important. By making it possible to divorce the message both from its sender and from the original circumstances of its formulation, writing cognitively relocates language in what has been called an "autoglottic" space (Harris 1989). That is to say, the text takes on a life of its own, which is ultimately independent of the life - or intentions - of its author. It becomes an "unsponsored" linguistic object, to which there is no parallel in a preliterate culture. And with this "de sponsoring" comes a fundamental change in the notion of meaning. Instead of tracing back meaning to the speaker or writer, as the authenticating source of the message, people come to regard meaning as residing in the words themselves. This prises open a conceptual gap between the sentence, on the one hand, and its utterance or inscription on the other. The sentence, being what lies behind and "guarantees" both utterance and inscription, is itself neither. And this requires a conception of language which is necessarily more abstract than any that is required in a preliterate culture.

41

Literacy and the future of writing

Furthermore, the availability of writing opens up new possibilities of metalinguistic discourse, which in tum have a powerful influence on how speech itself is conceptualized. As Nigel Love puts it: Developing a written counterpart to spoken language removes the difficulties attaching to a purely oral practice of metalinguistic discourse. For although type-token ambiguities may arise for writing as for speech, writing provides a firm anchorage for at least one dimension of type-token distinctions which is different from the medium in which the corresponding tokens are produced. It introduces a new level of clarity into attempts to show what the abstractions are, by providing a system of types in terms of which, in literate societies, utterances will henceforth be interpreted. (Love 1990: 110) All this amounts to a complex and far-reaching revolution, but one in respect of which it would be idle to start arguing, as some theorists are inclined to do, about what should be regarded as cause and what as effect. What can be said, however, is that the pivot of this revolution is a new view of language and of its potential manifestations in human affairs. The currently popular scenario for the future of writing gives pride of place to the computer and the development of techniques of word processing. It has already been claimed that our traditional concept of reading has been outdated by the advent of "dynamic text" and "hypertext". The basis of this claim is a transformation in the role of the reader from passive recipient to active participant in the process of information transmission. Instead of merely receiving a message pre-determined by the sender in respect of both form and content, the reader is now able to control and access whatever information and information-sequences are deemed relevant for particular communicational purposes. Text is presented not in a traditional mono linear format, but as a simultaneous configuration of choices, from which the reader must make a selection. Depending on that selection, further selections become available, as the reader explores possible ways through the maze of information available. In this process of exploration, individual readers construct their own text instead of accepting a text dictated to them. But this is not all. If it were, one might object that what it amounts to is simply a formalization and mechanization of reading strategies that have been available to the traditional reader for centuries. Indeed, Western culture developed a special form of book and a special form of verbal deixis based precisely on such reading strategies. The book is known as an encyclopedia, and the deictic device is known as cross-referencing. So a stronger claim must be made if we are to be convinced that there is anything new here other than the technological format. This stronger claim is based on the fact that programs may be set up in such a way that it may not be possible for the reader to retrieve a previously selected portion of text in exactly the same form as the first time: in other words, it may 42

Roy Harris

not be possible for the reader to turn back the page and find everything that was written on it before. Likewise, it may be impossible for the writer to foresee exactly the sequences and contextualizations that may arise as the result of selections made by different readers. As Jay David Bolter puts it: Unlike printing, which lends fixity and monumentality to the text, electronic writing is a radically unstable and impermanent form, in which the text exists only from moment to moment ... (Bolter 1989: 129) Under these conditions, we may ask, what exactly is the status of the text? Is it any longer a text at all? Whatever it may be, it is clearly no longer a static object; and in that sense it may be claimed that we are dealing with a genuinely new form of writing. For the primary characteristic of the writing process, as traditionally understood, was precisely that it produced a fixed form of words, available for inspection, re-inspection, interpretation and discussion as required. It was this fixity and quasi-permanence which was seen as contrasting with the ephemerality of speech. These non-ephemeral qualities were what made writing suitable for the recording of information and its transmission over space and time. It removed verbal communication from intrinsic dependence on the particular circumstances ofa face-to-face situation or the vagaries of memory and uncheckable repetition. Hence, from Biblical times onwards, not only the Ten Commandments but edicts and laws of all kinds were set down in writing. Writing became the guarantee of authenticity because and insofar as it guaranteed in turn the invariance of the text. While it was true that a written text was only as durable as the material surface on which it had been inscribed, and on some surfaces writing is easily erased or altered, nevertheless it was also true that until altered by material decay or human interference the text remained static. Again, however, if this were all it might appear that what electronic writing has done is simply reintroduce via technology a rapprochement between writing and speech; or rather has endowed writing with the ephemerality which was formerly treated as characteristic of speech. And a sceptic might well ask what the point is of employing the latest technological innovations in order to revert to a more primitive type of communication. Heim has argued that the psychological revolution brought about by electronic writing is to give priority to information over significance. This is a mistake. Or rather, if it is true, then the future of literacy must be reckoned a bleak prospect. But in order to make his case, Heim has to adopt a very restrictive definition of "information" and a very broad definition of "significance". Information, he says, is "a unit of knowledge which by itself has only a trace of significance" and "infomania," as he calls it, "erodes our capacity for significance". There is, he claims, a law of diminishing returns: the more injormation accessed, the less significance is possible (Heim 1989: 38) What exactly these sybilline pronouncements mean is far from clear, except that they are evidently meant to be full of foreboding. 43

Literacy and the future of writing

Heim seems to think that information is what electronic writing can handle, whereas significance is something only human experience can supply. This dichotomy collapses as soon as we ask how information can possibly inform in the absence of significance, or how significance could signify if it were devoid of information. And with that collapse the way is open for prognostications of a less gloomy nature. A text which is not a static object not only defeats the storage function generally regarded as one of the primary purposes of writing, but introduces something radically novel into our whole model of verbal communication, whether spoken or written; namely a discontinuity between the initial act of verbalization and its end product. And this discontinuity is of a different order from the material transformations which a message may undergo in its journey from, say, oral dictation to printed page. It may be helpful at this stage in the argument to introduce a distinction which will clarify and sharpen the point at issue. Let us adopt the term revision to cover all those transformations which a message may undergo via the traditional processes which intervene between initial verbalization and final publication. Let us at the same time adopt the term projection to cover the more radical possibilities which electronic writing renders potentially available. (The word potentially here introduces an important caveat, since at the moment there may be no computer programs to hand which actually implement all the possibilities envisaged. It is in that sense that we are discussing "the future of writing". Nor, one might add, is there any guarantee that they will ever become available, since it might not be profitable for technologists to develop them.) The essential difference between revision and projection is this. Revisional processes are refinements: the changes they introduce involve narrowing down the message. Projectional processes are speculative: the changes suggested are expansions. They involve opening up the message; or, more precisely, generating potential new messages from the original germ of an idea. Not that these two objectives are mutually exclusive. On the contrary, an optimal communication strategy might embody both. The point to be stressed here, however, is that hitherto literacy has been viewed as a skill or complex of skills exercised in the service of the expression of a given message within a given channel of communication. The psychology of literacy has been a revisionist psychology. And that is understandable - indeed inevitable - given that "writing" stood opposed merely to "speaking" and that the former had traditionally been treated as some kind of substitute for the latter. This revisionist psychology is still reflected in some of the word-processing packages that are currently marketed. A good example is the program that checks the writer's spelling. Programs of this kind run up against the difficulty that many writers are interested in writing about topics which the program has never heard of, and writing about them in a technical terminology which finds no place in current dictionaries. Implicitly, such programs define literacy as a negative accomplishment, having as its main objective to avoid mistakes. But let

44

Roy Harris

us for a moment stand this objective on its head and suppose there were a demand for programs which suggested unorthodox forms of spelling. (Why there should be such a demand is not what is at issue; nor whether English words could be spelled "better" than dictionaries currently advocate.) The point is that a projectionist program for orthography reverses the assumptions that are built into currently designed packages; and that the reversal is technologically just as easy. Technology as such is quite independent of semantic or pragmatic values. A machine can just as well provide us with unorthodox substitutes for orthodox items as with the reverse. Where the machine advances beyond human perversity is that it offers possibilities of projection, as opposed to revision, which would simply never occur to the human manipulator of text. And what goes for spelling goes for every other aspect of language which can be codified in terms of combinatorial possibilities. In his pessimistic novel Nineteen Eighty-Four Orwell anticipated machines programmed to write the lyrics of trashy pop songs. But we can just as easily reverse that anticipation and imagine word-machines designed to turn the tritest verse into a constellation of unimagined verbal pyrotechnics. To be sure, that is not what is traditionally subsumed under the rubric of "writing". But only, the integrationist will say, because our traditional concept of creative verbal processes is hamstrung by the doctrine of the primacy of speech. Speaking is only one of the ways in which we can do things with words. But to realize that limitation of speech is precisely the hallmark of literacy. Word processors, as everyone agrees, and as the term itself indicates, are essentially machines for manipulating verbal signs. Yet most of our current theories of verbal signs were formulated long before the advent of the computer. A serious question to be addressed is: are those theories now out of date? Has the semiology of the pre-computer era now been superseded? A preview of how writing by machine may alter our concept of literacy is already offered by the pocket calculator. What is expected of students in a threehour mathematics examination has changed dramatically since the days when all calculations had to be done by human brainpower. One can foresee analogous changes in the assessment of language skills when word processors automatically correct errors of spelling, grammar and punctuation. What is interesting about the word processor is that it provides a machine which enables the user to exploit systematically the potentialities afforded by the indeterminacy of the linguistic sign and the open-endedness of sign systems. Traditional writing does not do this because traditional writing is constrained by the writer's personal network of word associations, plus the orthological legacy of generations of educational convention. The word processor, on the other hand, is a piece of equipment for linguistic engineering. It can systematically invent new words, new paradigms, new constructions, new meanings and new languages if we wish. What has gone unnoticed is that a tool with the power and ubiquity of the computer has the potential of reversing the twentieth century's received wisdom

45

Literacy and the future of writing

on the basic relationship between language, speech and writing. There are various reasons for this. One is quite simply the sheer increase in the amount of written material generated. It is now confidently predicted that, with the internet explosion, written communication will quantitatively outstrip oral communication in the foreseeable future. If it does, that will certainly be a landmark in human history: speech will for the first time be the "minor" form of communication. But there are more important reasons which have to do with our grasp of the basic processes of verbal communication. In an oral or predominantly oral culture, no one intuitively thinks of saying something as a process of stringing together individual speech sounds in a certain order. Even in a highly literate culture, while I may appreciate intellectually that in order to say "Toronto" I have to select certain vowels and consonants and get them articulated in the right sequence, and while I am perfectly well aware that I may make a "slip of the tongue" and get it wrong, nevertheless I can usually count on doing it so effortlessly that this is not for me the "psychologically real" level of my speech performance. As a speaker, what I focus on is not the vowels and consonants but whether I want to say "Toronto" or not. I consciously choose saying "Toronto" in preference to saying, for instance, "London," "Oxford" or "Timbuctoo," because that is what I want to sayan that occasion. In short, my basic speech unit is "the word". Any pronouncement of this kind is quite likely to provoke howls of protest from linguists, psycholinguists and others. Controversies over what "a word" is have filled countless pages in the annals of linguistics and no definitive account of the concept has ever been given. But, as a fluent speaker of English, I do not need to wait for the results of academic research in order to understand what I am doing when I speak. I already know that my basic units are items like Toronto. And if any expert in neurolinguistics or experimental psychology wants to assure me otherwise, then I am afraid I have to say that the expert has got it wrong. For what we are talking about here is not a conclusion that has to be established in anyone's laboratory or as the result of cleverly devised tests. What we are talking about is a fact of my first-order linguistic experience. I do not need any kind of expert to tell me about that kind of fact, any more than I need a doctor to tell me when I am in pain or a dietician to tell me when I am hungry. This is not to say, however, that underlying my first-order linguistic experience there lurks a "theory of the word," much less that I must have "internalized" such a theory in order to be able to recognize and produce units like "Toronto". Even less is it to assume that I necessarily have an adequate metalinguistic vocabulary in which to discuss my first-order linguistic experience and its unit-recognition aspects. These are all matters of dispute and to go into them here would be a diversion. The point this mention of speech units is leading up to is somewhat different. It is, however, fundamental: my claim is that the computer imposes on us a different recognition of linguistic units than the everyday processes of speech.

46

Roy Harris

To type "Toronto" on the keyboard of my current word processor, I have no option but to decompose it into seven separate units and eight separate operations. Again, I need no expert to tell me this: I already know it from experience. The reason why the keyboard imposes this task on me is also no mystery. It operates with one particular type of writing system, namely alphabetic writing. In theory, it could operate with a number of other kinds of writing system, but for political and economic reasons it now seems likely that alphabetic writing and even one subvariety of alphabetic writing - will be imposed, via the computer, on all the peoples of the world. As in the past, the power to control writing will mean the power to control society. As a writing machine, then, the word processor is already redefining our concept of what a "word" is. The word is no longer a static lexical unit belonging to an inventory pre-registered in a dictionary. Implicitly, for a word processor, the word can be any symbol or symbolic unit which plays a role in the processing and can be controlled by well-defined keyboard operations. It is important to note that that role is not confined by the syntagmatic and paradigmatic relations which govern what we now recognize as the conventional words of ordinary language. Nor is it restricted by the conventional boundary lines which treat iconic symbols as non-words. Furthermore, such units and combinations can be invented by the writer as needed. The constraints on their invention are not conventions in the outside world, but constraints internal to the machine. Today, in short, technology puts us in the position of projecting "writing" as something altogether different from the ancillary system for recording, which was its traditional basic role. It opens up the possibility of treating writing as the essential creative process and speech as a marginal commentary on what has been written. That radical reversal of roles, we may reasonably speculate, will hold the key to the psychology of education in the next century. It is within this perspective that it becomes relevant to return to Pattison's argument about Homer. Pattison may be right in insisting that traditional literacy had to be defined in terms of pre-existing goals or models, which were not set by the technology of writing itself but were already in place before the technology became available. Even if this was true in the past, however, will it be true in the future? Arguably not: and this is where the word processor not only makes a crucial difference, but turns Pattison's proposition upside down. The limits of literacy will be set by the technology, not in the sense of restricting the title "literate" to those who read and write electronically, but by the exploitation of new linguistic possibilities which would not have been available without the technology to hand. And there is no way those limits can be set in advance; for we have as yet only an inkling of what computers of the future may make possible or even render commonplace. What effects these developments may have on human cognition is again a matter for speculation, but it seems possible to predict that the new literacy will almost certainly bring with it a new concept of knowledge. In an influential 47

Literacy and the future of writing

paper "On cognitive growth" published in 1966, Jerome Bruner distinguished three ways of knowing available to human beings. These three ways correspond to three distinct forms of representation, which Bruner called "enactive," "ikonic" and "symbolic". The enactive mode involves motor skills: knowing how to tie a knot or use a tool. The ikonic mode involves picturing: knowing what something looks like and the ability to recognize it from its appearance. The symbolic mode involves language: knowing what something is called, how to describe it or identify it from its description. In terms of this tripartite division, literacy of the traditional kind would be subsumed as a subdivision of the symbolic mode. But we may well hazard a guess that this location will no longer be appropriate for the new literacy of the future. If the computer makes available projectional word-processing on an unprecedented scale, that may require a concept of knowledge which cannot be pigeonholed in any of these three categories. For in order to know something in this new mode it would be necessary to situate that item in a network of possible representations, analogues, expansions, modifications, derivations, alternatives and opposites which only a computer could systematize automatically and make available for use. Indeed, the situation of the item in such a network would constitute knowing it in the fullest possible sense. This fourth mode of knowledge is intrinsically related to the ongoing redefinition of the word mentioned above. For the machine not only makes available a "word" that can function indifferently as definition, definiens, topic, subheading, summary, crossreference, and in a variety of other functions, but can systematically interrelate these multiple functions, making that interrelationship explicit and immediately manipulable in a way that no single book or set of books from the library ever could. The old literacy of the library era was in many ways valued precisely because it was a way of overcoming by sheer human effort the limitations of the individual book as a repository of knowledge. But it was a labour-intensive way and a somewhat haphazard way, because it depended on the training of the literate individual over a long period of time - indeed over a whole lifetime. In the history of human literacy, many books have taken a lifetime to write, and a few have occupied many readers' lifetimes in understanding, in study and research. Part of the problem with the traditional concept of literacy is that, because of the manner in which that kind of literacy has to be acquired, it tends to be psychologically backward-looking. Those civilizations which prized literacy most highly have always tended to be the most conservative as regards the forms and functions associated with literacy. Similarly, those who made the effort or were given the opportunity to acquire literacy in such societies often tended to regard it as a privilege to be jealously guarded rather than a benefit to be spread as widely as possible. These are the social macroconsequences of literacy based on the tablet, the scroll and the book. These traditional techniques of information storage answer to a concept of knowledge where what is known has to be recordable in conformity with some

48

Roy Harris

fixed set of communicational conventions, and set down in a static, quasipermanent form. Under this system, those who do not have access to the tablets, the scrolls or the books have no access to that kind of knowledge, and become by definition ignorant in a society where knowledge of that kind takes priority over other kinds. Thus it is that "illiteracy" and "ignorance" tend to become synonymous. For an integrationist, the developments sketched above are all predictable because they follow from recognizing that communication - and our concepts of communication - depend, as they always have done, on our integrational proficiency. In order to communicate effectively, according to integrational theory, we have to integrate our activities as individuals within parameters of just three kinds: biomechanical, macrosocial and circumstantial (Harris 1996). The advent of a powerful new writing-and-reading machine means an alteration of all three parameters. But this is not all. Traditional analyses of writing are all codebased. Integrationism reverses the priority of code over context. To come to terms with the literacy of the future will require a theoretical perspective which reverses the priority of code over context, because in the computer what we have available is the most powerful contextualization device ever known. Its capacity for creating and developing new contexts, visual and verbal, far outstrips that of the human mind. That is a far more important fact about the computer than its superhuman capacity for information storage. We are dealing with a machine which offers not only the possibility of integrating a simultaneous presentation of written, auditory and pictorial information, but of linking that information across languages and cultures as well. When we are accustomed to sitting at a keyboard and "typing" an audio-visual product that incorporates sounds, letterforms and pictures systematically interrelated, we shall have acquired a new concept of writing and a new concept of text. The point about the new literacy of the computer age is not that technology now makes possible a quantum leap in the sheer accessibility of the contents of libraries, important though that is in its own right. To speak of a fourth mode of knowledge is not to say that there is much more information available and to many more seekers. Nor is it to say merely that we are dealing with a new organization of what information is already codified. It is to say that knowing itself is construed as a matter of understanding a continuous development, an integrational process involving not only past but potential future contexts. Projectional knowledge, if that is what we wish to call this fourth mode, is knowledge of something not just as it is or was but as it has the potential to be. It is a dynamic, not a static concept of knowledge that the new literacy ushers in. People will doubtless say: that used to be called "imagination". But there is a significant difference between imagination and projectional knowledge. Imagination is hit or miss. It can project, but it cannot explore its projections systematically. The difference is apparent if we compare what the medieval architect could do with what the modern architect can do with assistance from the computer. The modern architect can actually see - on the computer screen -

49

Literacy and the future of writing

not only what it will look like to be inside the projected building, but what effect possible alterations in certain features will have. Alternative designs can be generated, calculated and examined in ways never before possible. We may still prefer medieval architecture to modern: that is our privilege as judges. But it does not alter the fact that modern architects now "know what they are doing" in a way that medieval architects never did. And through this knowledge they can do things that would never have been imagined or risked before. People who prefer to think still of computers as "mere" technology, practical tools for facilitating certain tasks but incapable of becoming determinants of the way we think about communication or any other human activities, are rather like those in the late middle ages who thought that the advent of explosives might perhaps make a difference to the winning or losing of a few battles, but could make no difference to anyone's thinking about warfare. The use of explosives eventually altered not only people's concepts of warfare, but their concepts of cities, politics and the state. Once computer-assisted communication has become the norm in schools and universities, then just as surely that will change society's notion of what education is and of what it is to be an educated human being. Acknowledgment

I am grateful to Richard Fargher for stimulating some of these thoughts about literacy. He will not be surprised to find how radically our views converge and diverge. References

Bloomfield, L. (1927) Literate and illiterate speech. American Speech, 2 (10). Bolter, J. D. (1989) Beyond word processing: the computer as a new writing space. Language & Communication, 9 (2/3), 129-142.

Bruner, J. S. (1966) On cognitive growth. In J. S. Bruner (Ed.), Studies in cognitive growth, New York: Wiley. Davis H. G., & Taylor, T. J. (Eds.) (1990) Redejining linguistics, London: Routledge. Goody, J. (1977) The domestication 0/ the savage mind, Cambridge: Cambridge University Press. Harris, R. (1980) The language-makers, London: Duckworth. - - (1989) How does writing restructure thought? Language & Communication, 9(2/3), 99-106. Reprinted in R. Harris & G. Wolf (Eds.) (1998), Integrational linguistics: A jirst reader, Oxford: Pergamon. - - (1993) La Semiologie de l'ecriture, Paris: CNRS. - - (1995) Signs o/writing, London: Routledge. - - (1996) Signs, language and communication, London: Routledge. - - (1998) Introduction to integrationallinguistics, Oxford: Pergamon. Harris, R., & Wolf, G. (Eds.) (1998) Integrationallinguistics: Ajirst reader, Oxford: Pergamon. Heim, M. (1987) Electric language, New Haven, CT: Yale University Press. - - (1989) The dark language of infomania. The Independent, 30.12.89.

50

Roy Harris

Love, N. (1990) The locus of languages in a redefined linguistics. In H. G. Davis & T. J. Taylor (Eds.), Redefining linguistics, London: Routledge. Olson, D. R. (1994) The world on paper, Cambridge: Cambridge University Press. Pattison, R. (1982) On literacy, Oxford: Oxford University Press. Wolf, G. (Ed.) (1999) Integrational linguistics in the context of 20th-century theories of language, Special Issue of Language & Communication, 19 (1).

51

4 The construction of mind and self in an interpretive communityl.2 Carol Fleisher Feldman

The present paper takes the notion of the interpretive community as its theme, and specifically the procedures through which such communities come into being and are reconstituted at times of change. I will approach this problem by considering three cases, but first I will begin with some background that may help to put the notion of the interpretive community into its scholarly context, and give a general account of it. Interpretive communities

Since the time in the 1970's when we thought of the human as a logician, many scholars have moved beyond this picture, as this conference itself attests, to a picture of persons situated in a human and artistic culture that informs many other ways of thinking, and is a target of them, too. My particular approach to understanding this suddenly much more complex human, has been to puzzle over forms of thinking that are not logically but narratively patterned, and that are not predictive explanations, but rather retrospective interpretations. I have been calling these forms of thought "interpretive" cognition. This kind of knowledge is shared among members of cultural group who know the same stories in the same ways, and hope to tell, or more precisely, to indicate to each other what things mean to them. Such a group according to Durkheim (1938), Saussure (1983) and Olson (1994), among others, is an interpretive community. According to Olson, the knowledge that members of an interpretive community share serves as something like a Peircean interpretant, mediating between symbol and referent, but for much larger congeries of symbols than Peirce (1952) had in mind, such as whole book-length texts. Thus, it is highly patterned by a culture's canonical literary forms - its genres, plots, myths, human archetypes, and so on, along with the rules for their construction. This interpretant-writ-Iarge is most obviously needed to unpack meaning-at-adistance, say a writer's intentions, rather than a speaker's. If they share an

52

Carol Fleisher Feldman

interpretive community, a reader can discover meaning not just from utterance and writer, who is often a shadowy figure, but rather from utterance, writer, and, the interpretive system shared by the two of them. Similarly, writers avoid the impossible requirement of having to orient to the particular knowledge of each reader, by orienting instead to the interpretive community of which the reader is a presumed member. Their interpretive community joins reader and writer through a Peircean tertium that the two have in common, composed of the canon of their common culture. Although the need for a common interpretant may be more obvious for writing, something like this process takes place in many oral communication situations as well, the more so the more formal they are. Thus, in general, in all kinds of communication situations, the interpretive community is the means by which understanding is possible. At a more theoretical level, the interpretive community presents solutions to the two otherwise baffling problems that are at the heart of the theory of meaning. I won't discuss them here, but I do want to mention them. First, it points to a solution to the question of how writing could ever bear pragmatist, use-based meaning. Second, and rather conversely, it points to how use-based meaning manages to rise above local idiosyncrasy to take on a general form. I've been talking as through interpretive communities just are, reifying them, but that's not the case at all. Rather, they are constantly evolving, and even appearing and disappearing. Roughly speaking, this is because interpretation yields results that are added to the old canon, a matter I will discuss much more below. The topic I will focus on in the rest of this paper is the ontogenesis of the interpretive community. I'll approach it by working over several examples of interpretive communities in evident flux. I'll tum first to the earliest lay readers, who were a brand new interpretive community on the face of the earth in modem times. Second, I'll report on an experiment to set up a new interpretive community in the laboratory, arid what we found. Third, I'll discuss the American national interpretive community at a moment of rapid change: before and after President Clinton's impeachment trial. Mechanisms of change

The possibility of transformation in interpretive communities depends on the fact that in the relation of any member to the discourse of such a community there is always a two-step process, the second step of which is transformational. The two steps are: first, the creation of new meaning, and second, interpretation. This process seems to me the normative pattern in which members interact within an interpretive community. The first step, meaning creation, begins with a kind of steady state - we could call it a stipulative steady state. At any specific moment when interaction between reader and writer begins, the interpretive community has an existing canon. For the reader of novels, for example, the literary canon of that moment

53

The construction of mind and self in an interpretive community

can be taken as given. The interpretive community relies on such a canon for communication and intersubjectivity - speakers refer to it in their indications to their listeners, and writers to their readers. In giving these indications writers and speakers anticipate a listener or reader's need for guidance about how to unpack meaning. The speaker or writer's allusions to the canon tell the listener or reader, where to look in the canon they all share for the relevant patterns; that is, the patterns that will help unpack this particular new meaning as the writer intended it. In this way the interpretive community makes use of prior canon, or convention, to support writers' inventions. Thus, at each historical moment in the life of an interpretive community, there is both an old canon, and the possibility of new meanings that it invites, new meanings that can be made communicable by their connection to the shared canon. According to Culler (1976), Saussure was particularly interested in this invitation to new meanings that lay within a shared canon. Unlike many other observers who have suspected that such a canon could obscure individuality, Saussure saw the canon as a liberator of creativity. He pointed to two main processes at work, after first noting the essential ambiguity of meaning within our system as a kind of precondition for engagement in the interpretive processes. The two processes are, roughly, struggle against constraints, and exteriorization. First, constraints. Culler says that for Saussure semiological systems constrain meaning, giving something to be resisted. This struggle imposes a high standard of articulation in the new, perhaps, that prevents new meanings from being facile and uninteresting, and in that sense, supports creativity. Second, exteriorization. The conventions of the canon allow us to get beyond ourselves, to bootstrap ourselves into new meanings beyond ourselves. In this way, recruiting the canon recruits cultural resources outside the self that support creativity. But after any new meaning is produced there is a second step - interpretation. For, once a new pattern is introduced, it provokes discussion among the readers, and a second activity begins; namely, the sharing of interpretations amongst them. There are some especially vivid examples of this two-step process found in the special genres of high art that are sometimes appear in strictly oral cultures. For example, Atkinson (1985) describes a poetic genre used by the Wana for making new political meanings. After the poet has spoken, the audience discusses what the words meant, a second step that is as much a part of these ritual events as the poetry reading itself. This collaboration for interpreting new meanings among what might be called the "receiving" members of an interpretive community has the effect of making public and social the shared interpretation arrived at by the community. It is easy to see that once there is a collective and public interpretation, this interpretation becomes, ipso facto, itself a part of the common knowledge of the interpretive community, and it feeds back into the canon, which is thereby updated. This feedback loop is the part of the process that gives the canon its dynamism. Thus, the dynamism obviously depends on the effort of the

54

Carol Fleisher Feldman

individual members to engage with new meanings and with the evolving interpretive patterns of their community, and perhaps a more generic desire for meaningful social connection in a community. It also depends on an educational system that gives the young an entry into culture, but that is another matter. It might be helpful consider the way this two-step process works in the very small, namely in a two-person conversation. Such an analysis might help us to get a handle on more diffuse groups where the processes are harder at first to see. At the outer reach, an analysis of conversation might eventually help us to understand better the glacial and global changes that we know occur in the interpretive system for important and enduring learned documents like the Bible, or the Magna Carta in British Common Law. In ordinary conversation, speakers first agree on a topic and some comments that indicate how this topic is to be approached in this conversation - whether it will be about gossip, personal feelings, useful information, and so on. We can think of this sharing of topic and comments at the start of a conversation as something like the canon for this tiny and fleeting interpretive community. At that point, the whole topic/comment compound they have agreed on becomes something like the given, and also boring. If the conversation is to continue it is time to say something new, but this has to be done in a way that presupposes the prior discourse. At this point, the old topic/comment compound is taken as given, or pushed down as a new topic, which, in tum, invites new comments on it. This is the moment when new meaning appears. I failed to see that after the new meaning appears there is an important second step - interpretation - when I first wrote about this process (Feldman, 1987). In the second step, the new comment is worked over by the two interlocutors and interpreted before the next round when it, in tum, is pushed down into the given. But it becomes obvious that this second step of agreeing on a socially shared interpretation, or interpretations, must be there when we think of the two discourse partners as a small and fleeting community. The two-step process - new meaning-making followed by interpretation might be fully general for communities of all sizes and durations. For example, small work-groups may recreate their canon - something like the story of this group, or what it's like to be a member of this group - over period of time, through a similar process. Bruner and Feldman (1996) reported on the differing canonical narratives of group identity told by members of three young New York theater companies. Obviously, what we caught in their differing genres was their story at the moment of our interviews. In light of the processes described above, it now seems a reasonable inference that those stories had evolved from earlier versions and were also bound to change. The stories they told us plainly had the flavor of a canon in each community. But, surely people were making new observations about it that were under discussion - new interpretations. And if so, the canonical story was bound to change. A few years ago I was surprised to hear how much one of the groups had changed in the year or two after we interviewed them. I had imagined them forever frozen in their unhappy narrative. But, their

55

The construction of mind and self in an interpretive community

restricting story apparently had the seeds of change built right into it. If the twostep analysis of the meaning-making process just described is right, it was inevitable that the canonical story should have led the group's members to new meanings, and shared interpretations of them, through which they worked out a new narrative of group identity. The realization of creative meanings in a socially communicable form is mediated, then, by the shared interpretations of the community at a moment, but also takes place in a context that is the history of canons in that community's history. And, correspondingly, any group identity built around a shared canon similarly takes place in a context of historical derivation. Thus, since many an interpretive community is in a constant state of evolution with time, the shared knowledge of these historical changes in its canon, and trajectories of change are, perhaps, also an essential part of the knowledge of the members of an interpretive community. Among the most dramatic examples of change are those that occur when a brand new interpretive community is born. In the next section of this paper, I will describe a case of this kind. It is the story of Carlo Ginzburg's Menocchio, one of the first well-documented members of the community of lay readers of exegetical religious texts, a community that came into existence along with the birth of the printing press. I think it can be treated a natural experiment, and it is in that spirit that I present it here. The birth of a new interpretive community: The 16th Century In general, the background condition of an interpretive community in the reading process is so pervasive that it can be hard even to see even that it is there. But, in the 16th Century, there was a disjunction: there was a new group of lay readers, more or less self-taught, who did not know the literary canon, or the procedures of interpretation. The conditions at this moment were a disjunction between merely technical reading skill and cultural knowledge that allows us to separate the interpretive system from the reading process with which it is normally intertwined. Menocchio, Carlo Ginzburg's (1982) protagonist in The Cheese and the Worms, was a self-taught reader who was eager to talk about the ideas he got from reading, and was eventually executed, with regret, by an Italian Inquisitional court for heresy, for they knew he knew not what he said, and yet, from the Church's point of view, he was dangerous all the same. Menocchio was a 16th Century miller from the Friuli who believed, inter alia, that the origins of mankind lay in spontaneous generation of the kind he thought he had observed when worms come from cheese. Hence the book's title. The book that Menocchio read, Fioretto della Bibbia, apparently had a fairly wide popular circulation, for we hear of it in other contexts. It seems to have been among the first group of books that ordinary people either bought, or borrowed, but anyway had in their hands for some while to read at home. It was a

56

Carol Fleisher Feldman

compilation of theological readings written in a style that was highly interpretive. Ginzburg says, "With an expository method based in the manner of the scholastics on the enunciation and successive refutation of erroneous opinions, it certainly contributed to the unleashing of his voracious intellectual curiosity. The doctrinal inheritance that the priest of Montereale presented as a compact and unassailable structure showed itself to be subject to various interpretations in the Fioretto" (p. 61). The expository method offered several ways to interpret a theological point, which must in itself have been a revelation to Menocchio, who had never before heard that there could be more than one right way to think of any such matter. Plainly the clergy among themselves had considered many possibilities, but what they had offered to the laymen in the churches was apparently always a singular, authoritative interpretation. Menocchio was goaded into action by the expository method, where he saw an invitation to consider what his personally preferred interpretation was among the many, whether authorized or not, and even to make up some of his own. But, not knowing the rules for constructing and interpreting texts, rules that prevailed even within the expository method, he often got things wrong. As Olson points out in The World on Paper, one should make the important distinction between sharing merely a text, and sharing an interpretive method. Minocchio shared a text, but not a method for interpreting it. Menocchio's story can be seen as an example of what happens when texts are read by someone from outside the interpretive community to which they are addressed. But, Menocchio lived at that great historical moment when printing first made books available to ordinary people, and merely technical lay readers first appeared, readers who were neither enculturated in the interpretive ways of the older literate communities, nor had they yet had time to form an interpretive lay community of their own. When they did in the next two Centuries, the lay readers of the Bible who followed Martin Luther, they developed a literalist interpretive procedure that searched for the one right, referential interpretation. Although this approach to meaning is often seen as "high," and attributed to the new sophistication about empiricism that accompanied the birth of science which occurred at about the same time, it is interesting to speculate that its origins may in fact lie in the convergence of its simplicity with the inevitable limitations of this new, and uneducated, community of lay readers. For a single literal interpretation perhaps does not require such an extensive interpretive canon or education in how to think with its patterns. As an unschooled reader, Menocchio often misunderstood the point of what he read. This amateurishness often led him to wrong readings, some of which just happened, almost like puns, to coincide with heresies worrying to the church. For example, he claimed that when the body dies the soul dies with it to be resurrected only on judgment day, and he denied the divinity of Christ. Reading outside of an interpretive community he found in the Fioretto support for the peasant materialism of his region, a notion of matter imbued with

57

The construction of mind and self in an interpretive community

divinity, even when the text he read denied it. The mention was enough for him. But, by using the text to support meaning-making, he was able to reformulate these folk ideas in new ways as a series of idiosyncratic claims that sounded like the better formulated heresies of more learned men. The unusual opportunity to study this early reader's ways of taking text for what it shows about the ontogenesis of the interpretive community, we owe most obviously to the verbatim record taken down by the Inquisitional Court, preserved, and subsequently found by Ginzburg. But we owe it too to a second factor, one partly responsible for his ever having been called by the Court, much less twice; namely, to Menocchio's burning desire to enter a social interpretive community as a writer in his own right, not just as a receiver of other people's interpretations. The fact is, though the Court asked, cajoled and begged, he simply wouldn't stop talking. Not to his peers and not to the Court. He was enchanted by the act of interpretation, and he was apparently drawn to conversation with the Court like a moth to a flame. What Menocchio's compulsion to share his interpretations illustrates is that it is not just the writer intending a meaning and the reader interpreting it that is at issue, but also the reader's desire to share their own interpretations with others, to enter into the interpretive community as writers themselves. This is what gives interpretation the trajectory that takes it back into the shared knowledge of the interpretive community that I described above, and what guarantees that the canons of such communities are never static. Thus, an essential element in the ontogenesis of the interpretive community is the social sharing of interpretations. Experiment with an artificial micro-community I now want to report on a study that perhaps allows us to see some of the processes of the ontogenesis of interpretive communities closer up. Undergraduates watched a Hollywood movie, Speed, then discussed a series of interpretive questions. The wrinkle in this study, designed to let us look at the construction of canons of interpretation, was that the subjects were together in little groups of 2 and 3. There were 54 groups. Half of the groups, or 27 groups, were friends to each other and the interviewer, and half strangers. As a second manipulation of community, we analyzed the text of their conversation dividing it into thirds - beginning, middle, and end of their roughly hour long conversations. This led to two relevant contrasts: the start vs. the end, and friends vs. strangers, and to four cells: friends at the start and the end, and strangers at start and end (see table I). Our idea was that friends groups would begin with a shared canon of some kind, while strangers would begin with none at all, but create one over the hour of the interview. Of these four cells, then, strangers at the start should have least in common, and friends at the end, the most - they are the extremes. Would this show up in the number of lexical items distinctive to them? We used the

58

Carol Fleisher Feldman

Table 1 Experimental design: Number of groups

Friends

Strangers

Start

27

27

End

27

27

chi-square to search for lexicon items distinctive to the groups in each one of these cells. In these chi-square applications, the groups are treated as subjects. 3 The distinctive words met two criteria: 1) they appeared significantly more in one cell than any of the three others (at p . CO ......,

....

.2 "0

o

.!: U

(/)

~ c

Q) Q)

....,

(/)

c

ro -,

Letters and pictures in seventeenth-century education

Having learnt to recognize letters, syllables, and the structures of printed words, schoolchildren might pass to learning to write. Before the more sophisticated technique of writing on relatively expensive paper with ink and pen, letters might be marked on a slate, a durable and erasable medium. Several of these hang on the wall in the far background, and the tall boy dressed in brown waiting by the teachers' desk has finished writing his project on the slate he carries in front of him, and has come up to have it checked. On the bench in the right foreground a slate lies at the bottom of the small pile of reading and writing materials next to the reading boy; the stylus and a small sponge for erasures and correction dangle from a string attached to one end. It is perhaps significant that none of the girls is shown engaged in the business of writing, although boys outnumber girls in the classroom as a whole. My own particular interest in the learning materials lying around this untidy classroom is in the various printed pictures we can see in the foreground. The printing press had two products, and two consequences: identical texts could be reproduced in multiple copies, but so could identical pictorial representations. Pictures, once unique productions, could be widely and fairly cheaply available in the form of prints. Woodcuts could be printed concurrently with moveable type in the same press, as they were, for example, in cheap ballads and broadsheets; the seventeenth century saw the increasingly common use of the rolling press, and as a result the more intricate and subtle images produced by engraving and etching in metal plates also came within the economic reach of a wider public. Through these media portraits of monarchs, generals, churchmen, and of other famous men and women circulated widely. Famous panel paintings and murals by distinguished artists were copied as prints, and thus had many more observers, spread over the whole of Europe, than previously had been possible - styles of art could therefore migrate internationally with considerable speed. News of political events could be illustrated with telling pictures: informative, emotive, polemical, and satirical. The printing press formed and influenced opinion in more ways than through text, and one need not have been literate to read a picture. But texts and pictures interacted. The satire was accompanied by verses, providing further edge and entertaining appeal; the picture of a battle carried an explanatory text or key to explain the stages of the fight, and to describe its notable episodes. The educational potential of printing words and pictures together was explored by a number of seventeenth-century theorists of teaching. So Steen's classroom, if hardly at the cutting edge of pedagogy, might be more advanced and enlightened than it may at first appear. Certainly various kinds of mechanically reproduced pictures lie about the classroom. Spilling out of the satchel in the extreme right foreground there are two: one appears to be a genre print, but the second is a portrait of the great humanist scholar Erasmus (Chapman 1996: 233). This detail is one of Steen's jokes: the eminent writer and educator, abandoned on the floor, stares up at the shoe soles of a rather lesser scholar. One might guess that Erasmus would have enjoyed the joke, but if the 100

John H. Astington

print is not there simply as a joke it is worth thinking about its function; I will return to this matter later. Erasmus, who died in 1536, is a famous Dutch figure from the past; the boy who kneels in front of the teachers' desk occupied about his school box holds a printed portrait of a famous contemporary at the time of the painting, Prince Willem III. The bust-length portrait, contained in an oval, appears to bear some kind of legend below it (Chapman 1996: 233-34). The most evidently pedagogical prints appear on and around the child's chair in the left foreground. These are a sheaf of coloured woodcuts of birds and animals, of quite large proportions, with a single print on each. Those we can see, from left to right, appear to show a heron, a cock, and a stag. Text appears below the printed frame on each of the two prints in plain view, and they appear to include printed letters within the confines of the pictures themselves. They are, in short, sheets from a printed alphabet series - a is for ape, b is for bear, and so on - and they represent another approach to teaching, less compact but more entertaining than the old hornbook. It is these pictures in particular which place Steen's schoolroom in the mainstream of contemporary educational theory. The leading international figure to write about early pedagogy in the first half of the seventeenth century was the Czech John Comenius, to give him his latinised name. Comenius stressed the direct connection between the real world on the one hand, apprehended through the senses, particularly the sight, and language on the other. Hence in learning to read language, or to learn a non-native language, the mind should apprehend objects as it learns the names of these objects, and relates the name to a written sign. The ideal classroom should be full of things, which could be seen, touched, smelt, and even tasted and heard, learnt through physical experience simultaneously with learning their symbolic representations: literacy was to be rooted in experience (Sadler 1968: 20-23; 38--40; 44--45). Such a system obviously has its limits in dealing with abstractions, let alone such words as army or city, say, although the latter can fairly easily be dealt with pictorially. And much early reading, at any time, deals with common concrete objects and common childhood experience, mostly readily communicated in pictorial form. Today early books for children are invariably illustrated, so that we find it odd to think of a time when this would not have been so. The period of Comenius's writings and Jan Steen's picture mark the beginning of the modern period in this respect. Comenius laid out more educational plans than he had time and opportunity to put into practice. So a general picture and word book for children under six was proposed but never produced; his famous achievement in this direction is the bilingual book for older children learning Latin, Orbis Sensualium Pictus, published in 1658, in which pictures of different subjects are followed by simple words and phrases related to the illustration. That Comenius's ideas about teaching were in the wider air by the time of Orbis Pictus - published in German and Latin at Nuremberg in 1658 - is demonstrated by its rapid translation into English. The translator was a Yorkshire-born schoolteacher working in London, Charles Hoole. Comenius had

101

Letters and pictures in seventeenth-century education

visited England and had a number of supporters there, notably the intellectual reformer Samuel Hartlib. 3 Comenius's theory, as expressed in the English edition of Orbis Pictus, was that the book was to "entice witty Children to it, that they may not conceit a torment to be in the school, but dainty-fare. For it apparent, that Children (even from Infancy almost) are delighted with Pictures, and willingly please their eyes with these sights". "[T]he senses," he continues, "(being the main guides of Child-hood, because the Mind doth not as yet raise up it self to an abstracted contemplation of things) evermore seek their own objects, and if they be away, they grow dull, and wry themselves hither and thither, out of a weariness of themselves: but when their objects are present, they grow merry, wax lively, and willingly suffer themselves to be fastened upon them, till the thing be sufficiently discerned. This book then will do a good piece of service in taking (especially flickering) Wits, and preparing them for deeper studies". "Let it be given to children into their hands to delight themselves withal as they please, with the sight of the Pictures, and making them as familiar to themselves as may be, and that even at home, before they be put to School" (Sadler 1968: 91-92; 94). Hoole, a practising schoolmaster, concurs in his translator's preface to Orbis Pictus: "Such a work as this I observe to have formerly been much desired by some experienced teachers, and I myself had some years since ... begun the like (having found it most agreeable to the best witted Children, who are most taken up with Pictures from their Infancy" (Sadler 1968: 10 1). He also quotes from an earlier work, the Gate to Science, published in 1641 by the schoolteacher Hezekiah Woodward: "If we could make our words as legible to Children as Pictures are, their information therefrom would be quickened and surer.... if we had Books, wherein are the Pictures of all Creatures, Herbs, Beasts, Fish, Fowles, they would stand us in great stead. For Pictures are the most intelligible Books, that Children can look upon" (Sadler 1968: 105). Before the English Civil War, then, no such picture books for children were widely available. Over the course of the next thirty years authors and publishers began increasingly to supply the needs of such teachers as Woodward, and of their pupils. In 1660 Charles Hoole wrote that he had published "a New Primar; in the first leafe, whereof I have set the Roman Capitalls (because that Character is now most in use, & those letters the most easie to be learn't) and have joyned therewith the pictures or images of some things whose names begin with that letter, by which a childs memory may be helped to remember how to call his letters; as A, for an Ape, B. for a Bear, &c. This Hieroglyphicall devise doth so affect Children (who are generally forward to communicate what they know) that I have observed them to teach others, that could not so readily learn, to know all the letters in a few hours space, by asking them, what stands A. for? and so concerning other letters backwards and forwards, or as they best liked" (Alston 1969: 9-10). Is this perhaps what is happening in the intent group of small girls at the lower left of Steen's picture? Thus children should be allowed, says Hoole, to get "the names of letters, and their several shapes withall, in a playing manner" (Alston 1969: 10), before 102

John H. Astington

being led on by similar teclmiques to distinguish upper and lower case letters, and then entire syllables. Pictorial alphabets were meant to be simultaneously entertaining and mnemonic, uniting two kinds of symbolism, and gradually shifting from concrete to abstract, from thing to word. Thomas Lye, whose reading book The Child's Delight was published in London in 1671, writes thus about the little engraved vignettes which appear at the start of it: "The symbolical alphabet in pictures stands in the front, not only to allure the sight, and assist the memory of little ones, but also to take off, or at least allay, that particular disgust, and prejudice, which their soft, and sportive souls have taken up against a Book, and Schoole" (Alston 1968: "Epistle"). We might doubt whether the quirky little engravings, made by Thomas Cross and set out in panels of eighteen vignettes to an octavo page, could "allure" for very long. A rather different layout is followed by Tobias Ellis in The English School, a spelling book first published in 1670, with plates which appear to have been added only in later editions; a surviving example is dated 1675, and bears the name of the printseller John Overton, rather suggesting that it might have been available as an independent single sheet, as well as part of a bound book (Alston 1969A: 1); many contemporary illustrations had such a double life. A simple symbolical alphabet, to use Lye's term, is followed on the right by an alphabetical list of monosyllabic words, arranged in three columns. In the final column on the right the Lord's Prayer is printed. The concurrent appearance of alphabet and prayer text is typical of the format of the hornbook, which remained a persistent model for many of the illustrated readers, as we shall see. Overleaf in the Ellis book appears a simpler, less crowded pictorial alphabet, which would seem to be the obvious starting point in teaching, rather than the more intimidating recto side. The illustrations in the book correspond with the progress from monosyllabic words to those of two syllables, which are printed with the Ten Commandments; overleaf, once again, is shown a plainer and more entertaining set of word and picture vignettes. The third text is the Apostles' Creed, accompanying trisyllabic words; once more the verso is far more alluring to soft and sportive souls. One other pictorial alphabet survives from 1668, but despite its title, "The New Horn-Booke", it was evidently sold as sheets of prints rather than as a bound book; the individual vignettes might be cut out and pasted on heavier paper, to use as flash cards. 4 The title plate at the top announces that the pictures constitute "a Pleasant Way to teach yeoung Childeren that can but speake in one hower to know these Lettors Viz. the English, Roman, Ittallian, and Secretary Great and Small. As aliso the Caide, Syrick, Arabeck, Hebrew, and Greeke Characters. By which they may aliso learne to write the afore sayd hands plact in every plate in the fore mentioned order. All which children haveing once seene will not be kept from learning of, who otherwis have bin constraind." This is evidently a fairly ambitious sales pitch, but it does alert us to the various uses to which engraved plates might be put. Pictorial alphabets might teach letters, and the style of these birds and animals has been copied from the prints of the 103

Letters and pictures in seventeenth-century education

Flemish artist Jan Collaert. 5 But engraved plates and woodcuts had for a century or more also been used in copy books, to show the correct forms of script in various kinds of writing. Those learning to write might copy - even trace - from a printed model, as these little alphabet plates include examples of large and small letters in a variety of English hands, as well as the other alphabets. They are brought together in this odd way, I think, as a publisher's venture, or whim. The same print publishers and printsellers made and stocked both alphabet pictures and copy books of letters, and they would have supplied schoolteachers as well as ambitious parents. John Garrett, the publisher of the rather oddly titled New Horn Book, has attempted to run the two kinds of publication together. Peter Stent, a contemporary printseller, also stocked and sold copy books, and in an advertisement sheet published in 1662 he announced that he had for sale "Twenty-five Plats of a Horn-book to teach children to read and write," which would have been a pictorial series similar to Garrett's (Globe 1985: 35). The commercial opportunism of printsellers in the active educational market of the 1660s and 1670s might be illustrated by the single-sheet print called "A New Book of all sorts of Beasts," published by John Overton in 1673 (O'Connell 1999: 33). The crude, cartoonish animals in this publication are certainly amusing, but they are hardly presented systematically, let alone alphabetically, and they could not be easily divided into a series. The cheerfully brazen claim that this sheet is "a pleasant Way to teach yeoung Children to reade almost as soone as speake" is unlikely to have misled experienced teachers. A cheap knock-off publication, it might have provided some delight, but little attendant profit of the didactic kind. Cheap pictorial alphabets continued to be produced; their appearance changed notably only with the advent of inexpensive colour printing in the nineteenth century. A print titled "The British Battledore," from the second half of the eighteenth century, maintains the format of the old hornbook, ready to be pasted to a wooden or cardboard paddle; it sold for a penny. It has pictorial alphabet vignettes running from top left anticlockwise around the perimeter, and in the central panel a lower and upper case alphabet, followed by vowels, and syllables formed with vowels. The Lord's Prayer takes up about half the central space. "The Royal Battledore", from 1775, places text and pictorial alphabet side by side, and adds verses pointing out the economic and social advantages of literacy: "He that ne'er learns his ABC / For Ever will a Blockhead be. / But he that learns these Letters fair, / Shall have a Coach to take the Air" (Tuer 1968). To return to Steen's schoolroom, Erasmus may be remembered, but the didactic methods of Comenius are also being followed, aided by the wider availability of relatively cheap printed material in the later seventeenth century. The animal prints on the little chair are part of a pictorial alphabet; their allure has been increased by having been coloured - by the children themselves, perhaps? - and they are not pinned or pasted to the wall, but have been handed out to be examined, sorted, and played with, as a direct stimulus to the senses and to visual memory.

104

John H. Astington

An emblematic association of word or letter with picture - a capital, the letters APE, and the picture of the animal occuring together within one frame or related visual field - was taken to be a first step in the mental connection of a limited number of written or printed symbols with the sonic variety of spoken language, and with the many ways in which language represents the world and human experience. Most ofthe examples I have cited seem to be a fairly low step in the staircase of literacy - more variety of illustration, for example, would seem to have been required to establish the idea that one letter can stand for a multitude of objects. But pictures are generally taken to be a preliminary to the more complex abstractions of written language - even Steen's picture might be read in terms of this progress. If our eyes move across the painting from left to right, in the direction of western literacy, they travel from the pictorial learning of the little group on the left, through the oral lesson-learning taking place at the teacher's desk - the little boy is saying his letters aloud as the dame points to them - and finally we move to the independent writing and reading going on at the table at the right. Casualties litter this trajectory along its margins, but I think we misread the balance of the picture if we take it to be satirically mocking the essential value of the educational enterprise, of the humane worth of literacy. Intriguingly, Steen begins and ends this trajectory, or arc, with a printed picture: the woodcut of the heron on the left and the engraving of Erasmus on the right. What may he be suggesting by doing so? Surely the open pages of one of Erasmus's books would have been a stronger symbol of the progress of literacy? That is to say that much of the theory of literacy in the seventeenth century speaks of the pictorial aids to reading as the sugar on the pill, an amusement to gain engagement with more complex intellectual operations. Comenius conceives of the relationship between language and visual impressions in a more profound way: both are operations of thought, and are connected. The little prints in Orbis Pictus may seem rather crude and awkward, but they illustrate little worlds, packed with a variety of meanings, dependent on the degree of attention one gives to them, and not all the visual information may be clear until it is explained by the text below. While pictures can create strong and immediate mental impressions, as might that of the bear representing the alphabetic b - they may also present a meaning which works fully only with the accompaniment of a text. Cartoons, for example, can be amusing entirely through visual means, but frequently include captions or dialogue - words and picture work together. Such mutual dependency is explored far more thoroughly by Comenius than by any other contemporary educator. His pictures are full of information, some of it immediately understandable, some not. The text which follows the picture explains the puzzling or obscure details; it gives us a narrative or controlling schema to comprehend the picture. The reader of the book thus becomes engaged in a process which includes satisfying one's curiosity from two complementary sources: visual representations and verbal commentary. Observation and interpretation are made to connect and interpenetrate.

105

Letters and pictures in seventeenth-century education

Among his illustrations of the activities of human culture in Orbis Pictus Comenius includes a picture of a schoolroom, less lively than that of Steen, but showing similar kinds of activities. The engraving depicts a plain room with two large windows, at one end of which sits a master in a chair, with a blackboard above him on the wall. A pupil stands in front of him reciting a lesson, and to the rear two others sit at a table writing. At the opposite end of the room a large group of pupils sit on benches, studying books. The picture seems plain enough, but the text, in English and Latin, set in double columns, surveys the entire significance of the variety of activities shown, keyed to numbers in the picture, which the simple head-title, "A School," actually contains. "A school is a shop in which Yong wits are fashioned to vertue, and it is distinguished into Forms. The master sitteth in a Chair; the Scholars in Forms; he teacheth, they learn. Some things are writ down before them with Chalk on a Table. Some sit at a Table, and write: he mendeth their Faults. Some stand and rehearse things committed to memory. Some talk together, and behave themselves wantonly, and carelessly; these are chastised with a Ferula and a Rod" (Sadler 1968: 307). A picture of a school made to be read by schoolchildren carries within it a little warning about the dangers of slacking. Steen's picture was not primarily addressed to children, although they are quite likely to enjoy it, but it carries a certain freight of similar warning, accompanied by Steen's characteristically tongue-in-cheek attitude to moralistic sternness. Comenius, then, conceives of active engagement with pictorial information as part of the entire intellectual development of which reading and writing form other parts. And as the words of the book Orbis Sensualium Pictus may be learned and copied, so, Comenius suggests in his introduction, may the pictures: "Let [children] be suffered also to imitate the Pictures by hand, if they will, nay rather, let them be encouraged, that they may be willing: first, thus to quicken the attention also towards the things; and to observe the proportion of the parts one towards another; and lastly, to practice the nimbleness of the hand, which is good for many things" (Sadler 1968: 95). Drawing and sketching, then, form part of Comenius's conception of educational development, but how widespread was this idea in practice? The Englishman Henry Peacham, an amateur gentleman artist of reasonable skill, writes of the ability to draw as one of the accomplishments of a fully rounded education, but admits that in his own youth his efforts in that direction were frowned on by schoolmasters as inappropriate frivolities. Peacham, born about 1578, would have been in school in the 1580s and early 90s. By the middle of the next century there had been some change in English outlook. John Dury's little book The Reformed School, published in 1651, is organized according to the different activities appropriate to the age of children. Dury, another associate of Samuel Hartlib, defines three periods of childhood: during the first "the Capacity of Children is none other but Sense and Imagination, with the beginnings of Memory". Later come the beginnings of reasoning, and thereafter full mastery of it, as well as of judgment and prudence. During the first stage, Dury writes, children may be taught a 106

John H. Astington

number of things: first to speak their own language clearly, then to read aloud, and then each pupil to write "his owne Mother-Tongue legibly; or any Tongue what soever, as to the forming of any Letters after a Copy". The next three appropriate accomplishments are described by Dury in these terms. "To draw all manner of Lines and Mathematicall Figures with a Ruler and compasses; and other Lines and Figures which are the rudiments of painting to know the lineaments and features of things. To know the signification of Numericall Figures; and to observe by the eye, eare and hand, the differences of Things in respect of their number, their parts, their quantities, their measures, their proportions and disproportions, and the like. To take notice of all Things offered to his Senses: to know their proper names, to observe their shapes; and to make circumstantiall descriptions thereof by word of mouth, and painting in black and white." This first stage of development Dury places before the age of eight or nine; thereafter, until thirteen or fourteen, both literacy and pictorial representation are to be encouraged concurrently: "the Children shall be exercised ... In writing faire and readily; and in drawing the pictures of Things whereof the impressions are to be fixed in their Memories" (Alston 1972: 52-55). This discipline, evidently influenced by Comenius's ideas, conceives of pictorial representation as related both to language - descriptions can be made equally in words or in drawings - and to mathematics and geometry. Attaining pictorial knowledge - representing the lineaments and features of things - is to increase one's conscious and rational grasp on experience. The elevation of painting and drawing from its commonly perceived status in earlier English culture as a mechanical craft rather than as an art with intellectual dignity is of considerable significance. It is appropriate that the educational role of graphic representation be celebrated by a painter, whose native culture gave more ready recognition to the visual arts than was the case in England, in either the sixteenth or the seventeenth century. Steen was the product of a middle-class urban upbringing, a graduate of a training in school Latin who went on to enroll at but not to graduate from Leiden University, abandoning traditional formal studies to take up painting (Chapman 1996: 25-37). It would be interesting to know something of Steen's own experiences as a child at school - he was born in 1626 - and whether like Henry Peacham he was initially deflected from developing a natural talent. In search of an answer to that question one would want to know more about the differences between English and Dutch educational practices, and whether the impressive achievements of the visual arts in the Low Countries were reflected or fostered in the schoolroom. Figures given by Rab Houston suggest that general literacy in the seventeenth century was more widespread in Holland than it was in England, as well as being more evenly distributed between the sexes (Houston 1988: 130-54). Concurrently there may also have been a more acute general consciousness about pictorial representations. In The Art of Describing Svetlana Alpers specifically singles out Comenius's theories as being particularly related to the Dutch context which produced so much wonderful 107

Letters and pictures in seventeenth-century education

graphic art. She writes as follows: "Comenius appeals not just to the appropriateness of pictures for children - though he did that - but to their appropriateness given the nature of language, perception, and our knowledge of the world. The argument might be put in two parts: perceptual and linguistic. First the path to understanding is through the senses (with sight prime among the rest). Since we store what we have seen in mental images, the stirring up of visual attention is basic to education. Second, language is essentially denotative. One must learn the names given to things and engage in acts of pointing or reference rather than acts of expression or statement. Since words are man's (not God's) to fashion, language is a representation of all the things in the world. Images clearly have a paradigmatic place in a view such as this, which privileges vision and presents language as a representation". She notes a connection between such theory and the close visual attention which Dutch painting frequently calls for. Especially is this true, she writes, in still-life pictures: "Each object is displayed not for use, or as a result of it, but for the attentive eye" (Alpers 1983: 95). The influence of Comenius's principles, however, was international, and they can certainly be seen at work in Steen's classroom. The picture itself invites a sensory exploration of a world and a mental ordering of its parts - or indeed a verbal commentary of the kind I've been attempting in this essay. Some of the traditional teaching it depicts is being mocked for its boring repetitiveness or its evident ineffectiveness - the picture of the unruly schoolroom is a genre subject going back to Pieter Breughel in the century before Steen - but Steen's schoolteachers are by no means entirely owl-like, locked in the darkness of the past. 6 Not only is early literacy being encouraged through pictorial means, in the light of the most liberal educational theory of the day, but the portrait prints among the homework projects of the older boys suggest a continued integration of the acquisition of literacy with other means of investigating and representing the world. Either the prints are for copying, "to practise the nimbleness of the hand" as Comenius puts it, or they are exemplary visual descriptions, in black and white, calling for corresponding descriptions in words: written reports on their pictorial subjects. Portraits in books were commonly accompanied by poems or epigraphs to comment upon them, as is Shakespeare's portrait, for example, in the 1623 Folio edition of his plays. Steen's amused survey of the beginnings of education appropriately reminds us of the importance of picturing in human intellectual culture. It also reflects a new approach to education made possible on a wide scale by the printing press, which could make not only cheaper written texts but also cheaper pictures, which a child might tuck into a school-bag. References

Alpers, S. (1983) The art of describing, Chicago, IL: University of Chicago Press. Alston, R. E. (Ed.) (1968) The child:s delight, Menston: Scolar Press. 108

John H. Astington

Alston, R. E. (Ed.) (1969) A new discovery of the old art of teaching schoole, Menston: Scolar Press. Alston, R. E. (Ed.) (1969) The English school, Menston: Scolar Press. Alston, R. E. (Ed.) (1972) The reformed school, Menston: Scolar Press. Chapman, H. P., Kloek, W. Th., & Wheelock, A. K., Jr. (Ed.) (1996) Jan Steen painter and storyteller, Washington, DC & Amsterdam: National Gallery of Art & Rijksmuseum. Globe, A. (1985) Peter Stent London Printseller circa 1642-1665, Vancouver: University of British Columbia Press. Houston, R. (1988) Literacy in early modern Europe, London: Longman. O'Connell, S. (1999) The popular print in England, London: British Museum Press. Sadler, J. E. (Ed.) (1968) Orbis Pictus, London: Oxford University Press. Tuer, A. W. (1968) History of the Horn-Book, New York: Blom (reprint of 1896 edition).

Notes

2

3

4 5 6

See, e.g., the woodcuts in numerous editions of Gregor Reisch's Margarita Philosophica, 1503 et seq. An owl wearing spectacles and surrounded by lights is the subject of the ninety-fifth emblem in Gabriel RoUenhagen's second book of Nucleus Emblematum (Utrecht, 1613); the picture was engraved by Crispijn de Passe the Elder, and re-used by George Wither in A Collection of Emblemes (London, 1635), p. 253. See G. H. Turnbull, Hartlib. Dury and Comenius (Liverpool, 1947); Charles Webster (Ed.), Samuel Hartlib and the Advancement of Learning (Cambridge, 1970); and Jim Bennett and Scott Mandlebrote, The Garden. the Ark, the Tower, the Temple (Oxford, 1998), pp. 32-41. The prints were sold by John Garrett. British Library, Bagford Collection, Harl. 5949. Animalium Quadrupedum, in nineteen plates (Antwerp, 1612). Schoolroom scenes were also displayed on the title pages of school books, mostly showing well-ordered classrooms, but occasionally depicting disorder. For examples from England in the second half of the sixteenth century, see Ruth Samson Luborsky and Elizabeth Morley Ingram, A Guide to English Illustrated Books 1536-1603, 2 vols. (Tempe, AZ, 1998); Vol. 1, pp. 606-8.

109

8 Pa inters a nd literacy William Blissett

This is an immense subject, and to do justice to it one would need ample space, copious illustration, wide reading and deep specialized learning. The investigator should take account of painters who were illiterate before and after the arrival first of writing, then of printing, some who were literate but never opened a book, or were casual readers of newspapers and ephemera, some who were constant and critical readers, some who could write lively letters or coherent essays about their art, the few who aspired to writing as a distinct pursuit, and the precious few who succeeded. This investigator would interrogate and poll the painters to the limit allowed by the records, pressing them hard; do you read at all? do you read with attention? do you possess what you read? can you allude to, quote from your reading? do you borrow, buy, amass, collect books? do you sign or annotate your books? do you associate specific texts with specific art-works, your own, or other? are you vitally concerned, or concerned at all, with the subject matter of art? and how much, and how well, do you write, and with how much pleasure and fulfilment? The questions have got out of hand; it is beyond my capacity to answer them. 1 All I propose to do here is to review the matter ofthe "Sister Arts" of painting and literature, to discuss at greater length the effect of general literacy on painting, and at the end to draw attention to the "Doubly Gifted" in the two arts. First let me introduce these three topics in three paragraphs. "Poetry is a speaking picture, painting a mute poetry": this gnomic utterance is ascribed to the early Greek poet Simoni des. Horace in his Art of Poetry (line 561) uses the phrase "ut pictura poesis," which Ben Jonson translates, "As Painting, so is Poesy", The currency and prestige of these two classical tags in the classical era from the Renaissance to the end of the eighteenth century led to a dignifying of the doctus pictor, or learned painter, as on the same plane as the doctus poeta, the learned poet. The painter Sir Joshua Reynolds asserted that he can never be a great artist who is grossly illiterate. From this context derives a long-continued problem and debate. Painting and Poetry have undoubtedly interpenetrated, in the intentions and performance of painters and poets, in the

110

William Blissett

arguments of critics, and in the experience of readers and viewers. But from the mid-eighteenth century on, there have always been champions to assert the integrity of the different arts and to protest the confusion of their special vocabularies. Any simple formula, for writing - make it pictorial, or for painting - make it literary, can hardly be defended. 2 What difference, if any, does literacy make to an artist? That question will occupy most of the ensuing discussion. About primitive, archaic, classical, and mediaeval artists the evidence is scanty, but when we arrive at the Gutenberg Era we can devise a sort of McLuhanite "probe". The printed woodcut, the easel painting, and the printed book arrive together and have parallel histories. And we have come to say, quite naturally, without any sense of border raid or even ofliving metaphor, "the way I read this picture ..." or "the way I look at this poem ..." And then as a last division, by way of postscript, there is the phenomenon, appearing grandly in Michelangelo, sporadically in many cultures, but with unique frequency and strength in England: the Doubly Gifted - the painter-writer whose achievements earn him a place in the record of either art. (Him or her, for our own Emily Carr belongs to Canadian literature as well as to Canadian art). Ambidextrous, does their right hand know what their left hand is doing?3 The sister arts Let us recall two celebrated pictures - "The Fall of Icarus" by Pieter Brueghel the Elder and "Et in Arcadia Ego" by Nicolas Poussin, in the Chatham House version. Both are highly "literary," the Brueghel in its influence, the Poussin in its conception. w.H. Auden's Collected Shorter Poems (1950) are arranged alphabetically by first line; its opening word is "about," and so "Musee des Beaux Arts"4 stands first in the book. It has been much anthologized and much discussed. It concludes with these lines: In Brueghel's Icarus, for instance: how everything turns away Quite leisurely from the disaster; the ploughman may Have heard the splash, the forsaken cry, But for him it was not an important failure; the sun shone As it had to on the white legs disappearing into the green Water; and the expensive delicate ship that must have seen Something amazing, a boy falling out of the sky, Had somewhere to get to and sailed calmly on. (19) A few years later, William Carlos Williams wrote his set of ten "Pictures from Brueghel" (1955); this is his "Landscape with the Fall ofIcarus":5 According to Brueghel when Icarus fell it was spring 111

Painters and literacy

a farmer was ploughing his field the whole pageantry of the year was awake tingling near the edge of the sea concerned with itself sweating in the sun that melted the wings' wax un significantly off the coast there was a splash quite unnoticed this was Icarus drowning (4) These two poems, the best known of some forty modem treatments of the subject, are often compared and regularly send their readers back to the painting, deepening and thereby changing their perception of it. Interpenetration is the right word. And yet we must apply the brakes a little. If we had only the poems, these two or all forty, could we reconstruct the visual, painterly qualities of the picture at all? No. And if we had only the picture, could we anticipate any of the literary as distinct from the discursive qualities of the poems - their very different diction and syntax, sound and movement? Again no. Poussin's picture is a haunting evocation of an ideal classical landscape with figures. Who are these people? A young woman and three men; if we have a classical education, we add, a shepherdess and two shepherds and the river-god Alpheus. Where are they? In a Virgilian pastoral landscape - not the stony actual Arcadia of northern Greece but Arcady, a landscape of the mind, a literary landscape of the heart's desire. What are they doing? They are doing what Virgil did, what two millennia of Virgilians, including Poussin and the viewers for whom he painted have done - reading. Reading an inscription, "Et in Arcadia Ego," I too have been in Arcady, even I was in Arcady. Though it has a ring of familiarity, it is not a classical text, just a profoundly classical sentiment. Whose? If the shepherds and the viewer peer more closely so as to read the meaning, they will discern, dark but at the centre of attention, a skull. The dead shepherd interred here sends a message to the living. The whole is utterly pictorial and at the same time utterly literary, though it illustrates no text. The

112

William Blissett

place, the figures, the tone of feeling, the inscription, the composition, all derive from literature. 6 There are countless examples of iconic poems like those of Auden and Williams, no end of literary paintings like Poussin's, and the vocabulary of one art is perpetually being borrowed, appropriated, kidnapped by the other. In his Laocoon (1766) Gotthold Lessing took a firm stand to repel these border raids, questioning the very right to exist of descriptive poetry or allegorical and narrative painting, on the ground that painting has no temporal dimension and must be confined to a single moment, whereas poetry loses its essential movement in time if it lingers over visual details of scenery or portraiture that it is ill-equipped to capture. He pushes his argument very hard, insisting that a picture is immediately not successively perceived. This is surely too rigid: there is a necessary delay (though not a pre-determined succession) in the recognition of a painting's subject matter and of the relation of its elements to the whole even of the "significance" of its significant form. He could not foresee how integral the cityscape would be to the literary art of Dickens or the landscape to that of Hardy, to say nothing of the vignettes of Theophile Gautier or the Imagists or the "still lifes" in literature as seen by Guy Davenport. Let us nevertheless concede that, while some painterly devices may profitably be compared to poetic devices, by way of metaphor, it is bootless to say of a painting, be it never so ambitious, that it is a tragedy or an epic. Carlyle, in describing Robespierre as "the sea-green incorruptible" gives a vivid and memorable, but highly metaphorical, "portrait," but his massive and exciting History of the French Revolution can hardly be said to "paint a picture," still less to "portray" that pictorially unimaginable sequence of events. An exclusive concentration on the study of imagery (assumed to be quasi-pictorial) in literature can lead to similar distortion. Some mixture of vocabularies is a conversational necessity and hence a journalistic and, uneasily, a critical necessity, but it is well that the whistle be blown from time to time. 7 The Muses are all daughters of Memory, now displaying, now concealing, a family likeness: in every art, through memory, the other arts are accessible, so that narrative action is latent in painting, descriptive regard latent in literature. 8 This interpenetration becomes inextricable with the advent of widespread literacy and its concomitants - the book, the easel painting, the illustrated book, the coffee-table art book, the "museum without walls," in the phrase of Andre Malraux. Leaving the question of the Sister Arts only tentatively posed and unresolved, we move on to the more pressing question of how literacy affects painting and the more distant question of how a literary talent subsumes, or subverts, a painter's gifts. The impact of literacy

What difference does literacy make to the artist? The question takes sharp focus in the twentieth century with the discovery of primitive, especially prehistoric art 113

Painters and literacy

and the art of young children and the insane. Three quarters of the time separating us from the Lascaux master had elapsed before the invention of pictographic writing, yet more before the invention of the alphabet. A bona fide illiterate in anybody's book, he may have his limitations, but nobody can dismiss his art as "literary". Consider in this context a drawing in pencil and bodycolour, "Merlin Land" (1931), by David Jones [Plate 1]. Its technique and composition have been claimed for the ancient pre-literate Celtic tradition, and that would suit its original title, "Merlin Appears to the Sleeping Arthur in the Form of a Young Boy," but for me its uncentred, frameless quality relates it to something even earlier than the "Welsh Thing," to a drawing on an uneven surface deep in a cave. It was in fact made by an artist who wrote four books of imaginative literature and two collections of historical and critical essays, an artist who is not merely literate but must take a high place in any informed list of the Doubly Gifted. 9 Place beside this a drawing [Plate 2] by an artist acclaimed by T.S. Eliot, when they were both at the beginning of their careers, as combining the utmost modernity with the energy of the cave man, Wyndham Lewis. To the uninstructed viewer it must seem aggressively physical, unreflective, unliterary - a product, perhaps, of that ideal illiteracy that Eliot pretended to want in the theatre audience. It is in fact from a portfolio of drawings on Shakespeare's Timon ofAthens; not only is it inspired by the tone and style of the play: it also in tum demands that we make a radical re-assessment of the play in the light of the drawings. Lewis wrote scores of books, including plays, poems, and fourteen works of fiction. Like Jones he belongs to the short list of the Doubly Gifted. We shall be returning to both again, noting here that the "primitive" quality of each is always that of a mature person, not of a child. Primitives and illiterates and artists, unless they are children, are not children. 10 From this we may be warned not to plunge recklessly into easy generalizations contrasting the work of illiterate and literary artists. The intermediate term in literacy. Literacy in classical Antiquity was widespread among citizens in Greece and even among some categories of slaves in Imperial Rome, but its relevance to art is hard to determine. Keats wrote his "Ode to a Grecian Urn" to celebrate a classical work of art on a closed strip, but there can be no "Ode on a Grecian Easel Painting" because none has survived, though we know from written records that they existed, were admired and collected. We know too that some extended passages of Latin literature were highly pictorial, to the detriment of the argument to be advanced by Lessing. Whether the fact that books in Antiquity were unfolded on scrolls served to dissociate them from the page-like panel painting and associate them with the frieze is a question to pose and leave. II The codex replaced the scroll in the early Middle Ages and will dictate the form of the early printed book. We have all looked at the pages of illuminated manuscripts and, from the same period, at stained glass windows, finding quite rightly a family resemblance between the two arts. The illuminator must have 114

William Blissett

been literate, proceeding if not letter by letter at least initial by initial, relating design to scriptural context, image to word (and, yes, sometimes allowing himself tangential liberties). The glazier need only have been told the scriptural story and reminded of traditional designs, motifs, and colours; being a reader would be no necessity or advantage to him. And so any close concern with literacy must concentrate on what, following Marshall McLuhan, we call the Gutenberg Era. I propose a kind of McLuhanite probe of the question of literacy and the artist, and I find myself, as Marshall regularly found himself, with more data, more unanswered questions, more inchoate notions, than can be squared-up-to in the usual manner of "writers like Blissett," to use one of his phrases. 12 Here is something that did not register with me until recently I re-read The Gutenberg Galaxy (with increased admiration), where it is found, properly ascribed to William Ivins. The woodcut printed from a block antedates the invention of printing from moveable type and was from the beginning highly popular - a low, mainly anonymous art. People were reading pictures before they read printed books (as children still do), and these pictures were printed on paper, on pages, available in multiple, identical copies. 13 The achieved painting of the Middle Ages and early Renaissance, being based mainly on the Bible or saints' legends ("legend," reading) or old auctores, derived from books but in no way resembled them. Paintings were murals or altar pieces or panels designed to be act in specific places as an embellishment of the architecture. Then, shortly after the appearance of the printed woodcut and the printed book, the easel painting began its rise to dominance, using religious themes from the first but allowing classical, secular, profane, and contemporary subjects. The easel painting could be privately commissioned or painted on speculation: it is portable, signed, suitable for hanging in the gallery of a prince, a nobleman, a magnate, in the homes of prosperous burghers and dilettanti and, in the past two centuries, the likes of you and me. It is the emblem of possessive individualism. 14 The size and shape of the earlier painting was accommodated to architectural considerations: Murals were unframed, altar pieces were enclosed usually in elaborate settings, a practice that has been continued in the provision of ornate frames for important paintings and elaborate bindings for important books. We may find this distracting, but they were clearly intended to glorify what they contain, like a monstrance. The size and shape of the easel painting, from the Renaissance to the mid-twentieth century, in contrast, has been dictated, sometimes by the patron and his commission, but very often by the artist himself, as one way of asserting his independence. Such paintings tend to be of two sizes - large enough to dominate a wall or small enough to take a place in a row, and to be of two shapes - upright or sideways. What may be called moderate rectangles vastly outnumber any round, square, or octagonal pictures. Frames have become appreciably plainer since the Impressionists opted for narrow light wooden frames.

115

Painters and literacy

Learned and pious people have long been readers. Jews, Christians, and Moslems alike are known as "people of the Book". But in the whole Gutenberg Era from Durer to Picasso an increasing number and proportion of the population would have frequent, even daily experience of giving their closest attention to the printed page. The independent framed picture is typically, if upright, the shape of a page of print; if sideways, the shape of the open book. Murals are seldom page-like, ceiling paintings never: the ceilings ofTiepolo are an assemblage of great vignettes, like the zodiac, that archetype of the overlapping, the unframed. The three-dimensional arts of sculpture and architecture are even further removed from the book, and although many Renaissance title pages are contrived to look like fa~ades, no fa aver dato questo nome a questo oggetto? (Where did it get its name?) Davide: Vuomo. (From human beings.) Cinzia: Quelli che 10 hanno creato. (From those who made it.) Matteo: (Secondo me un pupazzo e un pupo-pazzo. (A puppet is a crazy baby. [an Italian pun: pupazzo->pupo-pazzo=crazy baby]

206

lIaria Grazzani Gavazzi and Veronica Ornaghi

R.: Avrebbero potuto dargli un altro nome? (Could it have been called in another way?) All: Si (Yes) Matteo: Orsacchiotto. (Teddy bear.) R.: E potevano chiamarlo 'sole'? ( ... could it have been called 'sun'?) All: No!!! Davide: Perche il sole ha i raggi. (No, because of the rays.) Cinzia: No, rna magari come nome proprio. (No, but ... maybe as a 'proper name'.) Davide: Se 10 dici con l'espressione del nome proprio: ti presento 'Sole', il mio amico, si puo dire. (If you say Sun as a 'proper name', as when you say: I introduce Sun to you, my friend, ... you can.) Here we note Davide's change of point of view from an initial realist to a more nominalist position, since the discussion with Cinzia helps him to argue in the same correct direction. Another example comes from the following discourse sequence: Example 6 - Grade III (puppet)

Marco: Potevano decidere di chiamarlo 'macchinina' appena 10 avevano inventato, l'avevano appena inventato e potevano dare il nome che volevano. (They could call it 'little car' when it was invented, when just invented it could get every name.) Luca: Per ogni cosa ci vuole un bel po' di tempo per decidere il nome. (It takes time to give a name to things.) Gloria: Anche adatto deve essere no? (The name has to be appropriate too ... ) R.: Quindi quando 10 hanno inventato potevano chiamarlo come volevano ... anche 'bambino'? ( ... therefore ... once invented they could call it as they liked ... also 'child'?) Marco: Si. (Yes.) Gloria: Si. (Yes.) R.: Perche? (Why?) Marco: Se al bambino gli avevano dato un nome diverso da 'bambino', un nome al posto di 'bambino', un altro nome, allora si, potevano darglielo, se no ... che confusione!!! (If the child had a different name ~ not 'child' ~ another name, then ... yes, it [puppet] could have been called 'child', otherwise ... what a mess!!!) Interesting enough is Marco's awareness (8.7 yrs. old) that two different categories of object have to get different names in order to make communication work. In some other sequences of our corpus of data, children refer to names as 'just occupied,' as names that cannot be used for other objects.

207

Children's conceptions of name

The use of the marble-grabber (Type: Thing/invented, ad hoc constructed) created a situation for observing the process of giving/constructing names. Initially, children tended to attribute a name by guessing the function of the unknown object. When a 'meaningless name' was proposed ('rovegii') only one group of first graders accepted the language game of giving non-sense names. The third and fifth graders found it difficult to leave aside the 'meaning problem'. The following sequences show what happened when the researcher proposed to give the invented name 'rovegii' to the invented object:

Example 7 - Grade 11/ Jenny: Ma 'rovegii' pen so che sia un nome di persona. ('rovegii'. I think it is a person name.) Nicola: ... rovegii, non ha significato ('rovegii', it is meaningless) R.: E' un nome inventato ora. (It has just been invented.) Mattia: Secondo me i nomi sono dati perch6 hanno un significato, non e che si danno cosi, si sparano a caso ... Cioe hanno un significato. (In my opinion names are given because they have a meaning; one cannot give names superficially, randomly ... Names do have meanings.) R.: E tu, Nicola, come 10 chiameresti? (Nicola, how would you call it?) Nicola: 'Porta palline'perche ci metto Ie palline. (I'd call it 'small balls holder' because I put inside my small balls.) R.: E se 10 chiamo 'rovegii? Non ci posso mettere 10 stesso Ie palline? (But if I'd call it 'rovegii, couldn't I put my small balls in it anyway?) Nicola: Si rna 'rovegia' non sai il significato di 'rovegia' e una parola inventata, non sai che cosa vuol dire. (Yes, but ... 'rovegii, well, you don't know the meaning, it is an invented word.) Giulia: Si, rna se dai un significato a 'rovegii' puoi capirlo cosa vuol dire (Yes, but if we attribute a meaning to 'rovegii', you understand what it means.) Giulia concludes by taking a more nominalistic position, arguing that names are "conventions" and their meaning has to be shared. Note that Magda, the child of the first example (Grade I) tended to argue similarly: " ... rovegii, we don't know what that word means but if you do explain it, we will know". In the following conversation Giovanni recaps the other children's statements as Giulia did in the previous sequence:

Example 8 - Grade V R.: ... 10 pensavo di chiamarlo 'rovegii' ... (1 was thinking to call it 'rovegii ... ) Valeria: Che cosa vuol dire? (What does it mean?)

208

lIaria Grazzani Gavazzi and Veronica Ornaghi

R.: E' un nome inventato, non ha un significato particolare. Posso chiamarlo 'rovegia'? (It has been invented just now, no particular meaning ... can I call it 'rovegia'?) Giovanni: Si. (Yes.) Antonella: No. R.: Perch6 no? (Why not?) Antonella: Perche non ha senso. (Because it is meaningless.) Giovanni: Puo darsi che abbia proprio il significato di quell'oggetto ... (Maybe it has exactly the meaning of that object ... )

In the case of cat as a target object (Type: Animate Entity/Unknown) children are very clever, both in term of assigning a proper name and in term of the standard task ('cat' as a common noun). In fact, the language game consists at a certain point in assigning and changing the proper name of the cat.

Example 9 - Grade I Marco: Potremmo chiamarlo Baffo, come il gatto di una mia arnica. (We could call him Baffo, it is also the name of a cat of a friend of mine.) R.: II gatto, non chiamiamolo pili Baffo, rna chiamiamolo Pietro ... Possiamo, anche se e un nome da bimbo? (Let's change his name, let's call him Pietro ... instead of Baffo, can we do that?) All: Si. (Yes.) R.: Perch6 si puo? (Why can we do that?) Stefano: Perche i nomi ... (Because names .... ) Valentina: ... Si decidono ( ... are decided.) Stefano: Si decidono. (... are decided.) Example 10 - Grade I

R.: ... allora Dio i cani non poteva chiamarli 'gatti'? (Therefore God, could call dogs 'cats' ... ) Cristiana: No. Costanza: No. Federico: Si, poteva fare il contrario: il cane gatto e il gatto cane. (Yes, God could reverse names: calling dogs "cats" and cats "dogs".) Stefano: Poteva chiamarlo anche dinosauro (He could also call the cat 'dinosaur'. ) The construction of the name 'cat' is an easy task for fifth graders:

Example 11 - Grade V R.: Ma chi ha deciso che si chiama "gatto"? (Who decided the name "cat"'?)

209

Children's conceptions of name

Giulia: Hanno inoziato a chiamarlo cosi ... (At the beginning, someone ... ) R.: ... e se volevano chiamarlo in un altro modo? (Those persons, could they decide for another name?) Jenny: Si. (Yes.) R.: Per esempio? (For instance?) Jenny: Cane. (Dog.) R.: Potevano chiamarlo cane? (Could they have been called cat "dog"?) Giulia: Tu una cosa puoi chiamarla come vuoi, pero devi dirlo agli altri che quello eil nome. (You can call a thing as you like, but you have to make an agreement with the other people.) R.: Intendi dire che la cosa importante e mettersi d'accordo? (You mean that the important thing is the agreement?) Giulia Si! (Yes!!) Marco: Potevano anche mettersi d'accordo e chiamarlo "pesce". (It could have been called "fish" as well.) All: Sill (Yes!!)

Conclusions Summarizing, these findings which are only a part of the larger project carried out with a multi-method approach tend to conform to the findings obtained in other contexts and, at the same time, to suggest new lines of research. First of all, differences have been found as a function of the status of name (proper versus common nouns). Naming babies is a frequent practice of children's everyday life. Following Wittgenstein (1953) it is a "language game"a language family game - that allows children to catch easily the arbitrary feature of the proper nouns. For proper names even the youngest children are nominalist to at least some degree. However, names, for them, are drawn from a restricted base of real names. Common names are a complex class, and children seem to cope differently with these diverse research objects since differences emerged as a function of type of name. In order to catch the arbitrary nature of signs it seems easier to deal with unknown objects (such as a never-seen-before puppet) than with a well known object with a clearly established function (such as the ruler). Interestingly enough, children's fluctuations from one stance to another have been found as a function of the target name, since the same child can be realist in some cases and conventionalist in some others. There are actually two interpretations of the same question: one underlines the confusion induced, for instance, by calling "ball" a ruler; the other one - the metalinguistic interpretation - refers to the conventional nature of the sign. However, all the children share the same difficulties with the same target object. At the age of 8 years children frequently appeal to grammar and take into account different languages (e.g., English, the most common second language

210

lIaria Grazzani Gavazzi and Veronica Ornaghi

for Italian children) to make their arguments clear. Literacy, intended in a large sense (for instance, including the wide spectrum of verbal and nonverbal communicative practices for refering to different languages), seems to be an important factor to explain the metalinguistic awareness here implied. In addition, there are individual differences in the ability to deploy nominalist thinking. Such differences seem sometimes more important than the age factor itself. It remains to be seen whether they are explained by cultural factors (for instance, communication in Magda's family [example 3] was based on dialogue and conversation; TV was not present at home) and, if in the affirmative, which factors are to be added to traditional socio-demographic ones. Finally, arguing with peers, as many examples have shown, often allows children to modify their own point of view (see for instance Davide, [example 5]). In addition, discussion among children gives the researcher the chance to better understand the reasoning beyond words. The role of literacy in this development remains obscure and needs further analysis. Certainly, talk about words is enhanced by literacy but whether or not one sees them as conventional or as realist, true, God-given names may not be. After all it was only in the 17th century that the British philosopher John Locke defended the conventionalist view that words are arbitrary signs. And, as data by Homer et al. (1998) revealed, even adults in the 21st century don't totally share Locke's point of view. Perhaps the realist-conventionalist distinction, as we have discussed elsewhere (Grazzani Gavazzi and Ornaghi 2000), is not the most important dimension for characterizing this developmental process. We noted that even the youngest children in some contexts give nominalist, that is conventionalist answers whereas even the oldest often remain committed to realist answers, grounding their realism in the necessity for communicating with others. An alternative formulation would be to understand development in terms of a shift from realist to metalinguistic. In the former, children tie their understanding of the use of names to the properties of objects or classes of objects, or the need for communication. In the metalinguistic stance, they refer not to the reference of the terms but to their grammatical properties, the relations to other words, and the like. This reformulation would connect to the literacy hypothesis (Olson 1994) which suggested the link between literacy and metalinguistic development. References Bamberg, M. (Ed.) (1997) Narrative development: Six approaches, Mahwah, NJ: Lawrence Erlbaum. Berthoud-Papandropoulou, I. (1978) An experimental study of children's ideas about language. In A. Sinclair, ]. Jarvella & W. Levelt (Eds.) The child's conception o.flanguage, (pp. 55-64) Berlin: Springer. Bialystok, E. (1997) Effects of bilingualism and biliteracy on children's emerging concepts of print. Developmental Psychology, 33, 429-440.

211

Children's conceptions of name

Brockmeier, J., & Grazzani, I. (1999) Words, names, and intentional worlds. Redefining Literacy - Newsletter of the Language, Literacy, and the Mind Research Group of the Centre for Applied Cognitive Science, Toronto. Fall Edition. Brook, J. (1970) A test of Piaget's theory of "nominalism realism". The Journal of Genetic Psychology, 116, 165-175. Bruner, J. S. (1986) Actual minds possible worlds, Cambridge, MA: Harvard University Press. Downing, ]., & Oliver, P. (1974) The child's concept of a word. Reading Research Quarterly, 9, 568-582. Dunn, J. (1988) The beginnings of social understanding, Cambridge, MA: Harvard University Press. Ferreiro, E., (1997) What does it mean to study children's theories about the writing system? Unpublished manuscript. Ferreiro, E., Pontecorvo, c., Moreira, N., & Garcia Hidalgo, I. (1996) Cappuccetto Rosso impara a scrivere. Studi psicolinguistici in tre lingue romanze, Firenze: La Nuova Italia. Ferreiro, E., & Vernon, A. (1992) La distinction palabra/nombre en ninos de 4 y 5 anos. Infancia y Aprendizaje, 58, 15-28. Gombert, J. E. (1992) Metalinguistic Development, Chicago, IL: The University of Chicago Press. Grazzani Gavazzi, I., Brockmeier, J., Groppo, M., Carrubba, L., Confalonieri, E., & Ornaghi, V. (1997) Un aspetto della competenza meta linguistic a infantile: la conoscenza dell'origine dei nomi. Studi di Psicologia dell'Educazione, 16, 1/2, 167-181. Grazzani Gavazzi, I. (1999) La competenza metalinguistica: alcune posizioni teoriche e ricerche empiriche. Psicologia dell'Educazione e della Formazione, 112, 193-210. Grazzani Gavazzi, I., Carrubba, L., Ornaghi, v., Groppo, M., & Brockmeier, J. (1999) How children conceptualize their language: The origin of name. Paper presented at the Sixth European Congress of Psychology, Rome, July. Grazzani Gavazzi I., & Ornaghi V. (2000) Conceptions of names in Italian children: a meta linguistic study in a cross-cultural perspective. Paper presented at 30th Annual Meeting of Jean Piaget Society, Montreal, June. Homer, B., Brockmeier, J., Kamawar, D., & Olson, D. R. (1998) Children's meta linguistic understanding of words and names. Paper presented at the 28th Annual Symposium of the Jean Piaget Society, Chicago, June. Kamawar, D., & Homer, B. (1998) Children's metalinguistic awareness of words and names. Redefining Literacy - Newsletter of the Language, Literacy, and the Mind Research Group of the Centre for Applied Cognitive Science, Toronto. Spring Edition. Karmiloff-Smith, A., Grant, ]., Sims, K., Jones, M. c., & Cuckle, P. (1996) Rethinking metalinguistic awareness: representing and accessing knowledge about what counts as a "word". Cognition, 58, 197-219. Markman, E. (1976) Children's difficulty with word-referent differentiation. Child Development, 47, 742-749. Nelson, K. (Ed.) (1989) Narratives from the crib, Cambridge, MA: Harvard University Press. Olson, D. R. (1991) Literacy as metalinguistic activity. In D. R. Olson, & N. Torrance (Eds.) Literacy and orality (pp. 251-270), Cambridge: Cambridge University Press. Olson, D. R. (1994) The world on paper, Cambridge: Cambridge University Press.

212

lIaria Grazzani Gavazzi and Veronica Ornaghi Papandropoulou, I., & Sinclair, H. (1974) What is a word? Experimental study of children's ideas on grammar. Human Development, 17,241-258. Piaget, J. (1926) La representation du monde chez ['en/ant, Paris: Alcan. Pontecorvo, C. (Ed.) (1993) La condivisione della conoscenza, Firenze: La Nuova Italia. Pontecorvo, C. (Ed.) (1997) Writing development: An interdisciplinary view, Amsterdam & Philadelphia: John Benjamins. Resnick, L. B., Levine, J. M., & Teasley, S. D. (1991) Perspectives on socially shared cognition, Washington, DC: American Psychological Association. Siegal, M. (1997) Knowing children. Experiments in conversation and cognition, New York: Psychology Press (2nd Edition). Vygotsky, L.S. (1962) Thought and language. Cambridge: MIT Press (or. 1934). Werner, H. (1953) Einfiihrung in die Entwicklungspsychologie, Munich: Barth. Werner, H., & Kaplan, P. (1963) Symbol/ormation, New York: Wiley. Wittgenstein, L. (1953) Philosophische Untersuchungen, Oxford: Blackwell.

Notes Regarding the distinction between the two types of metalinguistic awareness Gombert writes (1992: 63): "Since the elements of the lexicon constitute minimal units of meaning, it is very difficult to dissociate the two aspects in the observation of real linguistic behaviour" - that is to say the metalexical from the metasematic awareness. 2 This research was conducted with 114 children, aged between 4 and 11, following in part the procedure by Homer, Brockmeier, Kamawar, and Olson (1998) who asked questions of the following types: Origin question (e.g., How did dogs get their name?), Standard Change question (e.g., Could dogs be called cats?) and Modified Change question (e.g., When dogs were first discovered, could they have been called plaps?). Whereas Piaget described a linear development from a realist to a nominalist model we found that children use both models from the very beginning in a combined way, as a function of different type of name; in fact, even 5 year old kids are nominalist when they are asked about proper nouns. 3 Kamawar and Homer (1998) proposed the following ontology of the status and type of possible 'objects': Type

Things

Animate Entities

Persons

Status

PN

CN

PN

CN

PN

Known

Child's school

Pencil

Mickey Mouse

Dog

Child's Friend

Unknown

Photo of lake

Lemon Zester

Invented

Toy Town Novel lego building

New Doll

Novel Creature

New Baby

CN

The questionable feature of this ontology is evident. For instance, a doll is not the same animate entity as a cartoon character; moreover, a doll becomes animate for a child on the basis of its affective and transitional properties. A possible, simpler classification could be made following the grammatical definition of what a noun is. The Devoto and Oli's Italian Dictionary offers the following definition: "Part of the

213

Children's conceptions of name

discourse that in the Western languages is generally opposed to the verb category ... Nouns are divided in common and proper nouns. The proper ones can be distinguished in names of person (e.g., John), place (e.g., Toronto) and people (e.g., Canadians). Common nouns are divided into the following types: concrete (e.g., horse), abstract (beauty), collective (e.g., Upper House) and singulative (e.g., senator)" (1996: 1283). On the basis of this definition research including several target objects has been conducted (Grazzani, Carrubba, Ornaghi, Groppo, & Brockmeier 1999).

214

14 The distinction between graphic system and orthographic system and their pertinence for understanding the acquisition of orthography Emilia Ferreiro

The classic way of studying children's productions in order to evaluate orthographic performance consists of classifying errors in three or four categories: omissions, substitutions, additions and transpositions or permutations. It is perhaps necessary to move away from this tradition in order to try to understand the multiple inaccuracies present in the texts of beginning writers, because the traditional classification permits viewing what has been written only in terms of its distance from current norms. It emphasizes inaccuracy and does not allow focusing on accuracy. Finally, it tells us nothing about the effective knowledge acquired nor the possible justification for inaccuracy. In the research reported here, I shall attempt to show the usefulness of distinguishing between graphic system and orthographic system in order to understand the development of the knowledge in this domain, as the children show it in their productions. The graphic system

Historically established writings that share the same Latin alphabet differ in the distribution of characters. The most ostensible differences are the following: length of written words, amount and type of diacritical marks, distribution of upper case and lower case letters, sequences of letters in initial or final positions, doubling of vowels or consonants, sequences from one to three letters that form frequently used written words. These graphic properties have a certain relationship with the phonological properties of the language in question, but are not totally determined by them. For example, the conjunction < y > in Spanish could be written as < i > ;1 the sound written in Spanish as < Ii > is expressed as < nh > in Portuguese and < gn > in Italian. (For instance, the well known Italian term gnocchi is written iioqui in Spanish and nhoqui in Portuguese.) From the graphic point of view, it is necessary to distinguish among the following:

215

The distinction between graphic system and orthographic system

(a) Graphic Repertory, i.e., the set of shapes-letters with their upper case variants and -perhaps- their typographical variants. Also necessary to include in the graphic repertory are diacritical marks, punctuation marks, word dividers at the end of the line, and perhaps other graphic devices that do not usually appear as part of the "graphic repertory," such as blank spaces (particularly blank spaces that define graphic words). In any case, it is necessary to keep in mind from the beginning that the graphic repertory is not exhausted by the question: "How many letters does a particular written language have?". (b) Combinatory Possibilities (Prohibited, obligatory or permissible combinations.) For instance, is impossible in Spanish but not in English. (c) Positional Restrictions to Combinatory Possibilities. Taking into consideration the spaces that limit graphic words - at the present state of obligatory rules of word segmentation - it is possible to determine that a certain combination of symbols can or cannot occur immediately after or before a space. (d) Rules of Correspondence Between Graphic Sequences and Sound Sequences. Depending on theoretical considerations, these rules of correspondence will belong to the graphic system or not. (Notice that I prefer to talk of the correspondence between sequences and not between isolated elements. In fact, I prefer to analyze the correspondence between elements as a limiting case of sequences: a sequence of one element.) There is no unanimous terminology to designate each of these variants. The term "ortographical" can cover them all, but loses specificity. I will refer to two sources of inspiration of my analysis. Both sources concern the study of written French. Gak (1976) made an important contribution by proposing to distinguish between the graphic system and the orthographic system. According to Gak, the graphic system refers to "the resources a language has available for expressing sounds," understood to be the set of letters and diacritical marks. (However, as stated before, I consider necessary to incorporate more elements as belonging to the graphic system.) The orthographic system is related to the rules that determine "the use of letters according to the circumstances". The possibilities of choosing among alternative letters for expressing the same sound are determined by the graphic system (the rules of correspondence), but the cases in which a certain letter overrules, impeding the occurrence of other equally possible letters, are within the orthographic domain. As a consequence, according to Gak, if an alphabetical writing were totally regular, that is, if it had bi-univocal correspondence between letters and sounds, such writing would have a graphic system, but talking about orthography would be impossible in such a written language. A similar distinction was independently established by Blanche-Benveniste and Chervel (1974: 117-120). They talk about the graphic material (materiau graphique) in their analysis of written French. They emphasize the need to

216

Emilia Ferreiro

describe by itself the graphic material "before dealing with the phonemegrapheme code". They also said that "many questions usually treated as phono-graphic are in fact exclusively related to the description of the graphic material" (Blanche-Benveniste and Chervel 1974: 122). But the problem remains on how to make "the description of the graphic material" and where to place the frontier between graphic and orthographic. Although I shall continue to be inspired by Gak, I shall put his analysis aside below, due to several reasons. First, in order to understand evolution we cannot use as a starting point the idea of correspondence between letters and phonemes. For several years, children work with graphic material and combinations before understanding letter-sound correspondences (Ferreiro and Teberosky 1979). On the other hand, it perhaps may be useful to consider that letter sequences receive at least as much attention as individual letters. This fact is plausible since the requirement for a minimum amount of letters, along with internal variety, and subsequently, the search for inter-relational variations (Ferreiro 1985, 1986) are a strong indication of precocious attention to letter sequences. Cassar and Treiman (1997) recently sustained the following, in a study concerning the consonants or vowels that can be duplicated in English, and their positional restrictions: " ... much evidence now points to the early emergence of orthographic knowledge. This knowledge about the written language is to some degree separate from phonological knowledge. Thus, orthographic knowledge may play a role in the letter choices that young children make when spelling" (p. 632)2. Cassar and Treiman refer to "orthographic knowledge" where I believe it would be more useful to speak of "graphic knowledge". Small children can express, in their production and in their interpretive choices, knowledge of possible or impossible combinations, without yet knowing about alternations at a given point in a graphic chain (a limiting condition of orthography). In other words, there can be knowledge of sequences as such (syntagmatic axis) without the consolidation of grapheme-phoneme correspondencies and especially, without the organization of alternatives in the paradigmatic axis (i.e., places of graphic alternation in a given sequence). This graphic knowledge undoubtedly is reorganized as orthographic knowledge progresses, but there is no reason to believe that it disappears. Corpus of data

I will refer in what follows to analysis made on a corpus constituted by 412 texts written in Spanish and 275 Portuguese texts. These texts were produced by 2nd and 3rd grade children. These texts were collected as part of a wide comparative project between Spanish, Italian and Portuguese languages (Ferreiro, Pontecorvo, Ribeiro Moreira, and Garcia Hidalgo 1996). Children were asked to write, in all these languages, a well known traditional story: Little Red Riding Hood. In addition, the Spanish texts were collected in the capital cities of three

217

The distinction between graphic system and orthographic system

Latinamerican countries: Mexico, Argentina and Uruguay. Portuguese texts were collected in two distant cities of Brazil: Curitiba, in the south and Fortaleza, in the northeast 3 . Under the conditions the texts in our corpus were produced, the children's attention was not directed at graphic or orthographic aspects, but at textual aspects (how to tell the story, what lexical choices to make, how to name the characters at successive moments in the story, how to link episodes, how to give the characters a voice, etc. 4 ) However, because of these same constrains in the production situation, the texts of our corpus can be considered illustrative, more for what does not appear than for what does. In effect, it seems that all the data relative to letter combinations not recorded (in a corpus with abundant orthographical errors) may tum out to be very relevant for the understanding of the (unconscious) acquisition of the combinatory principles pertaining to the written language these children are trying to learn. Research results

In an attempt to show the benefit of taking knowledge about the graphic system into account, I shall use some cases which I find privileged for reflection and which we can explore in our corpus: - letters impossible in the final position (i.e., letters followed by a blank space); digraphs (obligatory ordered combinations); - graphic sequences that correspond to explicit orthographic rules. I shall particularly emphasize the texts produced by second- and third-grade lower class children, since such texts show maximum of spelling problems and these children have few opportunities for interacting with books and other printed material outside of school. 5 1. Letters impossible in the final position. 6 In Spanish as well as in Portuguese, some consonants are impossible in the final position, i.e., followed by a blank space. For example, such is the case of < -c > . Table 1 indicates the number of occurrences of this letter in final position in Spanish and Portuguese texts of 2nd and 3rd grades (7-9 years old children). In this table and in all the other tables the groups of children are indicated by a combination of letters and numbers: the initial letter indicates the language (p = Portuguese; e = Spanish); the number that follows indicates the school grade (2 = 2nd; 3 = 3rd) and the final letter indicates social group of origin (m = middle class; b = lower class). As seen, 100% of occurrences of < -c > are onomatopoeias in Portuguese: toe-toe-toe / tue-tue-tue / doc doc. In Spanish, the same happens (with only two exceptions in 92 occurrences). The onomatopoeias in Spanish are similar: 218

Emilia Ferreiro

Table 1 Occurrences of < -c > in Spanish and Portuguese texts # Texts

# Occurrences

Onomatopoeias

p2b + p3b (N = 119)

08

18

18 (100%)

p2m + p3m (N = 156)

19

59

59 (100%)

e2b + e3b (N = 207)

07

17

16 (94%)

e2m + e3m (N = 205)

28

77

76 (99%)

toc-toc I tac lac tael toc,toc,toc. In all cases, these onomatopoeias correspond to the sound of knocking the door (the wolf's or Little Red Riding Hood's arrival at the grandmother's house) and they appear in twos or threes, with or without separators (dashes, commas). It can be seen that the exclusive use of a final < -c > for onomatopoeias suggests an interesting problem: since onomatopoeias are not words, it is thus possible to believe that only when writing something that is not a word in the language, can graphic restrictions be violated. Another (almost) impossible ending in Spanish is < -m > J The onomatopoeias are the following: pum pum I porn porn I tam tam. In all these cases, they correspond to someone knocking at the door. But we have also recorded: amm I Amml I mmm, at the time the wolf devours the grandmother or the Little Red Riding Hood. The only three occurrences that do not correspond to onomatopoeias were produced by a boy of e2b who wrote com two times instead of the preposition can, and by a child of e2m who wrote sacarom (a verb form) instead of sacaron.

Table 2 Occurrences of in Spanish texts # Texts

# Occurrences

e2b (N = 103)

05

08

06

e3b (N = 104)

06

14

14 (100%)

= 102)

02

03

02

e3m (N = 103)

01

02

02 (100%)

TOTAL (N = 412)

14

27

24 (89%)

e2m (N

219

Onomatopoeias

The distinction between graphic system and orthographic system

2. The case of digraphs composed of two different letters in Portuguese texts. In Portuguese as well as in Spanish there are digraphs consisting of the same repeated letter: < 11 > and < rr> in Spanish; < ss > and < IT > in Portuguese. Let us analyze first the case of digraphs composed of two different letters. It is useful to begin with the analysis of Portuguese, because it has a larger number of digraphs. We shall start with the digraphs consisting of a consonant followed by < h >: < ch > , < lh > and < nh >. All these digraphs were frequently used in the texts of the Portuguese corpus, because in the name of the main character, Chapeuzinho vermelho, we have these three digraphs. First we can ask ourselves the number of cases in which we observe a permutation in the order of these digraphs. The answer is conclusive: the sequences < hn > and < hI > were never produced. Concerning < hc >, we recorded only six permutations used in three words in a single p2b text: hcapeuzion (instead of chapeuzinho), hcega (instead of chegar) and hcamado (instead of chamado). These digraphs have positional restrictions: < lho > cannot begin a graphic sequence, while < lha > begins a graphic sequence only in three words of very limited usage: lhama, lhano, lhanura (from the Spanish nouns: llama, llano, llanura). The data are equally conclusive: no word was written with < lho > or < lha > in the initial position. The initial sequence < nho > is equally impossible (in fact, it occurs in only one word: nhoque, from the Italian gnocchi). In the entire Portuguese corpus we find no occurrence of < nho > in the initial position. Finally, the sequence < nha > appears only in the initial position in some terms of native origin and very limited usage. In the Portuguese corpus we recorded seven occurrences of this sequence in the initial position, but they deal exclusively with onomatopoeias: nhac or nhanc (= the wolf devouring the grandmother or Little Red Riding Hood). Notice that these onomatopoeias violate two graphic principles: initial < nh- > and final < -c > . What we stated previously applies to this case: only in sequences that are not words can principles of the graphic system be violated. (We must take into account that in terms of orthographic alternations, these three digraphs are not homogeneous: < ch > competes with < x > in the initial position; competes with . Only has no orthographic alternative.) 3. The uses of in Spanish. This digraph can be considered as solidly established since in 398 occurrences we never observed a permutation (i.e., no child wrote < hc >, and in only one occurrence it is replaced by < sh >, which is nonexistent in Spanish orthography: escusho (instead of escuchO). The grammatical category of words in which this digraph was utilized is quite variable: nouns (leche, cuchillo), verbs (escuchar, aprovechar), adjectives (chico, muchos), fixed expressions of farewell (chau). The extremely scarce deviations (7/398) are not concentrated in any of the school groups.

220

Emilia Ferreiro

Table 3 Occurrences of or in Spanish texts Total

Correct

Deviating

e2b (N = 103)

814

777 (95.5%)

37 (4.5%)

e3b (N = 104)

924

905 (98%)

19 (2%)

e2m (N = 102)

668

654 (98%)

14 (2%)

e3m (N = 103)

937

927 (99%)

10 (1%)

3343

3263 (98%)

80 (2%)

TOTAL (N = 412)

4. The use of < qu > in Spanish. In Spanish, the letter < q > can never appear alone, and for this reason, the sequence < qu > must be considered obligatory and can be treated as a digraph. In the entire sample, no permutation of the letters of this obligatory sequence was observed. That is to say, no child wrote < uq >. (There were 3343 occurrences of words with < qu >; the deviating productions never implied the use of : qe (= que); use of < qu > with value /ke/.

The only deviating response that has no equivalent in Spanish is one from the p2b group, which utilizes the orthographic alternative : ke (= que).

Table 4 Occurrences of < qu- > or < -qu- > in Portuguese texts Total

Correct

Deviating

p2b (N = 50)

230

223 (97%)

7 (3%)

p3b (N = 69)

411

409 (99.5%)

2 (0.5%)

p2m (N = 76)

300

300 (100%)

p3m (N =82)

515

512 (99%)

3 (1%)

1456

1444 (99%)

12 (1%)

TOTAL (N = 275)

221

The distinction between graphic system and orthographic system

6. Graphic (and orthographic) sequences that do not correspond to a single phoneme of the language, in Spanish texts. Until now, we have analyzed cases of digraphs that simultaneously form obligatory graphic sequences, and "particular cases" of the famous correspondencies between graphemes and phonemes (give that by definition a digraph is a sequence of two letters corresponding to a single phoneme). It is useful to compare these cases with others that appear to be purely orthographic, yet because of their obligatory character (without exceptions), children could also treat them as graphic sequences. Contrary to the previously mentioned cases, which referred (preferably) to initial or final combinations (beginning or end of graphic words), the case below deals with graphic sequences occurring within a graphic chain. One of the orthographic rules in Spanish that is incessantly repeated at school (and one of the few that adult users can recite without reflection) is the following: " is written before and ".8 The syllabic frontier of these words is located obligatorily between those two letters (for example, cam-po, tam-bien). Table 5 shows, out of the total number of words that should be written with < -mp- > , how many were written correctly and how many were written with < -np >, the non-orthographic sequence. The data are very clear (82% of correct responses), especially if we take into account that 21 of the 35 occurrences of written deviations (60% of them) are words that begin with the sequence < en- >, which forms an autonomous graphic sequence (a written preposition). It could be, in this case, a conflict between two graphic sequences: < en > and < -mp- >. In the sequential order of writing < n> would prevail over < m > because it follows < e > . Out of the total of 162 words that were correctly written, only 50 (31 % of them) begin with < em- >, thus reinforcing our hypothesis. The corpus of our story has fewer words that should be written with < -mb- > . In addition, in this case there is an orthographic alternative ( < v > instead of < b > ) that does not correspond to any difference in sound. The analysis should then take into account four alternatives: < -mb- > and < -nv- > (both correct);

Table 5 Occurrences of < -mp- > and < -np- > in Spanish texts

(correct)

(deviating)

e2b (N = 103)

32 (73%)

12 (27%)

e3b (N = 104)

46 (75%)

15 (25%)

e2m (N = 102)

36 (86%)

06 (14%)

e3m (N = 103)

48 (96%)

02 (4%)

TOTAL (N = 412)

162 (82%) 222

35 (18%)

Emilia Ferreiro

< -nb- > and < -mv- > (both incorrect). The scarce number of occurrences makes comparison risky. We observed the following oppositions, that need to be considered tentative and subject to further confirmation, but that correspond to our expectations: TOTAL (correct) = 71 (92%) versus TOTAL (incorrect)=6 TOTAL < -nv- > (correct) = 15 (75%) versus TOTAL < -mv- > (incorrect) = 5 7. The case o/Ihe contrast hetvl'een single and double in Spanish. This contrast is phonemic in Spanish (exemples: carrolcaro = car/expensive; perrolpero = dog/but). Tn initial position, only < r- > is possible, corresponding always to Irr/. It is thus an interesting case of a graphic rule that goes against a phonemelgrapheme correspondence. If graphic considerations prevails, few violations will be expected (i.e., we expect dominance of < r- > over < rr- > ). Data shows that it is the case: total < r- > represents 98.5% (1568 correct over 1592 occurrences; 97% of correct occurrences already in e2b). (See Table 6). In final position, only < -r > is possible, corresponding always to Ir/. It is thus a case where the graphic rule goes in the same sense as the phonemelgrapheme correspondence. The expectation is of a total agreement. However, data shows a similar situation with the initial < r- >: a total of 97.5% of correct occurrences (2934 correct occurrences over 3008 expectations; the majority of deviant occurrences correspond to omissions of final Irl in the case of infinitive verbs, according in fact to the phonic realization). However, the important fact is that we NEVER registered a graphic realization that goes against graphic and orthographic considerations (i.e., no child wrote a final < -rr> ). (See Table 7). So, it is easy to predict that all problems with the contrast < rrlr > will take place inside the graphic sequences (not at the beginning neither at the end), because the decisions are regulated exclusively by orthographic rules. In addition, in Spanish it is almost impossible to find a written word that twice presents double r (necessarily internal to a written string, as in ferrocarril or arrorrrJ); so, we predict that the most frequent case will be the transfonnatioJ1 of

Table 6 Occurrences of initial in Spanish texts Initial r

= 103) e3b (N = 104) e2m (N = 102) e3m (N = 103)

e2b (N

TOTAL (N = 412)

Total

< r- > (correct)

436

422 (97%)

14

379

374 (99%)

05

344

341 (99%)

03

433

431 (99.5%)

02

1592

1568 (98.5%) 223

< rr- > (deviating)

24 (1.5%)

The distinction between graphic system and orthographic system

Table 7 Occurrences of final in Spanish texts Final r

Total

(correct)

e2b (N =103)

721

694 (96%)

27

e3b (N =104)

850

829 (97.5%)

21

e2m (N =102)

605

589 (97%)

16

e3m (N = 103)

832

822 (99%)

10

3008

2934 (97.5%)

TOTAL (N = 412)

(deviating)

74 (2.5%)

Table 8 Occurrences of internal < -rr- > in Spanish texts Internal < -rr- >

Total

(correct)

e2b (N =103)

143

76 (53%)

67

e3b (N =104)

167

113 (68%)

54

e2m (N =102)

103

72 (70%)

31

e3m (N = 103)

123

101 (82%)

22

TOTAL (N = 412)

536

362 (67.5%)

174 (32.5%)

< -r- > (deviating)

an internal < -rr- > into a single < -r- > (as in carera, instead of carrera; agarar instead of agarrar, etc.). Data shows that it is the case: percentages decrease dramatically when we consider the normative use of internal < -rr- >. (362 correct over 536 occurrences = 67.5%). (See Table 8). However, what is more interesting is the following contrast: -

The transformation of a normative into was observed 174 times over 536 written words (32.5%). The transformation of a normative < -r- > into < -rr- > was observed only two times over 9995 occurrences (a child of e3m wrote prrimero instead of primero; another child of e2m wrote corrto instead of corto).

Final remarks It is perhaps useful to begin by briefly summarizing the main findings obtained

from the analysis of impossible graphic sequences, on one hand, and the obligatory graphic sequences, on the other. 224

Emilia Ferreiro

We have been able to show that even children from the unprivileged groups (p2b, p3b, a2b and a3b) are sensitive to graphic restrictions relative to impossible letters in the final position. In both Spanish and Portuguese written texts, < -c > was reserved to onomatopoeias. The same occurs with < -m > in the Spanish texts. As we stated, since onomatopoeias are not words, children seem to express through their productions the following rule: only when writing something that is not a word in the language, can graphic restrictions be violated. It is necessary to indicate that the beginning language textbooks in the four countries where these texts were produced (Brazil, Mexico, Argentina, Uruguay), make no mention of "impossible letters in the final position".9 This reinforced our idea that knowledge about impossible final sequences is not obtained by transmission, but through the active exploration of written material (in the language the children are trying to learn). The analysis of digraphs composed of two different letters was also extremely instructive. No permutations are observed in the order of the letters in these digraphs: no occurrence of < hn > of < hI > in Portuguese, and in a sole text of p2b we recorded < hc > ; there is no occurrence of < hc > or < uq > in Spanish. It is necessary to take into account that none of the texts incorporated into the corpus was accepted or rejected by spelling criteria. The criteria for rejection were of a narrative nature (absence of some of the main characters in the story, absence of the main episodes of the story or mixtures with other traditional stories) or of a graphic nature linked to figural aspects, such as extremely deficient letter formation or an almost total absence of graphic separation between words. Therefore, due to the dimensions of our corpus, the total absence of the phenomena we are reporting seems to be a crucial point. These digraphs (including < qu >, which was treated as a digraph for previously indicated reasons) cannot have been learned as letter additions. We believe there are graphic as well as phonic reasons for asserting this impossibility. In first place, if these sequences had been learned as one of many possible strings of letters, nothing would allow predicting the stability that approaches 100%. In second place, the phoneme associated with the phonic composition of these digraphs is not a result of the addition of individual graphemes. 10 Therefore, our data strongly suggest that these digraphs have been recognized and learned as such, and not as the addition of letters. However, if this is so, there are other consequences. In effect, if children began with the basic assumption of alphabetical systems (i.e., a one-to-one relationship between letters and phonemes), digraphs would be particularly difficult to assimilate, precisely because they violate the principles of bi-univocal relationships. Our data show that digraphs are not treated as exceptions to a general rule, but as "easily-understood" cases (perhaps because they are not exceptions to a children's rule, but to a rule for language analysts and teachers). It is possibly most natural (from the leamer's point of view) to begin with correspondencies between graphic sequences and sound sequences. But this idea

225

The distinction between graphic system and orthographic system

also has consequences in the traditional manner of describing learners' analytical work. The confrontation between graphic sequences and orthographic sequences ( < -mp- > and < -mb- > in Spanish) was also extremely instructive. These sequences are not digraphs, and therefore do not correspond to a single phoneme. Moreover, the syllabic frontier of words with this sequence passes precisely through the interior of the sequence (i.e., -mlp- and -mlb-, as we have shown with words from the lexicon of the story, like tam-poco and tam-bien). Even in e2b texts, near three quarters of the occurrences respect the < -mp- > sequence. The examples of deviations from orthographic rules seem to correspond to a conflict between autonomous graphic sequences (the preposition en) and orthographic rules (which, I insist, are perhaps recognized by children as obligatory graphic sequences).11 Lastly, it was a surprise for us to discover so many positive aspects in texts which, at first glance, seem full of "deviations" or "irregularities". The possibility of establishing a description "in positive terms" of beginning writers' texts has been one of our permanent concerns in this comparative project. We have achieved this goal in other aspects such as graphic segmentation and lexical repetitions (Ferreiro, Pontecorvo, Ribeiro-Moreira, and Garcia-Hidalgo 1996: Chapters 2 and 5). The current analysis prolongs the intentions present in Chapter 3 of the same work. I am convinced that the distinction between "graphic knowledge" and "orthographic knowledge" is essential for trying to understand the acquisition of orthography. The data presented here allow us to open the debate regarding the frontier between the two domains, as well as to sustain that combinations on the syntagmatic axis (the succession of graphemes) precede the consideration of possible alternations on the paradigmatic axis (i.e., possible letter replacement at a given point of the succession of graphemes). The points of alternations on the paradigmatic axis are places where grapheme/phoneme correspondence rules are just as present as the combinatory restrictions that determine the occurance of a certain letter to the exclusion of any other; i.e., the specific domain of orthography. The total absence of certain letters (or combinations) in certain positions - in a corpus in which any unsuspecting reader would find more inaccuracies than accuracies - is a fact that could not be neglected. In all tables differences in number are so big that no analysis of significance is needed. I am suggesting a research program with the following aim: to change the description of orthography acquisition by emphasizing accuracy and attempting simultaneously to find new parameters for analyzing inaccuracy. I am also suggesting that children can learn as graphic restrictions something that is usually viewed as an orthographic rule.

226

Emilia Ferreiro

References Blanche-Benveniste, c., & Chervel, A. (1974) L'ortographe, Paris: Maspero. Cassar, M., & Treiman, R. (1997) The begginings of orthographic knowledge: children's knowledge of double letters in words. Journal of Educational Psychology, 89,631-644. Ferreiro, E., & Teberosky, A. (1979) Los sistemas de escritura en el desarrollo del nino, Mexico: Siglo xxi Editores. (English translation: Literacy before schooling. Exeter, NH: Heinemann, 1982) Ferreiro, E. (1985) Literacy development: A psychogenetic perspective. In D. Olson, N. Torrance, & A. Hildyard (Eds.), Literacy, language and learning (pp. 217-228), New York: Cambridge University Press. Ferreiro, E. (1986) The interplay between information and assimilation in beginning literacy. In W. Teale & E. Sulzby (Eds.), Emergent Literacy (pp. 15-49), Norwood, NJ: Ablex. Ferreiro, E., Pontecorvo, c., & Zucchermaglio, C. (1987) Doppie 0 dopie? Come i bambini interpretano Ie duplicazioni di lettere. Eta Evolutiva, 27, 24-38. Ferreiro, E., Pontecorvo, c., & Zucchermaglio, C. (1996) Pizza or Piza? How children interpret the doubling of letters in writing. In C. Ponte corvo, M. Orsolini, B. Burge, & L. Resnick, Children s early text construction (pp. 145-163), Mahwah, NJ: Lawrence Erlbaum. Ferreiro, E., Pontecorvo, c., Ribeiro-Moreira, N., & Garcia-Hidalgo, I. (1996) Caperucita Roja aprende a escribir. Estudios psicolingiiisticos comparativos en tres lenguas, Barcelona: Gedisa (Simultaneous Italian edition, Firenze: La Nuova Italia Editrice, and Portuguese edition, Sao Paulo: Atica). Gak, V.G. (1976) L'ortographe dufran{:ais. Essai de description theorique et pratique, Paris: SELAE Sampson, G. (1985) Writing systems, London: Hutchinson.

Notes

2 3 4 5 6 7 8

In fact, the conjunction was written in this manner until recent centuries. The oscillations between < i > and < y > are one of the most interesting problems in the history of Spanish orthography. In 1987 we made, with C. Pontecorvo and C. Zucchermaglio, a comparative study Spanish-Italian - about the acceptance or rejection of double consonants. English translation 1996. For details concerning the conditions under which the texts were obtained see Ferreiro, Pontecorvo, Ribeiro-Moreira, and Garcia-Hidalgo, 1996, Chapter 1. It is possible that at the ages studied, orthographical correction could be accomplished only afterwards, at the moment of "revision of text". Far from being spontaneous, text revision needs a didactic intervention. In what follows, the analysis presented on Portuguese texts were provided by Nadja Ribeiro Moreira, co-author of the 1996 book already mentioned. In the following we shall utilize the notation used by Sampson (1985: 90). A short dash preceding a letter indicates the final letter of a written string; a short dash following a letter indicates an initial letter; a dash on both sides indicates an intermediate letter. Extremely rare exceptions, such as album or curriculum. The dashes to the left and right are ours, in dealing necessarily with interior sequences in the graphic word. In school terms, this rule is stated as: "Hay que escribir erne antes de pe y de be" ("M must be written before P and B").

227

The distinction between graphic system and orthographic system

9 During a preliminary survey carried out in Mexico, the teachers showed surprise when asked which letters cannot appear at the end of a word. Their surprise and hesitation seem to indicate that we are not dealing with "scholastic knowledge" that is immediately available for transmittal. 10 For example, in Spanish, the letter < c > , associated with the unvoiced velar phoneme /k/, added to the "mute" letter < h >, cannot lead -except by magic- to the unvoiced fricative palatal phoneme that corresponds to the digraph < ch > . 11 Another example of overlapping between graphics and orthography appears in the case of the final < -n > in Portuguese. An impossible ending in principle, but possible when associated with the plural morpheme. Various frequent words in the Portuguese corpus are normatively written with in singular but with in plural (for example, homemlhomens, umluns). The possibility of isolating a graphic phenomenon, in contrast to an orthographic one, is not always available in a particular written language.

228

15 Children's analysis of oral and written words Sofra A. Vernon

Introduction

For more than twenty years, mainstream psychological research has been based on the assumption that writing represents speech as a sequence of phonemes. Thus, in order to learn to read, children must learn to analyze the phonological components of their oral language, which will in tum enable them to establish one-to-one correspondences between letters and sound units (phonemes). The enormous quantity of research that shares this point of view is called phonological awareness. Although it is not the purpose of this paper to summarize phonological awareness research, we will point out some basic assumptions and open questions, each one of them related to the others: (a) Linguistic units and experimental tasks: Because alphabetic scripts are considered to represent phonemes, children's awareness of this linguistic unit has been the main focus of attention. It is clear that children can "operate" with syllables and intrasyllabic units (onset and rime]) at around four years of age, before learning to read or write (Kirtley, Bryant, Maclean and Bradley 1989, Maclean, Bryant, and Bradley 1987). Phonological segmental analysis, however, develops at a later age, and some reading instruction or specific training seems to be needed in order to develop such a segmental skill. It has been found that even adults who can read a non-alphabetic script, and pre-literate adults are unable to segment phonemes (Morais, Cary, Alegria, and Bertelson 1979, Read, Zhang, Nie, and Ding 1986). However, since pre-literate children can be trained to do segmental analysis of words and pseudo-words (Lundberg, Frost, and Petersen 1988), the debate on whether segmental phonological awareness is a pre-requisite or a consequence of learning to read is still open. In other words, sensitivity to syllables and intrasylIabic-syIlabic units can develop without knowledge of a writing system, whereas awareness of phonemes seems to require some experience with an alphabetic writing system (Treiman and Zukowski 1996). 229

Children's analysis of oral and written words

A great variety oftasks has been used to investigate and train phonological skills: Segmenting, counting, blending, deleting, identitying linguistic units in certain positions. These tasks have varying degrees of success among children, depending on the linguistic unit that has been chosen (syllable, intrasyllabic units, such as onsets or rimes, or phonemes), its position in the oral string used as stimulus, and the cognitive activity the task requires the child to perform (Yopp 1988). (b) Development: As has been mentioned above, research results show that there is a development of at least some phonological skills. Syllables, and onsets and rimes have been considered as more natural, since children do not need any specific training to identify or analyze them. Phonemes, on the other hand, have been considered psychologically unnatural, since explicit training is required to be able to operate at this linguistic level. Phonological analysis of words is a difficult task for young children (Liberman, Shankweiler, Fischer, and Carter 1974). In this sense, the phoneme does not seem to be a basic notion, and oral linguistic development alone may not lead to the construction of this abstract unit. For some authors, intrasyllabic awareness precedes and predicts children's progress in learning to read, whereas segmental phonological awareness appears later, partly as a consequence ofiearning to read (Goswami and Bryant 1990). Even if these two types of awareness have been considered as strong predictors of reading success (Mann 1991), several authors have questioned whether syllabic and intrasyllabic awareness naturally leads to segmental phonological awareness (Cary and Verhaeghe 1994, Nation and Hulme 1997, Seymour and Evans 1994), or that it is actually correlated with the learning of reading and spelling (Muter, Hulme, Snowling, and Taylor 1998) at the early stages. (c) Effects of language and script: Although in the last few years phonological awareness research has been directed at different languages (Morais 1995), most studies have been carried out with English-speakers. Experimental designs and results from studies concerning English-speakers have been generalized to other languages (Carrillo 1994, Defior and Tudela 1994, Jimenez-Gonzalez 1993, 1997) even when phonological structure of these other languages is very different from English. Not only do languages have varying sets of phonemes, but they also have different syllabic and rhythmic patterns. Different phonological structures must, by logic, produce some differences in phonological awareness. Phonological awareness has become important in research because it has proven to be an important skill for reading acquisition. Beginning reading, in tum, has been considered mainly as a word identification task (Stahl and Murray 1998). Effective word identification seems to depend upon decoding (Share and Stanovich 1995) since alphabetic scripts are thought to represent mainly phonological units of language. Phonological awareness is a good predictor for reading success, and good readers get better scores in 230

Sofia A. Vernon

phonological awareness tasks than poor or disabled readers (Rosner and Simon 1971, Torneus 1984). Together with phonological awareness skills, knowledge of letter names and sound-correspondences has also been considered an important factor involved in reading acquisition (Ehri 1998, Stahl and Murray 1998, Treiman 1998). Simultaneous training in both phonological awareness and letter-sound correspondences has proven beneficial in some studies (Ball and Blanchman 1991, Bradley and Bryant 1985, Treiman 1998), probably because, as Stahl and Murray (1998: 84) suggest, "alphabet knowledge seems to be implicated in the development of phoneme identity. Thus, what we have been calling 'phonological awareness' may indeed be something akin to the process of learning letter-sound relations and using them to decode words". Also, the fact that letter-sound correspondences are not straightforward in such written languages as English or French, has strengthened the search for other mechanisms through which speech is represented, and other sound units, such as onsets and rimes, have gained importance in phonological awareness and spelling research. Goswami and Mead (1992), for instance, claim that onset-rime awareness is connected to making analogies between the spelling patterns representing the rimes in words which cannot be read through the establishment of letter-sound correspondences alone, as in light, sight and fight. Treiman (1998) has stated that spelling helps children understand the alphabetic principle, and that it helps their reading skills, because writing enhances children's analysis of words into smaller units. Some studies suggest that children may be able to write alphabetically even before they can use the alphabetic principle in reading (Bryant and Bradley 1980). The findings that full segmental awareness may be reached only when people have had some experience with alphabetic scripts or specific training, that segments are not self-evident, the difficulties in explaining reading as decodification in many alphabetic scripts, and the evidence of the contributions of spelling for the construction of the alphabetic principle makes it surprising that phonological awareness has been tested almost exclusively in oral contexts. Most phonological awareness tasks demand children to analyze oral words or pseudo-words and results are then correlated with standardized reading tests. Very little research has been conducted concerning the effects writing may have upon phonological or other kinds of metalinguistic awareness. Other psychological data

In recent years, some linguists, anthropologists and psychologists have strongly argued that even glotographic scripts (i.e., the scripts that represent language) are not the ideal representation of speech. In fact, Harris (1986, 1995) has stated that the reason why linguists have centered their attention upon speech and the sound system is precisely because we write with an alphabet. Harris (1995: 98) has pointed out that "phonemes themselves are in the first instance theoretical 231

Children's analysis of oral and written words

idealizations extrapolated from letters of the alphabet". Even if alphabetic scripts do, up to a certain point, represent phonological units, they also reflect many other aspects of oral language (Vachek 1989). Harris (1995: 97) points out, for instance, that even in a simple phrase such as "a book," the first letter represents, simultaneously, a sound, a morpheme, and a word. Also, writing makes use of elements that cannot be said to transcribe oral speech: Blank spaces between words, typographic differences between letters, the use of "silent" letters, etc. Olson (1994, 1996) has suggested that writing, in fact, is a model that enables us to think about the structure of spoken language. Children may learn about phonemes because letters appear in writing as separate, discrete units. Since letters, words, and sentences all provide different types of information to the reader, children may learn a variety of concepts pertaining to language form and use through writing. In psychological research, any task requires the child to center his/her attention upon a given phenomenon. Several studies have shown that even pre-literate children have many co-existing hypotheses around the differences between writing and orality. Different types of analysis may arise depending on the task, the context, the kind of text that has to be produced or interpreted, the child's developmental level, etc. In metalinguistic tasks, children can be asked to perform at many different linguistic levels (phonological, syntactic, semantic, pragmatic), depending upon the researcher's aims. These are only some examples of the different types of analysis young children can make when their attention is drawn to different linguistic or metalinguistic phenomena: 1. Ferreiro and Vernon (1992) studied the meaning 4 and 5 year old Mexican children gave to metalinguistic terms "word," "name" and "is called," using pictures and objects. Some of these had printed text and some had not. Pencil and paper were always available. Children were asked, for each object or picture "What is this called?," "Does it have a name?" and afterwards "Do you know what a word is? Give me an example". Results showed that, for most children, "is called" refers to an oral word, whereas "name" is reserved to a string of letters that can be interpreted and verbalized (i.e., a written word or phrase). The term "word" was unknown to many of the children, or had an ambiguous meaning: When asked to give an example of what a "word" was, some children told a whole story (orally); others referred to a single letter without a meaning, or to an oral word. This study heightened the importance of writing in children's understanding of metalinguistic terms. The differentiation children made between oral and written words was as strong as it was unexpected. One of the interviewed children, for instance, argued strongly that an oral word could be called a "word," but when it was written down, it would become a "name". 2. Ferreiro (1997 a, 1997b), asked second graders who could write alphabetically to count the words in a proverb, then to write the proverb and count the

232

Sofia A. Vernon

written words. Children had difficulties in matching the words they had identified in both tasks, even if both units were called "words". 3. Ferreiro (1978) showed a written sentence to kindergartners, and then read it to them with normal intonation. She then proceeded to ask children what was written in different parts of the text. Although most of them believed the nouns had been written down, they did not believe that articles, pronouns and other function words actually appeared in the written sentence. Children established a difference between what is written in a written sentence and how the same written material could be verbalized or interpreted. 4. Ferreiro and Gomez-Palacios (1982) showed pre-alphabetic first graders a picture of a duck in a pond, asked them to write, and then interpret, the word "duck". Later, they showed the picture of four ducks (in an horizontal arrangement), and asked children to write the plural "ducks". Some children merely wrote more letters for the plural than for the singular word. However, many children constructed what the authors called a "plural through juxtaposition" (i.e. the repetition of the singular form four times - one for each depicted duck). Surprisingly, even children with advanced invented . spellings (closer to conventional writings), such as ao or ato for pato did this. So, on the one hand, children could produce the singular by representing syllables or phonemes, and on the other hand, these same children expressed plurality in a cumulative way, similar to what Harris (1995) and Olson (1994) have called "tokens". In spite of this, children "read" the juxtaposition not as "duck," "duck," "duck," "duck," but as the plural word "ducks". In this case, juxtaposition may have been suggested by the spatial arrangement of the pictures. Nevertheless, the fact is that children use different types of analysis in producing a written text, and that they proceed in a different way than in oral speech. 5. Even when tasks demand a centration upon the sound pattern of language, the effects of writing are evident. For example, Ehri (1984) reports an experiment where fourth graders were asked to identify the number of syllables in words where writing suggested there was one more syllable than in the oral word (such as interesting). Children who could spell words correctly usually detected one more syllable than those who did not. In another study, Ehri and Wilce (1980), asked fourth graders to segment words into phonemes. These researchers selected seven words whose spellings had an extra letter. For example, pitch, which contains a T which could be pronounced if the reader were aware of the letter. All words were matched with another word matched phonemically but which did not have the extra letter (pitch/ rich). Children frequently detected the extra phonemes in words whose spellings represented those phonemes, but rarely did so when they were not represented. Scholes (1998) asked university students to delete sounds from words and then write the results. In some items, the sound to be 233

Children's analysis of oral and written words

deleted did not correspond to a letter, as in "What word do you get if you take the 3rd sound away from the word 'faxed'?". When the deleted sound corresponded to a letter, 90% of the responses were correct. In the other condition, only 12%. These research results suggest that differentiating oral and written language might involve having several different kinds of linguistic units in mind at the same time, and finding out which might be convenient to use for different tasks. Although awareness and use of these different types of units seem to be hierarchically ordered, developmental growth may not be a mere substitution of responses, but rather having the means to progressively construct new linguistic units and knowing in which contexts certain units are more relevant than others. Writing as a research variable

Several assumptions and hypothesis arise from what we have said so far. We agree with Olson and Torrance (1991: 266) when they say that "writing 'fixes' language in such a way as to make it an object of reflection". We have conducted research to explore the effects of writing on the analysis children can make of words. Our hypothesis was that writing can "objectivize" language in two related, but different, ways. (a) First, the physical presence of writing may give relevant information even to the pre-literate child. The number and types ofletters, for instance, may lead children to find, in speech, the number of parts that are included in the word as a whole. (b) Second, Ferreiro, and Teberosky (1982) have shown that pre-literate children have different hypothesis about how writing works, and have described the evolution of these conceptualizations that eventually lead to the comprehension of alphabetic script. Children at different moments of this development establish diverse relationships between letters and sound units. We expect, therefore, that children's performance in the tasks we have given them will vary according to their writing development. (c) Third, although phonological awareness studies have categorized children's responses to the tasks as incorrect and correct, we believe that phonological awareness develops in close relationship with writing development. It might be useful to analyze children's incorrect responses when they attempt to analyze oral or written words during their development. (d) Fourth, incorrect responses may not all be equivalent. Differentiating between different kinds of incorrect responses lead to a better understanding of children's acquisition processes. We examined these hypothesis in two studies which shall be reviewed briefly. 234

Sofia A. Vernon

Study 1:

One hundred Spanish-speaking kindergartners (mean age, 5 years, 8 months) who had not had any formal reading or writing instruction were first given a writing task, using Ferreiro and Teberosky's (1982) procedure: Children were asked to write six different common nouns. These words were the following: mariposa (butterfly), camote (sweet potato), pepino (cucumber), sopa (soup), tuna (a popular cactus fruit) and pan (bread). Children wrote the words without a model. Both the production process and the interpretation of the product were taken into account. This task was used mainly to classify children into different levels of development, as described by Ferreiro and Teberosky (1982), Ferreiro (1991) and Vernon (1993). These levels, from more to less advanced, were: 1. Alphabetic writings: Children make systematic phoneme-letter correspondences, although their spellings may not be conventional. These children are very novel writers, and have an incipient knowledge of conventional spellings. 2. Syllabic-alphabetic writings: Children use an apparently mixed system of writing. Sometimes they represent each syllable with a letter, but they also start representing intrasyllabic-syllabic units when they write. As all items (except the monosyllable) were composed of the most common type of syllable in Spanish (CV), the result was a mixture of representations of syllables and phonemes such as MAIOSA for mariposa (butterfly). This kind of writings is commonly identified in English as invented spellings. 3. Strict syllabic writings with the use of pertinent letters (PL): Children systematically use one letter for each spoken syllable. In addition, they use a pertinent letter for most of the syllables they represent. That is, they write the vowel or (less frequently) the consonant that belongs to the syllable. For instance, a typical syllabic writing for mariposa (butterfly, a four-syllable word) would be AIOA. Consonants are sometimes used when the name of the letter is the same as the syllable. Children usually count the number of syllables before writing or say each syllable aloud before writing it. 4. Strict syllabic writings without the use of PL: These children also make a strict one-to-one correspondence between the oral syllables and the letters they write. However, the letters they use are still not pertinent, in the sense that any letter can represent any given syllable. For example, the word camote (a three-syllable word) could be spelled EIA. 5. Pre-syllabic writings: At this level, children make no attempts to establish letter-sound correspondences of any kind, either when they write or when they are asked to read their own written productions. That is, they "read" the target word in the written string without analyzing it. Needless to say, the letters they use are not pertinent for the word they are trying to represent. An example of this kind of writing is OliAOT for the word pepino (cucumber). 235

Children's analysis of oral and written words

Twenty children from each writing level were individually interviewed. All children were given two different deletion tasks. One of them follows the usual procedure: Oral words are presented as stimuli, and children were asked to delete the first sound. Two examples were given first. The interviewer would say "I will say a word, and you say the same word, taking the first sound away, like this: pato .... ato". After the examples, two trials with feedback were used. The experimental items followed. The other task used written stimuli: Children were shown, one at a time, cards with printed words (composed with upper-case letters, type Times New Roman, size 24). The interviewer showed the card, and checked if the child was able to read it. Only some children at the alphabetic writing level could read some of the words. If the child was unable to read the word, the interviewer would read the card aloud saying, "In this card it says __," then covered the first letter with a piece of paper, leaving the rest of the word visible. For the two examples, she would then say, "I covered the first letter. Now, it says __". She would then show again the uncovered word, said the word aloud, then covered the first letter, and produced the result orally. For the two trials with feedback and all the experimental items, she would show the word, read it aloud, cover the first letter and ask, "What does it say now?". Half the children started with the oral word task, and then proceeded to the written word task. The other half had the reverse order. Items for both deletion tasks had the same initial phoneme or letter, and the same syllabic structure (see appendix 1). Three kinds of word-nouns were included: CV-Cv, CVC and V-Cv. These included the most common syllabic structures and patterns for content words in Spanish. Results show that the more advanced conceptualizations children have of their writing system, the greater the chance they have of giving correct answers in both deletion tasks. That is, their writing level makes a substantial contribution. Also, for all children, the presence of written stimuli enhances the production of correct responses. So, for example, children at the alphabetic level get 99% of correct responses for the written word task, whereas they get only 73.1 % of correct responses for the oral word task. The least advanced children, on the other hand, get less than 2% correct answers for both tasks. This seems to point out that phonological awareness develops, that there is a strong relationship between writing development and phonological awareness, and that the presence of writing in phonological awareness tasks enhances the child's ability to solve the task we used. Results suggest that writing influences how even pre-literate children hear and think about words. (See Table 1). Study 2 In another study (Vernon 1997, Vernon and Ferreiro 1999), 54 kindergarten children were also given first a writing task in order to classify them according to their writing level. The writing levels were the same as the ones presented in the previous study, except an extra level (initial syllabic writings) was added, in 236

Sofia A. Vernon

Table 1 Percentages of correct answers for both deletion tasks (oral Vs stimuli) across writing levels Writing Level

Written Words

Oral Words

Alphabetic

99.0%

73.1%

Syllabic-alf.

84.3%

34.4%

Syllabic with the use of PL

55.3%

19.0%

Syllabic without PL

26.0%

18.4%

l.7%

l.3%

Pre-syllabic

order to mark the transition between presyllabic and strict syllabic writings without the use of pertinent letters. Thus, the levels of writing were the following, from more to less advanced: alphabetic, syllabic-alphabetic, strict syllabic with the use of pertinent letters, strict syllabic without the use of pertinent letters, initial-syllabic and presyllabic. In initial-syllabic writings, children write without trying to establish letter-sound correspondences. However, once they have written a word, they try to justify the number of letters they have used, by trying to "read" their productions in a syllabic way. As they have not tried to control the number of letters beforehand, they cannot usually match the syllables they verbalize with the letters they have written. Thus, the letter-sound correspondences they establish are asystematic. Sometimes they read one letter per syllable, and sometimes they read several letters in the written string for a syllable. The 54 children were divided into six groups, based on their writing level, nine children in each. Words in the writing task were common nouns with CVand CVC syllables (the most common in Spanish): mariposa (butterfly), gusano (worm), venado (deer), perico (parrot), toro (bull), sapo (toad) and pan (bread). Children were given two phonological awareness tasks. In the first one, they were asked to segment words orally. The interviewer gave the children a set of cards with drawings, without any printed letters or words. The child was asked to say the names of these pictures "in little bits," so that the experimenter would guess the word. The game started with the experimenter explaining that words could be said in different "bits," and that some were easy to guess, whereas some were very difficult. To show this, she would give several examples, using syllabic, intrasyllabic-syllabic and phonological segmentations of common words. For example, she would say: "This is a Ikaxal (box); if I say Ika-xa!, it is very easy to guess. But if I say Ika-x-a/, it is not so easy to guess. But if I say Ik-a-x-al it is very hard to guess. You have to say the words in little bits so I have a real hard time guessing. But you have to say all the bits for each". 237

Children's analysis of oral and written words

In the other task, children were given, one by one, some written words, which were read aloud by the interviewer. She would then ask the children to say the word "in little bits," one for each letter. The interviewer would give two examples. Of course, in this case, each letter represented a phoneme. In both tasks, the interviewer would try to get the most analytical answer from the child. In the guessing game, for example, if the child made a syllabic segmentation, she would say: "That's very easy. Try making it harder for me to guess". In the written task, some children jumped some letters, or said only one segment for several letters. The interviewer asked the child to try to point to, and verbalize a "bit" for each letter. Whenever children gave more than one response for the same item, only the most analytical one was taken into account. Half the children started with the first task, and then proceeded to the second task. For the other half, the order was reversed. Appendix 2 shows the words used as stimuli for both tasks. As we have pointed out earlier, in the first task, several models of performance (syllabic, intrasyllabic-syllabic and phonological) were provided by the experimenter, whereas for the written word task, the model that was provided was phonological. In spite of the models, phonological responses were by no means the most predominant response. Table 2 shows the percentage of phonological segmentations for each writing level in both tasks. Data suggest that phonological segmentations can be accomplished only by children at the most advanced levels of writing (Syllabic-alphabetic and alphabetic), and that written stimuli enhances this kind of analysis. Since children segmented the words in various ways in both tasks, we decided to classify the segmentations they produced. Responses for both tasks were the following: (a) No segmentation: In spite of feedback and attempts on the part of the interviewer, children would say the complete word.

Table 2 Percentages of phonological responses at each writing level for oral guessing game and written words

Writing Level

Written Words

Oral Words

Alphabetic

88.9%

64.4%

Syllabic-alf.

55.6%

34.5%

Syllabic with the use of PL

3.7%

0

Syllabic without PL

0

0

Initial syllabic

0

0

Pre-syllabic

0

0

238

Sofia A. Vernon

(b) Syllable-based responses: Children divide the word into syllables, such as me-sa, or ga-to. For monosyllabic words, children divide the word into two parts, using the same vowel as the syllabic nuclei for both parts, as in so-of or pa-an. In this paper, we have grouped two other syllable-based responses in which the child isolates one of the phonological segments. This only happens before or after he/she produces the complete syllable. Children can do partial isolation of a vowel (as in me-e-sa, or, in monosyllabic words, so-o-ol), or, less frequently, partial isolation of a consonant (as in me-s-sa or so-I-ol). In these two cases, children do not seem to be able to segment a phoneme without using the whole syllable as a starting point. Syllabic segmentations appear very frequently, especially at the pre syllabic and all the syllabic levels. In relative terms, 65% of these three responses correspond to normal syllabic segmentations, 26.3% to partial isolation of a vowel and 8.7% to partial isolation of a consonant. (c) Complete isolation of the phonemes of the second syllable, or the coda2 of a monosyllabic word: Children seem to be able to segment phonemes in Spanish for the first time towards the end of the word. Segmentations of this kind are, for instance, me-s-a or so-I. (d) Onset-rime and isolation of phonemes in the first syllable: Children divide monosyllabic words into the two main constituents of the syllable (as in s-ol), or the first syllable of bi-syllabic words, as in m-e-s-a. (e) Full phonological segmentation: This response shows the ability to segment words into phonemes, as is m-e-s-a or s-o-I. Again, in this study, we found that the strongest variable is the children's level of writing and, secondly, the type of stimuli (pictures or written words) used for the task. Table 3 shows the percentages of responses for these 5 types of response. Children progress gradually towards phonological segmentations as they achieve a more advanced level of writing and tend to give more analytical responses when they can see a written word than when they operate with oral stimuli. As Tolchinsky and Teberosky (1997) have pointed out, the syllable seems to be the preferred segmentation unit for Spanish-speaking children. Although syllable-based segmentations tend to decrease as writing level advances, children at the alphabetic and syllabic-alphabetic levels still use it, especially when they work in an oral context. Also, onset-rime segmentations cannot be found frequently at any level. Rather, the isolation of the coda or the phonemes in the second syllable seems to be a transitional response that mediates between syllabic and phonological responses. This suggests that some differences in phonological awareness can be found in different languages. In conclusion, pre-literate and novel writers seem to be extremely sensitive to the presence of writing in the phonological awareness task presented here. In fact, writing gives cues to children about how many "parts" or "bits" they can find in a given word, and of the kind and the position of the sounds in each word. Also, and more important, the children's conceptualizations of the writing 239

Children's analysis of oral and written words

Table 3 Percentages of responses at each writing level (0 = oral guessing game; W = written cards). Row percent = 100% for each task

Stimuli

0

Alph.

3

2

1 W

0

W

W

0

W

0

W

1l.5

2.5 18.4

3.7

5.7

4.9 64.4 88.9

26.4

4.9 29.9 32.1

6.9

7.4 34.5 55.6

l.2 77.9 56.8 1l.6 25.9

7.0

Syl-alph.

2.3

syl PL

3.5

Syl without PL

7.1

89.4 77.8

Initial Syl.

9.2

9.9 88.5 85.2

Pre-syl.

0

5

4

30.7 18.5 69.3 80.3

12.3

3.5 18.5

3.7

2.3

l.2

3.7

3.7

l.2

system (what we have called "writing level") determines the way they analyze oral language. On the whole, phonemes seem to be segmentable only for children with the most advanced levels of writing (alphabetic or syllabic-alphabetic), especially when they have writing as a visible aide. When analyzing an oral word, even advanced children tend to make more syllabic-based segmentations (at least in Spanish). Developmentally, phonological responses, probably attained through the acquisition of writing, do not eliminate other, more holistic responses. Children are able to analyze speech in a variety of ways at any given time of their development. Although they come to understand some levels of analysis later (such as segmental analysis), that does not stop them from using the units that have proven to be useful at earlier moments. Syllabic segmentation, for instance, continues to be part of the strategies Spanishspeakers use, even when they become conventional writers. Children seem to progressively have a clearer notion of which kind of analysis pertains to the written or oral domains. The knowledge children acquire through orality and writing seem to interact in order to accomplish several different ways of analyzing their language for different purposes. Finally, results suggest that phonological awareness and writing development interact from the start. Our hypothesis is that learning to write and to interpret print creates cognitive problems that can only be solved through the acquisition of knowledge about how speech is represented in an alphabetic script, and the scope and limitations of such a notation. These studies, as well as others referred to at the beginning of this paper, may indicate that children start working with many different kinds of linguistic (phonological, morphological and syntactic) units from the moment they start trying to understand what written marks represent, and that the written form provides a guide to their analysis of the spoken form.

240

Sofia A. Vernon

References Ball, E., & Blachman, B. (1991) Does phoneme awareness training in kindergarten make a difference in early word recognition and developmental spelling? Reading Research Quarterly, XXVI, 1, 49-66 Bradley, L., & Bryant, P. E. (1985) Rhyme and reason in reading and spelling, Ann Arbor, MI: University of Michigan Press. Bryant, P. E., & Bradeley, L. (1980) Why children sometimes write words which they do not read. In U. Frith (Ed.), Cognitive processes in spelling (pp. 355-370), London: Academic. Carrillo, M. (1994) Development of phonological awareness and reading acquisition: A study in Spanish Language. Reading and Writing: An Interdisciplinary Journal, 6, 279-298. Cary, L., & Verhaeghe, A. (1994) Promoting phonemic analysis ability among kindergartners. Reading and Writing. An Interdisciplinary Journal, 6,251-278. Defior, S., & Tudela, P. (1994) Effect of phonological training on reading and writing acquisition. Reading and Writing: An Interdisciplinary Journal, 6, 299-320. Ehri, L., & Wilce, L. (1980) The influence of orthography on readers' conceptualization of the phonemic structure of words. Applied Psycholinguistics, I, 371-385. Ehri, L. (1984) How orthography alters spoken language competencies in children learning to read and spell. In J. Downing & R. Valtin (Eds.), Language awareness and learning to read (pp. 119-147), New York: Springer Verlag. Ehri, L. C. (1998) Grapheme-phoneme knowledge is essential for learning to read words in English. In J. L. Metsala & L. C. Ehri (Eds.), Word recognition in beginning literacy (pp. 3--40), Mahwah, NJ: Lawrence Erlbaum. Ferreiro, E. (1978) What is written in a written sentence? A developmental answer. Journal of Education, 160,4,25-39. Ferreiro, E. (1991) Psychological and epistemological problems on written representation of language. In M. Carretero, M. Pope, R. Simons & J. I. Pozo (Eds.), Learning and instruction: European research in an international context, 3, 157-173. Ferreiro, E. (1997a) La nocion de palabra y su relacion con la escritura. [The notion of word and its relationship with writing]. In R. Barriga & P. Butragueno (Eds.), Varia Lingiiistica y Literaria. 50 afios del CELL, Mexico: El Colegio de Mexico. Ferreiro, E. (1997b) The word out of (conceptual) context. In C. Pontecorvo (Ed.), Writing development: An interdisciplinary view (pp. 47-59), Amsterdam & Philadelphia: John Benjamins. Ferreiro, E., & Gomez-Palacios, M. (1982) Analisis de las Perturbaciones en el Proceso de Aprendizaje de la Lecto-Escritura. Fasciculo 3. Las Relaciones entre el Texto y la Imagen. [Analysis of perturbations in the reading and writing learning process. Volume 3. The relationships between text and images]. Mexico: SEP-OEA. Ferreiro, E., & Teberosky, A. (1982) Literacy before schooling, Exter, NH: Heineman Education. Ferreiro, E. & Vernon, S. (1992) La distincion palabra/nombre en ninos de 4 y 5 arros [The distinction between word/name in four and five year-old children]. Infancia y Aprendizaje, 58, 15-28. Goswami, U., & Mead, F. (1992) Onset and rime awareness and analogies in reading. Reading Research Quarterly, 27,2, 153-162. Harris, R. (1986) The origin of writing, London: Gerald Duckworth.

241

Children's analysis of oral and written words

Harris, R. (1995) Signs of writing, London: Routledge. Jimenez-Gonzalez, ]. E., & Ortiz, M. (1993) Phonological awareness in learning literacy. Cognitiva, 5, 2, 153-170. Jimenez-Gonzalez,]. E. (1997) A reading-level match study of phonemic processes underlying reading disabilities in a transparent orthography. Reading and Writing. An Interdisciplinary Journal, 9, 23-40. Kirtley, C., Bryant, P., Maclean, M., & Bradley, L. (1989) Rhyme, rime and the onset of reading. Journal of Experimental Child Psychology, 48,224-245. Liberman, I. Y., Shankweiler, D., Fischer, F. W., & Carter, B. (1974) Reading and awareness of linguistic segments. Journal of Experimental Child Psychology, 18, 201-212. Liberman, I., Shankweiler, D., & Liberman, A. M. (1992) The alphabetic principle and learning to read. In D. Shankweiler, D. & Liberman, I. Y. (Eds.), Phonology and reading disability, (pp. 1-33), Ann Arbor, MI: The University of Michigan Press. Lundberg, I., Frost, J., & Petersen, O. P. (1988) Effects of an extensive program for stimulating phonological awareness in preschool children. Reading Research Quarterly, 23, 263-284. Maclean, M., Bryant, P. E., & Bradley, L. (1987) Rhymes, nursery rhymes and reading in early childhood. Merril-Palmer Quarterly, 33, 255-282. Mann, V. A. (1991) Phonological abilities: Effective predictors of future reading ability. In L. Rieben & C. A. Perfetti (Eds.), Learning to read: Basic research and its implications (pp. 121-133), Hillsdale, NJ: Lawrence Erlbaum. Morais, J., Cary, L., Alegria,]., & Bertelson, P. (1979) Does awareness of speech as a sequence of phones arise spontaneously? Cognition, 7, 323-331. Morais,]. (Ed.) (1995) Literacy onset in romance language. Reading and Writing: An Interdisciplinary Journal. 7, 1. Muter, V., Hulme, c., Snowling, M., & Taylor, S. (1998) Segmentation, not rhyming, predicts early progress in learning to read. Journal ofExperimental Child Psychology, 71, 3-27. Nation, K., and Hulme, C. (1997) Phonemic segmentation, not onset-rime segmentation, predicts early reading and spelling skills. Reading Research Quarterly, 32(2), 154-167. Olson, D., & Torrance, N. (1991) Literacy and orality, New York: Cambridge University Press. Olson, D. (1994) The world on paper: The conceptual and cognitive implications of writing and reading, London: Cambridge University Press. Olson, D. (1996) Towards a psychology of literacy: on the relations between speech and writing. Cognition, 60, 83-104. Read, c., Zhang, Y., Nie, H. & Ding, B. (1986) The ability to manipulate speech sounds depends on knowing alphabetic writing. Cognition, 24, 31-44. Rosner, J., & Simon, D. (1971) The auditory analysis test: An initial report. Journal of Learning Disabilities, 4, 384-392. Scholes, R. (1998) The case against phonemic awareness. Journal of Research in Reading, 21, 3, 177-188. Seymour, P. H., & Evans, H. M. (1994) Levels of phonological awareness and learning to read. Reading and Writing. An Interdisciplinary Journal, 6, 221-250. Share, D., & Stanovich, K. (1995) Cognitive processes in early reading development: Accommodating individual differences into a model of acquisition. Issues in Education, 1, 1, 1-57. Stahl, S. A., & Murray, B. (1998) Issues involved in defining phonological awareness and its relation to early reading. In ]. L. Metsala & L. C. Ehri (Eds.), Word recognition in beginning literacy (pp. 65-87), Mahwah, NJ: Lawrence Erlbaum.

242

Soffa A. Vernon

Tolchinsky, L., & Teberosky, A. (1997) Explicit word segmentation and writing in Hebrew and Spanish. In C. Pontecorvo (Ed.), Writing development: An interdisciplinary view (77-97), Amsterdam & Philadelphia: John Benjamins. Torneus, M. (1984) Phonological awareness and reading: A chicken and egg problem? Journal of Educational Psychology, 76, 1346-1358. Treiman, R. and Zukowski, A. (1996) Children's sensitivity to syllables, onsets, rimes and phonemes. Journal of Experimental Child Psychology 61, 193-215. Treiman, R. (1998) Why spelling? The benefits of incorporating spelling into beginning reading instruction. In J. L. Metsala & L. C. Ehri (Eds.), Word recognition in beginning literacy (pp. 289-313), Mahwah, NJ: Lawrence Erlbaum. Vachek, J. (1989) Written language revisited, Amsterdam & Philadelphia: John Benjamins. Vernon, S. (1993) Initial sound/letter correspondences in children's early written productions. Journal of Research in Childhood Education, 8, 1, 12-22. Vernon, S. (1997) Escritura y conciencia fonol6gica en niiios hispano-parlantes [Writing and phonological awareness in Spanish-speaking children]. Infancia y Aprendizaje, 81, 105-120. Vernon, S., & Ferreiro, E. (1999) Writing development: A neglected variable in the consideration of phonological awareness. Harvard Educational Review, 69, 395-415. Yopp, H. K. (1988) The validity and reliability of phonemic awareness tests. Reading Research Quarterly, XIII, 2, 159-177.

Appendix 1 Items for both phonological awareness tasks Oral Words

Written Words

FOTO

FOCA

lAGO

lATA

CANA

CAMA

SOPA

SOPE

PllA

lUPA

META

PINO

lUNA

MESA

PAN

PAZ

SAL

SOL

GOl

GAS

ECO

ESO

OSA

OlA

HUMO (pronounced /umo/)

UNO

AJO

ARO

HIPO (pronounced (/ipo/)

IVA

243

Children's analysis of oral and written words

Appendix 2

Words used as stimuli in the second study Guessing game (oral task):

Bisyllabic CV-CV nouns: luna (moon), foca (seal), taco (taco), mesa (table), sapo (toad) and gato (cat). Monosyllabic: gis (chalk), pez (fish), sol (sun), and pan (bread). Written word task:

Bisyllabic CV-CV nouns - sapo (toad), peso (name of Mexican coin), mesa (table), luna (moon), taco (taco), gato (cat). Monosyllabic: (pan (bread), luz (light), and gis (chalk). Notes Each syllable can be devided into two main parts: The onset, which is constituted by one or more consonants that precede the vowel, or syllabic nucleous, and the rime, which is the vocalic nucleous and any sounds that may follow it. For example, in cat, the single sound /k/ constitutes the onset, and the rest is the rime. 2 The syllable can be devided into an onset and a rime. The latter is constituted by an obligatory nucleus, and in some cases, a final consonant, which is called the "coda".

244

16 Young children's "clever misunderstandings" about print Janette Pelletier

Those of you whose work or personal lives take you into the world of young children's early experiences with print will need no convincing that the scratches, letter-like figures, and iconic forms they produce prior to schooling are in fact manifestations of tremendous competence. Some of the rest of us might claim that children go through "stages" of scribble development, but we likely all feel confident in saying that once in school, they will eventually learn to "write" correctly. In this chapter I will provide more evidence to the growing body of literature that demonstrates how children's paths to literacy begin much earlier than at school age and how these paths are filled with remarkably clever misconceptions about print. During the preschool period, young children in literate societies begin to take notice ofliterate activities such as adults' reading and writing. Children participate in literate activities themselves when they listen to stories. A family member may point to printed marks on the page under the picture and say something related to that picture. On other occasions, perhaps a child may be given a writing tool and may begin to make marks on a page. These experiences are shared cultural activities between adult and child, yet interpreted quite differently by the two parties. The literate adult has his perspective on the connection between the "very words" (Olson 1996: 149) that appear in print and what is meant by those words. The adult knows that the print stands for the words - what is said. In contrast, the young child knows that the printed words stand for the picture - what is meant (Olson and Pelletier 2000). What is said and what is meant are not incongruous. Thus, the adult and child sharing the story understand the same meaning. However, the young child sees something different in the print than does the adult. The child sees marks that relate to meaning - perhaps in the print she looks for the shape of the object, perhaps the size, or perhaps a name of someone or something tied to the story. The child's challenge in her mind, is to find patterns in the print that somehow tie to the message conveyed by the adult's words. In this chapter, I aim to discuss this challenge from the perspectives of various interpretive frameworks including psychogenetic, historical and socio-cultural.

245

Young children's "clever misunderstandings" about print

The assumption is that these frameworks are not mutually exclusive, but rather, are considered complementary. First, we might consider a psychogenetic explanation that builds on the notion of cognitive-developmental shifts in understanding. That is, the ways in which children derive patterns in the print may be related to stages in development. Children appear to pass through remarkably consistent stages in their understanding of print, across languages and across cultures (see for example, Ferreiro and Teberosky 1982, Ferreiro and; Vernon 1992, Levin and Korat 1993, Levin, Korat, and Amsterdamer 1996). The second way in which I aim to explain children's growing understanding of the relation between language, script in particular, and how children come to read and write in adult-like fashion, is by looking at at historical evidence. A compelling suggestion is that there are parallel stages between psychogenetic and historical writing development (Olson 1994), based on Harris' study of the history of writing (Harris 1986). In this chapter, I explore in more detail the ways in which the conceptions of print held by young English-speaking Canadian children show some evidence of this parallel development and I suggest that the "stages" may not be as clear as they were in history, but rather, show signs of overlap or transition. This overlap may be considered a stage in itself or considered a product of culture-specific or language-specific factors. The third explanatory basis is a socio-cultural one; that is, it examines developmental shifts from a predominantly contextual perspective. The patterns that children find in print may be influenced by the emphases given to the significance of print and to the features of print across cultural contexts. In this vein, one would argue that over the course of the preschool and early school years, the child gradually learns to disentangle the language itself from its meaning, a task that Vygotsky has claimed is very difficult to do, since language and thought are at once interconnected (Vygotsky 1962). This explanation contextualizes the child within the culture of literate institutions, a culture of documents, where text is "fixed," allowing it to be the subject of reflection (Olson 1994; Olson and Pelletier 2000). One of the issues to consider is the interplay among children's developmental level, their experiences with print, and the impact of schooling. This interplay, like Vygotsky's language and thought interplay, makes it difficult to pinpoint the contributions of the various factors on children's movement from misconceptions to correct conceptions about print. However, we acknowledge that schooling in general and literacy in particular bring a new level of awareness about print because they both treat words as linguistic entities, or objects of reflection (Olson 1994). Furthermore, literacy is not only about print, but is also about cultural practices. Children grow up in what Brockmeier (2000) terms the "symbolic space of literacy". These three explanatory bases, that is psychogenetic development, historical parallels, and cultural transmission, will be explored in this chapter in relation to the empirical work I have been carrying out over the past few years. The work has focused on how young children move from childhood misconceptions to

246

Janette Pelletier

adult-like conceptions of print and has examined from various perspectives how this move may be interpreted. I will just make a few more general comments to set the tone. Think about your own or your friend's two or three-year-old child who picks up a marker and "writes" something on a page. In the child's mind, "writing" is not yet perceived as correct or incorrect, but rather as a means to record or to communicate a message or tally, or perhaps simply as an activity for which intentions are unknown on an explicit level. Why would a child think to perform such an activity? One could perhaps make a case that as in the evolution of writing there are no concepts such as sentences, words or phonemes prior to the onset of writing. Thus, any "message" or "tally" is simply a direct one-to-one correspondence between object and its graphic representation, such as we see in history, for example, a stone drawing of one's herd of oxen. This idea seems strikingly similar to children's initial attempts to write - two horses are just that, two graphic marks each representing a horse (see Figure 1). Reciprocally, then, reading is the one-to-one interpretation of the graphic representation. Two marks on a page are just that, two things, events, or names. As Gunther Kress described so well, in alphabetic scripts, letters just stand for pictures as in a tableau (Kress 1997). They are not a model of linguistic properties. In children's development, as in history, once a writing system is known, at least in an alphabetic writing system, the concept of linguistic structures such as words and phonemes are known as a model of speech (Olson 1994). In order to participate in the culture that uses this model, one must learn the code. In what follows, I consider first some of the important work that has pioneered the area of young children's early literacy development and discuss this work in light of the theoretical perspectives introduced earlier. I then turn to my own ongoing research in the area and offer theoretical interpretations that pertain to this work. Finally, the chapter concludes with an applied focus - specifically, on the relation between early literacy development and schooling.

Figure 1

247

Young children's "clever misunderstandings" about print

What young children bring to the reading experience

Emilia Ferreiro, her many colleagues and followers comprise one group of researchers whose seminal work has spurred much of the current discussion and the fascinating research on the psychogenesis of literacy. To briefly review, they document stages of understanding of print that children bring to reading. In the very early stages, children distinguish drawing from writing, cleverly knowing that drawing can not capture something important that letters can capture, for example, one's name. Later, children may employ strategies to examine whether writing looks different depending on the context, such as whether it links to the picture or to the spoken word. For example, one would need to have more letters for a big object, an older person or a group of things. Children look for a one-toone correspondence between the graphic sign and the sound segments. This stage gives way to a focus on syllables, the stage Ferreiro terms, "the syllabic hypothesis" stage, in which children attempt to explain print by the number of syllables in the utterance. Thus, in the example, "Mommy has the keys," the word "has" would be read "me" because it would be considered the second syllable of "Mommy". Gradually, in the face of evidence that becomes too difficult to refute, such as the conflict between the syllabic hypothesis and the graphic strings they encounter, children crack the alphabetic code and learn that printed words represent what is said and not the objects or the events therein (Ferreiro and Teberosky 1982). Ferreiro acknowledges a psychogenetic explanation of this development: "Children possess conceptualizations about the nature of written language long before the intervention of systematic instruction. These conceptualizations are not arbitrary but possess an internal logic, making them understandable from a psychogenetic point of view" (Ferreiro and Teberosky 1982: 89). Following a similar psychogenetic approach employing Piaget's clinical inquiry methodology, an explanation of early reading "stages" was given by Yaden, Smolkin, and MacGillivray (1993). They employed a case study approach with a small number of preschoolers who were shown storybooks by their parents. When asked about the representational content of the letters, children appeared to move through three stages: associating letters with a pictured object, associating letters with selected oral words and finally associating letters with initial sounds of the printed word. Kress deals less with internally-driven stage-like progressions that children make in reading and sees the child as one who "transforms" text based on his or her own experience (Kress 1997: 58). Thus reading is making internal signs by transforming the form of the text into the meaning; form mirrors meaning. What sparks the interest underlying this transformation? According to Kress, the interest is context-driven. Is one is dealing with logographic or alphabetic scripts? Are there cues available in the environment? Does the child understand that there is a maker of the sign and if so, what the sign maker is trying to communicate? Children simultaneously draw on whatever is "available" to make sense of print,

248

Janette Pelletier

for example, the visual shape of the letters, the layout of the representation between speech and writing, the significance ofthe page itself and where the words are located, the correspondence between either the letters and the sound or the drawing and the sound, and others (Kress 1997: 15). Children learn to participate in a literate culture through their own transformative means of representation. In their review, Schieffelin and Cochran-Smith (1984) stress cross-cultural influences on the importance given to literacy and literacy experiences in early childhood. For the Kaluli people of Papua, New Guinea, literacy has been tied to one's life mission. If this does not include Christianity and work in missionaries, then literacy is not seen as important for children. Similarly, when families immigrate to North America from places where early literacy experiences are not even known, their children must somehow come to understand that literacy is a valued enterprise in their new country, that they need to learn about reading and writing and that they need to transmit information about the new culture and its focus on literacy to their parents. In one case, children do not become literate because it is not pertinent to their lives. In the other case, children become literate once they enter the school system where there is a focus on becoming literate. These examples highlight socio-cultural and context-related explanations of literacy development. What young children bring to the writing experience Just how random is scribble? What exactly do children "mean" when they begin to put print on a page? Do they "mean" what they want to "say?" These are a few of the questions that guide this section. Although Gunther Kress (1996, 1997) might well attribute all marks on a page to a child's attempt to make meaning, others may be less inclined to do so. Consider the example of Emily's two written productions, both carried out on the same morning in her junior kindergarten class (4-year-old kindergarten) (see Figures 2 & 3). In the case of Figure 2, Emily was observed that morning busily filling blank pages with row upon row of letters. Many of the letters Emily chose to write are those found in her own first name, MIL Y, and last name D SON. Was Emily writing a message, or was she simply practising writing letters? In the case of Figure 3, Emily was "practising writing like Mommy". Was Emily making meaning? One of my former graduate students, now a junior kindergarten teacher, asks the same kinds of questions about the children's writing experiences in her class. She observes, 'Another example is when I ask children what they have written, and they won't tell me - they either laugh it off, "I don't know" or they tell me to read it. Sometimes I will read what they have written, literally, and they laugh about the words they have made up. Sometimes I draw it out of them and see if there was a story brewing, but usually there isn't. For example, one girl (4 years, 11 months) wrote:

249

~~re3

250

Janette Pelletier

RERENEN ENYYVR ENYENEI I asked her what she had written. She just smiled and walked away. But she was writing. Another example of a girl writing but not necessarily making meaning was a girl who wrote random letters, spaced apart as "words," told her mother that she couldn't read it back to her because, "I don't have my glasses on" (she doesn't wear glasses). So she knows that writing is supposed to say something but that her writing didn't say something (Edward 2000). However, these examples are not to suggest that children do not make meaning when they first begin to write. Consider Figures 4 and 5. In both Figures 4 and 5, children were playing at the home centre in their junior kindergarten class. The child who wrote the message in Figure 4 was writing out an emergency message for 911. I knew this because I asked him what he had written. In Figure 5, the child was creating a shopping list before she and

MESSAGE FOR M____~~~~~~~~~-----WHILE YOU WERE OUT

M____________________________ OE~

__________________________

PHONE NO RETURNED YOUR

TEl.EPHCNED

CAU.

CALLED 10 SEE YOU wr.NTS 10 SEe YOU

PLEASECAU.

I

W1U. CAU. ..........

MESSAGE: _ _ _ _ _ _-,,~f"'tr----:l~ .......- - -

00 \ DATE ________

TIME - - ; ; o o c - - - - -