The City and the Sign: An Introduction to Urban Semiotics 9780231892544

Citation preview

The City and the Sign

THE CITY AND THE SIGN An Introduction to Urban Semiotics Edited by M. Gottdiener and Alexandras Ph. Lagopoulos

Columbia University Press New York 1986

Library of Congress Cataloging-in-Publication Data Main entry under title: The City and the sign. Bibliography: p. Includes index. 1. Cities and towns—Psychological aspects. 2. Semiotics. 3. Signs and symbols in architecture. I. Gottdiener, Mark. II. Lagopoulos, Alexandras Ph. H T 1 5 3 . C 5 6 1986 307.7'6'oi4i 86-2277 ISBN 0-231-06146-3

Columbia University Press N e w York Guildford, Surrey Copyright © 1986 Columbia University Press All rights reserved Printed in the United States of America This book is Smyth-sewn and printed on permanent and durable acid-free paper.

Book design by Ken Venezio

To my mother, Arlene —M. G.

To Karin, Kyveli and Panayiotis —A. Ph. L.

CONTENTS

Acknowledgments Introduction M. Gottdiener and Alexandros Ph. Lagopoulos I. Theoretical Perspectives

23

1. For a Topological Semiotics Algirdas ]ulien Greimas 2. Function and Sign: Semiotics of Architecture Umberto Eco 3. Semiology and the Urban Roland Barthes 4. Introduction to the Semiotics of Space Pierre Boudon 5. Speech and the Silence of the City Raymond Ledrut II. Applied Urban Semiotics

135

Section A: Theory in Practice

137

6. For a New Semiological Approach to the City Richard Fauque

viii

Contents

7. Urbanism and Semiology Françoise Choay

160

8. Semiotic Urban Models and Modes of Production: A Socio-Semiotic Approach Alexandros Ph. Lagopoulos 9. Culture, Ideology, and the Sign of the City M. Gottdiener Section B: Case Studies

176 202

219

10. The Images of the City Raymond Ledrut 1 1 . Urbanism in Question Françoise Choay 12. Semiological Urbanism: An Analysis of the Traditional Western Sudanese Settlement Alexandros Ph. Lagopoulos 13. Recapturing the Center: A Semiotic Analysis of Shopping MaUs M. Gottdiener 14. Re-writing of a City: The Medina of Tunis Pierre Boudon

219 241

259

288 303

Glossary

323

Index

329

Acknowledgments

The editors would like to thank the following for permission to reproduce or translate copyrighted material: Editions Anthropos, Paris, for permission to translate into English selections from Les images de la ville (1973), by Raymond Ledrut; "Parole et silence de la ville," Espaces et Sociétés 9 (July 1973), by the same author; and, "Pour une nouvelle approche sémiologique de la ville," Espaces et Sociétés 9 (July 1973), by Richard Fauque. The editorial board (Groupe Expansion) of L'Architecture d'Aujourd'hui, Paris, for permission to translate "Sémiologie et urbanisme" by Roland Barthes (December 1970-January 1971). Georges Borchardt, Inc., New York, for permission to translate selections from L'urbanisme: Utopies et réalités. Une anthologie (Paris: Editions du Seuil, 1965), by Françoise Choay; "Pour une sémiotique topologique," in Sémiotique et sciences sociales (Paris: Editions du Seuil, 1976), by Algirdas Julien Greimas; "Introduction," Communications 27 (Paris: Editions du Seuil, 1977), by Pierre Boudon. Professor Françoise Choay, Institut d'Urbanisme, Université de Paris VIII, for the permission to publish her "Urbanism and Semiology" which first appeared in translation in Ch. Jencks and G. Baird, eds., Meaning in Architecture (London: Barrie and Jenkins, 1969). Professor Umberto Eco for the permission to publish selections from "Function and Sign: Semiotics of Architecture," first published in English in Via 1 (1973), a version of which

x Acknowledgments also appeared in G. Broadbent, R. Bunt, and Ch. Jencks, eds., Signs, Symbols, and Architecture (Chichester, New York, Brisbane, Toronto: Wiley, 1980). Hutchinson Publishing Group, Ltd., London, for "Semiological Urbanismi An Analysis of the Traditional Western Sudanese Settlement," in P. Oliver, ed., Shelter, Sign, and Symbol (London: Barrie and Jenkins, 1975), by Alexandras Ph. Lagopoulos. Mouton Publishers, Amsterdam, for "Semiotic Urban Models and Modes of Production: A Socio-Semiotic Approach," Semiotica (1983), 45(3/4), by Alexandras Ph. Lagopoulos; and, for permission to translate into English "Réécriture d'une ville: La médina de Tunis," Semiotica (1978), 22 (1/2), by Pierre Boudon. The editors would also like to thank Professor Algirdas Julien Greimas, Ecole des Hautes Etudes en Sciences Sociales, Paris; Professor Umberto Eco, Università di Bologna, Italy; and Professor Thomas A. Sebeok, Indiana University, Bloomington, Indiana, for their warm interest in our project, as well as Editions du Seuil for their help. Translation is a taxing job in itself, and the translation of highly sophisticated semiotic texts from French into English is especially demanding. This chore was undertaken by Dr. Lily Stylianoudi and Ms. Kitty Rouanet, who carried out the initial translations; Mr. D. J. Connolly, who acted as language consultant; and Professor Karin Boklund-Lagopoulou who supervised and revised the translations; all of them we thank warmly. The editors enjoyed considerable support from Columbia University Press and would like to thank, in particular, Maureen MacGrogan and David Diefendorf for their help with the manuscript. Initial work on this project was aided by a grant from the Riverside Foundation, Riverside, California. M. Gottdiener wishes to thank the Council for the International Exchange of Scholars for support as a Fulbright Research Fellow that enabled collaboration in Greece. Finally, we wish to thank Ms. Eleni Mavromatidou for the preparation of the manuscript, Mr. Konstantinos Victor Spyridonidis for the drawing of the figures which were reproduced for this edition, and Ms. Elefthenia Deltsou, who prepared the Index.

INTRODUCTION M. Gottdiener and Alexandras Ph. Lagopoulos

SEMIOTICS AND THE UNIVERSE OF SIGNS Among European analyses of culture, especially contemporary modalities of thought, semiotics represents one of the principal approaches. In the United States bits and pieces of the semiotician's lexicon, such as "signifier," and "text," have crept into common academic usage. Much of the important recent work in semiotics, however, has remained inaccessible to Anglo-American researchers because it has appeared in languages other than English. Our reader helps remedy this deficiency by presenting in translation some of the key European works published in the growing field of urban semiotics, that is, the inquiry into the social signification of urban forms, or, more generally, forms of settlement space, such as villages, tribal camps, and the like. In addition, the editors have prepared the present introductory chapter on urban semiotics, which helps to situate this approach within the context of other, more conventional modes of inquiry. Quite clearly, the study of signification in the built environment is not an object of analysis unique to urban semiotics. Other ap-

2 Introduction proaches more familiar to the Anglo-American scientific public, such as cognitive geography, cognitive psychology, and the now dated socio-cultural trend of human ecology have long addressed the topic of environmental perception and even symbolic life within the city. As we shall demonstrate below, however, urban semiotics represents a unique and perhaps improved way of studying the social role of signification in settlement space. Before proceeding with a discussion of the semiotic approach to space, it is necessary to convey a sense of how this field has developed, especially through certain contentious issues addressed by continental semioticians. The foundation of all semiotics is the concept of the sign. It is most commonly defined following the work of the Swiss linguist Ferdinand de Saussure. For Saussure the sign is composed of a signifier, or its cultural manifestation—physically, verbally, and so on, i.e., through some social act—and the signified—the underlying concept which the signifier is asked to convey. Saussure's approach differs from the semiotics founded by his contemporary, the American Charles Sanders Peirce, a distinction which cannot be pursued here (see Eco 1976; Hervey 1982). Yet both definitions of the universe of signs lead to the following components which we shall consider the common background against which the domain of semiotics should be defined: (1) Semiotics studies signs relating to the recognition of the social and natural environment of an individual and of his/her internal world. These signs constitute the denotative languages which relate, not only to natural languages, but also to other cultural sign systems, such as spatial or fashion systems of signification. (2) Systems of signification can be understood and elaborated upon through metalinguistic operations, that is, through access to a secondary level of discourse possessing the ability to comment on the object language of the primary level. These metalanguages range from purely or mainly subjective modes of discourse to objective, formal, or scientific languages. (3) Systems of signification themselves encompass denotative signs as well as the values socially ascribed to them, i.e., the connotative codes of culture. This property is similar to the distinction made by the formal study of ideology—a topic to be discussed below.

Introduction 3 In short, the universe of signs includes: the non-physiological part of perception; conception; scientific modes of discourse; and the value systems, or the socially constituted world views of social subjects, which are a function of social interaction. Certain researchers have also extended the field of semiotics to include the behavior of animals and plants, resulting in the emergence of such fields as zoo-semiotics. This is not the place to discuss the legitimacy of such approaches. We shall restrict ourselves to the observation that the epistemological and theoretical problems created by such extensions compel us to insist on the anchoring of semiosis solely within human societies. Thus, semiotics overlaps with sociology, social anthropology, phenomenology, human geography, Marxist theories of ideology, psychology, and the study of religions. For urban semiotics in particular, material objects are the vehicles of signification, so that the symbolic act always involves some physical object as well as social discourse on it. In the case of urban semiotics these objects are the elements of urban space, for example, streets, squares, buildings, and facades. Semiotic analysis can also be extended to include codes of property ownership, written texts of planning, the plans of designers, urban discourse by the users of the city, and real-estate advertising. Despite the general ground of agreement concerning the universe of signs, important differences exist among semioticians and these can even be said to constitute separate schools of thought. Linguistics, from which structuralism and European semiotics emerged, studies the signs of the natural languages. This domain, therefore, addresses itself to all three systems of signification referred to above, given that linguists are interested in language as a whole. Thus, some linguists see semiotics as merely a special case of the study of language itself. A contrasting approach was initiated by Peirce, who conceived of semiotics as the domain of logic (cf. American journal of Semiotics 1983). Peircian semiotics involves what the Danish linguist Louis Hjelmslev calls "scientific metasemiotics," an approach belonging to the fields of logic and mathematics rather than to what Hjelmslev considered as the field of "semiology" (Hjelmslev 1961: 120). A third approach follows from the work of the French literary critic Roland Barthes (1967). He was interested in the social ideology underlying literary texts and other cultural forms, especially those

4 Introduction associated with modernity. For Barthes, semiotics studies the processes of connotation or the social significations of cultural objects relating to ascribed values. Finally, Greimas defines semiotics as the theory of all systems of signification, thereby encompassing all the preceding approaches (Greimas and Courtis 1979). From our perspective the main issue addressed by the differences among these separate schools of thought and the debate touched off between them focuses on the question of whether or not there is a general, all encompassing approach to systems of signification. At issue is the possibility of unifying in one and the same domain the whole range of inquiry into the phenomenon of signification, ranging from the study of linguistics through that of cultural systems, such as body gestures and art, to the study of the formal metalanguages of all sciences, that is, to logic and epistemology. Thus, the issue framed by the articulated differences in approaches to semiotics in general becomes whether or not a unified perspective can be found to integrate at once all the social sciences and psychology, on the one hand, and logic and epistemology, on the other, by some general theory of semiotics - and whether biology, ecology, and ethology can be included in the synthesis, as certain semioticians suggest. The search for a unified approach to semiotics, however, is not our concern here. Furthermore, this overarching issue defining the field does not prevent more specialized inquiries into the properties of specific cultural systems of signification. In fact, the larger task may even be facilitated by the latter effort. Our approach starts from component three above and asserts that semiotic systems contain, besides denotative codes, socially constructed values or ideologies which operate as connotative codes inseparable from denotation. This position implies that the object of semiotic analysis is ideology, or rather ideological systems, which are observed empirically and which are culturally specific. Our choice to define the semiotic approach in this manner, however, bears an implication pertinent to the issue of grand synthesis discussed above. In particular, the application of semiotics to the study of ideology asserts that the objects of analysis can also include the culturally specific value systems operating behind all kinds of metalanguages—religious explanations, philosophical discourses, and even scientific

Introduction 5 theories. Thus, our approach drags semiotics back into a relation with logic and epistemology but in a manner which is different from the attempts at synthesis discussed above. In our case we conceive of the semiotic inquiry in a sense similar to that of Barthes, namely, as a debunker of all apparently and socially sustained "objective" modes of discourse. The orientation of semiotic inquiry toward the study of ideology corresponds to what Greimas (see paper 1 in this volume) has called "socio-semiotics" which includes both the study of "mythical epistemologies," namely, the ideologies with respect to signs, and the various social ideologies and languages (Greimas and Courtés 1979). The latter involves social value systems and modes of discourse which can be identified as historical, social products. Following Greimas' distinction, then, there appear to be two separate approaches to urban semiotics. The first is a purely semiotic one and focuses on spatial systems of signification proper which are considered independently from their social contexts. The second perspective, the one closer to that of the editors, links such systems with their social contexts through the study of the ideology incorporated in them. Socio-semiotics, therefore, studies both systems of denotation and metalinguistic systems in relation to the culturally specific systems of connotation operating behind them. The specific epistemological position we propose asserts that the socio-semiotic approach should not be limited within the universe of signs, as Greimas believes, but involves as well the material, social processes accounting for systems of signification. We also think that codified ideology which acts as the system of connotation precedes the denotative system in urban space, i.e., denotative codes in the urban milieu derive from connotative codes. In what follows, then, "socio-semiotics" will refer to the approach of the editors which extends Greimas' work to link up with non-semiotic, social processes. By way of contrast, while the object of Kevin Lynch's research referred to as the "image of the city" is a semiotic object of analysis, or rather could be one, it is limited by his perspective to the denotative level. Thus Lynch ignored the connotative level, a subject to be discussed below. In the opposite sense, Françoise Choay has focused her work on a typology of ideological discourses on urban space. This is also a semiotic object of analysis, however it

6 Introduction is confined to the connotative level. In both these cases, therefore, a more integrated approach is called for in order to understand the process of signification in the urban environment. The need for a socio-semiotic approach to settlement space can be understood best by comparison with the mainstream perspective which currently dominates thought on this subject, namely, cognitive geography. This field borrows heavily from the early work of Lynch and has united with techniques employed by cognitive psychologists. Let us briefly turn our attention to the work of Lynch and its subsequent development in cognitive geography before proceeding to our discussion of urban socio-semiotics. COGNITIVE GEOGRAPHY AND MENTAL MAPPING The Legacy of Lynch. When considering the urban environment as a meaningful entity, most American social scientists seem immediately to recall first the classic study by Lynch (i960) on the "image" of the city (see e.g., King and Golledge 1978:241). Only later would other equally significant early studies come to mind, such as those by Firey (1945), Tolman (1948), Wright (1947) and the early work of Wohl and Strauss (1958). Consequently, it might be surprising to learn just how limited Lynch's approach really is. To be sure the reduction of urban imagery to the physical form of the city and the stressing of legibility and spatial elements, such as paths, nodes, edges, and so on, represents something of a watershed in environmental planning and even a mode of architectural analysis of the city. Yet, the conversion of this first step into a kind of tradition in planning launched without benefit of serious auto-criticism has been a mixed blessing. More importantly, Lynch's attitude of arriving at the meaning of the urban experience through the acquisition of mental maps has become the cornerstone of cognitive geography and represents therefore the principal means at present for researching signification in the city. It has long been recognized that Lynch's methods themselves are questionable, such as the sampling and questionnaire techniques used to acquire images of the city (Porteous 1977). But even these concerns leave the essential epistemological nature of cognitive mapping intact. It is this foundation for approaching the study of signification within the city which must be called into question.

Introduction 7 There is little argument that the work of Lynch has led to a more human approach to urban design; one that explicitly recognizes the role of users in fathoming urban space. Yet the underlying premises of this tradition require re-examination. In particular, cognitive mapping research relies on a methodological individualism which accepts unquestioningly intra-subjective pictures of the environment as the basis of urban behavior. Thus, cognitive approaches arrive at the signification of the city through the percqjtion of its inhabitants rather than their conception. The socio-semiotic approach just introduced is in some respects fundamentally opposed to cognitive imaging methods precisely because the crucial object of analysis is conception rather than perception. In the tradition launched by Lynch the understanding by city inhabitants of signification in the urban environment is reduced to a perceptual knowledge of physical form. Furthermore, the physical features which Lynch came to stress, namely the now famous fivefold distinction of paths, edges, nodes, and so on, reduce the use of urban environments to the activity of movement. Because Lynch focused on studying signification associated with getting about the city, he ignored other meaningful aspects of the environment. This critique is expressed well by Ledrut (1973:25-28, 1 1 1 - 1 3 ) . According to him Lynch's approach is strongly psycho-biological, since for Lynch the image is part of the adaption to the environment. Ledrut finds that this approach does not differentiate enough between humans and animals caught in a maze. In contrast Ledrut searches for the symbolic, connotative level of the image and asserts that it is from this level of signification that denotation derives (a position which we have also adopted). In this manner the relation of people to the city goes beyond perceptual recognition and introduces the role of ideology. In short, the inhabitant of the city does not adapt to an environment, rather, residents play a role in the production and use of the urban milieu through urban practices. Over time the weaknesses of limiting the image of the built environment to Lynch's elements have progressively surfaced. On the one hand, the central reliance of Lynch on imageability has been successfully questioned by a variety of researchers (Goody et al. 1971; Sieverts 1967; and Smith 1974). On the other, research has downgraded elements such as paths to roles which support but do

8 Introduction not act as stimuli of behavior (Golledge and Zannaras 1973:88). Finally, at least one element, that of "edge," seems to be principally an artifact of cognitive mapping methods themselves rather than an active category of spatial perception (Pocock and Hudson 1978:51). Although analysts following Lynch note the continued importance of his ideas for planning theory (Appleyard 1970), it has become clear that actual benefits are slight. As Pipkin observes (1983:53), The implied modifications of urban structure are relatively modest, amounting to boundary marking, legible routing, "land marking" and the like. These proposals are unexceptionable, but, in the context of American urban problems they are hardly incisive.

Despite such weaknesses research in the Lynchian tradition has uncovered some important means by which inhabitants of the city organize their behavior. Chief among these is the realization that conceptual stimuli in the environment play a more fundamental role than mere formal perception, so that physical forms are assigned certain significations which then aid in directing behavior. Urban structures act as stimuli because they have become symbols and not because they support behavior by facilitating movement. Thus we can say that the image of the city is a conceptual rather than perceptual one. As Pocock and Hudson contend (1978:33), Man [sic] then, is very much a symbolic animal. Moreover, from the point of view of image formation if, as has been shown, man symbolizes the environment which he engages, it is logical to assume, given acquaintance or involvement, that those symbols themselves may become stimuli to which he then responds, this is a further reason why the physical stimuli per se do not exert an influence on the image directly proportional to their objective prominence. Accordingly, the concept of imageability cannot be defined in specifically physical terms.

In sum, the perceptual approach to the image of the city falls short as a means of analyzing the symbolic nature of that image and its results point to the position adopted by socio-semiotics. Contemporary approaches in the tradition following Lynch have uncovered that the focal conceptual element in the urban image is that of use, especially functional use (Harrison and Sarre 1975, 1976; Smith 1974; Moore 1983; and Appleyard 1970). That is, the conception of the uses of urban places by inhabitants provides one of the compo-

Introdtiction 9 nents of the image of the city. Despite this realization, the main approach to signification in the city, namely cognitive geography, has left this feature relatively undeveloped. In contrast, the semiotic approach to the study of signification in objects begins with what Barthes refers to as the sign-function, i.e., the understanding that the signification of an object is closely related to its function or use, so that the object use is itself converted into the signified of the object (cf. Eco paper 2 in this volume). We adopt here the early position of Barthes (1967: 106-7) with regard to the genesis of signification (see also Krampen 1979). Before examining how the sociosemiotic approach proceeds from this fundamental realization, let us summarize briefly the limitations of cognitive geography, most especially its reliance on the research technique of mental mapping. This method persists in hampering mainstreamers' efforts at studying the city image. The Mental Map as the City Image. Research on mental mapping is the dominant mode of investigating environmental knowing. At its core it relies on intra-subjective judgments related to perception rather than conceptualization. Because people understand the urban environment largely on the basis of what they do there, i.e. in functionalist terms, and on the basis of their own symbolic worlds, rather than imageable perception per se, cognitive geography asks the wrong questions and researches the wrong data. The underlying assumption of cognitive mapping research is that a mental image of the environment provides the support for behavior in the city. Consequently, the understanding of urban activities comes through the identification of mental maps. As Pocock and Hudson (1978:14) suggest, the importance of this work is its conceptual break from previous approaches to urban behavior which hypothesized the presence of rational, utilitarian mechanisms of individual decision-making. Research on environmental cognition has produced findings which contribute to our understanding of the subjective component relating to socio-spatial behavior. Yet, that very knowledge is constrained by the decision to subsume the individualistic epistemology of cognitive psychology. As Pipkin observes (1983), it is hazardous at the least to rely on conceptual constructs as primary data which are products of

io Introduction intra-subjective memory and consciousness. More specifically, there are serious problems presented in the inter-subjective comparison of cognitive psychological research, due to the fact that commonly held images of the city are not the simple aggregate of individual memory. According to Pipkin, there is no sufficient basis for moving from the cognitive maps of individuals to hypothetical group modes of conceptualizing space. In short, the cumulative result of years of cognitive research substantiates the social basis for urban imaging. This is sometimes referred to as the "constructivist assumption" (Moore 1983:21). There is sociologically speaking no individual image of the environment. Rather, the city is imaged differently by different people according to the group experience of social life. As Tuan (1977:5) suggests, for example, humans are bom with only a rudimentary notion of time and space. The conception of the environment is a social product which must be learned. For this reason the important variables in understanding the composition of commonly held urban images are those which grasp the nature of social experience. According to Moore (1983:36), Recent evidence from sociology and social geography indicates that different social and cultural groups conceive of the environment in entirely different ways, that is, that to explain differences in environmental cognition, three of the most important constellations of explanatory variables may be social group values, lifestyle and culture.

Also, Lagopoulos (1973:115-20) has concluded that the denotative urban image and its connotations, the symbols attached to it, change as a function of social groups and classes. The social basis of the environmental image and its variation according to aggregate categories of group organization has been well documented for differences between classes (Fried and Gleicher 1961; Rainwater 1966; and Duncan 1973), and between different subcultural groups within the city (Downs and Stea 1973, 1977; Golledge and Ruston 1976; and Horton and Reynolds 1971). These results call into question the epistemological foundation of cognitive mapping studies because the latter centers on individual data divorced from an examination of the group processes which produce and sustain certain specific urban images.

Introduction 1 1 At present the limitations of the cognitive paradigm flowing from the constraints above are widely acknowledged. In fact, critics assert that this approach does not fulfill its intended research objective, namely, the explanation of urban behavior, because the study of mental maps has become progressively divorced from the social basis of conception. Cognitive research has established connections between macro and micro levels of analysis only with extreme difficulty. The fundamental ideological bias of mental mapping is its unwillingness to recognize that the primary data of its research is itself an ideological product. The image of the city in cognitive maps is not the city itself, nor is it some reflection of fundamental innate processes of spatial perception, because we know that most of the ways in which people perceive space are socially learned and experientially based. The cognitive map is so much a product of social interaction that even individuals living near each other in the very same neighborhood will hold different conceptions of their area as a product of separate social networks (King and Golledge 1978:251). The mental map, therefore, depicts the imaginary city. It is itself a part of the urban ideology. This imaginary picture of what each individual believes to be the surrounding social reality constitutes, according to Althusser, the basis of the ideological representation of social process. In terms similar to Althusser's concept, Strauss captures the crucial fact of the ideological mode of environmental perception (1961:6), Not only does the city-dweller develop a sentiment of place gradually, but it is extremely difficult for him even to visualize the physical organization of his city, and, even more, to make sense of its cross-currents of activity. Apparently, an invariable characteristic of city life is that certain stylized symbolic means must be resorted to in order to "see" the city.

Mental mapping of the environment is, in short, an example of ideological representation par excellence. The failure to recognize this feature severely limits the ability of researchers in cognitive geography to explain the dominating symbolic processes taking place within the city. In sum, cognitive geography locates the production of spatial meaning within the minds of individuals. As we have seen, over time research now points to the dependency of urban imaging on func-

u Introduction tional uses of places and socially conditioned judgments about urban space. This suggests that signification is derived from cultural codes originating outside the individual and emerging from processes of social interaction—both areas of investigation which are generally ignored by cognitive approaches. As Harrison and Howard (1972) have suggested, cognitive mapping research relies solely upon individual accounts of signification. More importantly, through the elicitation of maps this understanding is reduced to the identification of physical features which are divorced from their social symbols (Pipkin 1983:66; and Pocock and Hudson 1978:32). This functions to conceal rather than extract the symbolic richness of urban life anchored as it is within an environmental milieu. Thus, the socio-semiotic approach to culture, in contrast to cognitive geography, explicitly assumes that signification is a social product dependent for its sustanance on the interaction among individuals in society and between social groups and cultural codes. The ascription of signification to urban objects frames a distinct problem which has only been resolved weakly by mainstream approaches. The strength of the semiotic perspective lies in its explicit recognition of the need to resolve the issues inherent in such an inquiry before the analysis of the city image can proceed. In short, as Krampen observes (1979:2), urban semiotics begins by asking the question: What kind of meaning is connected to the city and by what kinds of mechanisms? . . . i.e., what conditions must be fulfilled to enable one to attribute communicative functions to objects in general and to architectural objects in particular?

Let us, therefore, consider the semiotic and socio-semiotic approaches to space in more detail. URBAN SOCIO-SEMIOTICS Semiotics and Socio-Semiotics. For semiotics every architectural or urban object is transformed, at the level of denotation, into a signified of its own function. But beyond its conceptualized functional use, the object has another function as well, which is symbolic. Thus, it signifies on a second level also, that of connotation. As

Introduction 1 3 indicated above, it is this second aspect which, for Ledrut (see papers 5 and 10) dominates the conception of the city. Furthermore, and according to the socio-semiotic perspective this time, connotative codes are social products produced by groups and classes involved in urban practice. Through such distinctions, the socio-semiotic approach avoids the pitfalls of cognitive geography, such as psychologism, behaviorism, and individualism, which also appear in certain branches of semiotics. The presence of connotative codes, the dependence of both connotative and denotative signification on cultural codes, and the significant role of both these levels with respect to urban space suggest that the urban milieu is a rich source of signification and symbols indissolubly connected with urban life. Historically, the first approach to the semiotics of space was that of architectural semiotics. The latter has proven over the years to contain several serious limitations, which prevent it from becoming a model for the semiotics of urban space. Among these limitations we can enumerate the misunderstanding of semiotic theory and the unwarranted extension of semiotic operations onto non-semiotic social processes; the slavish attachment to the linguistic model; an identification of the architectural language with its poetic function alone; the opposite reduction of signification merely to the denotative level; and the neglect of the articulation between semiotic and non-semiotic processes in the social life of the city. This last point is indicative of a detachment from sociology which is quite typical in semiotics. Thus, architectural semioticians will often assert a monolithic view of city inhabitants by ignoring the social stratification of signification and by clustering together finance capitalists, real estate developers, the working class, and teenage graffiti sprayers as the same group of inhabitants in the study of the conception of the city. These limitations are not inherent to architectural semiotics, but follow from specific historical phases in the development of the field of semiotics as a whole, and are present to an extent in urban semiotics as well. Thus, the development of socio-semiotics proceeds in part through a critique of earlier semiotic approaches to space. At this point we should not like to give the impression of being sociosemiotic imperialists. We believe that both architectural semiotics and cognitive geography can offer useful insights for the study of spatial signification. We only hope that the shock of an encounter

14

Introduction

with socio-semiotics might help to restructure these perspectives. Alternatively, we have no doubt that urban semiotics can benefit from these fields as well. For example, despite the sociological aspirations of urban semiotics, the field has remained so far without the benefit of much contact with social theories developed by other social sciences, which is not the case for human geography. Also, the refined and highly developed quantitative techniques of cognitive geography have yet to penetrate the research repertoire of semiotic inquiry (some notable exceptions, such as Ledrut, paper 10 in this volume, and Krampen 1979, do, however, exist). Thus, it is hoped that this reader will initiate a three-way dialogue between mainstreamers, pure semioticians, and socio-semioticians. Having stated our position vis-à-vis other approaches as clearly as possible, we can now acknowledge that the difference between semiotics and socio-semiotics remains centered on the latter's need to overcome the limitations of the former. More to the point, sociosemiotics, as we understand it, is a materialist inquiry into the role of ideology in everyday life. It seeks to account for the articulation between semiotic and non-semiotic social processes in the ideological production and conception of space which are not only materialistically based but are also determined by the nature of specific historical modes of production. Going back to the origins of semiotics, we can detect some major traits which were to delimit the scope of this domain up to the present moment. Ferdinand de Saussure's structural linguistics, for example, relied heavily on langue, language as a synchronic system. For Saussure, the diachronic transformation of langue is due to parole, that is, a certain individual (cultural) practice, while language performance is ruled by the synchronic system of language. By focusing attention on the linguistic system as such to the neglect of everyday cultural process, however, he treated language as autonomous from reality. Thus, Saussure was not only a positivist but an idealist as well and the only reality we can know in his system is the reality of our minds (Saussure 1971). Such idealism along with empiricism became the foundation for the theory of signs proposed by the other founder of semiotics, Charles Peirce (see American Journal of Semiotics 1983). In fact, for Peirce, as indicated above, semiotics is a branch of formal logic.

Introduction

15

There is a continuous thread in European social thought from Saussurean structural linguistics to structuralism, and from these inquiries to semiotics as we know it today. Claude Lévi-Strauss, the founder of structuralism, is at once a neo-positivist and a neo-Kantian idealist. He considers all social systems as systems of communication and thus that the whole of social anthropology is a branch of semiotics. Lévi-Strauss believes that there may be isolated social facts exterior to communicative systems, but, like Saussure, he focuses on semiotic structures, which are for him the only structures that exist and which are synchronic. For Lévi-Strauss diachronic transformations are ruled by semiotic factors. Thus, structural anthropology denies the material part of social anthropology, a fact which is also clear in his analysis of settlement space. According to Lévi-Strauss, the settlement is a complex semiotic text obeying universal semiotic structures. European semiotics has not diverged much from these precursors over the years. It has remained focused on an idealist inquiry into synchronic semiotic structures, despite some attempts at incorporating weak notions of process into static, categorical modes of analysis. Recently, European semiotics also underwent an evolution and fragmentation similar to that encountered by structuralism as a whole. It came into contact with psychoanalysis, and adopted a Marxist vocabulary, interpreting Marxist concepts very loosely in this process, as in the work of Lévi-Strauss or Julia Kristeva. Yet, in the end, it acquired an increasingly idealist orientation. The culmination of this intellectual inquiry at present is deconstructionism, which denies altogether any reference of semiotic systems to an exterior, non-semiotic reality. Thus, the paradigm founding deconstructionism has revealed itself to be related to Peircian semiotics and idealist philosophy (cf. Norris 1982). The formal semiotic approach to urban space (see for example Boudon, papers 4 and 14 below) differs from the editors' socio-semiotics, because it limits analysis to the discovery of generative grammars underlying spatial structures. Urban semiotics then becomes the study of spatial structures derived from internalized grammars of patterns and designs which become externalized through semiosis. The semiotic perspective studies these spatial grammars independently from non-semiotic social processes. Thus, for example,

16 Introduction Greimas (paper 1 in this volume) considers space as both produced and conceived through the mechanism of culturally determined spatial languages. These are studied by what he calls "topological semiotics," one branch of which is urban semiotics. Greimas places analytical emphasis on systems of signification pertinent to the production and conception of space divorced from their non-semiotic, sociological context. The semiotic content of the built environment is thus studied through the isolation of spatial grammars or languages. In the approach of Greimas, therefore, the non-semiotic dimension of society is neutralized and treated in an internalized aspect contained within the subjective, the mental, and the discursive (see also Boudon, papers 4 and 14; Barthes, paper 3; Fauque, paper 6; and Choay, papers 7 and 11). We have seen that all three contemporary modes of European social thought, i.e., structural linguistics, structuralism, and semiotics, approach the study of structures and systems of communication by neglecting the relation between systems of signification and the non-semiotic, material processes of the social world. In this way they encourage or integrate themselves into idealism. As already indicated, the articulation between semiotic and material processes is, for us, the focal point of analysis in socio-semiotics. In order to explain our approach as distinct from the mainstream European tradition, it is necessary to specify explicitly the concept of the sign which harmonizes best with the requirements of socio-semiotics as we have defined it. As indicated above, our perspective differs from the way in which Greimas conceptualizes sotio-semiotics, but, in the interests of controlling the explosion of new terms, we have retained the same label. In what follows we shall also explain the manner by which research applying our perspective can be carried out. Socio-Semiotics and the Urban Sign. A socio-semiotic analysis of urban space consists of a series of combined aspects of study. They relate to the decomposition of spatial signs according to the model of Hjelmslev (1961). For Hjelmslev, the semiotic sign is a "bifacial unit" consisting of signifier and signified which accounts for the way in which sources of signification grounded in more general cultural circumstances are united with material artifacts. This approach fits well the socio-semiotic perspective because it identifies explicitly the

Introduction 17 social linkages to ideological systems and the non-semiotic environment. Thus, both Hjelmslev's "expression" and "content," which correspond respectively to the signifier and the signified in Saussure's terminology, can be broken down further into two levels corresponding to their "form" or to their "substance." The "substance of the content" of the sign consists of a level of non-codified signification situated in the larger culture which makes the sign itself possible. The signifieds themselves, then, would be the "form of the content," and would consist of a codified ideology manifested in space in the case of urban semiotics. At the level of the "expression" of the sign, in contrast, the substance would be that part of material space which bears signification, which is not a sign itself, and of which semiotics studies those qualities that make it a vehicle of signification. Finally, the "form of the expression" would be for urban semiotics the spatial signifiers or the morphological elements used in the social construction and organization of settlement space (see figure A). In short, and according to Hjelmslev's model, the urban sign can be decomposed into four separate levels conforming to the form and substance of both the content and expression, with only the formal components being what is usually understood as relevant to the analysis of the sign by other semioticians. Following this model, signifiers are not only attached to signifieds but both signifiers and signifieds are themselves connected relationally and functionally to their respective cultural and material contexts. From our perspective, and following the distinction of Ledrut (see Ledrut, paper 5 in this volume), urban space is not a text but a "pseudo-text," because it is produced by non-semiotic processes as well as semiotic ones and because there is not always a sender in the historically conditioned built environment. A socio-semiotic analysis of an urban sign system or "pseudo-text" would then procede as follows. On the one hand, observational data would be collected on both the substance and form of the expression. In the first case (substance), a description of material urban space invested by signification would be obtained, while in the second (form), attention would be given to the specific spatial elements which are the vehicles of signification. On the other hand, cultural research is required to document the

18 Introduction Figure A. A decomposition of the city sign Signified Signifier

Content Expression

Substance Form Form Substance

Non-codified ideology Codified ideology materialized in space Morphological elements Material objects of space

forms and substance of the content. Such a task requires, firstly, attention to historically and culturally established signification, realized through research into the general cultural traits of the society within which the settlement space is embedded. Secondly, considerable case study research is required to document the codified ideology structuring the signifieds of space. The socio-semiotic approach is divided further according to two distinctions which we refer to as the semiotic "production" and "consumption," or "conception" of settlement space. Spatial organization is a social product and the production of space is governed by social processes characteristic of specific social formations. These involve political and economic as well as cultural activities, in short, non-semiotic as well as semiotic processes. Through the differentiation of material elements social practices mold and interpret physical space according to codified ideology. Socio-semiotic analysis at the level of the production of space, therefore, involves identifying the role of connotative and denotative codes as mediators of spatial production. The built environment is thus the end result of the intersection between non-semiotic and semiotic processes which are mediated through spatial ideologies. At a different level of society, however, it is necessary to speak of the users of space and not the social groups of producers. Ideological connotative and denotative codes are brought into play again when the users of space enter the scene. In fact, most of what is known by mainstreamers as the "image of the city" can be placed within this approach, that is, the semiotic conception of space. Here it is quite clear that both the semiotic production and the conception of space merge with the operations of traditional social science, because they are studied through interview and survey analysis. In particular, the inquiry into the conception of space analyzes its multi-coding by the socially stratified groups of synchronic users. In summation, semiotic systems are generated by symbolic social

Introduction

19

relations which, in turn, constitute only one aspect, the mediating dimension, of urban social practices. Semiotic systems do not appear "by themselves." They are rooted in the material practices of society. That is, we assert that systems of signification follow from semiotic practices and these, in turn, follow from the generalized social, political, and economic practices of society. Thus, if we want to explain urban semiotic systems rather than merely describe them, it is necessary to specify the manner by which such systems are constituted socially. It is this articulation between the semiotic and the non-semiotic social processes which constitutes the socio-semiotic approach theoretically and methodologically (see Boklund-Lagopoulou and Lagopoulos 1984).

CONCLUSION Attention to differences among semiotic systems due to and explained by differences in the social position of the corresponding social agents is the hallmark of the socio-semiotic approach when contrasted wtih conventional urban and architectural semiotics. Differences in the ideologies of the agents of production have repercussions on the semiotic systems of production, and further, these differences can also be related or linked processually to differences in the semiotic conception of urban space. Such contrasts are discovered through the empirical work associated with the extended Hjelmslevian model of the sign discussed above. From the socio-semiotic perspective, the material processes of social production and reproduction, and the ideological function of systems of signification are the dominant factors in the appearance and structuration of semiotic systems. There is no doubt that the production and functions of urban space are not the same as those of architecture, although they share points in common. These differences, due, for example, to their differing social nature or to the political framework, lead to differences in the ideologies of the agents of production. In turn, these differences have repercussions on the semiotic systems of cultural production. They also lead to differences manifested in the conception of space by the stratified population of users. In short, by attention to the kinds of questions discussed above, we arrive at the central aspiration of this introduc-

20 Introduction tìon, the specification of a (relatively) independent field of inquiry, urban semiotics, investigated by a materialist, sociologically aware semiotics, a socio-se miotics. This reader presents many of the major contributions applying a semiotic approach to settlement space which are based on the two different epistemologica! positions outlined above and which have played their part in the development of urban semiotics and urban sodo-semiotics respectively. All of thè papers are preceded by introductory remarks presenting the authors, their work, and their respective approaches to semiotics in a critical and comparative manner. In this way, the texts contained within serve to clarify the main issues raised here as they illuminate the application of semiotics to the analysis of signification in settlement space. The whole constituted by our introduction and these major papers, along with their introductory texts, will aid the reader in obtaining a coherent understanding of the emerging field of urban semiotics, and, in addition, should help to stimulate the further development of the field. It is also hoped that the present collection, the first reader of its kind, will provide a means for the influence of semiotics on other nonsemiotic modes of urban analysis. The reader is divided into three parts. The first part, "Theoretical Perspectives" includes texts reviewing and commenting upon the semiotic approaches to urban space. We have included papers of a theoretical nature which raise the main issues of urban semiotics and offer important theoretical insights involved in the constitution of the field. The second part, section A, "Theory in Practice," presents theoretical discussions relating in a direct manner to empirical material or aiming at the analysis of such material; most of the papers here are of a socio-semiotic orientation. Section B, "Case Studies," provides a series of empirical studies and practical applications of method deriving from both of the perspectives we have discussed. Nearly all of this work is accompanied by theoretical discussions. Due to the number of new terms specific to semiotic analysis, a glossary is also provided for the reader's benefit. In addition, the further exploration of the issues found here can be pursued through the comprehensive bibliographic materials supplied by the individual papers. In short, the present reader is a compilation

Introduction

21

w h i c h represents the complete spectrum of work in urban semiotics through the medium of many of its most important papers.

REFERENCES American journal of Semiotics (1983), 2 (1-2). Special issue on Charles Sanders Peirce, edited by Kenneth Ketner. Appleyard, D. 1970. "Styles and M e t h o d s of Structuring a City." Environment and Behavior 2: 100-17. Barthes, R. 1967. Elements of Semiology. N e w York: Hill and Wang. Boklund- Lagopoulou, and A . Ph. Lagopoulos. 1984. "Social Structures and Semiotic Systems: Theory, Methodology, Some Applications and Conclusions." I n T . Borbé, ed., Semiotics Unfolding, 1: 431-38. The Hague: Mouton. D o w n s , R., and D. Stea. 1973. Image and Environment. Chicago: Aldine. 1977. Maps in Minds. N e w York: Harper and Row. Duncan, J. 1973. "Landscape Taste as a Symbol of Group Identity." Geographical Review 63: 334-55. Eco, U. 1976. A Theory of Semiotics. Bloomington: Indiana University Press. Firey, W. 1945. "Sentiment and Symbolism as Ecological Variables." American Sociological Review 10: 140-48. Fried, M., and P. Gleicher. 1961. "Some Sources of Residential Satisfaction in an Urban Slum." journal of the American Institute of Planners 27: 305-15. Golledge, R., and G. Ruston, eds. 1976. Spatial Choice and Spatial Behavior. Columbus: Ohio State University Press. Golledge, R., and G. Zannaras. 1973. "Cognitive Approaches to the Analysis of Human Spatial Behavior." In W . Ittleson, ed., Environment and Cognition. N e w York: Seminar Press. G o o d y , B., et al. 1971. City Scene. Research Memorandum 10. Center for Urban and Regional Studies, University of Birmingham, England. Greimas, A . , and J. Courtés. 1979. Sémiotique: Dictionnaire raisonné de la théorie du langage. Paris: Hachette. Harrison, )., and W. Howard. 1972. "The Role of Meaning in Urban Image." Environment and Behavior 4: 389-411. Harrison, J., and P. Sarre. 1975. "Personal Construct Theory in the Measurement of Environmental Images." Environment and Behavior 7: 3-58. 1976. "Personal Construct Theory, the Repertory Grid, and Environmental Cognition." In G. Moore and R. Golledge, eds., Environmental Knowing. Stroudsberg: D o w d e n , Hutchinson and Ross. Hervey, S. 1982. Semiotic Perspectives. London: Allen and Unwin. Hjelmslev, L. 1961, 1969. Prolegomena to a Theory of Language. Madison, WI: University of Wisconsin Press. Horton, F., and D. Reynolds. 1971. "Effects of Urban Spatial Structure on Individual Behavior." Economic Geography 47: 36-48.

22

Introduction

King, L., and R. Golledge. 1978. Cities, Space, and Behavior. Englewood Cliffs, N.J.: Prentice Hall. Krampen, M. 1979. Meaning in the Urban Environment. New York: Methuen. Lagopoulos, A. Ph. 1973. Structural Urbanism: The Settlement as a System (in Greek). Athens: Technical Chamber of Greece. Ledrut, R. 1973. Les images de la ville. Paris: Anthropos. Lynch, K. i960. The Image of the City. Cambridge: MIT Press. Moore, G. 1983. "Knowing About Environmental Knowing." In J. Pipkin, M. LaGory, and J. Blau, eds., Remaking the City, pp. 21-50. Albany, N.Y.: SUNY Press. Noms, Ch. 1982. Deconstruction: Theory and Practice. London and New York: Methuen. Pipkin, J. 1983. "Structuralism and the Uses of Cognitive Images in Urban Planning." In Pipkin, LaGory, and Blau, pp. 77-100. Pocock, D., and R. Hudson. 1978. ¡mages of the Urban Environment. London: Macmillan. Porteous, J. 1977. Environment and Behavior. Reading, Me.: Addison-Wesley. Rainwater, L. 1966. "Fear and the House-as-Haven in the Lower Class." journal of the American Institute of Planners 32:23-31. Saussure, F. de. 1971. Cours de linguistique générale. Paris:Payot. Sieverts, T. 1967. "Perceptual Images of the City of Berlin." In E. Brill, ed., Urban Core and Inner City, pp. 282-85. University of Leiden. Smith, P. 1974. The Dynamics of Urbanism. London: Hutchinson. Strauss, A. 1961. Images of the American City. New York: Free Press. Tolman, E. 1948. "Cognitive Maps in Rats and Men." Psychological Review 55: 189-208. Tuan, Y. 1977. Space and Place. Minneapolis: University of Minnesota Press. Wohl, R., and A. Strauss. 1958. "Symbolic Representation and the Urban Milieu." American Journal of Sociology 63:523-32. Wright, J. 1947. "Terrae Incognitae: The Place of Imagination in Geography." Annals of the Association of American Geographers 3 7 : 1 - 1 5 .

1 FOR A TOPOLOGICAL SEMIOTICS Algirdas Julien Greimas

Greimas is one of the leading figures in semiotics, and his general semiotic theory, founded on Saussure's and Hjelmslev's linguistics, is closely related to his structural semantics and narrative theory. Several theoretical clarifications are necessary to help the reader understand better Greimas' approach to urban semiotics. Greimas' theory is a comprehensive approach to social action from a semiotic perspective. It is therefore, a major contribution to social theory in general. For Greimas every social situation can be considered as a narrative. This narrative is, according to him, composed of utterances and regulated by a surface syntax, studied by an "actantial grammar." (Actant is Greimas' term for a formal role in a sequence of action, analogous to the position of subject or object in a phrase. The various structural positions of actants are filled by concrete actors engaged in these roles.) There are two forms of elementary utterances: the utterances of state, and the utterances of doing, referring to transformations between states. The elementary utterances are constituted Translated, from the French, as published in A.J. Greimas, Semiotique et sciences sociales (Paris: Seuil, 1976). This text formed the introduction to the collection of papers given at the colloquium on the Semiotics of Space organized by the Institut de l'Environnement, Paris, in May 1972, and published as Semiotique de Vespace (see Greimas 1974).

26

Algirdas Julien Greimas

by a relation between two syntactic functions or "octants," which are beings or objects related to an act. These utterances follow the modalities of having to "do" or "be," wanting, being able (to do or be), and knowing. The individual subject has a competence, which makes the act possible, while its performance is a "doing" producing utterances. The production of utterances presupposes the existence of enunciation, which regulates the passage from competence to performance. In short, Greimas' approach when applied to the city leads to the discovery of generative topological languages which link conception with action in relation to space. We have no doubt that this paper by Greimas is the most global approach to urban semiotics and extends into socio-semiotics. As becomes clear, the author transfers the concepts used for the analysis of narrative to that of social action. The author holds the idealist position that the referent, and by extension reality, is a metaphysical concept and that only the universe of signs exists. In this case, we would have to accept that material social relations, independent of the consciousness of social subjects, do not exist and that social action is reduced to the conception of that action in the minds of these subjects. On the other hand, we could object that the Greimasian model seems too general and a- historical to analyze concrete social processes. Furthermore, the acceptance of the existence of social reality and material social relations leads to the conclusion that social relations evolve both on the non-semiotic, material level and on the semiotic level, and that all the semiotic relations in society are inseparable from material relations (the latter being primary in Marxian analysis). We would also disagree with an attempt at reconciling Greimas' position with our own by, for example, accepting the reality of material relations but by considering them as reflecting the consciousness of the social subjects. This kind of approach is widely diffused in, for example, cognitive geography. We believe, on the contrary, that there are no sociological grounds for admitting that the semiotic and material social relations reflect each other, especially when we take into account the effects of ideology, which can mask material social reality frequently to the point of inversion. To conclude, while the author's deep ideological model at work in the production of urban space claims to account for the whole of the form of urban space, it only accounts for its semiotic dimension and cannot in our opinion account for the material forces shaping urban space (cf. Ledrut, paper 5). Despite this limitation as a socio-semiotics, Greimas' model is the most comprehensive semiotic model of space.

For a Topological Semiotics 27 INTRODUCTION 1. If it is the case that every knowledge of the world starts by the projection of the discontinuous on the continuous, we may perhaps provisionally return to the old opposition: expanse vs. space in order to say that the expanse, taken in its continuity and its plenitude, filled with natural and artificial objects made present to us by all the sensory channels, can be considered as the substance which, once informed and transformed by man, becomes space, that is, form, capable (through the fact of its articulations) of serving the purpose of signification. Space as form is thus a construction which in order to signify selects only certain properties of "real" objects, only some of its possible levels of pertinence: it is evident that every construction is an impoverishment and that most of the richness of the expanse disappears with the emergence of space. However, what is lost in concrete and lived plenitude is compensated for by multiple gains in signification: by establishing itself as signifying space it simply becomes a different "object." 2. In examining not so much the origins of space—which is pointless—but its most simple articulations, we will first remark that any place can be understood only by fixing it in relation to another place, that it is defined only by that which it is not. This first disjunction can either be undefined and appear as: here vs. elsewhere or else take precise shape as: enclosed vs. enclosing In any case, the appropriation of a topia is possible only by postulating a heterotopia: it is only from this moment forward that a discourse on space can be instituted. For space thus founded is nothing but a signifier, it is there only to be taken in charge and it signifies something other than space, namely, man, who is the signified of all languages. Given this, the contents, varying with the cultural contexts and differentially established (owing to this distance of the signifier), are of little importance. Whether nature is excluded and

28

Algirdas Julien Greimas

opposed to culture, the sacred to the profane, the human to the superhuman, or, as in our desacralized societies, the urban to the rural, in no way alters the status of signification of the mode of articulation of the signifier with the signified which is at the same time both arbitrary and motivated. Semiosis is established as a relation between a category of the signifier and a category of the signified; a necessary relation between categories simultaneously indetermined and fixed in a given context. It is obvious, also, that the binary articulation of these categories is here suggested only to exemplify the minimal conditions of signification, that intermediary spaces (the suburb for instance) can be instituted, or that a transcendent space such as in the pictorial representation of Saint George founding the city, can be conceived by opposition to immanent space. What matters is to see that the conditions are present for considering space as a form capable of establishing itself as a spatial language, allowing us to "speak o f " things other than space, just as the natural languages, although they are sonorous languages, do not have as their function "speaking o f " sounds. 3. If we make use of the naive distinction according to which every object can be considered and studied either as what it is or as what it signifies, we can say that space will appear different according to w h e t h e r it is c o n s t r u e d a s scientific form or as semiotic form, the first

form recording only articulations of discriminatory character, the second form attempting to found discontinuities in signification. Two topologies, the first one mathematical, the other semiotic, are thus possible. In order to avoid the terminological difficulties and ambiguities we encounter, w e might as well designate by the name of topological semiotics the description, the production, and the interpretation of spatial languages. 4. Whilst maintaining the principle that at least a binary articulation of space is necessary in order for a minimum of "spoken" meaning to appear through it, one should nevertheless recognize the existence of a phenomenon of focalization: when we distinguish, for example, a space here and a space elsewhere, it is from the point of view of the here that we establish the first articulation (the here of the citizen not being the here of the nomad looking at the city). Consequently, every topological study is obliged to choose beforehand its point of observation, distinguishing the place of enunciation from the

For a Topological Semiotics

29

place uttered and specifying the modalities of their syncretism. The topic place is at the same time the place of which we speak and from inside of which we speak. 5. Spatial language appears thus, at a first stage, as a language by which a society is signified to itself. In order to do this, it operates first by exclusion, by opposing itself spatially to all that is not it. This fundamental distinction, defining it only negatively, then allows us to introduce internal articulations which enrich it in signification. The social organization of the village (as Lévi-Strauss has shown) thus appears signified spatially. Spatial language is, however, far from being the only means of expression of this social morphology. It is not by chance that sociolinguistics, when it attempts to establish a classification of "social languages" (sub-languages employed by a single society whose differences vary from simple stylistic variations to the use of distinct natural languages), rediscovers more or less the same categories: sacred private external superior masculine

vs. vs. vs. vs. vs.

profane public internal inferior feminine

which seem operational, in order to establish a typology of buildings inside a pre-industrial city or in order to account for the distribution of spaces inside a spatial complex. We are here deeding with a static social morphology which attempts to manifest itself through all languages, or rather which has been established as signification thanks to these languages. 6. Not until the arrival of commercial and industrial societies do we see these stable social morphologies shaken and gradually replaced by the dynamics of mobile social groupings. We realize, as a result, that these taxonomies, spatially or linguistically manifested and perceived as systems of signification, have been replaced by socio-semiotic syntaxes developing in particular discourses spoken and heard within the context of a system of communication. The city, which once conceived itself and signified for itself, is now conceived as an object. The city, which once constructed itself, is constructed by an individual instance, distinct from the city.

30

Algirdas Julien Greimas

Two kinds of Utopia then arise from the fact that the city as conceived can be conceived as a sick city and that the space which covers it and signifies it is considered as a negative space. Above the diachronic transformations proper to every semiotic system, a meta-discourse of protest establishes itself in order to call into question existing human space—a discourse which rejects space as the signifier of a social signified. Be it More or Le Corbusier, the end purpose of the meta-semiotic project is the same. METHODOLOGICAL APPROACHES 1. These few proceeding remarks are destined, as is apparent, only to present, in intuitive and simplified terms, the problematic of an eventual topological semiotics, only to answer, by successive approaches, the question posed naturally by every researcher, namely: how and where do we begin the exploration of a promising but vast and complex domain. They can now be specified and developed. 2. Everything happens as if the object of topological semiotics were double, as if its project could be defined at the same time as the inscription of society in space and as the reading of this society through space. Two dimensions that we have provisionally named spatial signifier and cultural signified appear thus to be constitutive of this semiotics; dimensions which are susceptible to being treated autonomously, but whose correlation alone permits the construction of topological objects. (a) The spatial signifier considered in itself is co-extensive with the natural world, also called the world of common sense: it is through this world that we read an infinity of significations appearing in the form of figures of the world, as objects external to our perception of them. The definition, within this vast spatial whole, of a zone of the signifier which would be proper to topological semiotics is possible only if, parallel to it, a specific signified is postulated. Moreover, this spatial signifier can be used not only for the establishment of the categories of the world, for the construction of a world of objects such as it appears (in its translation into natural languages) in the form of an inventory of lexemes like "forest," "meadow," "road," "house," "roof," "window," etc. It is capable of establishing itself as a veritable spatial language, as a "spatial

For a Topological Semiotics 31 logic," both natural and formal, allowing us to speak "spatially" of things bearing no apparent relation to spatiality. We know, for example, the particularly rich semantic investments to which spatial categories such as high vs. low, or right vs. left are susceptible, or the multiple semantic articulations of the cardinal points in what LéviStrauss calls concrete logics: the contents manipulated by spatial categories exceed by far the limits of the field of signification that we would assign to topological semiotics. (b) If the spatial signifier appears as a genuine language, one understands that it can be made the vehicle for signifying and above all for signifying the presence of man in the world, his activity which informs substance, transforms the world. Thus, we risk a poor beginning for the semiotic project if we start, for example, with the rather widespread idea that the architect's production aims at using space in order to "create beauty." All human behavior, be it only the "digging of a hole," for example, is doubly significant: firstly for the "subject of doing" and then for the "spectator of this doing." All social practices organized in programs of doing carry signification in them as project and result, and inversely: every transformation of space can be read as significant. However little these partial practices are organized into systems of competence, the question of the global end purpose of doing—or of what is done—cannot but be posed, entailing a reflection (unconscious or conscious, collective or individual, it does not matter) on the value of this latter. Exactly as in the case of the spatial signifier that we have seen developed into an autonomous natural logic, the immediate signified, present in the very process of the transformation of space, detaches itself from its signifier, receives new articulations, and sets itself up as autonomous discourse in order to speak of space. This discourse can use spatial language as its signifier—as when a builder "builds" his new city—but it can also go beyond this signifier and use other languages of manifestation: pictorial language, cinematographic language, most of all the natural languages, in order to "think" the signification of human space. The place of ideologies and mythologies is thus founded: myths of the origins and of the destiny of the city, various diagnostics and therapeutics for "healing" the city by applying a "treatment" to space. It is thus that, starting from a language of action informing space

32

Algirdas Julien Greimas

and giving meaning to it, two autonomous discourses develop, discourses more and more distant from each other, whose correlation is nevertheless necessary for the constitution of topological semiotics. 3. Once the general framework of a topological semiotics has been outlined, it is nonetheless true that the conception and the construction of particular topological objects are subject to two types of constraint, the one formal, the other cultural. (a) Formally, the first definition of the topological object is negative: in order to consider a given space, we have to oppose it to an anti-space, the city as opposed to its surrounding country. On the other hand, spatial focalization, that is, the identification of the subject of the enunciation with the uttered space, is necessary in order to ensure the positive determination of this space: urban semiotics is as possible as rural semiotics. (b) Culturally, the appropriation and exploitation of space by man are characterized by a sociological relativism of such a nature that a general model that would account for the whole set of possible topological objects and at the same time comprise rules of restriction justifying their cultural typology seems at present both necessary and impossible. In order to constitute topological semiotics, a reflection on the status and structure of topological objects in general is needed: however, the only hope of constituting it seems to be to fragment it into a great number of particular semiotics which would instead treat of this or that particular class of topological objects, while subordinating their explorations to a unitary semiotic project as the only guarantee of an ulterior comparative approach. 4. That these particular semiotics, urban semiotics, for example, are not concerned with particular objects (the city of Tours or of Carcassonne) but only with classes of topological objects, seems to be self-evident: there cannot be a grammar for every utterance or even every discourse. We should start, then, by the recording of invariants, recognizable either on the syntagmatic plane as recurrent phenomena or on the paradigmatic plane by the identification in two parallel objects of comparable phenomena. Here is an example of methodological extrapolation, of those "borrowings" from linguistics which some are suspicious of, but which are part of the general epistemology of the sciences. Actual topological objects are often complex and ambiguous, if

For a Topological Semiotics 33 only because of the fact of the durable solidity of their signifier, or the fact that their "message," like Egyptian writing engraved on a stone, is the product of a mediating communication, as opposed to immediate speech. The result is a historical stratification of the object, many substratums and superstratums coexisting with the dimension of the present: a "real" topological object thus pertains not only to one but to many models; it is the product of many grammars, as we say today. Here is a second reason for not confounding urban semiotics with the study of particular cities, canonical cities with real cities, the organization of actually occurring objects with the construction of topological objects. AN IDEOLOGICAL MODEL OF THE CITY 1. It is then only by adopting a certain scientific strategy, which would allow us to proceed in concert with the construction of particular semiotics and the elaboration of a common methodological conceptualization, that we can hope to arrive one day at a conception of a general topological semiotics and to specify the limits of its project which appears now too vast—if it is extended to the totality of human behavior that transforms space—or too limited—if it includes only artificial and secondary codes of signalization (arrows, insignia, shop windows, etc.) which cover with their overdeterminations spaces which are already significant. 2. Let us think, as an example, of the topological objects called "cities" as appertaining to a particular semiotics that we can call urban semiotics. It is evident that we are here in the presence of a complex and polysemic object which is immediately conceivable only as a global effect of meaning and that its reading can only be conceived as the disarticulation of a whole into its constitutive parts. And yet, the effort attempting to decompose the city into an infinity of objects that fill its space, would in no way further the analysis: these partial objects would appear in their turn complex and polysemic, and for two reasons. Firstly, an object by itself is not the conceptual object of semiotic and/or scientific study: a topological set is not constituted of objects but of their common properties. Secondly, a Dogon lock, for example, is a global object, in other words multifaceted and undifferentiated as long as the cultural context in

34

Algirdas Julien Greimas

which it is inscribed does not question it by placing it in different isotopies of possible readings. It is only as placed before us, surrounded by other objects pertaining to our familiar space, that it can be asked whether it it is beautiful, or good and useful, questions to which our answers risk being most often false because they are supported by our implicit Euro-centrism. 3. It is the same for cities as it is for locks. Perceived as global objects, our cities, since the epistemological revolution which we have defined in one of its aspects as the substitution of a discursive syntax for a socio-semiotic morphology, are subject to a multi-isotopic reading. The phenomenon is in fact particularly visible on the level of the mythical conception of the city: conceived formerly as a euphoric unit-object of which only the origins and the destiny were problematic, the city of today is conceived in terms of a profane mythology which articulates it, on the general axis of: euphoria vs. dysphoria in a triple discourse on the true, the good, and the beautiful. This sociological triad serves as a point of departure for the establishment of the principal isotopies of the reading of a city. It also haunts the dreams and the reflections of producers (or would-be producers) of cities, threatening at every moment to transform descriptive semiotics, which only tries to make explicit significations inherent in its object, into normative semiotics. Articulated in positive and negative values according to the category euphoria vs. dysphoria, the three systems: the aesthetic (beauty and ugliness) the political (social and moral "health") the rational (efficacy of functioning, economy of behavior, etc.) produce, on the syntagmatic plane, three distinctive isotopies allowing us to group together the partial objects constitutive of urban space and to disambiguate polysemic objects susceptible to being treated in turn on many isotopies. 4. A new category is added in order to complicate this multi-plane reading of the modern city: it comes from the relatively recent oppo-

For a Topological Semiotics

35

sition of two concepts endowed, it is said, with a quality of universality, that of: society vs. individual. Admitting that the archaic village is the spatial expression of its social organization, the fact that we can consider this latter from the point of view of a social statics or dynamics (as in the Pilou-Pilou ceremonies) in no way alters the fact that the concept of "community" is there coextensive with exploited space. This is no longer the case in our modern cities where the opposition society vs. individual ceases to be isomorphic with the ancient morpho-semiotic category of public vs. private, even if we consider it as enriched with new significant sub-articulations and manifested either by the opposition of places as public and private according to the criterion of their occupation (walls, stairwells vs. buildings, apartments), or by a typology of spaces corresponding to that of behavior (places of work, of leisure, of habitation). On the other hand, the individual, in his opposition to the urban community, is not to be taken as a cipher constitutive of a sum which would be society, nor as a unique and "lived" event, irreplaceable in space and in time (although political mythologies take pleasure in "degrading" him or in "exalting" him in one or the other of these directions): considered as an a epistemological concept, the individual is comparable to the ideal Weberian type in sociology, the idiolectical universe in semiotics. Individual and society, individual universe and cultural universe appear to us to be co-extensive concepts, to be virtual uttered places, open to the same semantic investments: as in the definition of topic space starting from the opposition of here vs. elsewhere, the "point of view" alone, in other words the co-occurence of the place of the utterance and of the place of the enunciation, will determine the type of discourse which will be held on the city, as this latter can be considered either as "urban culture" or as a life-style of the city-dweller. If, by the fact of its abstract and purely differential character, the opposition society vs. individual can give occasion, due to varied semantic investments, to multiple ideological games (society for the individual or the individual for society?) and thus can produce a rich urban mythology, it can equally be used as an epistemological cate-

36

Algirdas Juliett Greimas

gory dichotomizing the discourses which can be held on the city: euphoric or dysphoric discourses, ordered on aesthetic, political, or rational isotopies, which can have as subject either society (the urban community) or the individual (the citizen). In fact, exactly as the city can be beautiful, happy, and functionally organized, the individual inscribed in the city is also capable of feeling beauty, of being happy or unhappy, of seeing his needs fulfilled according to the law of least effort. 5. With the help of two semantic categories: society vs. individual euphoria vs.dysphoria and of three axiological isotopies: the aesthetic the political the rational we can compose a grid of reading and set up an inventory of the elements in the combinatorial matrix of production of a city. This grid and this inventory obviously have nothing of the exhaustive or the necessary about them, but they can give us an idea of how we might construct, for a historically and geographically determined area, an ideological model of the city, a model generative of multiple modern mythologies, but also producing, under certain conditions of the manipulation of the spatial signifier which are yet to be specified, topological objects appertaining to urban semiotics. Certan comments seem to us to be useful in order to specify the status of this model. (a) It must not be considered uniquely as a model for reading the city, but also as an abstract and deep structure which can be used to generate an infinite number of regular urban forms; not being a normative model, that is, not appertaining to a science of the good, the true, and the beautiful, it must be able to predict the conditions of generations of ugly as well as beautiful cities, of happy and unhappy cities, functional or dysfunctional, realized or only possible. (b) Since the model is situated on the level of deep structures, the categories constituting it are to be considered as formal categories, that is, as categories susceptible simultaneously to semantic invest-

For a Topological Semiotics

37

ments variable from one cultural context to another, and to subarticulations of invested contents permitting the apparition of veritable axiological micro-universes. Without speaking of aesthetic or political categories in general, whose relativity seems self-evident, cultural differences are manifested at all levels and by all channels: thus the thermic euphoria experienced in an inhabited space will be different for an Englishman and an American, the auditory or olfactory euphoria for the inhabitant of an oriental city will be judged as dysphoric by an occidental. It is this relativity of semantic investments and of their articulations which allows us to consider this kind of model as a grammatical model. (c) In addition to its taxonomical organization, one can see that the model implies a small number of rules capable of orienting the actualization of its combinatory matrix. Thus, in addition to the compatibility of two social or individual euphorias or dysphorias, of community culture and individual life-style, a rule of the dominance of the one over the other can also be formulated and applied. The same holds for the rules of priority to be given to the various isotopies of the construction of cities—rules the application of which can produce cities predominantly functional, political, or aesthetic.

PROJECT FOR A GRAMMAR: THE CITY AS UTTERANCE 1. The model we have proposed above should be considered as hypothetical, for two reasons. Although it is based on the dominant epistemt it is nonetheless constructed intuitively, starting from the superabundant preoccupations of the urban planners. As a model organizing the form of the content at an abstract level, it remains detached, without predictable revelation with the plane of spatial expression, whose parallel articulations alone can validate it. In effect, it is through the spatial language that the categories constitutive of this model must be manifested and/or read. This in turn is possible only if, firstly, an equivalence, whose nature remains to be specified, can be postulated between the articulations of deep content and those of the language of manifestation and if, further, the distance separating them can be bridged by procedures of generation and levels of construction progressively joining the postulated model to the spatial manifestation. We must now attempt to foresee eventual solutions to this problem.

38

Algirdas Julien Greimas

2. Of the different approaches allowing the analysis of a topological object as complex as a city, application of the model of communication seems to be among the most productive. Within the framework of this elementary structure, consisting of a sender-producer and a receiver-reader, we can inscribe the city as an object-message to be decoded either by imagining the procedures preceding this message and leading to the production of the city-object, or by paraphrasing the procedure of the reader trying to decode the message with all its allusions and all its presuppositions. In both cases, the city can be considered as a text whose grammar we will have, at least partially, to construct. This text, however, far from taking the appearance of surfaces and volumes that can be represented by drawings or scale models, must at first approach be imagined as an agglomerate of beings and of things, among which the semiotician will attempt to perceive relations allowing him to construct a meta-text, in the form of either inventories or a series of utterances, the subjects of which would be the human beings (the users of the city), and the grammatical objects of which would be the things (the things the subjects are in contact with and manipulate). The recognition of the isotopic levels of organization on which the objects can be treated and especially the recurrence of observable relations connecting the subjects to the objects will then allow the semiotician to establish lists of regular utterances and of their semantic investments. 3. This would be, in its general lines, the simplified procedure leading to the construction of a textual grammar of the city considered as a global utterance. Since the city-text which we would analyze is manifested in the spatial language which is, as we have seen, the language allowing the reading of the world of sensible qualities, the objects entering in relation with the subjects recognized in this text will not interest them as such but only because of certain of their sensible properties: visual, auditory, thermic, olfactory, etc., space itself being conceived at this stage only as the concept totalizing all these qualities, and the user in turn being defined as the interpreter of urban space. We see clearly that such an approach allows us to return, at the level of sensorial perception, to the category of euphoria vs. dysphoria which, applied first to the city as a whole, can also be applied to the

For a Topological Semiotics 39 individual in so far as he is in euphoric or dysphoric relation with urban space. We also see that the vague and indefinable terms such as "live," "feel," "perceive," which we frequently use, can be reduced to this relation of the subject to space, to this "use of space" of which w e are unable to say if it is conscious or unconscious, thought or experienced, but which is, in one word, meaningful. 4. All that we have said above can be applied to any space acting on man; consequently, urban space must be defined in its specificity not only by the "qualities" it produces and communicates to man but also by the constructed (though not necessarily "built") nature of the objects serving as support to these qualities. The analysis of the spatial language into pertinent features, into minimal phemic units (the "qualities of the world") thus constitutes a level both pertinent to and insufficient for the description of the signifier of urban space: pertinent because it is at this level that the world is significant for man, but insufficient because the process of the production of a city cannot be described without the prior emplacement of the constructed objects and systems of constructed objects which support and condition the establishment of the sensorial isotopies. 5. It is thus that the utterances of state allowing the formalization of the relation of the subject to the world presuppose the existence of utterances of doing susceptible of accounting for the production and/or the transformation of these states. Thus, in order to create a state of thermic euphoria, the subject is presumed to obtain wood, light the fire, etc., in other words, to execute a whole program of behavior aiming at the production of a thermic state. His final bodily behavior is thus a program of signification. It may be characterized by the fact that it is a stereotyped program, at the same time recurrent and executable by any subject considered as a syntactic role (and not as an individual of flesh and blood); and also by the fact that these are programs in which human subjects can be replaced, partially or entirely, by automata. There is no need here to go into sociological considerations describing the process of industrialization, showing how, starting from the tool as an extension of the hand, humanity has arrived at the construction of automata replacing human action, presupposing in their turn other somatic or mechanized programs of doing and thus instituting new forms of social organization functioning through

40

Algirdas Julien Greimas

successive mediations and substitutions. This would mean trespassing on other disciplines, more precisely on a certain sociology, only the research results of which could be exploited by topological semiotics. From this point, though, the recognition of this process of the replacement of segments of somatic doing by automized programs concerns semiotics, to the extent that this "objectification" of social practices facilitates the segmentation of the urban text into autonomous and isotopic instances of doing. We see thus that the semiotic manifestation of urban space, which can be roughly illustrated as the relationship of: (thermic signifier) to (euphoric signified) presupposes a certain doing of the subject (which may be no more than pushing a button) effected on a support-object (central heating apparatus), a localized substitute for a somatic program, but that this individual instance of doing presupposes in its turn a new collective instance with a new support-object (urban network for the distribution of gas or electricity), manipulated by a collective subject (gas company or public service company). We are thus faced with two types of support-objects, on the basis of which we can distinguish two forms of subject participation in urban space which for the purposes of our analysis will constitute two autonomous syntactic instances: individual and collective. 6. From this point of view, the individual instance appears as constituted by the set of relations of the individual in relation to the objects surrounding him, making him the center of this relational network; the collective instance on the other hand appears as the set of networks (electricity, gas, water, sewage, telephone, mail, subway, streets, etc.), the terminals of which constitute so many individual instances. Two kinds of doing—individual and social—are linked to these instances; they consist of ensuring the maintenance and the functioning either of individual networks or of collective networks. To these two types of support-objects and of programs of doing correspond, consequently, two types of subjects, considered no longer as individuals but as syntactic roles corresponding to the programs. Just as objects interest semiotics only by virtue of certain of their properties allowing us to group them in topological sets, so subjects can

For a Topological Semiotics 41 be decomposed into roles according to the programs they are supposed to execute. It is only through this double "destruction" of objects and of subjects that a semiotic syntax is possible. 7. The grammatical approach that we have just outlined possesses numerous advantages. The principal one is probably the integration of human subjects in the text of the city: whilst producing a semiotic interpretation of the "user of the city" it permits us in a sense to make our representation dynamic, allowing us to conceive of the city as a set of interrelations and interactions between subjects and objects. Furthermore, by allowing the distinction of two canonical forms of transcription of relations between subjects and objects—utterances of state and utterances of doing—it specifies two distinct places for the manifestation of meaning. Since it is at the interior of the utterances of state that man "experiences" space by linking himself to the qualities of the world, the existence of the utterances of doing, whose function is to produce states, manifests a new instance of signification: the doing of man is, in fact, meaningful for himself and for others. Nevertheless, this significant doing, which consists in the production, the emplacement, and the manipulation of the objects for the purpose of the constitution of significant states, is in our modern cities largely replaced by programs effected by automata: this results in a proliferating desemantization of cities which, to the extent that it is felt as dysphoric, appears as an alienation. We should however not confuse these two concepts of desemantization and of alienation, the first of which is simply an observed fact of life, while the second also involves an axiological judgment. Desemantization is a general semiotic phenomenon; we can say that our life is occupied in part in replacing our significant behaviors by desemantized programs made into automatisms. The economical, rational exploitation of our bodily activity can result in the abolition of meaning but this may be felt as either euphoric or dysphoric. Thus, to return to the problematic of urban semiotics, we can say that semiotic programs—surviving or replaced by automata—are not always and not necessarily endowed with more meaning than programs of substitution, which are also capable of producing euphoric states; that the functionalization of the city, situated on the isotopy of the rationality of individual or collective life, is neither good nor bad

42

Algirdas Juliett Greimas

in itself. Given this, the phenomenon of desemantization—as well as that of re-semantization (the reintroduction, for example, of fire places in addition to central heating)—appears as a semiological fact susceptible of being treated outside of ideological considerations. 8. The major disadvantage of this grammatical approach resides, it seems to us, in its inability to give a clear image and a satisfactory meta-textual representation of the collective component of the city. It is true that in this approach the collective instance is clearly distinguished, since it is both presupposed by the individual instance and characterized by a particular class of support-object manifested in the form of multiple urban networks governed by autonomous forms of organization. But although this (partial) image of the city, built upon the localizations, in the form of networks, of supportobjects considered as substitutes for the real activities of men, can give an idea of the structures of maintenance and functioning of a city, w e do not quite see how, starting from this point, a grasp of the "meaning of the city" can be achieved. Everything happens as if such a grammar, centered on the term "individual" from the epistemological category society vs. individual, were incapable of changing its point of view and accounting for the social dimension of the city, as if another grammar and another sociology were necessary in order to define the relation of man to urban space, now not in terms of "life-style" but in terms of "urban culture." It is true that in examining the doing of citizen-subjects, w e were able to recognize in addition to individual roles, social roles through which individuals participate in the accomplishment of collective tasks. Consequently, w e can say that these social roles are "lived" in some w a y , that these social activities are significant for the individual. But an analysis of this kind cannot be pushed very far, if only because social activities are participative, each role and every program being inscribed within the context of a collective doing which surpasses them. Evidently, social roles can be reduced to the individual w h o will interpret them in terms of fatigue, boredom, etc. But the problem is to know if and how the individual as social role "lives" his participation in a common doing, what meaning he attributes to himself and to his doing as part of a whole. From the problematic of the individual [and purely formal] "actant," w e thus pass to that of collective octants. Being familiar with

For a Topological Semiotics

43

the concepts of "society" and "class" and the anthropomorphic attributes such as "class consciousness" with which we endow them, we may ask if a grammaticalization of these collective entities and the representation of social groups and social organizations under the form of collective subjects might not equip urban semiotics with the methodological tools allowing us to account for the modes of existence of "social man," in other words, of man engaged in social doing and man participating in social being. A certain number of studies in narrative semiotics tend to show that it is not impossible to describe economic and social organizations, cultural and political institutions, as collective "actants" endowed with modalities of wanting, of being able, and of knowing, and invested with axiological contents which are experienced as such by the participants in this "moral person." Urban social organization could thus be decomposed into different collective actants and [concrete] actors whose descriptions, first partial, later comparative and holistic, would furnish us with an equal number of conceptions of communal meaning, while the syntactic models thus obtained would serve as the framework for a semantic analysis of the "collective representations" of the city. 9. The advantage of such an approach (inspired by the provocative paper by Renier 1974) is that it gives a clear definition of the object of urban semiotics: rejecting the traditional points of view according to which the city is a thing, a complex of objects experienced and perceived by men, it sets in their place a conception of the city-text, made up of people and of things, of their relations and interactions. The human subjects whose presence in the test is the only element that can account for its significant character, are thus distinguished from the subject of the enunciation, the producer of the city; and the grammar of the city-utterance can be supplemented by a grammar of the enunciation, the more easily now that the instances of hierarchical generation, each step of which presupposes the preceding one, have already been anticipated in the theory. In short, it is only a matter of inverting the process which, starting from the conjunction of the individual with the qualities of urban space, was led to postulate support-objects constructed on several levels, by a reverse process, descending as it were, which would start by demonstrating the collective mechanisms in order then to

44

Algirdas Julien Greimas

pass to the objects, constituting the immediate environment of the individual. PROJECT FOR ANOTHER GRAMMAR: THE ENUNCIATION OF THE CITY 1. In spite of the specificity of the object considered—"the architectural whole" is only an auxiliary aspect of the problematic of urban semiotics—it is still the grammatical program of Castex and Panerai (1974) which can best serve as an example to illustrate this generative approach. It is evident that the analysis of so limited an object is valid only to the extent that it makes explicit a priori all the limitations that have allowed it to define its purpose and scope, that is: (a) that the architectural whole is treated in isolation, as an enclosed object whose enclosing environment is for the purpose of analysis provisionally ignored; (b) that the object thus defined is considered in only one of its isotopies, the visual one, or rather only in the sub-isotopy of forms, to the exclusion of the isotopies of color and light; (c) that the description concerns only the level of the signifier of this objectid) that the approach as a whole is inductive and generalizing, attempting from the description of empirically occuring objects to constitute an inventory of forms and a list of the derivation rules which at a later stage would serve as material for the construction of a grammar of the production of regular "architectural wholes." 2. Given these limitations, the project of Castex and Panerai occupies, in the general economy of urban semiotics, the place reserved to one of its sub-components, namely the individual instance, filled with objects and systems of support-objects constructed in view of signification for the user of space, and excluding all signification which would derive from his participation in the life of the community. Also from that perspective, the organization of spatial forms appears as the final instance of a generative path situated just before their manifestation as "built space," for the procedure as a whole can only be conceived as starting from an ideological instance in a deep structure of which architectural wholes are only surface realiza-

For a Topological Semiotics 45 tions. This ideological model of a semantic nature can only be spatially manifested by first being linked to the signified of the language of spatial manifestation, a signified which will find its expression thanks to the "phonological" component, that is, thanks to the architecture of spatial forms which will finally be attributed to it. Castex's project, situated by these few remarks in relation to the textual grammar we have outlined above, appears a good example of the generative approach, if only in the fact that it sees the architectural text as the result of the expansion and combination of one or more elementary structures of signification linking up with the elementary spatial articulations, and that it is only later, through successive sub-articulations and overdeterminations, that the topological object becomes complex, taking the form of some architectural whole. 3. Such a construction implies a first difficulty residing in the choice of units and the levels of analysis, for this initial choice determines the strategy of description as a whole. In the example of the George Barton House chosen by Castex we see that three possibilities are open to the describer, three spatial structures can be considered as structures ab quo starting from which the procedure of the generation of the building can be undertaken. Admitting for simplicity's sake that the description can be made from the plan rather than from a scale model, and that architectural space can be treated, by a reductive transposition, as surface and not as volume, these three structures are: (a) first the cross ("six units of space grouped in the form of a cross") (b) then, on a lower level, the square (the unit of which the cross is only an expansion) (c) finally, at least the straight line and the right angle, which are the constitutive features of the square. Starting from these three types of spatial units and from these three distinct levels of analysis, we need to choose the type of basic units and the optimal level of analysis to begin the process of description: it is not obvious at first approach, what criteria of pertinence would oblige the describer to consider this or that unit or level as his starting point.

46 Algirdas Julien Greimas In theory, it is the level of distinctive features, that of phemic categories, such as (by way of suggestion only): straight vs. curved right (right angle) vs. non-right (angle) categories disengaged by typology, that science "both approximate and rigorous" (Guilbaud), which permit the production of squares, triangles, and circles by the combination of their terms—it is this level that must be considered the elementary level. In fact it is at this level that Baroque architecture is partially defined, by the category concave and convex. It is equally at this level, with the oppositions of straight and curved, horizontal and vertical lines, that the first articulations of signification appear, isomorphic with spatial oppositions. Everything happens as if the grammar of the production of spatial forms ought to start from these elementary categories in order to engender, through the rules of the combinatorial matrix which it would acquire, spatial figures such as the square and the triangle. It is only if the architectural corpus to be treated is, as in our case, relatively limited that the level of spatial figures can perhaps be chosen as a starting point for the construction of a limited number of topological objects: we see, in effect, that the choice of the square as basic figure best satisfies the rule of simplicity of the description, because it is from the square that the highest number of derivation rules can be formulated in the simplest fashion. This rule, however, is only pragmatic and for this reason subordinated to the principles of coherence and exhaustivity. If on the other hand the spatial categories are able to generate spatial figures, these can in their turn produce composite figures or configurations, such as the cross of the George Barton House which is coextensive with the actually occurring architectural whole described: thus a hierarchy of spatial units is established, following the decomposition and the avercomposition of the figure chosen as basic structure, justifying in this manner, to a large extent, the strategic choice of the describer. 4. Nevertheless, the recognition of this third level of overcomposition, although it appears to be an expected extension of the process that generates the architectural whole, does not fail to create difficulties in the interpretation of the objects thus constructed: an "archi-

For a Topological Semiotics 47 tectural whole"—which is defined only intuitively—can be produced either by configuration (= the cross) or by arresting the generation at the level of simple figures (= a square building) or, finally, by the coordination of two figures (= two square buildings juxtaposed). Instead of speaking, for lack of a definition of the "whole," of contradictory models or of the "exasperation of an architectural code," we might see there rather the simple effect of the passage from the figurative level to the configurative level, from a phrase grammar to a discourse grammar: if derivation is a process of decomposition of the utterance (the figures could be assimilated to semantic utterances), the overcomposition of basic units produces configurations corresponding to the expansions of utterances into discourse, it being understood that, once the discursive level is recognized, the utterance-figure is already a discursive unit, substitutable for the entire discourse in expansion. The rules of a discourse grammar— treating composition of architectural wholes but also of objects much more complex—should, consequently, be anticipated independently of those of the elementary grammar. From another viewpoint, and without excluding the theoretical possibility of contradictory codes which would imply the production of a topological object from at least two autonomous elementary structures, we can imagine the existence of objects characterized by the complementarity of spatial figures of which one group, for example, would be constructed from straight lines and the other from curved lines (cf. the Pantheon in Paris). In this case, the strategic choice of the level of figures as point of departure for the generation of architectural forms could be maintained only by adding new rules of transformation to the derivation rules, by postulating, for example, that at such or such a stage of derivation, square figures are transformed into circular figures, a principle which would presuppose a strict hierarchization of the different types of figures and would inevitably limit the field of application of the process of description we have imagined. Only the process of production of forms from spatial categories can guarantee the coherence of the description, at the risk of appearing less economical than the one we have just examined. However, the interest of the proposed formalization extends beyond the context of the architectural whole and the examples analyzed. We see that the same approach and the same procedures can

48

Algirdas Juliert Greimas

be applied to the more extensive topological objects and spaces studied by urban semiotics, on condition that a clear separation is established between the articulations of the figurative and phrasic level, and those of the configurative and discursive level. The recognition of many semiotic levels of organization of forms allows, on the other hand, a more confident approach to the problems of the topological signified, by showing that spatial language, as a consequence of the fact that it can be manifested at all recognized levels of articulation, disengages multiple and graduated significations (that the straight line, the square, and the cross are significant separately and each in its turn), the arrangement of which produces an effect of global meaning. SENDER AND RECEIVER OF THE URBAN MESSAGE 1. Of the two possible methodological approaches—interpretative and generative—corresponding to the two poles of the structure of communication, the city considered as a global utterance that can be read by the receiver, and the city as uttered by the sender, it is this latter—and for reasons that are not all of a scientific nature—which is favored by architects who desire to approach the urban problematic from a semiotic perspective. It is as if the urban planner, in adopting this approach, finds himself naturally in a familiar ideological landscape: in identifying himself with the sender/enunciator of the city, he is transformed in his own eyes into producer of the city, in a travesty—to use an accurate expression overheard in the debates—of his own profoundly individualistic and reactionary ideology: we know how tenacious and self-satisfying is the myth— which only dates from the eighteenth century—of the creative individual. The subject of the enunciation, a semiotic place which can be legitimately privileged by the adoption of a specific methodological approach, is as a result transformed into an abscess of fixation for all sorts of ideological, aesthetic, and sociological uneasiness, delaying in the same measure the constitution of an urban semiotics. For we too often forget that the model of communication which facilitates the semiotic conception of the city is first of all a formal model which only institutes the instances of production and of reading as empty places, that, on the other hand, urban semiotics has

For a Topological Semiotics 49 the task of describing neither real cities nor their producers in flesh and blood but canonical objects and syntactic "actants." Sociological investigations allowing a contextual analysis of the "actant"-producer are nonetheless necessary, although they are subject to sociocultural relativism: if we can say summarily that in the case of the Bororo village the same population plays simultaneously or successively the syntactic roles of producer and of reader of its own topological space, it is of another order of difficulty to answer the question of who has constructed the city of Paris. Let us not be deluded by the commissioning of this or that architect to construct a new city: Mr. Aillaud certainly did not, on the aesthetic isotopy, choose the form of the triangle, generally read as dysphoric, for the contour of his city of Grigny, nor on the political isotopy did he choose the model of the dormitory city. He is only one of the actors—one whose role deserves to be specified—in the complex collective "actant" which if analyzed would manifest economical and political components much more powerful than the architect/urban planner. 2. An urban socio-semiotics could here find one of its objects. Thus, for example, to the extent that the producer can be conceived as the subject of the enunciation, endowed with a competence, it should be useful to analyze this competence into a being-able-to-do, a wanting-to-do and a knowing-to-do of the producer: the planner, possessing no real power, would find himself in part absolved or, at least, he would not confuse the two syntactic roles he may be called upon to play. The structure of the collective actant is not made up only of the disposition of the modalities of being able, wanting, and knowing, it also comprises an investment of ideological content: the study of the process by which the various separate wills constitutive of this "actant" accomplish the blending of sometimes contradictory values giving rise to the appearance of an ideological model of the city to be constructed—and which, because it is implicit, corresponds very little to what the architect thinks and above all to what he does— would allow us to describe the decision-making mechanisms which lead to the construction of cities based on the three isotopies already examined. Because if we know—or we think we know—more or less the problems relative to the political finalities of planning, such studies would equally allow us to correctly situate aesthetic prob-

50

Algirdas Julien Greimas

lems by describing the various systems of constraints imposed on the architect: the so-called natural constraints, the pressure of actors who are part of the composition of the collective "actant," but also the self-censorship exercised by that imaginary model of reading, implicitly known and accepted, of "the preferences of the user." Finally, a third type of analysis is possible: it would consist of the syntagmatic decomposition of the global program of the production of a city into collective and individual actors or into automatized substitutes. Adopting the generative form, such a description would appear as the inverse course of the procedure already outlined above in proposing the model of a textual grammar: in its generic form, the description would aim at giving a representation of processes and programs actually realized by various actors and ending in the construction of an actual city. 3. The methodological difficulties increase as we abandon the point of view of the sender for that of the receiver. The very terms usually used to designate this instance: reader, user, consumer, pertain to different disciplines and ideological attitudes, with the consequence that they are constantly used in a metaphorical or analogical sense. We must add to this that even the semiotic conception of the city as object-message is not without ambiguities. We are too much in the habit of interpreting communication in linguistic terms not to have some difficulty in imagining that meaning might be communicated without the intermediary of natural languages. In effect (and we have already insisted on the point), to receive spatial messages is not—or not only—to perceive them, it is what we vaguely call "experiencing" the city by reacting in a significative way to all spatial stimuli. Such an interpretation of the signification of non-linguistic messages, if it seems clear at the moment of its formulation, nonetheless requires delicate use in practice: it requires that the "primary" meaning of the city should be confused neither with conscious thought nor with the discourses held on the city; it implies that the frontiers between that which is conscious and that which is unconscious in the way in which one experiences the city should be abolished or at least suspended. It is only at this price that the concepts of reading and of use of the city can be considered as synonymous, and that the consumption of the

For a Topological Semiotics

51

meaning of the city, while retaining its metaphorical bent, ceases to be exclusively a reference to the consumer society. 4. To live in a city means for the individual, as we have seen, that he is the place in which all spatial messages converge, but also that he reacts to these messages by involving himself dynamically in multiple programs and mechanisms that appeal to him and constrain him. It is thus in principle by elaborating a model of living, a semantic representation of what is meant by the life style of the citizen, that we can hope at least partially to apprehend the structure of the content of the "actant"-sender. Nonetheless, we can see that such a model can only be typological, that together with a certain number of constants it will inevitably comprize a quantity of variables corresponding to both the social stratification and the historical relativization of urban communities. We can go still further, by introducing new variables and multiplying the number of possible readings of the city: we can oppose, for example, along the category external vs. internal, the reading of the user of the city to that of the visiting guest, we can distinguish a particular reading proper to the social category of tourists, even make a typology of it, examine the aesthetic attitudes of "elites": architects or interpreters of their aesthetic aims, etc. Just as in the analysis of the instance of the sender, that of the receiver surpasses the preoccupations of topological semiotics properly so called, and becomes involved in considerations of the social structure in correlation with the collective axiological systems. 5. The concept of life-style, we would say, does not exhaust all the possible significations of the city, if only because the whole of "immediately lived experience" that it tries to subsume is at every moment surpassed by the constructions of the imaginary that man projects around him. The filmic space "beyond the frame" which is progressively constituted during the projection on the screen of partial visual spaces in order to finally make them merely metonyms of a global imaginary referent, can give an idea of our mediated conception of urban space. Whether we imagine the citizen as a walker stocking away partial views of the city, as a user experiencing the commodities or the lack of comfort the city offers him, or as a social being engaged in the different activities constituting his life—he

52

Algirdas Julien Greimas

forms a global image of the absent city, an image accepted as the place of his inscription in space. It is of little significance what psychological or physiological status w e will attribute to these mediated representations of topological objects: the existence of the city as global imaginary referent appears to be beyond doubt. This global referent is clearly consolidated by meta-semiotic transpositions of all kinds: maps of the city, postcards, roadsigns marking panoramic views (Alençon, city of façades!) etc., not to speak of the innumerable discourses on the city; on the other hand, it is nourished by other ideologies formed in other circumstances (alienation, pollution, promiscuity); it serves as a pretext for multiple secondary elaborations manifested under the form of various urban mythologies (Paris, City of Light): thus a whole architecture of significations is raised above urban space, determining to a great extent its acceptance or its refusal, the happiness and the beauty of urban life or its intolerable squalor. Consequently, w e w o u l d be wrong to imagine the receiver of the city as a naive reader, as a kind of tabula rasa on which the sender would be the first to inscribe his spatial hieroglyphics: on the contrary, he appears as a structure of reception, possessing a code for deciphering messages which is complete but not necessarily identical to the code of the sender which was used for the production of the message—just as in linguistic communication two actants face to face are assumed to ensure the transmission and the reception of messages which are full of possible misunderstandings.

O N T O P O L O G I C A L DISCOURSES Since space does not need to be spoken in order to signify, the terms of message, of discourse, or of text that we are induced to apply in connection with space are only denominations of semiotic concepts which are assumed to be defined—as structures and not as terms— on the level of an epistemological language establishing the principles of treatment of all semiotic systems. In relation to this first "spatial text," all discourses on space are always secondary: whether they are more or less faithful transpositions of the spatial language into other languages, or autonomous manifestations of original modes of the construction of space or, most often, both at once, the

For a Topological Semiotics 53 discourses on space, verbal, graphic, pictorial, or cinematographic, are always situated at a distance from spatial discourse proper. Verbal discourse, the dominant character of which does not need to be emphasized—in effect, all other languages are translated into, and can be compared thanks to, verbal discourse—constitutes the primary concern of the semiotkaan. A double and paradoxical task is his: he must at the same time recognize the distance separating spatial discourse from the discourses paraphrazing it, but also— since he himself holds his discourse in natural language—he must try to suppress this distance or to cancel its effects. Recognizing the distance is, in the first place, to distinguish the properties of significant space from the properties characterizing verbal discourses treating space, because discourses are not defined, whatever we may say, by the contents they manipulate (to speak of political, social, religious discourses is to proceed to a typology of value systems) but by the forms of their organization. Being of grammatical nature, the typology of discourses is then a problem of general semiotics and the discourses on space pertain to it, without thereby constituting a class apart. Thus, the discourses—Utopian, protesting or futuristic, descriptive or normative—that we can hold on space would easily find homologies in semantic domains other than space. To cancel the effects created by the distance separating the "discourse of things" from the discourse held on this discourse is, first of all, to make explicit the conditions of the scientificity of the latter, in order to then submit the semiotic discourse we attempt to construct to rules allowing us to satisfy these conditions. Thus, contrary to what happens during the production of non-scientific discourses where, for example, the temporalization and the spatialization of models is a normal procedure of enunciation, semiotic models are considered as a-chronic, realizable in all times and all spaces, but independent of their realization. Thus, contrary to what happens in pre-scientific times, which have created theoretical models sometimes very perceptive and capable of being taken up again later, semiotic models must satisfy the principle of adequation, in the sense that scientific discourse should be equivalent, in a manner, to the primary discourse which it transposes and by this fact be susceptible to being validated through indispensible complementary procedures

54

Algirdas Julien Greimas

or discourses. But the principles of the validity of discourse and the procedures of its validation also pertain to the general epistemology of science.

REFERENCES Castex, J., and Ph. Panerai. 1974. "Structure de l'espace architectural." Notes Méthodologiques en Architecture et en Urbanisme (Sémiotique de l'espace), 3/4: 39-63. Greimas, A.J. 1974. "Pour une sémiotique topologique." Notes Méthodologiques en Architecture et en Urbanisme (Sémiotique de l'espace), 3/4: 1 - 2 1 . Renier, A. 1974. "Nature et lecture de l'espace architectural: Essai de définition de l'espace architectural et étude de son système de lecture." Notes Méthodologiques en Architecture et en Urbanisme (Sémiotique de l'espace), 3/4: 23-32.

2 FUNCTION AND SIGN: SEMIOTICS OF ARCHITECTURE Umberto Eco

Eco is, like Greimas, one of the prominent figures in semiotics. Educated as a philosopher, his approach to semiotics shows influences from European linguistic and semiotic theory, structuralism, Peircian semiotics, and the theory of semantics. His La struttura assente (1968) is a seminal theoretical work and includes a long section on architectural semiotics which is not found in the 1976 reformulation of this work in English (A Theory of Semiotics). The reading below is virtually a complete translation of this section. It is one of the best contributions to architectural semiotics and offers many insights into urban semiotics as well (Eco treats the tivo fields together). The first section of the essay refers to the two fundamental aspects of architecture: functionality and communicativeness. Eco differentiates communication from stimulation within the context of built forms. In the second section, the author begins by discussing the theoretical models underlying Published in Via (Structures Implicit and Explicit), 1: 1 3 1 - 5 3 (Philadelphia: University of Pennsylvania Press, 1973). A version of this essay also appeared in G. Broadbent, R. Bunt, and Ch. Jencks, eds., Signs, Symbols, and Architecture (Chichester, New York, Brisbane, Toronto: Wiley, 1980). This material first appeared as chapter C, "La funzione e il segno: Semiologia dell' architettura," in Eco, La struttura assente (Milan: Bompiani, 1968).

56

Umberto Eco

different semiotic approaches. Commenting on the semantic triangle of Ogden and Richards (which is composed of the "symbol" [cf. signifier] and "thought" or "reference" [cf. signified]—left side of the triangle—and the "referent," that is, the reality to which the symbol refers) the author accepts the existence of the referent, but considers it as falling outside the domain of semiotics (see also "The Anthropological System" below). He thus asserts a non-idealist epistemological position to which the editors also subscribe as indicated previously. For Eco, a sign is the union of a sign vehicle and a culturally codified signification. He also distinguishes with respect to use objects two layers of signification: a denotative signification, the "primary function," which refers to the function of these objects, and the connotative signification, the "secondary function," which is of a symbolic nature. In the third section of this essay, the author examines how signification changes in history and the modes of these changes. This adds a necessary dynamic component to what could have remained a static semiotic analysis. In the fourth section, the author discusses the issue of architectural codes, while in the fifth he analyzes the differences and similarities between architecture and the messages of mass communication, and presents three possible attitudes of the designer in respect to the existing social codification of the architectural system. Finally, in the last section of his essay Eco defines architecture narrowly as limited to spatial forms and focuses on what he considers to be the external source of its codes, i.e., cultural systems. He also relates his views on architecture to Hall's proxemics. Eco's conclusions are that, because of the rhetorical nature of architecture as a message of mass communication, its design languages are very conventional, because a building must be used by everyone. Yet, aver time, new denotative and connotative meanings of all kinds are ascribed to architectural objects by social processes. These often conflict with the inflexibility of existing architectural forms. Consequently, Eco suggests that architects should create structures which, while functioning as forms in their times, are open to new processes of signification appearing in the future.

ARCHITECTURE AND COMMUNICATION Semiotics and Architecture. If semiotics, beyond being the science of recognized systems of signs, is really to be a science studying all

Function and Sign: Semiotics of Architecture

57

cultural phenomena as if they were systems of signs—on the hypothesis that all cultural phenomena are, in reality, systems of signs, or that culture can be understood as communication—then one of the fields in which it will undoubtedly find itself most challenged is that of architecture. It should be noted that the term architecture will be used in a broad sense here, indicating phenomena of industrial design and urban design as well as phenomena of architecture proper. (We will leave aside, however, the question of whether our notions on these phenomena would be applicable to any type of design producing threedimensional constructions destined to permit the fulfillment of some function connected with life in society, a definition that would embrace the design of clothing, insofar as clothing is culturalized and a means of participating in society, and even the design of food, not as the production of something for the individual's nourishment, but insofar as it involves the construction of contexts that have social functions and symbolic connotations, such as particular menus, the accessories of a meal, etc.—a definition that would be understood to exclude, on the other hand, the production of three-dimensional objects destined primarily to be contemplated rather than utilized in society, such as works of art.) Why is architecture a particular challenge to semiotics? First of all because apparently most architectural objects do not communicate (and are not designed to communicate), but function. No one can doubt that a roof fundamentally serves to cover, and a glass to hold liquids in such a way that one can then easily drink them. Indeed, this is so obviously and unquestionably the case as it might seem perverse to insist upon seeing as an act of communication something that is so well, and so easily, characterized as a possibility of function. One of the first questions for semiotics to face, then, if it aims to provide keys to the cultural phenomena in this field, is whether it is possible to interpret functions as having something to do with communication; and the point of it is that seeing functions from the semiotic point of view might permit one to understand and define them better, precisely as functions, and thereby to discover other types of functionality, which are just as essential but which a straight functionalist interpretation keeps one from perceiving.1

58 Umberto Eco Architecture as Communication. A phenomenological consideration of our relationship with architectural objects tells us that we commonly do experience architecture as communication, even while recognizing its functionality. Let us imagine the point of view of the man who started the history of architecture. Still "all wonder and ferocity" (to use Vico's phrase), driven by cold and rain and following the example of some animal or obeying an impulse in which instinct and reasoning are mixed in a confused way, this hypothetical Stone Age man takes shelter in a recess, in some hole on the side of a mountain, in a cave. Sheltered from the wind and rain, he examines the cave that shelters him, by daylight or by the light of a fire (we will assume he has already discovered fire). He notes the amplitude of the vault, and understands this as the limit of an outside space, which is (with its wind and rain) cut o f f , and as the beginning of an inside space, which is likely to evoke in him some unclear nostalgia for the womb, imbue him with feelings of protection, and appear still imprecise and ambiguous to him, seen under a play of shadow and light. Once the storm is over, he might leave the cave and reconsider it from the outside; there he would note the entryway as "hole that permits passage to the inside," and the entrance would recall to his mind the image of the inside: entrance hole, covering vault, walls (or continuous wall of rock) surrounding a space within. Thus an "idea of the cave" takes shape, which is useful at least as a mnemonic device, enabling him to think of the cave later on as a possible objective in case of rain; but it also enables him to recognize in another cave the same possibility of shelter found in the first one. At the second cave he tries, the idea of that cave is soon replaced by the idea of cave tout court—a model, a type, something that does not exist concretely but on the basis of which he can recognize a certain context of phenomena as "cave." The model (or concept) functions so well that he can now recognize from a distance someone else's cave or a cave he does not intend to make use of, independently of whether he wants to take shelter in it or not. The man has learned that the cave can assume various appearances. Now this would still be a matter of an individual's realization of an abstract model, but in a sense the model is

Function and Sign: Semiotics of Architecture

59

already codified, not yet on a social level but on the level of this individual w h o p r o p o s e s and communicates it to himself, within his o w n mind. A n d he w o u l d probably be able, at this point, to communicate the model of the cave to other m e n , by means of graphic signs. T h e architectural code w o u l d generate an iconic code, and the " c a v e principle" w o u l d become an object of communicative intercourse. At this point the d r a w i n g of a cave or the image of a cave in the distance becomes the communication of a possible function, and such it remains, e v e n w h e n there is neither fulfillment of the function nor a w i s h to fulfill it. What has h a p p e n e d , then, is w h a t Roland Barthes is speaking about w h e n he says that " a s soon as there is a society, every usage is converted into a sign of itself" (Barthes 1968:41). To use a spoon to get food to one's mouth is still, of course, the fulfillment of a function, through the use of an artifact that allows and promotes that function; yet to say that it " p r o m o t e s " the function indicates that the artifact serves a communicative function as well: it communicates the function to be fulfilled. Moreover, the fact that s o m e o n e uses a spoon becomes, in the e y e s of the society that observes it, the communication of a conformity by him to certain u s a g e s (as o p p o s e d to certain others, such as eating with one's h a n d s or sipping food directly from a dish). The s p o o n promotes a certain way of eating and signifies that way of eating, just as the cave promotes the act of taking shelter and signifies the existence of the possible functions; and both objects signify even when they are not being used. . . .

THE ARCHITECTURAL SIGN . . . With this semiotic f r a m e w o r k , one is not obliged to characterize a sign on the basis of either behavior that it stimulates or actual objects that w o u l d v e r i f y its meaning: it is characterized only on the basis of codified meaning that in a given cultural context is attributed to the sign vehicle. (It is true that even the processes of codification belong to the realm of social behavior; but the codes do not admit of empirical verification either, for although based on constancies

6o Umberto Eco inferred from observation of communicative usages, they would always be constructed as structural models, postulated as a theoretical hypothesis.) That a stair has obliged me to go up does not concern a theory of signification; but that occurring with certain formal characteristics that determine its nature as a sign vehicle (just as the verbal signvehicle stairs occurs as an articulation of certain "distinctive units"), the object communicates to me its possible function—this is a datum of culture, and can be established independently of apparent behavior, and even of a presumed mental reaction, on my part. In other words, in the cultural context in which we live (and this is a model of culture that holds for several millennia of history as far as certain rather stable codes are concerned) there exists an architectural form that might be defined as "an inclined progression of rigid horizontal surfaces upward in which the distance between successive surfaces in elevation, r, is set somewhere between 5 and 9 inches, in which the surfaces have a dimension in the direction of the progression in plan, t, set somewhere between 16 and 8 inches, and in which there is little or no distance between, or overlapping of, successive surfaces when projected orthographically on a horizontal plane, the sum total (or parts) falling somewhere between 17 and 48 degrees from horizontal." (To this definition could of course be added the formula relating r to t.) And such a form denotes the meaning "stair as a possibility of going up" on the basis of a code that I can work out and recognize as operative even if, in fact, no one is going up that stair at present and even though, in theory, no one might ever go up it again (even if stairs are never used again by anyone, just as no one is ever going to use a truncated pyramid again in making astronomical observations). Thus what our semiotic framework would recognize in the architectural sign is the presence of a sign vehicle whose denoted meaning is the function it makes possible . . . . The semiotic perspective that we have preferred—with its distinction between sign vehicles and meanings, the former observable and describable apart from the meanings we attribute to them, at least at some stage of the semiotic investigation, and the latter variable but determined by the codes in the light of which we read the sign

Function and Sign: Semiotics of Architecture

61

vehicles—permits us to recognize in architectural signs sign vehicles capable of being described and catalogued, which can denote precise functions provided one interprets them in the light of certain codes, and successive meanings with which these sign vehicles are capable of being filled, whose attribution can occur, as we will see, not only by way of denotation, but also by way of connotation, on the basis of further codes. Significative forms, codes worked out on the strength of inferences from usages and proposed as structural models of given communicative relations, denotative and connotative meanings attached to the sign vehicles on the basis of the codes—this is the semiotic universe in which a reading of architecture as communication becomes viable, a universe in which verification through observable physical behavior and actual objects (whether denotata or referents) would be simply irrelevant and in which the only concrete objects of any relevance are the architectural objects as significative forms. Within these bounds one can begin to see the various communicative possibilities of architecture. Architectural Denotation. The object of use is, in its communicative capacity, the sign vehicle of a precisely and conventionally denoted meaning—its function. More loosely, it has been said that the first meaning of a building is what one must do in order to inhabit it— the architectural object denotes a "form of inhabitation." And it is clear that this denotation occurs even when one is not availing oneself of the denoted inhabitability (or, more generally, the denoted utility) of the architectural object. But we must remember from the outset that there is more to architectural communication than this. When I look at the windows on the facade of a building, for instance, their denoted function may not be uppermost in my mind; my attention may be turned to a window-meaning that is based on the function but in which the function has receded to the extent that I may even forget it, for the moment, concentrating on relationships through which the windows become elements of an architectural rhythm—just as someone who is reading a poem may, without entirely disregarding the meanings of the words there, let them recede into the background and thereby enjoy a certain formal play

62

Umberto Eco

in the sign vehicles' contextual juxtaposition. And thus an architect might present one with some false windows, whose denoted function would be illusory, and these windows could still function as windows in the architectural context in which they occur and be enjoyed (given the aesthetic function of the architectural message) as windows.2 Moreover windows—in their form, their number, their disposition on a facade (portholes, loopholes, curtain wall, etc.)—may, besides denoting a function, refer to a certain conception of inhabitation and use; they may connote an overall ideology that has informed the architect's operation. Round arches, pointed arches, and ogee arches all function in the load-bearing sense and denote this function, but they connote diverse ways of conceiving the function: they begin to assume a symbolic function. Let us return, however, to denotation and the primary, utilitarian function. We said that the object of use denotes the function conventionally, according to codes. Let us here consider some of the general conditions under which an object denotes its function conventionally. According to an immemorial architectural codification, a stair or a ramp denotes the possibility of going up. But whether it is a simple set of steps in a garden or a grand staircase by Vanvitelli, the winding stairs of the Eiffel Tower or the spiraling ramp of Frank Lloyd Wright's Guggenheim Museum, one finds oneself before a form whose interpretation involves not only a codified connection between the form and the function but also a conventional conception of how one fulfills the function with the form. Recently, for example, one has been able to go up also by means of an elevator, and the interpretation of the elevator involves, besides the recognition of the possible function—and rather than being disposed to the motor activity of moving one's feet in a certain way—a conception of how to fulfill the function through the various accessory devices at one's disposal in the elevator. Now the 'legibility ot these features of the elevator might be taken for granted, and presumably their design is such that none of us would have any trouble interpreting them. But clearly a primitive man used to stairs or ramps would be at a loss in front of an elevator; the best intentions on the part of the designer would not result in making the thing clear to him. The designer may

Function and Sign: Semiotics of Architecture

63

have had a conception of the push buttons, the graphic arrows indicating whether the elevator is about to go up or down, and the emphatic floor-level indicators, but the primitive, even if he can guess the function, does not know that these forms are the "key" to the function. He simply has no real grasp of the code of the elevator. Likewise he might possess only fragments of the code of the revolving door and be determined to use one of these as if it were a matter of an ordinary door. We can see, then, that an architect's belief in form that "follows function" would be rather naive unless it really rested on an understanding of the processes of codification involved. In other words, the principle that form follows function might be restated: the form of the object must, besides making the function possible, denote that function clearly enough to make it practicable as well as desirable, clearly enough to dispose one to the actions through which it would be fulfilled. Then all the ingenuity of an architect or designer cannot make a new form functional (and cannot give form to a new function) without the support of existing processes of codification . . . . A work of art can certainly be something new and highly informative; it can present articulations of elements that correspond to an idiolect of its own and not to preexisting codes, for it is essentially an object intended to be contemplated, and it can communicate this new code, implicit in its makeup, precisely by fashioning it on the basis of the preexisting codes, evoked and negated. Now an architectural object could likewise be something new and informative; and if intended to promote a new function, it could contain in its form (or in its relation to comparable familiar forms) indications for the "decoding" of this function. It too would be playing upon elements of preexisting codes, but rather than evoking and negating the codes, as the work of art might, and thus directing attention ultimately to itself, it would have to progressively transform them, progressively deforming already known forms and the functions conventionally referable to these forms. Otherwise the architectural object would become, not a functional object, but indeed a work of art: an ambiguous form, capable of being interpreted in the light of

64

Umberto Eco

various different codes. Such is the case with "kinetic" objects that simulate the outward appearance of objects of use; objects of use they are not, in effect, because of the underlying ambiguity that disposes them to any use imaginable and so to none in particular. (It should be noted that the situation of an object open to any use imaginable—and subject to none—is different from that of an object subject to a number of determinate uses, as we will see.) One might well wish to go further into the nature of architectural denotation (here described only roughly, and with nothing in the way of detailed analysis). But we also mentioned possibilities of architectural connotation, which should be clarified. Architectural Connotation. We said that besides denoting its function the architectural object could connote a certain ideology of the function. But undoubtedly it can connote other things. The cave, in our hypothetical model of the beginning of architecture, came to denote a shelter function, but no doubt in time it would have begun to connote "family" or "group," "security," "familiar surroundings," etc. Then would its connotative nature, this symbolic "function" of the object, be less functional than its first function? In other words, given that the cave denotes a certain basic utilitas (to borrow a term from Koenig), there is the question whether, with respect to life in society, the object would be any less useful in terms of its ability, as a symbol, to connote such things as closeness and familiarity. (From the semiotic point of view, the connotations would be founded on the denotation of the primary utilitas, but that would not diminish their importance.) A seat tells me first of all that I can sit down on it. But if the seat is a throne, it must do more than seat one: it serves to seat one with a certain dignity, to corroborate its user's "sitting in dignity"—perhaps through various accessory signs connoting "regalness" (eagles on the arms, a high, crowned back, etc.). Indeed the connotation of dignity and regalness can become so functionally important that the basic function, to seat one, may even be slighted, or distorted: a throne, to connote regalness, often demands that the person sitting on it sit rigidly and uncomfortably (along with a scepter in his right hand, a globe in the left, and a crown on his head), and therefore seats one "poorly" with respect to the primary utilitas. Thus to seat one is only

Function and Sign: Semiotics of Architecture

65

one of the functions of the throne—and only one of its meanings, the first but not the most important. So the title function should be extended to all the uses of objects of use (in our perspective, to the various communicative, as well as to the denoted, functions), for with respect to life in society the "symbolic" capacities of these objects are no less "useful" than their "functional" capacities. And it should be dear that we are not being metaphorical in calling the symbolic connotations functional, because although they may not be immediately identified with the "functions" narrowly defined, they do represent (and indeed communicate) in each case a real social utility of the object. It is clear that the most important function of the throne is the "symbolic" one and clearly evening dress (which, instead of serving to cover one like most everyday clothing, often "uncovers" for women, and for men covers poorly, lengthening to tails behind while leaving the chest practically bare) is functional because, thanks to the complex of conventions it connotes, it permits certain social relations, confirms them, shows their acceptance on the part of those who are communicating, with it, their social status, their decision to abide by certain rules, and so forth.3 ARCHITECTURAL COMMUNICATION AND HISTORY Primary Functions and Secondary Functions. Since it would be awkward from here on to speak of "functions" on the one hand, when referring to the denoted utilitas, and of "symbolic" connotations on the other, as if the latter did not likewise represent real functions, we will speak of a "primary" function (which is denoted) and of a complex of "secondary" functions (which are connotative). It should be remembered, and is implied in what has already been said, that the terms primary and secondary will be used here to convey, not an axiological discrimination (as if the one function were more important than the others), but rather a semiotic mechanism, in the sense that the secondary functions rest on the denotation of the primary function (just as when one has the connotation of "bad tenor" from the word for "dog" in Italian, cane, it rests on the process of denotation).

66

Umberto Eco

Let us take a historical example where we can begin to see the intricacies of these primary and secondary functions, comparing the records of interpretation history has left us. Architectural historians have long debated the code of the Gothic, and particularly the structural value of the ogive. Three major hypotheses have been advanced: (1) the ogive has a structural function, and the entire lofty and elegant structure of a cathedral stands upon it, by virtue of the miracle of equilibrium it allows; (2) the ogive has no structural value, even if it gives the opposite impression; rather, it is the webs of the ogival vault that have the structural value; (3) the ogive had a structural value in the course of construction, functioning as a sort of provisional framework; later, the interplay of thrusts and counterthrusts was picked up by the webs and by the other elements of the structure, and in theory the ogives of the cross vaulting could have been eliminated. (For a bibliography on this question see Frankl 1960.) No matter which interpretation one might adhere to, no one has ever doubted that the ogives of the cross vaulting denoted a structural function—support reduced to the pure interplay of thrusts and counterthrusts along slender, nervous elements; the controversy turns rather on the referent of that denotation: is the denoted function an illusion? Even if it is illusory, then, the communicative value of the ogival ribbing remains unquestionable; indeed if the ribbing had been articulated only to communicate the function, and not to permit it, that value would, while perhaps appearing more valid, simply be more intentional. (Likewise, it cannot be denied that the word unicorn is a sign, even though the unicorn does not exist, and even though its nonexistence might have been no surprise to those using the term.) While they were debating the functional value of ogival ribbing, however, historians and interpreters of all periods realized that the code of the Gothic had also a "symbolic" dimension (in other words, that the elements of the Gothic cathedral had some complexes of secondary functions to them); one knew that the ogival vault and the wall pierced with great windows had something connotative to communicate. Now what that something might be has been defined time and again, on the basis of elaborate connotative subcodes

Function and Sign: Semiotics of Architecture

67

founded on the cultural conventions and intellectual patrimony of given groups and given periods and determined by particular ideological perspectives, with which they are congruent. There is, for example, the standard romantic and protoromantic interpretation, whereby the structure of the Gothic cathedral was intended to reproduce the vault of Celtic forests, and thus the preRoman world, barbaric and primitive, of druidical religiosity. And in the medieval period, legions of commentators and allegorists put themselves to defining, according to codes of formidable precision and subtlety, the individual meanings of every single architectural element; it will suffice to refer the reader to the catalogue drawn up, centuries later, by Joris Karl Juysmans in his La cathédrale. But there is, after all, a singular document we could mention—a code's very constitution—and that is the justification Suger gives of the cathedral in his De rebus in administratione sua gestis, in the twelfth century (Suger 1867, 1946). There he lets it be understood, in prose and in verse, that the light that penetrates in streams from the windows into the dark naves (or the structure of the walls that permits the light to be offered such ample access) must represent the very effusiveness of the divine creative energy, a notion quite in keeping with certain Neoplatonic texts and based on a codified equivalence between light and participation in the divine essence (see Eco 1956, 1968). We could say with some assurance, then, that for men of the twelfth century the Gothic windows and glazing (and in general the space of the naves traversed by streams of light) connoted "participation" (in the technical sense given the term in medieval Neoplatonism); but the history of the interpretation of the Gothic teaches us that over the centuries the same sign vehicle, in the light of different subcodes, has been able to connote diverse things. Indeed, in the nineteenth century one witnessed a phenomenon typical of the history of art—when in a given period a code in its entirety (an artistic style, a manner, a "mode of forming," independently of the connotations of its individual manifestations in messages) comes to connote an ideology (with which it was intimately united either at the moment of its birth or at the time of its most

68

Umberto Eco

characteristic affirmation). One had at that time the identification "Gothic style = religiosity," an identification that undoubtedly rested on the other, preceding connotative identifications, such as "vertical emphasis = elevation of the soul Godward" or contrast of light streaming through great windows and naves in "shadows = mysticism." Now these are connotations so deeply rooted that even today some effort is required to remember that the Greek temple too, balanced and harmonious in its proportions, could connote, according to another lexicon, the elevation of the spirit to the Gods, and that something like the altar of Abraham on the top of a mountain, could evoke mystical feelings; thus one connotative lexicon may impose itself over others in the course of time and, for example, the contrast of light and shadow becomes what one most deeply associates with mystic states of mind. A metropolis like New York is studded with neo-Gothic churches, whose style (whose "language") was chosen to express the presence of the divine. And the curious fact is that, by convention, these churches still have (for the faithful) the same value today, in spite of the fact that skyscrapers—by which they are now hemmed in on every side, and made to appear very small, almost miniaturized— have rendered the verticality emphasized in this architecture all but indistinguishable. An example like this should be enough to remind us that there are no mysterious "expressive" values deriving simply from the nature of the forms themselves, and that expressiveness arises instead from a dialectic between significative forms and codes of interpretations; for otherwise the Gothic churches of New York, which are no longer as distinctively attenuated and vertical as they used to be, would no longer express what they used to, while in fact they still do in some respects, and precisely because they are "read" on the basis of codes that permit one to recognize them as distinctively vertical in spite of the new formal context (and new code of reading) the advent of the skyscraper has now brought about. Architectural Meanings and History. It would be a mistake, however, to imagine that by their very nature architectural sign-vehicles would denote stable primary functions, with only the secondary functions varying in the course of history. The example of ogival ribbing has already shown us a denoted function undergoing curi-

Function and Sign: Semiotics of Architecture

69

ous fluctuations—it was considered by some effective and essential, but by others provisional or illusory—and there is every reason to believe that in the course of time certain primary functions, no longer effective, would not longer even be denoted, the "addresses" no longer possessing the requisite codes. So, in the course of history, both primary and secondary functions might be found undergoing losses, recoveries, and substitutions of various kinds. These losses, recoveries, and substitutions are common to the life of forms in general, and constitute the norm in the course of the reading of works of art proper. If they seem more striking (and paradoxical) in the field of architectural forms, that is only because according to the common view one is dealing there with functional objects of an unequivocally indicated, and thus univocally communicative, nature; to give the lie to such a view, there is the story—its very currency puts its authenticity in doubt, but if untrue it is in any case credible—about the native wearing an alarm clock on his chest, an alarm clock interpreted as a pendant (as a kind of "kinetic jewelry," one might say) rather than as a timepiece: the clock's measurement of time, and indeed the very notion of "clock time" (see Bergson), is the fruit of a codification and comprehensible only on the basis of it. One type of fluctuation in the life of objects of use can therefore be seen in the variety of readings to which they are subject, regarding both primary and secondary functions . . . . 4 ARCHITECTURAL CODES What Is a Code in Architecture? Architectural signs as denotative and connotative according to codes, the codes and subcodes as making different readings possible in the course of history, the architect's operation as possibly a matter of "facing" the likelihood of his work being subject to a variety of readings, to the vicissitudes of communication, by designing for variable primary functions and open secondary functions (open in the sense that they may be determined by unforseeable future codes)—everything that has been said so far might suggest that there is little question about what is meant by code. As long as one confines oneself to verbal communication, the notion is fairly clear: there is a code-language, and there are certain

70 Umberto Eco connotative subcodes. But when, in another section of this study, we went on to consider visual codes, for example, we found we had to list a number of levels of codification (including, but not limited to, iconic and iconographic codes), and in the process to introduce various "clarifications" of the concept of code, and on the different types of articulation a code may provide for.5 We also saw the importance of the principle that the elements of articulation under a given code can be syntagms of another, more "analytic" code, or that the syntagms of one code can turn out to be elements of articulation' of another, more "synthetic" code. This should be kept in mind when considering codes in architecture, for one might be tempted to attribute to an architectural code articulations that belong really to some code, either more analytic or more synthetic, lying outside architecture. We can expect some problems, then, in the definition of the codes of architecture. First of all, from the attempts there have been to date to spell out aspects of architectural communication, we can see that there is the problem of neglecting to consider whether what one is looking at is referable to a syntactic code rather than a semantic code—that is, to rules concerning, rather than the meanings conventionally attributed to, individual sign-vehicles, the articulation of certain significative structures separable from these sign vehicles and their meanings—or for that matter to some underlying technical convention. Catchwords like "semantics of architecture" have led some to look for the equivalent of the "word" of verbal language in architectural signs, for units endowed with definite meaning, indeed for symbols referring to referents. But since we know there can be conventions concerning only the syntactic articulation of signs, it would be appropriate to look also for purely syntactic codifications in architecture (finding such codifications and defining them with precision, we might be in a better position to understand and classify, at least from the point of view of semiotics, objects whose once denoted functions can no longer be ascertained, such as the menhir, the dolmen, the Stonehenge construction). Then, too, in the case of architecture, codes of reading (and of construction) of the object would have to be distinguished from codes of

Function and Sign: Semiotics of Architecture

71

reading (and of construction) of the design for the object (admittedly we are considering here only a semiotics of architectural objects, and not a semiotics of architectural designs). Of course the notational codes of the design, while conventionalized independently, are to some extent derivatives of the codes of the object: they provide ways in which to "transcribe" the object, just as to transcribe spoken language there are conventions for representing such elements as sounds, syllables, or words. But that does not mean a semiotic investigation of the architectural design would be without some interesting problems of its own—there are in a design, for example, various sytems of notation (the codes operative in a plan are not quite the same as those operative in a section or in a wiring diagram for a building), 6 and in these systems of notation there can be found iconic signs, diagrams, indices, symbols, qualisigns, sinsigns, etc., perhaps enough to fill the entire gamut of signs proposed by Peirce. Much of the discussion of architecture as communication has centered on typological codes, especially semantic typological codes, those concerning functional and sociological types; it has been pointed out that there are in architecture configurations clearly indicating "church," "railroad station," "palace," etc. We will return to typological codes later, but it is clear that they constitute only one, if perhaps the most conspicious, of the level of codification in architecture. In attempting to move progressively back from a level at which the codes are so complex and temporal—for it is clear that "church" has found different articulations at different moments in history—one might be tempted to hypothesize for architecture something like the "double articulation" found in verbal languages, and assume that the most basic level of articulation (that is, the units constituting the "second" articulation) would be a matter of geometry. If architecture is the art of the articulation of spaces (Zevi i960, 1957), then perhaps we already have, in Euclid's geometry, a good definition of the rudimentary code of architecture. Let us say that the second articulation is based on the Euclidean crtoixeia (the "elements" of classical geometry); then the "first" articulation would involve certain higher-level spatial units, which could be called cho-

72

Umberto Eco

rentes, with these combining into spatial syntagms of one kind or another.7 In other words, the angle, the straight line, the various curves, the point, etc., might be elements of a second articulation, a level at which the units are not yet significant (endowed with meaning) but are distinctive (having differential value); the square, the triangle, the parallelogram, the ellipse—even rather complicated irregular figures, as long as they could be defined with geometric equations of some kind—might be elements of a first articulation, a level at which the units begin to be significant; and one rectangle within another might be an elementary syntagmatic combination (as in some window-wall relationship), with more complex syntagms to be found in such things as space-enclosing combinations of rectangles or articulations based on the Greek-cross plan. Of course solid geometry suggests the possibility of a third level of articulation, and it could be assumed that further articulative possibilities would come to light with the recognition of non-Euclidean geometries. The trouble is that this geometric code would not pertain specifically to architecture. Besides lying behind some artistic phenomena—and not just those of abstract, geometric art (Mondrian), because it has long been held that the configurations in representational art can be reduced to an articulation, if perhaps a quite complex one, of primordial geometric elements—the code clearly underlies the formulations of geometry in the etymological sense of the word (surveying) and other types of "transcription" of terrain (topographic, geodetic, etc.). It might even be identified with a "gestaltic" code presiding over our perception of all such forms. What we have here, then, is an example of one sort of code one can arrive at when attempting to analyze the elements of articulation of a certain "language": a code capable of serving as a metalanguage for it, and for a number of other more synthetic codes as well. So it would be better to pass over a code of this kind, just as in linguistics one passes over the possibility of going beyond "distinctive features" in analyzing phonemes. Admittedly such analytic possibilities might have to be explored if one had to compare architectural phenomena with phenomena belonging to some other "language," and thus had to find a metalanguage capable of describing them in the same terms—for instance, one might wish to "code" a

Function and Sign: Semiotics of Architecture 73 certain landscape in such a way as to be able to compare it with certain proposed architectural solutions, to determine what architectural artifacts to insert in the context of that landscape, and if one resorted to elements of the code of solid geometry (pyramid, cone, etc.) in defining the structure of the landscape, then it would make sense to describe the architecture in the light of that geometric code, taken as a metalanguage.8 But the fact that architecture can be described in terms of geometry does not indicate that architecture as such is founded on a geometric code. After all, that both Chinese and words articulated in the phonemes of the Italian language can be seen as a matter of amplitudes, frequencies, wave forms, etc., in radioacoustics or when converted into grooves on a disk does not indicate that Chinese and Italian rest on one and the same code; it simply shows that the languages admit of that type of analysis, that for certain purposes they can be reduced to a common system of transcription. In fact there are few physical phenomena that would not permit analysis in terms of chemistry or physics at the molecular level, and in turn an atomic code, but that does not lead us to believe the Mona Lisa should be analyzed with the same instruments used in analyzing a mineral specimen. Then what more properly architectural codes have emerged in various analyses or, recently, "semiotic" readings of architecture? Varieties of Architectural Code. It would appear, from those that have come to light, that architectural codes could be broken down roughly as follows: 1. Technical codes. To this category would belong, to take a ready example, articulations of the kind dealt with in the science of architectural engineering. The architectural form resolves into beams, flooring systems, columns, plates, reinforced-concrete elements, insulation, wiring, etc. There is at this level of codification no communicative "content," except of course in cases where a structural (or technical) function or technique itself becomes such; there is only a structural logic, or structural conditions behind architecture and architectural signification—conditions that might therefore be seen as somewhat analogous to a second articulation in verbal languages, where though one is still short of meanings there are certain formal conditions of signification.9

74

Umberto Eco

2. Syntactic codes. These are exemplified by typological codes concerning articulation into spatial types (circular plan, Greek-cross plan, "open" plan, labyrinth, high-rise, etc.)/ but there are certainly other syntactic conventions to be considered (a stairway does not as a rule go through a window, a bedroom is generally adjacent to a bathroom, etc.). 3. Semantic codes. These concern the significant units of architecture, or the relations established between individual architectural sign-vehicles (even some architectural syntagms) and their denotative and connotative meanings. They might be subdivided as to whether, through them, the units (a) denote primary functions (roof, stairway, window); (b) have connotative secondary functions (tympanum, triumphal arch, neo-Gothic arch); (c) connote ideologies of inhabitation (common room, dining room, parlor); or (d) at a larger scale have typological meaning under certain functional and sociological types (hospital, villa, school, palace, railroad station). 10 The inventory could of course become quite elaborate—there should, for instance, be a special place for types like "garden city" and "new town," and for the codifications emerging from certain recent modi operandi (derived from avant-garde aesthetics) that have already created something of a tradition, a manner, of their own. But what stands out about these codes is that on the whole they would appear to be, as communicative systems go, rather limited in operational possibilities. They are, that is, codifications of already ivorked-out solutions, codifications yielding standardized messages—this instead of constituting, as would codes truly on the model of those of verbal languages, a system of possible relationships from which countless significantly different messages could be generated. A verbal language serves the formulation of messages of all kinds, messages connoting the most diverse ideologies (and is inherently neither a class instrument nor the superstructure of a particular economic base). (For Joseph Stalin's well-known views on linguistics see Stalin 1951.) Indeed the diversity of the messages produced under the codes of a verbal language makes it all but impossible to identify any overall ideological connotations in considering broad samplings of them. Of course this characterization might be chal-

Function and Sign: Semiotics of Architecture

75

lenged, for there is some evidence to support the theory that the very way in which a language is articulated obliges one speaking it to see the world in a particular way (there might be, then, ideological bias and connotation of some kind inherent in the language). (That language determines the way in which one sees reality, see Whorf 1956.) But even given that, on the most profound, ultimate level, one could take a verbal language as a field of (nearly absolute) freedom, in which the speaker is free to improvise novel messages to suit unexpected situations. And in architecture, if the codes are really those indicated above, that does not seem to be the case. The point is not that in articulating a church, for example, the architect is in the first place obeying a socio-architectural prescription that churches be made and used (about this sort of determinant we will have more to say later). And in the end he would be free to try to find and exploit some way in which to make a church that while conforming to its type would be somewhat different from any that had yet appeared, a church that would thereby provide a somewhat unaccustomed, "refreshing" context in which to worship and imagine the relationship with God. But if at the same time, in order for it to be a church, he must unfailingly articulate the building in manifold conformity to a type ("down to the hardware," one might say), if the codes operative in architecture allow only slight differences from a standardized message, however appealing, then architecture is not the field of creative freedom some have imagined it to be, but a system of rules for giving society what it expects in the way of architecture. In that case architecture might be considered not the service some have imagined it to be—a mission for men of unusual culture and vision, continually readying new propositions to put before the social body—but a service in the sense in which waste disposal, water supply, and mass transit are services: an operation that is, even with changes and technical refinements from time to time, the routine satisfaction of some preconstituted demand. It would appear to be rather impoverished as an art, then, also, if it is characteristic of art, as we have suggested elsewhere, to put before the public things they have not yet come to expect (Eco 1968: ch. A.3).

76

Umberto Eco

So the codes that have been mentioned would amount to little more than lexicons on the model of those of iconographic, stylistic, and other specialized systems, or limited repertories of set constructions. They establish not generative possibilities but ready-made solutions, not open forms for extemporary "speech" but fossilized forms—at best, "figures of speech," or schemes providing for formulaic presentation of the unexpected (as a complement to the system of established, identified, and never really disturbed expectations), rather than relationships from which communication varying in information content as determined by the "speaker" could be improvised. The codes of architecture would then constitute a rhetoric in the narrow sense of the word: a store of tried and true discoursive formulas. (That is, they would constitute a rhetoric in the sense of the term discussed in Eco 1968: par. A.4.2.2.) And this could be said not only of the semantic codes, but also of the syntactic codifications, which clearly confine us to a certain quite specialized "grammar" of building, and the technical codes, for it is obvious that even this body of "empty" forms underlying architecture (column, beam, etc.) is too specialized to permit every conceivable architectural message: it permits a kind of architecture to which civilization in its evolving technologies has accustomed us, a kind relating to certain principles of statics and dynamics, certain geometric concepts, many of them from Euclid's geometry, certain elements and systems of construction—the principles, concepts, elements, and systems that, proving relatively stable and resistant to wear and tear, are found codified under the science of architectural engineering. ARCHITECTURE AS MASS COMMUNICATION? Mass Appeal in Architecture. If architecture is a system of rhetorical formulas producing just those messages the community of users has come to expect (seasoned with a judicious measure of the unexpected), what then distinguishes it from various forms of mass culture? The notion that architecture is a form of mass culture has become rather popular, 11 and as a communicative operation directed toward large groups of people and confirming certain widely subscribed to attitudes and ways of life while meeting their expecta-

Function and Sign: Semiotics of Architecture

77

tions, it could certainly be called mass communication loosely, without bothering about any detailed criteria. But even under more careful consideration/1 architectural objects seem to have characteristics in common with the messages of mass communication. To mention a few: —Architectural "discourse" generally aims at mass appeal: it starts with accepted premises, builds upon them well-known or readily acceptable "arguments," and thereby elicits a certain type of consent. ("This proposition is to our liking; it is in most respects something we are already familiar with, and the differences involved only represent a welcome improvement or variation of some kind.") —Architectural discourse is psychologically persuasive: with a gentle hand (even if one is not aware of this as a form of manipulation) one is prompted to follow the "instructions" implicit in the architectural message; functions are not only signified but also promoted and induced, just as certain products and attitudes are promoted through "hidden persuasion," sexual associations, etc. —Architectural discourse is experienced inattentively, in the same way in which we experience the discourse of movies and television, the comics, or advertising—not, that is, in the way in which one is meant to experience works of art and other more demanding messages, which call for concentration, absorption, wholehearted interest in interpreting the message, interest in the intentions of the "addresser."13 —Architectural messages can be interpreted in an aberrant way, and without the "addressee" being aware of thereby perverting them. Most of us would have some sense of being engaged in a perversion of the object if we were to use the Venus de Milo for erotic purposes or religious vestments as dustcloths, but we use the cover of an elevated roadway for getting out of the rain or hang laundry out to dry over a railing and see no perversion in this. —Thus architecture fluctuates between being rather coercive, implying that you will live in such and such a way with it, and rather indifferent, letting you use it as you see fit. —Architecture belongs to the realm of everyday life, just like pop music and most ready-to-wear clothing, instead of being set apart like "serious" music and high fashion.

78

Umberto Eco

—Architecture is a business.M It is produced under economic conditions very similar to the ones governing much of mass culture, and in this too differs from other forms of culture. Painters may deal with galleries, and writers with publishers, but for the most part that has to do with their livelihood and need not have anything to do with what they find themselves painting and writing; the painter can always pursue painting independently, perhaps while making a living in some other way, and the writer can produce works for which there is no market, perhaps with no thought of having them published, but the architect cannot be engaged in the practice of architecture without inserting himself into a given economy and technology and trying to embrace the logic he finds there, even when he would like to contest it. . . . EXTERNAL CODES Architecture as Based on Codes External to It. To recapitulate: (a) we began with the premise that architecture would, to be able to communicate the functions it permits and promotes, have to be based on codes; (b) we have seen that the codes that could properly be called architectural establish rather limited operational possibilities, that they function not on the model of a language but as a system of rhetorical formulas and already produced message-solutions; (c) resting on these codes, the architectural message becomes something of mass appeal, something that may be taken for granted, something that one would expect; (d) yet it seems that architecture may also move in the direction of innovation and higher information-content, going against existing rhetorical and ideological expectations: (e) it cannot be the case, however, that when architecture moves in this direction it departs from given codes entirely, for without the basis of a code of some kind, there would be no effective communication. . . . It goes without saying, for instance, that an urban designer could lay out a street on the basis of the lexicon that embraces and defines the type "street"; he could even, with a minor dialectic between redundency and information, make it somewhat different from previous ones while still operating within the traditional urbanistic system.

Function and Sign: Semiotics of Architecture

79

When, however, Le Corbusier proposes his elevated streets (closer to the type "bridge" than to the type "street"), he moves outside the accepted typology, which has streets at ground level or, if elevated, elevated in a different fashion and for different reasons—and yet he does so with a certain assurance, believing that this new sign, along with the rest of his proposed city, would be accepted and comprehended by the users. Now whether such a belief is justified or not, it would have to be based on something like this: the architect has preceded architectural design with an examination of certain new social exigencies, certain "existential" desiderata, certain tendencies in the development of the modern city and life within it, and has traced out, so to speak, a semantic system of certain future exigencies (developing from the current situation) on the basis of which new functions and new architectural forms might come into being. In other words, the architect would have identified (a) a series of social exigencies, presumably as a system of some kind, (b) a system of functions that would satisfy the exigencies, and that would become sign vehicles of those exigencies, and (c) a system of forms that would correspond to the functions, and that would become sign vehicles of those functions. From the point of view of common sense, this means that to produce the new architecture Le Corbusier was obliged, before thinking like an architect, to think like a sociologist, an anthropologist, a psychologist, an ideologist, etc., and we will return to that shortly. But first we might consider the peculiarity of the phenomenon from the semiotic point of view. Only at the last level, the level of (c), do we find forms that could be understood as "architecture." So while the elements of architecture constitute themselves a system, they become a code only when coupled with systems that lie outside architecture. . . . What about architecture, then, if we accept the hypotheses above? Let us use X for the system of architectural forms, Y for the system of functions, and K for the system of social exigencies, or the anthropological system—an x might be a table of a certain width, which permits and signifies a certain function y (to eat at a considerable distance from one another, let us say), which in turn allows

8o

Umberto Eco

the realization of an anthropological value k ("formal" relationship), whose sign vehicle that function has become. Then the units in X, as spatial forms, admit of several kinds of description—two dimensional (through a set of drawings or a photograph), verbal (through an oral or written description), mathematical (through a series of equations), etc.; the units in Y, as functions, admit of either verbal description or representation in terms of some iconic (cinematographic, for example), kinesic, or other kind of system for "transcribing" functions; and the units in K, as anthropological values, can be described verbally. Now it is clear that while a form x is being used it might seem (to the user) quite closely tied to a function y and an anthropological value k— just as closely as a meaning seems (to the speaker) tied to a verbal sign vehicle. But from the point of view of semiotics, it is possible to describe the units of each of these three systems independently, without, that is, having recourse to the units of either of the other two. This is something that was never envisaged by those who have considered the notion of meaning suspect, because up to now studies in semantics have been conducted inside the circle of verbal "interpretants." So above and beyond what else it offers, semiotics shows us the possibility of investigating systems of signs where the planes of expression and content are not inseparable—or at least where they can be more successfully separated. The Anthropological System. But in introducing this K, this anthropological system, have we jeopardized the semiotic framework behind everything we said before? Having said that architecture has to elaborate its sign vehicles and messages with reference to something that lies outside it, are we forced to admit its signs cannot, after all, be adequately characterized without bringing something like referents back into the picture? We have argued that semiotics must confine itself to the left side of the Ogden-Richards triangle—because in semiotics one studies codes as phenomena of culture—and, leaving aside verifiable realities to which the signs may refer, examine only the communicative rules established within a social body: rules of the equivalence between sign vehicles and meanings (the definition of the latter being possible only through interpretants or other sign vehicles by means

Function and Sign: Semiotics of Architecture

81

of which the meanings may be signified), and rules regarding the syntagmatic combination of the elements of the paradigmatic repertories. This means not that the referent is nonexistent, but that it is the object of other sciences (physics, biology, etc.): semiotics can, and must, confine itself to the universe of the cultural conventions governing communicative intercourse. If for architecture, then, or for any other system of signs, we had to admit that the plane of content involved something that did not belong to the semiotic universe, we would be faced with a phenomenon confounding semiotics, or at any rate confounding all the notions we have elaborated, here and elsewhere, on semiosis.15 So it is not casually that we have been referring to an anthropological "system"; we have been referring, that is, to facts that while belonging to the universe of the social sciences may nevertheless be seen as already codified, and thus reduced to a cultural system. . . . To put it differently, let us say that the architect has decided to restructure the urban fabric of a city (or the "shape of landscape" in a certain area) from the point of view of the perceptibility of its "image."' 6 He might then base his operation upon rules of a code concerned precisely with phenomena of image-recognition and orientation (a code that could be elaborated on the basis of data from interviews and basic research on perception, and perhaps even take into account exigencies of commerce or circulation, medical findings on factors contributing to stress, etc.). But then the validity and significance of the operation, based on that code, would depend upon confining oneself to that particular point of view. As soon as it became necessary for the architect to relate his architecture to some other system of social phenomena as well—the one dealt with in proxemics, let us say—the code concerned with image-recognition and orientation would have to be broken down and integrated with a code concerning proxemic phenomena; and since there would no doubt be more than just these two external systems to relate to, it would become necessary to find the relations between a number of different systems tracing them all back to an underlying Ur-code common to all of them, on which elaboration of the new architectural solutions would ultimately have to be based.' 7 So the architect, in practice, is continually obliged to be something

8a

Umberto Eco

other than an architect. Time and again he is forced to become something of a sociologist, a psychologist, an anthropologist, a semiotician . . . And that he can rely in this to some extent on teamwork—that is, on having experts in the various fields working with him—does not change the situation very much, even if teamwork makes it seem less a matter of guesswork. Forced to find forms that will give form to systems over which he has no power, forced to articulate a language that has always to express something external to it—we said there were possibilities of the poetic function and self-reflexiveness in architecture, but the fact remains that because of its very nature (and even though it has traditionally been understood as a matter of pure "arrangement," regarding only its own forms) these can never "take over" in it, as they can in other types of discourse, such as in poetry, painting, or music—the architect finds himself obliged in his work to think in terms of the totality, and this he must do no matter how much he may seem to have become a technician, a specialist, someone intent on specific operations rather than general questions. Conclusion. One might at this point be left with the idea that having the role of supplying "words" to signify "things" lying outside its province, architecture is powerless to proceed without a prior determination of exactly what those "things" are (or are going to be). Or one might have come to a somewhat different conclusion: that even though the systems of functions and values it is to convey are external to it, architecture has the power, through the operation of its system of stimulative sign-vehicles, to determine what those functions and values are going to be—restricting men to a particular way of life dictating laws to events. These both go too far, and they go along with two unfortunate ideas of the role of the architect. According to the first, he has only to find the proper forms to answer to what he can take as "programmatic" givens; here he may accept on faith certain sociological and ideological determinations made by others, which may not be well founded. According to the second, the architect (and we know what currency this delusion has enjoyed) becomes a demiurge, an artificer of history. This alternative to these varieties of overconfidence has already

Function and Sign: Semiotics of Architecture

83

been suggested: the architect should be designing for variable primary functions and open secondary functions.

NOTES 1. See Norberg-Schulz 1965: ch. 5. Cf. Dorfles 1959, 1962, esp. ch. 5; Langer 1953, the chapters on virtual space; Brandi 1956, i960, 1968; Bettini 1958: 191-203; and Choay 1965. 2. In this case it is the aesthetic function that is predominant in the architectural message, what Roman Jakobson, speaking of acts of verbal communication, has termed the poetic function (1966: 350-77). But architectural messages display also the five other communicative functions listed by Jakobson: architecture involves communication that is connative (or imperative, making one inhabit it in a certain way), emotive (think of the calm of a Greek temple, the turbulence of a baroque church), phatic (obviously in the many attention-getting devices of architecture— the phatic function might be found to be predominant, then, in such messages as obelisks, arches, and tympana—but also at the level of urban fabric, where "channels" are opened and established for architectural messages, as in a piazza's ensuring continued attention to the facades of the buildings that surround it), metalingual (where, for one example, to relieve any confusion about the code for interpreting the message architecture assumes a self-explaining, or "glossing," function—think of the benches built into certain otherwise inhospitable American plazas), and of course referential (what we will be concerned with here for the most part—that is, the denotations and connotations of architectural objects). 3. We should note that the symbolic value of forms was not entirely ignored by the theorists of functionalism: see Sullivan 1947: 202-13; and Fusco (1964) shows that their symbolic value was important not only to Sullivan but also to Le Corbusier. On the connotative value of forms at the level of urban design—turning to the relational forms in the fabric of large urban areas—see Lynch, i960: esp. p. 91: cities are to be given forms that can stand as symbols for urban life. 4. See Argan, et al. 1965, particularly the title essay, where the notion of works that remain "open" (the "opera aperta") is applied to architectural design. One way of understanding the "openness" of architectural and urbanistic objects is suggested in Roland Barthes' "Semiology and Urbanism" [see paper 3, this volume). Aggreeing with certain views held by Jacques Lacan—discussed in Eco 1968: ch. D. 5, "La struttura e I'assenza"—Barthes believes that with regard to the city the question of meaning becomes less important than a detailed analysis of abstracted "signifiers." Thus, "in [an] effort to approach the city semantically, we must try to understand the play of the signs, to understand that any city is a structure but one need never try, or even wish, to fill this structure," for "semiology at present never posits the existence of a final signified," and "any cultural (or, for that matter, psychological) complex confronts us with infinite metaphorical chains, in which the signified is always deferred or becomes itself a signifier." Now it is true that any city confronts us with phenomena of enrichment (and

«4

5. 6.

7.

8.

9. 10.

11. 12. 13.

Umberto Eco substitution) of meaning, but the semantic value of the city emerges not only when one sees it as a structure that generates meaning: it emerges also when, in experiencing it, one is filling it with concrete significations. Indeed, to oppose to the concrete process of signification—in the light of which the city is designed— the notion of a free play of pure sign-vehicles might be to empty the activity of architecture of much of its creative thrust. For if this notion were carried to an extreme, and the significative power of a city considered really infinite—as infinite as the significative power of verbal languages, which in spite of the fact that man has little say with regard to their constitution and laws still permit him to be adequately "spoken"—then there would no longer seem to be any point in designing a " n e w " city: in any existing city there would already be the elements of an infinite number of possible combinations, permitting every type of life within that form. In reality, the problem of architecture is that of defining the limit beyond which an existing form no longer allows the type of life one has in mind, the limit beyond which the architectural sign-vehicles that pass before one appear no longer as a matrix of freedom but as the very image of a domination, of an ideology that imposes, through the rhetorical forms it has generated, various modes of enslavement. Eco 1968: ch. B . I - 3 , "I codici visivi," "II mito della doppia articolazione," and "Articolazioni dei codici visivi." Through the use of the wrong code, then, a plan might be read as a section or vice versa; see the amusing situation described in Klaus Koenig 1964: ch. 8, 1967: 107-17. The term choreme is derived from x 1 and such that at least one (b) is the contrary of the other signs. But we are also faced with syntagmatic rules capable of organizing these basic elements into definite sets since we have: a given element (a) which can only be associated with the element (b') for instance. We would say then that (b') saturates (a) in the same way that (dog) saturates (barks) in linguistics. If we agree to free ourselves from all prejudices, we will admit that in such a case we are dealing with a system of signs. One will recall that it was indeed when Barthes, using an aesthetic-architectural criterion, stated that the column of a building, on the one hand might be in an actual relation of contiguity with other parts of the building (the architrave, for example)—a syntagmatic relation, i.e., a string of "details" on the level of the whole building—or, on the other hand, if this column were for example Doric, it would call for a comparison with other architectural orders, Ionic, Corinthian, etc., and this would be a virtual relation of substitution—associative or paradigmatic—corresponding to the variations of style, for instance, in a single element of a building. We note that this view of the object does not exclude other approaches. A construction engineer could very easily demonstrate that this architectural object pertains in the same manner to a technological logic in the sense that given an element (a), considering the limitations imposed by the properties of the materials (resistance, span, etc.), it can only be associated with a certain other element (b). As for the economist, he could point out that budgetary reasons have led to a serialization, a "fragilization", an "impoverishment" of the architectural object. Thus in this manner the different logics cancel each other out. However that may be, as long as we are on this level of research, let us note that the work being done is of a structural orientation,

A New Semiological Approach to the City

141

although we cannot really say that the city speaks a discourse, at least in the habitual sense of the word. Simply, we are dealing with a plurality of logics (operating from aesthetic, functional, etc., criteria), the rules of combination of which (on the syntagmatic and paradigmatic levels) are capable of generating various spaces or, if we prefer, of arranging into coherent sets, and according to specific rules, a limited number of elements which the selected criterion permits us to identify. Why should we not develop our researches in this direction? Because, of course, by doing so we would simply be proposing a new formulation of a set of items already known. The architect, the technologist, and the aesthetician could, if it aided them in their researches, rearrange their knowledge so as to make their approaches fit into new settings (cf. Castex and Panerai 1974; and Boudon 1973). But this is not our problem. THE " S E M A N T I Z A T I O N " OF THE CITY In the case of the process of semantization of a city, the criterion for the segmentation of the minimum significant units is thus provided by discourse and not by aesthetics or technology, etc. We may specify that we are faced with two possibilities: the semantization process can take place either on the up-side (on the side of the senders) or on the down-side (on the side of the receivers) of the act of communication, something which, incidentally, makes the direction of this communication rather problematical. 1. The Semantization of Urban Space on the Up-Side of Communication. In the West, as Choay remarks, starting from the fifteenth century that urban space is written about, described, and drawn before it is constructed, so that, from this time on, the criterion of semantization is provided us by the architects, the landscape architects, the engineers, etc., who are no longer content just to "construct." Sylvia Ostrowetsky (1972, 1974) considers that an "ideological attitude" ("technicist," democratic, determinist, technocratic, liberal, culturalist) in a sense forms the basis from which the process of the construction of the city will develop. The revelation of such ideological atti-

142 Richard Fauque tudes is possible thanks to the rationalizing discourse which always accompanies the doing. "I have placed these fountains here because it is the center and. . . . " This ideological enterprise thus writes itself into—or tries to write itself into—the urban fabric. Hence the contents of the urban fabric can serve as the criterion of segmentation of the urban continuum which is the city-project. We note that this discourse uses a substitutive sign system, according to the expression of Eric Buyssens, since it is relayed by plans, drawings, scale models, etc., which in themselves constitute autonomous systems. Let us next remark that this ideological discourse is not pure. It depends on a certain number of constraints imposed by the functional, economic, technological, aesthetic, etc., discourses (we meet them again) which are unceasingly obliterating it by superimposition.4 In that case, how is it possible to isolate the ideological discourse from all the other discourses which are themselves intervening in its elaboration? This is why it seems to us perhaps premature to want to situate research in urban semiology from the start on the up-side of this (supposed) communication, thus on the side of the senders. Doubtlessly this is actually where we run the risk of meeting the greatest difficulties. 2. The Semantization of Urban Space on the Down-Side of Communication. Here, it is not the senders' discourse but the reading of the city by the receivers that offers us the means of segmentation. However, as there are different levels of possible readings, we will have to make new distinctions: (a) In effect, there is first and foremost an honest, conscientious, and explicit communication which occurs between sender and receiver each time we deal with a signalization, regardless of its nature: traffic lights, the " N o entry" sign, but also commercial signs (the green cross of the chemist's, the red cigar of the tobacconist's, the golden coat of arms sign over a notary's office, the horsehead of a butcher's, the scallop-shell of the Shell gas stations, etc.); ideograms of which the symbol is immediately recognizable (the male and female silhouettes of the public W.C.); arbitrary ideographic signs (the stars indicating the category of a hotel); graffiti; advertising or political posters, etc. In all these cases, we confront a dis-

A New Semiological Approach to the City

143

course belonging to a semiology of communication, according to Mounin's expression. 5 But is this an interesting point of view for us? No . . . and yet, yes, since in fact these systems of communication are not limited to conveying only the significations that we ordinarily attribute to them. Thus, some months ago the traffic of Toulouse was thrown into a certain disorder: " N o entry" sings, one-way streets, etc., blossomed almost everywhere, changing in no time the physiognomy of the traffic and provoking numerous reactions amongst the driving population. Consequently, the "No entry" sign, the one-way street, no longer constituted merely an interdiction or prescription. In fact, any motorist could add and every Toulousian could be caught thinking something along the lines of: "Honestly you never know what they're going to imagine next," or "one day it's yes the next day it's no. Will they ever agree on what they want?" Some significations thus show quite clearly: "aggressivity" and "gratuitousness" directed toward the users in the first case; "instability" and "confusion" in the second. To put it briefly, we have two overlapping systems: /

• —

* —

= ^

signifier

—

/ »

'

"no entry" V

V signified

I

signifier

'

/

signified "honestly w e don't k n o w w h a t they're g o i n g to imagine next . . . " w h o s e semantic features are: "aggressivity" "gratuitousness"

(b) Without elaborating here, since it will be examined in the second part, we mention here that "all the rest" comes next: the monuments, the public buildings, the private architecture, all the roads

144

Richard Fanque

(street, avenue, boulevard, pavement), the squares, the rivers, the streams, the canals, the gardens, the parks, the fountains, the "city gadgets" (benches, bus shelters, litter boxes, etc.), all of which, we suppose, carry signification the status of which it is not easy to define. Actually, we may note that we shall never have proof that the city speaks to us; on the other hand it will never be proved that it is the users that make it speak. We may think we shall find such proof in the fact that there is a certain homeomorphism between the different readings, considering thus the variations in readings as the consequence of a certain freedom of interpretation on the part of the individuals. In other words, w e consider that in order for such a homeomorphism to appear, the city in itself must be a text. However, let us note that we could just as well deduce the opposite from the same facts: if the various readings seem to belong to the same family, it may well be that this does not result form the text as such, but from the fact that the various perceptions are ideologically related.

HYPOTHESES A N D METHODS Preliminary Remarks. So we now know, even in a very approximate way, what we are going to talk about. However we do not yet know the way in which we shall have to talk about it. Nonetheless we could say that: (1) The object of our investigations should lead us to confirm or deny the thesis that the city constitutes a system (or a plurality of systems) of signs. (2) We have a double means of establishing this, in referring either to (a) the observation of urban behaviors: Which itinerary is chosen for a sweetheart's walk? Where do we take a member of the family or a friend visiting the city? Why? That is, do we show them what is most beautiful? Most interesting? Most picturesque? Most typical? or to (b) the reactions of persons surveyed in relation to what they see, feel, breathe, etc., while strolling around the city. Strictly speaking, we should follow the same individuals going to

A New Semiological Approach to the City

145

and coming from their place of work, walking, shopping, etc. In effect there are some urban itineraries that are rather automatic, followed without thinking because they are daily, regular journeys: place of work place of residence. On the other hand, there are some itineraries which demand a complete availability ("shopping"). We walk around at random without having any preconceived idea of where we want to go. At one extreme, this kind of walking is done in the ludic mode. But this is not all: the nostalgic walker, the unhappy lover, the walker in love, all of them have a characteristic way of perambulation, and we would bet that each time it corresponds to a different reading of the city. Would one imagine that Rastignac, come to conquer Paris, has the same reading of the city as—let us say—Thérèse Desqueyroux, ending up there accidentally with a heart full of bitter memories? Finally, the motorist and the pedestrian, for instance, do not have the same perception: the flow of the visual and auditory images is not the same. Our exteroceptive impressions vary according to whether we are the prisoner of a traffic jam and shut in the car, or whether w e are pushed, squashed, and bruised by a mass of people at the bus station during the rush hour. (1) Thus we have, on the side of the reader, three series of variables which can explain the variations in reading: (a) the quality of the itinerary: daily vs. exceptional, for instance; (b) the variations which refer to the reader's affective situation: for instance happy vs. unhappy; (c) those which refer to the reader's "material" situation: pedestrian vs. motorist for instance. (2) On the other side, the side of the city, we have: (a) auditory images Jackhammer, horns, sirens; the squeak of tires, the chirping of sparrows . . . associated with the semes: "crazy world" (b) static visual images open or closed view open view closed view

146

Richard Fauque

(big avenue) (sinuous alley) luminosity, lights . . . associated with the semes: "intimacy," "majestic" . . . (c) dynamic visual images movement of crowds, traffic flow . . . associated with the semes: "aggressivity," "confusion" . . . (d) olfactory images sulphur emanations, exhaust gas . . . associated with the semes: "pure" (e) exteroceptive images (tactile, thermic, algesic) hustle, wind, rain, snow, cold, heat, etc. associated with the semes: "suffocation," "liberty," "gaity," "sadness." We remark that: (1) The visual, for instance, can serve for the constitution of an autonomous, significant set: the advertising which, in the city, with few exceptions is presented in the form of visual images. (2) The visual and the auditory, for instance, whose signifiers of different perceptual natures can cover an identical or at least an equivalent signified: the siren of a fire engine

the yellow-orange lights of the fire engine

the siren of an ambulance

the blue light of the ambulance cars

(3) Many signifiers can be involved in the same global process of signification: (a) either in a redundant manner: this is the case of the siren and the light of the ambulance; it is also the case of an alley of which during our research interviews it was said that it "went well with the rain because it was narrow, sinuous, and badly lit"; (b) or in such a manner that the various elements combine together to form the signified. This is the case of a pedestrian waiting for a bus in a crowd:

A New Semiological Approach to the City olfactory image

musty smell of perspiration + strong perfume of a coquettish woman nearby

147

+ exterocep + auditory + interoceptive + kinesthetic image tive tactiimage image lo-thermoalgesic image stamping of the crowd

roar of the traffic

difficulty breathing

impossibility of moving

All the foregoing obviously shows that no classification of signifieds based on the nature of signifiers is possible. Or to put it another w a y , the different types of signifiers do not share the semantic space in a univocal way. We do not have:

B semantic space corresponding to the visual signifiers

but in fact:

semantic space corresponding to the olfactory signifiers

148 Richard Fauque This allows us to become aware of a point that is quite important from a methodological point of view, namely, that we do not have to prejudge the relations existing between the nature of the signifiers and the semantic categories to which they may correspond. What would one think of a linguist who tried to reconstitute the semantic system of a language by referring to its different modes of actualization (oral language, writing, morse or braille alphabets, etc.) in order to discover the meaning of words? At best he would achieve a certain taxonomic organization of the different kinds of signifiers, but what more would this bring him? Likewise we remark that we should not a priori consider as signifiers the façade of a building, a fountain in a square, an odor. For here again, what would one think of a linguist who would try to discover the meaning of words by referring to the denotatum (or to the referent)? The only real problem in fact consists in asking oneself how can the minimal elements of signification be located. Let us not forget that it is the signified which constitutes the only possible criterion for the segmentation of the level of expression. Once again, to insist on this methodological element is to insist on the fact that in order to operate this segmentation, we can refer to no criteria other than semantic: aesthetic, logical, functional, or sensory. This of course does not mean that the signification process cannot reorganize the whole of these dimensions (aesthetic, logical, etc.), but then it is meaning which is the criterion of segmentation for the urban continuum conceived as the object of a sensory perception on several levels: visual, olfactory, auditory, etc., and with multiple purpose: functional, aesthetic, etc. Prosaically, this means that nothing authorizes us to consider a priori a square, a street, or whatever other urban element as a minimum element of signification. To consider them as such is to consider the problem of semantization as resolved. In fact, if we can locate the signifiers in the speech chain of urban discourse without having recourse to the string of signifieds parallel to it, this would mean that the city signifies independently of the people who live in it, who read it, etc., and that assumption, at least within the limits of our hypothesis, does not make any sense. 6 But if it is the different readings that individuals do of the city which

A New Semiological Approach to the City 149 constitute the only criterion of segmentation of the urban continuum, we will not have one system of signs, but a plurality of sign systems (as many in fact as there are possible readings of this urban continuum). We will thus have a plurality of semiotics, if by this term we designate the system brought to light by a semiological approach. In other words, to quote Metz, urban semiotics would be to urban semiology what natural languages are to linguistics (Metz 1966). Finally, we note that, strictly speaking, we cannot say that in order to operate the segmentation of the urban continuum we start from the signifieds themselves, but from the meanings revealed by our transcriptive or descriptive metalanguages: our natural language (French), our significant behavior (lovers follow one itinerary, a mother taking her child for a walk another itinerary). Moreover, let us remark that nothing prevents us from asking the people in question the reasons why they follow this or that itinerary in each case: "with my girlfriend I like walking here, along the canal . . . there are flood gates and then there are trees and in Autumn there is a carpet of dead leaves on the bank." AN ACCOUNT OF METHODS AND FIRST RESULTS It is time now to reveal our method with more precision by showing the results which it has permitted us to reach. To begin with, let us remark that in order to have a systematic organization of the elements of language, discontinuities must be made apparent. Moreover, the distinctive features must be pertinent. Let us borrow an example from the phonological study of languages. In French the phoneme b is voiced as opposed to a bilabial and unvoiced p. In Arabic, the phoneme b is only bilabial; this is because there is no p in this language and this makes the voiced character of b a non-pertinent trait. This leads us to the following observation: in order to have a differential perception it is necessary: (1) that the two phonemes (in the case above) have something in common, (2) that the two phonemes present a difference. We have then a double movement, both conjunctive and disjunctive.

150

Richard Fauque

In French, the bilabiality common to b and p allow these two phonemes to be the object of a comparison. But, in another sense, the feature voiced vs. unvoiced is disjunctive, and it will then be a pertinent feature if it carries a difference of meaning. In matters of urban semiology, we will have a meaning support (the selected examples have a heuristic value but not a demonstrative one): road + x, and a feature:y+ or y~, respectively, long vs. short: +

short y-

A-

A + and A' can be compared since they have in common the element x. A+ and A" are different because they possess the term y, the one marked (y + ) and the other unmarked (y - ), which carries a difference of meaning if for a reader A+ ~ A". But we can also have, associated to the same support of communication, a different feature: Z + — Z", respectively, wide vs. narrow; then: way

+

x

^ B+

.

narrow

.

z-

y




bilabiality + voicedness

V b

+0 "AT" "beau"

/bO/

In urban semiology we shall have: x + y*

x + y*

What does this mean? That perhaps we have a first articulation in the sense that from a reduced number of minimal significant elements it is possible to form a rather large number of messages. Still we have to remark that these elements are continually subject to "effects of meaning" (Greimas 1966). When analyzing the interviews, we even often have the impression that the "spaciousness" of a street or its "narrowness," for instance, do not have an immediate autonomous signification, as if their meaning derived from their

152

Richard Fauque

juxtaposition with other features. These mutations of sense are most often occasioned by special categories of features, such as location (centrality ~ peripherality; proximity ~ distance), or quality (density ~ sparsity). 7 This shows how much the process of signification can vary according to its associated features. A street, for instance, because it is in the center, will refer to signifieds marking euphoria and gaity. Another street will refer to the same signifieds simply because it is busy. Reciprocally, the presence of traffic, depending on what street it is associated with and where the street is located, may refer now to signifieds marking euphoria, now to signifieds marking dysphoria. Likewise the absence of traffic will qualify one street as being calm and pleasant, but another, because it is outside the center, as being dull, even dangerous. An excellent example of this phenomenon is given by Ostrowetsky when she remarkes that: what?

where?

4

4

a tower

at Aubervilliers —> does not signify power

but that: a tower

at La Défense

—* signifies power

In other words, the articulation of these two elements, which can themselves be reduced to a certain number of other elements, creates a new meaning. The tower, combined with centrality, means something different from what it means when it is combined with peripherality. Combined with a low-income neighborhood, with a commercial neighborhood, with a residential neighborhood, with a business neighborhood, etc. the tower means something different each time: neighborhood

tower

I

density

A

I

verticality

b'

I

b"

centrality

+

I

b'"

popularity commerciality

COLUMBIA/Gottdiener 110-J Final Page No. 1 5 2

=

x

specific signified

A New Semiological Approach to the City

153

We can now approach the problem of the organization of the signifieds into semantic tree-structures. We have many semantic axes, each one of which is articulated, according to Jakobson, into two semes designated as (marked vs. unmarked), i.e.: (1) presence of a seme vs. absence of a seme (2) presence of a seme s vs. presence of a seme not s. We can say, then, that an urbeme, the "place du Capitole" in Toulouse, is the collection of its qualities (of its semes) or a specific semic constellation. Which means that all its semes will be found again in the formation of other urbemes, but differently combined, creating thus a plurality of other urbemes. So that a street, a square, an avenue, even if they do constitute urbemes, 8 for urban semiology they are only a collection of semes. If x is such an urbeme, we may define it by the sum of its semes: x = X a + b + b' + z Nevertheless, let us add that an urbeme cannot be reduced to the simple sum of its semes. In fact, we are dealing with a set of semes connected by hierarchical relations, a fact which authorizes the term of semic tree-structure: —

absence of x or presence of non x

absence of y or presence of non y

y

z

etc.

"etc.

etc.

154 Richard Fauque If we prefer another way of visualizing the phenomenon, since it concerns progressive specifications by adjunction of new disjunctive features, we may represent it by:

o

In the case where we do not have a continuous hierarchization of various semes, it is because we are dealing with two semic constellations: if there is no successive encasing it is because there is juxtaposition: 2 1

and

^ ^ ^ ^ ^

and not

1+ 2 In other words, if we take up again the example of the tower, we have: (tower) vs. (non-tower) or absence (of tower) (density) vs. (sparsity) (etc.) vs. (etc.)

(centrality) vs. (peripherality)

(select) vs. (popular) (business (residential neighborhood) vs. neighborhood) (etc.) vs. (etc.)

A New Semiological Approach to the City

155

and a collision between the first and the second semic constellation, which gives as result an effect of meaning. As Greimas shows, we can have: idiom

effects of meaning

part where hair is growing bare headed he got it into his head mind break your head bony part this herd is composed organism as discreet unit of 100 heads life he paid with his head human being headmaster —We have a lexeme (head) which has various meanings. —We suppose that for all these meanings there is a stable semic minimum which we define as Ns (semic nucleus). —But if Ns is invariant, the variations in meaning can only come from the context. Therefore there are contextual semes or classemes: Cs. We thus have effects of meaning or sememes: Sm = Ns + Cs. But neither Ns nor Cs are simple: Ns for example = S, + S a which means that it is composed of two semes: (extremity) + (superativity). The same occurs in urban semiology as shown by the previous examples. Until now we have tried to show that a set of pertinent units constitutes a kind of lexicon. Further, we have seen that these units—formed hierarchically—could be articulated with each other in order to engender new significations. But we have not yet approached the problem of their articulation in its specificity. These units do not have a temporal dimension, as is the case for natural languages. We must in fact abandon this preconception that the urban text posseses the dimension of time. This illusion, of which we can be the victims in strolling through a city, is based on the fact that we forget that it is not the urban discourse which unfolds itself before us, but instead we who move about in it. Given this, it is easy to understand what the problem of articulation might be for an urban semiology: it is spatiality and not temporality. This means that, consequently, in this domain, the manner of articulation is no longer constituted by succession but by juxtaposition. Thus the problem of knowing how the different urbemes are articulated has become the problem of knowing how people under-

156 Richard Fauque stand these juxtapositions. Which are the juxtapositions that we find natural, or happy and pleasant, and which do we on the contrary find odd and unexpected or unpleasant? The elements could be: above vs. below. Take a street in a neighborhood on a hillside, for instance, and different types of construction, buildings at one or the other end of the street. Which is the organization of these elements that we find the most natural or the most pleasant? it

I

:high constructions

I

:low constructions

Above, below, but also next to, in front of, one behind the other, as extension of, etc. For example, take an urbeme x defined by its location (centrality vs. peripherality), and by its qualities (such as spaciousness vs. narrowness, etc.). What do we expect?

what?

.

I /

y

what? what?

I /

what? what? what?

^ I / etc.

A New Semiological Approach to the City

157

But things can be less simple: given (x'), we may have beside it (y') or (y") or (y'") we can have behind it (z') or (z") we can have in front of it only (r') In other words, given an element (x') of a paradigmatic string, it can be articulated with certain other elements of other paradigmatic strings in such a way as to have: 1st paradigmatic string

2nd paradigmatic string

where

signifies behind

where

— s i g n i f i e s in front of

3rd paradigmatic string

4th paradigmatic string

In the case (x') —* (r')we say that (r') saturates (x') if for the interviewed person it is the only element that can be articulated with (x') in the relation noted as —*. In the case (x')^(z') if the paradigmatic class (zi) (composed of these two elements) has n' chances of appearing, and if, when it appears, (z') has n " chances and (z") n ' " chances of appearing, we can through a calculation of composite probabilities affect (z') and (z") with a co-efficient of saturation of (x') when they are found articulated to (x') in a relation noted as —*. Having done this, we could generate a city similar to that which has served us as a corpus for our description.9 And this other city, if it were in fact constructed, though it would have nothing in com-

158

Richard Fauque

mon with the previous one, would nonetheless share with it a certain family resemblance. This city would not shock us, in the sense that it would produce the same reactions in people; at bottom, we would not find this city foreign to us.

NOTES 1. This paper is an account of the preliminary results of the first stage of a research project currently being carried out for D.G.R.S.T.—Supplementary coordinated research project on "urbanization" N. 72.7.0648, "Urban Space and the Everyday Life of City Dwellers." 2. Cf. the colloquium on the semiotics of space organized by the Centre de Mathématiques, Méthodologie et Informatique at the Institut de l'Environnement, Paris, May 24-26, 1972 (published as Notes Méthodologiques en Architecture et en Urbanisme 3/4' 1974)3. The term "logic" is continually used by the researchers themselves. 4. Or again, it pertains to the domain of Utopia, as in Thomas More, Plato, etc. 5. Cf. "Une étude sémiologique du code de la route," in Mounin 1970. 6. Raymond et al. 1971; M.G. Raymond 1968; H. Raymond 1968; and EkambiSchmidt 1972. 7. We note, following the semanticists, that there is no reason to be shocked by the use of neologisms (such as peripherality). They are part of a meta language and are to be preferred to the terms of ordinary language since they show clearly that they have a purely functional existence. 8. Once again we cannot decide this a priori. Besides, one realizes that where the mapping nomenclature of the Highway Department recognizes the presence of a unit, the "Place Esquiral" for example, semiology will reveal several urbemes. This is simply to observe to what extent the units of the system are not to be confused with the empirical units of the Highway Department maps. 9. Let us note above all that we would be able to use the results obtained to construct more pleasing cities. It would be enough not to repeat certain articulations which the analysis would have revealed as unpleasant.

REFERENCES Boudon, Pierre. 1973. "Recherches sémiotiques sur le lieu." Semiotica 7 (3):i89-225. Castex, J., and P. Panerai, 1974. "Structure de l'espace architectural." Notes Méthodologiques en Architecture et en Urbanisme (Sémiotique de l'espace), 3/4: 39-63. Ekambi-Schmidt, J. 1972. La perception de l'habitat. Paris: Editions Universitaires. Greimas, A.J. 1966. Sémantique structurale. Paris: Larousse. Metz, Christian. 1966. "Les sémiotiques ou sémies." Communications 7:146-57.

A New Semiological Approach to the City

159

Mounin, G. 1970. Introduction à la sémiologie. Paris: Minuit. Ostrowetsky, Sylvia. 1972. "De l'urbain à l'urbain." Cahiers Internationaux de Sociologie 52:93-110. 1974. "Une topologique du lieu." Notes Méthodologiques en Architecture et en Urbanisme (Sémiotique de l'espace), 3/4: 185-99. Raymond, H. 1968. "Analyse de contenu et entretien non directif: Application au symbolisme de l'habitat." Revue Française de Sociologie 9(2):i67-79. Raymond, H., et al. 1971. L'habitat pavillonnaire. Paris: Centre de Recherche d'Urbanisme. Raymond, M. G. 1968. "Idéologies du logement et opposition ville-campagne." Revue Française de Sociologie 9 (2): 191-210.

7 URBANISM AND SEMIOLOGY Françoise Choay

Françoise Choay, a philosopher and critic of architecture, became the precursor of urban semiotics with the publication of her book, L'urbanisme: Utopies et réalités (i965; see paper 11), and the present paper. Her work centers around the semiotic analysis of urban discourses in texts written by architects, urban planners, and social visionaries. This is a textual analysis because it focuses on writing about urban objects rather than the objects themselves. The author differentiates between two types of urban systems: the "pure" systems corresponding to pre-Renaissance societies, and modern or "mixed" systems. Her perspective on the first relies heavily on the views of LéviStrauss. Spatial signification derives, for Choay, from the fact that space is homologous with a series of cultural systems that are also homologous to each other. Because these systems cover all aspects of social life, they make space "hypersignificant," i.e., richly endowed with symbolization. In contrast, the space of the modern city is "hyposignificant," i.e., poor in spatial significaFirst published in English in Ch. Jencks and G. Baird, eds., Meaning in Architecture (London: Barrie and Jenkins, 1969); based on a previous version in L'Architecture d'Aujourd'hui (June-July 1967). The article was republished, with revisions, in French, in Le sens de la ville (Paris: Seuil, 1972). These revisions are indicated in brackets in the present text.

Urbanism and Semiology 161 tion. Consequently, the modern city has become re-semanticized with the help of supplementary systems of signification, such as traffic signs and commercial advertising. Both de-semantization and partial re-semantization have been going on in cities since the industrial revolution, because urban space in industrial societies derives from the technological and economic aspects of a new system of production. The logic of the functioning of this system is, according to Choay, economic efficiency. This logic is considered by the author to be the signification underlying the production of the city, an assertion that can be questioned because scientific explanation and social signification do not necessarily coincide. Choay links the new, one-dimensional, instrumental signification of urban space with the proliferation of specifically urban discourses of metalinguistic nature. In her later work she specifies that this kind of discourse on the city is found as early as the second half of the fifteenth century in Italy, and has its roots in Greek antiquity. Choay calls these occidental discourses on urban space "inaugurative" discourses and her book (see paper 11) focuses on them as they have changed from the nineteenth to the twentieth century. In her later work Choay has partly revised her approach to the semiotics of "pure" systems. She now considers that, unlike the assertion of the present reading, signification in the production of pre-capitalist, pre-Renaissance settlement space was not independent from discourse on space. Referring to the work of scholars of pre-capitalist urban space, such as Lagopoulos, Choay observes that there is an entire class of discourse which is indirectly spatial and involved in the production of space, and which she calls "connotative" discourse. The paper presented here was first published in English in 1969 (first French edition: 1967), but a revised version of it appeared in French in 1972. In accordance with the author's wishes communicated to the editors we have incorporated the main aspects of these revisions in the text presented here. The most recent revisions appear in brackets, while important aspects of the original text which are not included in the 1972 version are presented below in double brackets. Minor changes, such as changes in syntax and sequence or minor additions to the text, have not been included.

[In the revised version of her paper, the author has rewritten the introduction. Thus, we are not including here the introduction of

i62

Françoise Choay

the original English text. The revised introduction begins with the remark that the first planners, the architects-theoreticians of fifteenth-century Italy, refused specialization, in contrast to modern city planners who, being out of touch with the sciences of man, neglect the linguistic and semantic dimensions which have come to cover the whole of social practice. The central issue the author raises is the possibility of considering the city as a non-verbal system of signifying elements, with structures relating to those of other systems of social practices, and thus of studying the city according to a method derived from general linguistics. This issue is explicitly related to planning practice and to a possible new form of it, which would resemanticize urban space.] i. PURE SYSTEMS Before we come to the problem of the city or urban environment, we must on a more general basis show that some built-up agglomerations are relevant material for semiological analysis1; that the crude materiality of their elements and their practical destination do not make them unfit for welcoming systems of meaning, unfit for elaborate symbolic expression. Evidence of this has been given by Claude Lévi-Strauss in some memorable pages devoted to the "villages" of a few South American [[a-historical]] societies (Lévi-Strauss 1955: 233-56, 1958: chs. 7-8). I shall briefly remind the reader of his example of the Bororo village (figure 7.1). This settlement is composed of a number of cabins laid out in groups of three, so as to form a circle around a vacant space the center of which is occupied by a bigger cabin called the men's house. Two imaginary axes N - S and E-W divide the central circular surface into twice two halves, each of which is given a special name. The various elements of the village (cabins, house, and the parts of the central space) together form a system: each of which, while lacking individual autonomy, still becomes meaningful precisely because of its mutual connections. These connections (or relationships) as determined by a series of rules which refer to and involve all the aspects of the Bororo life, from cosmology and religious rituals to social obligations, work organization, systems of kinship, and rules of marriage. Among the Bororo, the position of my cabin in the village circle determines once and for all the nature of my economic

Urbanism and Semiology

163

Figure 7.1. Plan of the Bororo village. Baddogebo

East

Xobuguiugue Clan

Paivoe Clan

Bocodóri

Apiboregue

Clon

Clan

Quie Clan

Baddogéba Xebeguiúgue Clan

West

Ivuaguddudogue Clan

activity, my participation in religious ceremonies, and my possible choice of a mate. Moreover, the "structure of the village not only permits the subtle play of institutions, it also renews and assures the intercourse between man and the universe, between society and the supernatural and between the living and the dead" (Lévi-Strauss 1955:229). In a word, it involves and determines the totality of behavior; the constructed system is saturated with meaning. The proof of this is given in a negative way by the example of what the Salesian missionaries do to the Bororo when they wish to convert them to Christianity. They do nothing more than lay out the round village in a rectilinear plan. Lévi-Strauss' analysis is clear: "Disoriented from their contact with the cardinal points, stripped of their plan which furnished an argument for understanding, the natives rapidly lose the sense of their traditions as if their social and religious systems had become too complicated to accept the scheme, which had been rendered obvious by the plan of their village, the outlines of which the daily gestures prepetually refreshed' (Lévi-Strauss 1955:241).

164

Frangoise Choay

Thus, full volumes and empty volumes (emptiness as we conceive it does not exist, it is significant in itself and by its relatedness) combine in a way quite fascinating for us, citizens of industrialized countries, because they give roots to the inhabitants. It doesn't matter that the Borroro village is precarious and demolished every three years (after the surrounding land is exhausted) and rebuilt futher in the forest on virgin ground; its slight structure succeeds in linking the inhabitants to the ground in a way we have never been able to acheive, in spite of our monoliths which sink much deeper physical roots in the earth. A built-up system such as the Bororo's thus appears to us as both homologous [(since it is a vehicle of signification)] and antithetic [(since it creates roots)] to the system of writing: the latter actually tends to cut the individual's roots and liberates him from the grip of his physical surroundings. Convincing as this may sound, the question remains as to whether we can transfer such an analysis to the case of societies which are involved in [a more rapid] historical process. I shall now proceed to analyze a few examples of so-called "cities" which actually, in spite of their common denomination, belong to quite different historical contexts. [Sixth Century Greece.] The first example is the Greek City at the end of the sixth century B.C., 2 the moment when the agora appears and the individual hearth disappears ("the ancient incommunicability of the hearth rooted in a definite place in the ground"). The agora holding the Hestia Koine is the sanctified place, the center, the major element which gives meaning and in connection to which a new type of organization is given to the minor elements, the houses (which are from then on identical and loaded with an identical semantic meaning). This structure of the city system is the same as that of the political system: i.e., isonomia (the citizens' juridico-political egality) and is related to and conditions its functioning. Vernant speaks of "the organization by the city of a homogeneous political space, the center of which alone owns a privileged value because in their connection to it all the various positions occupied by the citizens appear systematic and reversible." He also describes the Hestia Koine as the new hearth which "from then on will express the center as the common denominator of all the houses constituting the polis" (Vernant 1974:161-62).

Urbanism and Semiology

165

But space organization does not refer to political custom alone. Not only is it connected with religious ritual but it also refers to a pattern of knowledge, to a mathematical experience which has then elaborated the concepts of egality, symmetry, reversibility, and to a cosmology derived from Ionian philosophies. "In spite of their see ling disorder, the earth, the seas and the river are in their figuration on the map, organized in a pattern following strict rules of correspondence and symmetry" (p. 169), writes Vernant. And he also points out that the homology of these various structures is confirmed by the use, "both in the physical and political thought, of the same vocabulary, the same conceptual tools" (p. 167). [The Medieval Syntagm.] A second example can be borrowed from the medieval city which was the framework of a cruder social and political organization. [The system of the medieval city can be defined especially by closure (inside its surrounding walls) and by the differential relations between two types of elements: functional cellular mini-elements (individual houses), and semantically charged maxi-elements (cathedral or church, palace, squares). The former are opposed both to the latter (in a relation of transcendance) and to each other by distinctive features, in particular the slope and pattern of their roofs, the windows of their facade:] their heterogeneity displays itself along the street in a relationship of proximity that will here be called syntagmatic. 3 This analysis is of course nothing but a gross schema which does not take into consideration the various types of medieval cities which have appeared in time and space; furthermore, the medieval city has yet to be studied from a semiological viewpoint. 4 Nevertheless it is clear that the medieval urban system gave its framework to daily life both by crystallizing the forms of church and feudality and by organizing in a syntagmatic structure the emotional relationship of proximity which is still alive today in so many Western villages. In spite of their striking differences, one still has to note a number of traits which these two types of historical agglomerations and the Bororo village have in common. [Whether or not they belong to societies involved in "hot" history,] they can all be considered as closed systems [with a slow evolution. They signify through the play of their proper elements, without any help from verbal or

166

Frattfoise Choay

graphic systems of supplementation. This is w h y w e shall call them pure.] These built-up systems are as such specific and pure, but they interfere and are interfered with, condition and are conditioned by, an important number of other systems. Thus they refer to global behavior including mental and spiritual responses, and therefore will here be called hypersignificant. In such built-up systems any empty space is nonetheless meaningful. Evidence of this is given by the Greek language, which has no word corresponding to our abstract concept of space. There exists only the word place, topos, which means, as will become clearer, that only "places" can provide pure semiological systems.

2. MIXED SYSTEMS The above examples have served two purposes: first, they were meant to illustrate the concept of semiology as applied to built-up agglomerations, and secondly to demonstrate the richness and abundance of meanings which can become embedded in built-up areas—their social power, the w a y they help the individual to integrate himself in a society the structures of which are redrawn or reactivated in space. . . . N o w let us turn to examples closer in relevance to our o w n situation and study the fast evolving, open systems which appear to be the cities of "modern times." We immediately notice that, contrary to the former, there is not much to be deciphered in the n e w urban developments, that they are hyposignificant and that they have lost their former purity [in that they presuppose the use of exterior verbal and graphic codes]. The acceleration of history reveals the vice inherent in all built-up systems: a permanence and a rigidity which make it impossible for them to continually transform themselves [imperceptibly, easily, like all the other,] less "rooted" systems such as language, technology, clothing, or painting. Unable to change at the same pace or with the same subtlety as the other social structures, the urban system is threatened in its very existence (i.e., its openness to meaning) and hence partly doomed to continual anachronism. Against this permanent threat, the modern city's o w n means allow

Urbanism and Semiology

167

it but one meager defence: partial restructuring, or in other words the inclusion of worn-out syntagms in the new, up-to-the-minute syntagms. Concerning such a process, the architectural history of Paris is relevant. For instance, the Portes St. Denis and St. Martin have lost their former [joyful connotations linking them to the royal entries, to feasts and parades; they have also lost their value as borders, becoming merely two alien presences in an urban fabric where they serve at most as rather absurd landmarks—while in contrast Ledoux's barrières have had a different destiny]. The Parc Monceau rotunda, once a barrière for the payment of taxes on imported goods, has lost its economic meaning and become the gate to the Haussmann-designed park and has hence become essentially connected with leisure. In the same manner, a sequence of small, varied houses may have their street level unified by exterior painting and inside renovation, in order to mirror the transformations of capitalism and the exigencies of the new, giant corporations. For the same reason also, in the nineteenth century the fortifications of Paris were transformed into boulevards, and thus a system of defense became a traffic system. Often this restructuring has made unconsciously (as it is in language) but the opportunity remains for planners to consciously envisage such a process. However, in the best of cases, the use of such a method is restricted mostly to ancient and anachronistic parts of the urban fabric. Actually the fundamental weapon that modern society has at its disposal against the obsolescence threatening its cities belongs to [the systems of supplementation], such as traffic codes and graphic signs, which began to invade the Western cities at certain stages of their development. 5 To some present theorists like Kevin Lynch they appear as the means of giving the city a new legibility. But the fact that the legibility of our present urban agglomerations is mostly due to the efficiency of such graphic systems (whether designed or not) must not hide the bare, inescapable fact that from now on the builtup systems in Western society have lost their [semantic] autonomy: if left to themselves and their specific elements, they do not carry symbolic weight any more. But here we are driven in a circle: the more the built-up system makes demands upon supplementary systems, the more it proves itself obsolete and its former task of information is carried via print-

i68

Frangoise Choay

ing and telecommunications through other systems of information, the development of which appears at the same time as cause and result of the obsolescence of the built-up system. It is precisely at the time when this system ceases to refer to the whole range of social behavior that the transition from place to space occurs. The first stage of this may be found in the Baroque city. Here the spatial organization is contaminated by pictorial space which corresponds to an analysis of sight. In the process of acquiring this aesthetic quality and becoming a show, the city also acquires a play dimension, and is lived at a distance which is the exact opposite of the intimate and constraining relationship characterizing the pure systems. [However, the Baroque city was not lived as a show by all of its inhabitants, but only by a privileged fraction of them. For in societies with rapid evolution, the built-up system does not simply tend to become anachronistic, it is also a source of anachronism, as can be seen in the great modern cities where whole sections of built-up systems survive that belong to past eras, preindustrial or dating from the beginnings of industrialization. These places are inhabited by elements of the population who for various reasons make no use, or make insufficient use, of urban systems and exist in urban space in an anachronistic manner, charging it with various semantic values. This is why, when we analyze the behavior of the inhabitants of a modern city in respect to urban space, and the signification which this space has for them at the same time, we discover that the notion of synchrony is an illusion; a cultural diachrony is evident within the synchrony of the calendar.] Regarding another symbolic system in our society, the system of scientific thought, the French philosopher Gaston Bachelard was struck by the variety of levels at which science could be approached at one and the same time by the members of the same society; he also noticed that an individual referred to different levels according to his particular scientific interest. To describe this situation, he coined the expression niveaux épistémologiques (Bachelard 1962, 1967). To describe the homologous situation which has arisen in the field of urban space, I will use a corresponding concept, niveau politologique: "politological level." The existence of these various semantic levels among various

Urbanism and Semiology

169

classes of a population is fundamental for the understanding and tackling of urban planning (showing that as far as urban semiology is concerned, we cannot borrow the concept of synchrony from general linguistics since the physical synchrony actually displays a semantic diachrony). Presently, we get clear evidence of these various levels in two crucial situations—situations which may be clarified by the concept of politological levels. First is the case of urban renewal when it deals with minorities (economic or ethnological) still living in syntagmatic systems, the pattern of which is given by the village. In a recent and remarkable study of the renewal of the thirteenth arrondissement of Paris (Coing 1966; compare with Schorr 1963), the investigators notice that the word "village" came back, like a litany, again and again, in the declarations of the "renewed" population when they were questioned about their arrondissement before renewal. For the same reasons, Herbert Gans called his book on American slums The Urban Villagers. And this situation was also foreseen by Spengler when he remarked: "each large city has conserved corners where there are living, in the lanes like the country, the fragments of a humanity almost rural which is maintained on that side of the street by an intercourse which is almost rustic" (Spengler 1948, 2:91). In every western town we could find such villages which, in Paris, are called rue Harvey, rue Clisson, L'Impasse de L'Avenir. The second situation occurs when underdeveloped countries become industrialized. We are then dealing with cultures or microcultures which possess their own built-up systems linked to the totality of their cultural structures and customs. Town planning, in the Western manner, deprives these populations of their behavioral framework; it leaves them at a loss. Their only way of coping with such a situation is to master the Western symbolic systems, insert themselves in its cultural structures, learn to use its abstract, supplementary systems of information and communication, and to disengage themselves from place and its non-verbal spell—which is by no means an easy task. It is equally hard for the Western planner who not only has to confront a pre-industrial system, but one belonging to an alien culture—and thus has to learn to decipher it like a foreign language. This is the case with some French sociologists and planners who are at present studying the urban communities of Magreb. Such Moroccan cities cannot

170

Françoise Ckoay

be understood except as a social framework of egocentric space (there are no plazas, no avenue, in a word, no projective space) itself linked to a circular structure of time. . . . In spite of apparent similarity, this example has no connection with the medieval syntagm. There are analogies in the town plans, although the mosque or the castle (Kasba) bear a smaller semantic load. But the fundamental difference consists in the lack of differences among the Moroccan facades. A facade, once built, will never be altered or repaired even if it falls to pieces. Poor and rich have the same facades and doors, so that a foreigner has no means of identifying the district where he strolls (he would have to be able to decipher the symbolic positions of the women standing on the thresholds of the houses). This visual homogeneity corresponds to the fact that in Arab society there is no pattern of authority comparable to the one working in our societies or in the WASP societies . . . There is nothing but "chains" of influence connected with polyfunctional individuals whose powers and abilities are numerous and not clearly defined. The city . . .protects itself from country people by its labyrinth: no axis, no monuments; . . .meaning here accompanies motional habits.6

[3. THE PRIMACY OF ECONOMY] Now let us return to our present so-called cities. I have characterized them by a series of concepts such as hyposignificant and semantic reduction which may lead to some misunderstanding. But by these words I do not intend any judgment of value. They only seemed to describe an historical situation. It must be understood that they have been written without the slightest touch of nostalgia. Hyposignificant does not mean without signification, but only that the built-up system no longer refers to the totality of cultural behavior. . . . More precisely, hyposignificant means that the built-up system has retained a precise and limited meaning; since the industrial revolution it has been exclusively linked with the new type of production, both under its technological and economic aspects. Thus in the course of the nineteenth century, the new type of production determined the split of industrial towns into two parts, one pertaining to the economic function of production and the social class of producers, the other to the economic function of consumption and the social class of consumers. [In this way, urban space in fact becomes the vehicle for almost nothing but the economic sector.] One cannot give enough emphasis to the semantic reduction implied: it was so drastic (even with respect to the forerunning Bar-

Urbanism and Semiology

171

oque city) that it must be considered as a mutation. [We shall present briefly here two consequences of this reduction, both on the level of consciousness and that of the unconscious.] [The Urban Imaginary.] [On the level of the collective unconscious], frustration was suppressed and concealed behind an imaginary construction, an "imaginaire urbain" which we can trace back to Rousseau.7 From the end of the eighteenth century the imaginaire urbain gave birth to a new mythology in which the city played the part formerly played by nature. In pre-industrial times, the built-up agglomeration—hamlet, village, town—was experienced as the reassuring element where man discovers himself and builds himself in opposition to nature. Since the industrial revolution, the urban fabric itself, the big cities, conurbations, etc., have appeared to the collective consciousness as another nature, a mysterious wilderness, threatening man's existence. How else can we explain the violence of the new images and myths? According to the descriptions of Marx, the human world of artifacts has truly become a new nature; but the urban environment as artifact has lost its previous implicit (if not unconscious) intelligibility. We are here facing the fundamental contradiction of our modern cities—hence Nietzsche and Baudelaire (and so many others), hence the famous Spenglerian descriptions, hence the Corbusian metaphors ("the modern city is a flourishing cancer"). [The Contamination of Language.] On the other hand, at a less emotional level, the proliferation of urban agglomerations expressing nothing but the elementary, single-meaning new order gave birth to a compensating phenomenon: an endless rationalization of urban space to which discursive analysis attributed a diversity of purposes and significations. Thus a commentary (i.e., logos) inserted itself between the urban agglomerations and its inhabitants, like a screen, creating between them an insurmountable distance. The semantic reduction finally meant an irreversible contamination of the built-up system by the verbal language, the logos, and the complete and definitive loss of its former purity. The very idea of a non-verbal system (which the appearance of supplementary systems still permitted to subsist) was finally destroyed by this metalanguage. This

172 Françoise Choay inescapable verbal screen drawn between us and our present urban space explains the abstract quality of the latter much more than does the privilege given to sight, alleged by McLuhan. [The proof of this can be seen in the evolution of our vocabulary of the city beginning in the nineteenth century. . . . These lexicological transformations are only a symptom: they originate in the conceptual enterprise undertaken in the course of the nineteenth century to construct a theory of the urban system, and provide it with a verbal foundation which might compensate for the brutal semantic reduction which followed the industrial revolution.] I have shown elsewhere8 that the rationalization, the process of theorizing about the urban system, developed from the very first along two lines of thought, following two basic "models." The "culturalist" one, inspired by nostalgia, aims at recreating a syntagmatic system. The other, the "progressivist" one, is actually less Utopian than it appears. Its aim is to promote a new building pattern; inside the framework of a neutral, meaningless space the elements of urban space will be classified and associated following functional similarities. I shall here call this structure paradigmatic or metaphoric. This metaphoric approach has progressively substituted for the original bipartite nineteenth-century industrial town, a "more rational" tripartite classification which tends to eliminate from the meaning of the space system social references, in favor of an operative analysis of the economic field. Social segregation ceases to be a principle of structuring; the opposition between living and working areas is transcended by a "center" [which today characterizes all urban plans]. The three metaphoric poles of the new system, housing, industry, center, are linked by a complex set of "circuits" the meaning of which is efficiency. The urban system then becomes an operational space. The information given by this network of connections—its hidden spirit—is the transformation of nature, its metamorphosis in the course of a production process, the sequences of which are inscribed in the ground. [Also, while the system of progressivist urban planning is not conscious of this fact, and in spite of its references to a naive humanism—which we should question today—it has the same signification as the "disorder" of urbanization. If it is distinguished from this "disorder" through its efficiency—its vocation—it is also, nonetheless, monosemic9 and possesses only

Urbanism and Semiology

173

one semantic dimension: the economic.] But such monosemy is in harmony with our fast evolving world and a society geared to production and consumption. There is no point in criticising or even fighting this monosemy in the name of so-called humanism. [[The times of polysemic, built-up systems, referring to an immediate "total man," are gone for Western man. His world is and has been for some time en miettes, crumbled into pieces, and his fate may well be to face up to this situation. If he does not, he cannot expect any aid for his lost unity in the tendency of present built-up systems.]] [Choay has added a new conclusion to her paper, "Conclusion: The Game," and a postscript, "P.S. (1972)," which are summarized here. She finds that the growth of consciousness due to the use of semiotics in urban matters is already a (negative) form of action. We have to stop demanding from urban space a discourse it cannot give us. Semiotic analysis shows that the economy is today the only referent of the urban system, and thus redresses a certain kind of functionalism and helps define the planning field. On the other hand, an important part of the population remains unsatisfied with operational efficiency alone. Semiotic analysis can lead to the elaboration of signification systems adapted to the population's politological level. Semiology can even help in constructing new systems which, however, to the extent that we have lost our global engagement with urban space, can be lived only in the dimension of play, as is already the case with the everyday objects surrounding us and the system of fashion. Choay writes that between the original and the latest French versions of the present paper six years have passed, during which her thinking has evolved. She indicates four theoretical points clarified by this evolution. First, there has been a gradual replacement of the "language 0/ the city" by the "language on the city." Thus, urban space creates and is created by a metalinguistic discourse which can be approached from two theoretical directions: an epistemological critique of it, which can open new perspectives, and a structural analysis, which would confront a heterogeneous domain consisting of differing ideological structures. Thirdly, the built-up system is also articulated with important psychological and other structures.

174

frangoise Choay

Finally, the author herself no longer sees urban space as reduced to purely economic signification, and feels that a resemantization of urban space is both necessary and possible.]

NOTES 1. F. de Saussure, the founder of structural linguistics, coined the concept of "semiology" (from the Greek semeion = sign): "a science which studies the life of signs in the middle of social life." For him la langue as a "system of signs expressing ideas . . .and thus comparable to writing, to the deaf and dumb's alphabet, symbolical rites, forms of politeness . . .is nothing but the most important among these various systems." But this "particular system" is a kind of ideal-type, from the study of which research in other fields of symbolic activity may be stimulated and developed. Saussure foresaw that "linguistics might become the general model of any semiology." And such was, after him, the point of linguists like Benveniste or jakobson. Some non-verbal semiological systems have been studied in such a light by Lévi-Strauss and Barthes. In his Elements of Semiology (1967a) Barthes indicates that "the aim of semiological research is to restore the functioning of systems of meaning, other than language." (See Saussure 1971; Jakobson 1963; and Barthes 1967a, 1967b. 2. See Martin 1951, 1956. For the interpretation of archaeological evidence, see Vernant 1969, 1974, as well as Lévéque and Vidal-Naquet 1964. 3. The word has been borrowed from the linguists who, following Saussure, have differentiated two fundamental kinds of relations between linguistic elements: i.e., spatial contiguity and similarity (which schematically correspond to two forms of mental activity, conjunction and association). Syntagm corresponds to the relationship of contiguity, paradigm to the relationship of similarity. "To implicitly speak of the selection of certain linguistic entities and their combination in linguistic units of a much greater complexity . . .the concurrence of simultaneous entities and the concatenation of successive entities are the two modes according to which we, speaking subjects, combine the linguistic constituents," says Jakobson in his Essais. From this double process, which one finds in all semiological systems, this author has taken (in borrowing from the field of rhetoric) the concepts of "metaphor" and "metonymy" which he uses to characterize the styles in aesthetic systems (literature, music, and painting). Regarding my own use of the word syntagmatic (I could have used "metonymic"), if "placed in a syntagm, a term acquires its value only in opposition to preceding or following terms, or both." 4. The first stage of such an approach has been given for the Gothic architectural system by Panofsky ir» his Gothic Architecture and Scholasticism (1951), in which he brilliantly displays the homology between the two different structures: architecture and philosophical thought. 5. First stages of this progress: the numbering of houses and the naming of streets with special signs. 6. Excerpts from an unpublished letter from the French sociologist J. F. Clément.

Urbanism and Semiology

175

For some interesting comments regarding the connections between space structures and cultural behavior (language, work, family structure) in Moslem North African countries, see also Bourdieu (1965), dealing with Algerian southern villages. 7. Rousseau was among the first writers to use the word urban in the sense of "belonging to the town or city" as opposed to "belonging to the country." In his Confessions, in particular, he evokes the "urban crowd" from which he escaped to a rustic retreat. 8. Choay (1966). Not only has the invasion of verbal languages in urban matters to be studied under its systematic, theoretical aspect, as I did in the latter book, but also an analysis remains to be made of the transformations which urban terms have undergone in the West since the nineteenth century. As far as Latin languages are concerned, we see the progressive obsolescence of concrete words coupled with the growth of a taxonomy full of neologisms (see CERDA, a fascinating book in this respect) and the whole development of the French espace-vert (a legacy from the times of Napoleon, popularized by Corbusier and the CLAM), now the most common French word in daily vocabulary. It has taken the place of the former jardin. Its very composition espace-vert, shows the meaninglessness of the areas to which it is attached. 9. With a single meaning (opposed to polysémie).

REFERENCES Bachelard, G. 1962. La philosophie du non. Paris: PUF. 1967. La formation de l'esprit scientifique. Paris: J. Vrin. Barthes, R. 1967a. Elements of Semiology. London: Cape. 1967b. Writing Degree Zero. London: Cape. Bourdieu, P. 1965. Le déracinement. Paris: Minuit. Choay, F. 1966. L'urbanisme: Utopies et réalités. Paris: Seuil. Coing, H. 1966. Rénovation urbaine et changement social. Paris: Editions Ouvrières. Jakobson, R. 1963. Essais de linguistique générale. Paris: Minuit. Lévêque, P., and P. Vidal-Naquet. 1964. Clisthène l'athénien. Paris: Les Belles Lettres. Lévi-Strauss, C. 1955. Tristes tropiques. Paris: Pion. 1958. Anthropologie structurale. Paris: Pion. Martin, R. 1951. Recherches sur l'agora grecque. Paris: Boccard. 1956. L'urbanisme dans la Grèce antique. Paris: Picard. Panofsky, E. 1951, 1957. Gothic Architecture and Scholasticism. Cleveland: World. Saussure, F. de. 1971. Cours de linguistique générale. Paris: Payot. Schorr, A. L. 1963. Slums and Social Insecurity. Washington: USGPO. Spengler, O. 1948. Le déclin de l'Occident, vol. 2. Paris: Gallimard. Vernant, J. P. 1969. Les origines de la penseé grecque. Paris: PUF. 1974. Mythe et pensée chez les Grecs. 2 vols. Paris: Maspero.

8 SEMIOTIC URBAN MODELS AND MODES OF PRODUCTION: A SOCIO-SEMIOTIC APPROACH Alexandros Ph. Lagopoulos

We can distinguish two periods in the work of the author. During the first period he is oriented toward the urban semiotic system and shows limited interest in material social factors involved in the structuring of space (see, for example, Lagopoulos, paper 12). In the present period of his work, to which this article belongs, Lagopoulos extends his semiotic approach to socio-semiotics. Believing that semiotic systems are produced by material social processes, which he calls "exo-semiotic," he turns toward the articulation of semiotic systems with these non-semiotic social processes, without abandoning the study of semiotic systems. The author's epistemology, and theory of society and ideology are founded on historical materialism. The sociological concept which is of basic operational value for the following paper is that of "mode of production," a fundamental Marxist concept. For Marxism, class and individual ideology and, in general, the whole of A preliminary version of this paper was presented at the colloquium of the Hellenic Semiotic Society, Thessaloniki, June 22-23, >979- First published in Semiotica (1983),

45(y4):275-96-

Semiotic Urban Models and Modes of Production

177

culture are produced by material social labor. The epistemological principle behind this assertion is a causality operating beyond the observational level which, as a concept, is found in no other epistemological paradigm, and which radically differentiates Marxism from, for example, functionalism and systems theory. For Lagopoulos, the causality of the mode of production does not exclude an inverse influence of culture on the exo-semiotic, and a relative autonomy of semiotic systems. The mode of production is identified with or closely related to the material causal processes of a society, and accounts for the fundamental laws governing its structuring. Thus, this concept can be used to group apparently different societies, which are, however, identical in their fundamental structure. This is what the author is doing here in order to study urban semiotic systems across a variety of societies. Usually semioticians define and organize their bodies of work with purely semiotic criteria. But, if one accepts that the main properties of semiotic systems are socially produced, then this way of operating leads to results which run the gamut from approximate to meaningless. In contrast, the author is led to adopt a socially based typology of cultures as the basis for his comparative analysis of the semiotics of settlement space.

EPISTEMOLOGICAL CONSIDERATIONS The entirety of a society, its global social structure or (economic and) social formation, may be thought of as resulting from the interrelationship (articulation) between two general 'planes,' the first of which constitutes the social base and the second, the superstructure (Marx 1957: 4). The relationship of the global social structure with the biotic and abiotic environment leads to a new structure, on the scale of the ecosystem. This new structure is dominated by the global social structure (figure 8.1) (Scheibling 1977: 54-55)The base can be analytically divided into two principle "planes": the productive forces (infrastructure) and the social relations of production (structure); while the superstructure can be analytically (and provisionally) divided into: a demographic "plane," institutions, ideologies, and other social products. Depending on the mode of production (see below), it is possible for a part of the superstructure to be functioning within the base (see figure 8.1) (Godelier 1973: 18; Lefebvre 1971: 355-57).

178

Alexandros Ph. Lagopoulos

The interrelationships between and within the different planes of the global social structure are dialectical, that is, contradictory in nature; the terms of these interrelations constitute a unity. If w e formulate the relation of the relations between the planes of the global social structure, we describe the laws that govern it. The nucleus of these laws is based in the mode of production, a term which in its narrow sense is identical to the base, and in its wider sense is identical to the base together with a section mainly of the politicalideological ensemble belonging to the superstructure. According to an orthodox Marxist viewpoint, the global social structure is in the final analysis determined by its productive forces (see also Godelier 1973: v, vii, viii, 16, 18, 69, 88, 2 1 1 - 1 2 , 2 1 4 - 1 5 , 2 4 1 44). In my opinion, this observation can be extended; for the structure of the productive forces can be determined by their relationship with the productive forces of one or more global social structures constituting the social environment of the first global social structure. Thus w e can see that the global social structure is in the last analysis determined by the productive forces and/or the productive forces existing in its social environment; nonetheless, when the determination comes from the exterior, it is "filtered" through the productive forces of society in question. This determination linked to the productive forces is the structural cause of the dominance (hegemony) of one or more of the planes of the global social structure; the dominance presupposes that the plane or planes in question function as the social relations of production. Organized, or organized and constructed urban space belongs to that plane of the superstructure that we have called "other social products." This space is, in my opinion, the projection of spatial functions (socially defined uses of space), which also belong to the superstructure, into geographical space, followed by the construction of shells for functions and of circulation networks. Thus, functional urban space is articulated with the whole of the globa) social structure and its "ecosystemic" environment in the way shown in figure 8.1. If we consider this schema with a view to different approaches to the analysis of urban space, this space may be analyzed as resulting from and, considering the nature of dialectical relations, as also influencing, three processes:

Semiotic Urban Models and Modes of Production

179

(a) one process beginning in the base and directly influencing space—this is the socio-economic production of real space; (b) a second process beginning in institutions—this includes the political production of space; (c) a third process beginning in ideologies—this is the ideological production of space. The first process is fundamental for the constitution of space. The second includes, among others, the process of state planning. Finally, the third process includes, among other things, the plane of pertinence to the semiotic production of urban space. Ideology serves as mediator between the global social structure and the semiotic model of space. All three processes are subject to the constraints of the ecosystemic environment (see also Lagopoulos 1978: 103, 142). Figure 8.1. Position of functional space in relation to the various planes of the global social structure and its "ecosystemic" environment. fundamental dialectical influence < possible extension at the base. e co -

i»s 'emit env1ren. neni

••

4• '

ins tituiiens « __ __

Î

vs

J

_ 1 ideologies F t- b•o_se —« .-S52. - - - ~r > i otUer social prodocts-.functioftaP urban space,. ..J

In addition to the above-mentioned semiotic approach to space, space can be studied semiotically as such, that is, as a set of signif i e s . Finally, urban space, once it exists, passes into the consciousness of social classes, groups, and subjects. The signifieds of the mental "image" of urban space, of the semiotic conception of urban space, related to this process, belong to the ideological "plane." If pSa = presignifier (substance of expression)—that is, real space, the space resulting from the articulation of an urban semiotic system with the construction materials (urban form) and geographical space (urban organization and form); Sa = signifier—urban organization and form; Se = signified—meaning of the Sa; d = denotation—plane of functions; and c = connotation—plane of ideology par excellence—then the passage, as seen by an analyst or a social subject, from ideology to real space can be described, in dialectical terms, in the following fashion:

180

Alexandros Ph. Lagopoulos

{n|n = S e J ^

-- Se d } ^ ^ ^ ^

{1|1 {1|1 = Sa d }

Semiotic production of urban space

[k|k = pSa} (ipr)

where =>: primary dialectical influence and 106-7, " 4 / « 7 - 1 8 , 123).

Semiotic Urban Models and Modes of Production

187

The Greek polis connotes the new, relatively desacralized, cosmological conceptions of physical space. Their geometrical basis is a circle organized around its center. The polis as a pSa is organized around its economic, civic, and political center, the agora. In the agora is the center par excellence of the polis, the public hearth, hestia koine. Polis, agora, and hestia have as Se c "center of the world," a Se c belonging to the cosmological code. The cosmological code is grafted onto another, the socio-political code, which has gradually come to dominate it. In fact, the agora also constitutes a center in the sense that all citizens should theoretically be located at an equal distance from it, due to the principle of isonomy, civic equality. The process of desacralization during the period of the polis is reflected in the fact that the political code came to dominate the cosmological code, a fact that can be clearly seen in the case of the hestia koine (see also Vernant 1974, 1: 179-80, 183-86, 206, 2 1 0 - 1 1 , 215). The functioning of rational thought within the semiotic urban model results in the appearance of some new characteristics, which differentiate it from the models of the communal, Germanic, and Asiatic modes of production. The identification of the settlement with the world gradually disappears, in favor of the analogy between them, an analogy often founded on geometrical form and abstract number, and tending toward an allusion only. This urban model, which gives an example of the models produced by the antique and esclavagist modes of production, was realized in all greek poleis. Thus, Athens in the sixth century B.C. (figure 8.4) was centerd on its agora, containing among other buildings the hestia koine (Old Bouleuterion) and the Altar of the Twelve Gods, which had as its Sec "center of the world." The street network presented a radial form. During this century, the social and spatial organization of the city by Clisthenes was tripartite and concentric. The principal code used by him was political, and his principal aim the achievement of isonomy; this code was probably accompanied by a cosmological code (Leveque and Vidal-Naquet 1964: 1 0 - 1 5 , 83, 9 1 93, 100, 107, 123, 128, 132). Feudal Mode of Production. In spite of its more developed aspects, the feudal mode of production appears, in certain respects, to be a

i88

Alexandros

Ph.

Lagopoulos

Figure 8.4. The central part of Athens in the sixth century B.C. (according to Travlos 1971).

The Agora and the Acropolis in the sixth century B.C.: 1. Late Helladic graves. 2. Submycenaean graves. 3. Geometric graves. 4. Grave enclosure. 5. Blaute. 6. Aphrodite Pandermos. 7. Ge Kourotrophos. 8. Demeter Chloe. 9. Kyloneion. 10. Bouzygion. 1 1 . Aglauros. 12. Anakeion. 13. Thesmophorion. 14. Eleusinion. 15. Prytaneion. 16. Boukoleion. 17. Epilykeion. 18. Thesmotheteion. 19. Heliaia. 20. City wall of the sixth century B.C. 21. Altar of the Twelve Gods. 22. Leokoreion. 23. Shrine of Zeus. 24. Temple of Apollo. 25. Temple of the Mother of the Gods. 26. Old Bouleuterion. 27. Prytanikon. 28. Boundary stone of the Agora. 29. Triangular shrine. 30. Theseion. 31. Enneakrounos. 32. Dromos. 33. Orchestra. 34. Tyrant-slayers. historical r e g r e s s i o n . It h a s points of resemblance w i t h the Asiatic m o d e , h e n c e the t e n d e n c y to c o n f u s e the t w o on the part of m a n y sociologists

and

historians.

In m a n y

societies u s i n g the f e u d a l

m o d e of p r o d u c t i o n , this last is a transformation of the Asiatic m o d e of p r o d u c t i o n that p r e c e d e d it. W e will concentrate on European feudalism.

Semiotic Urban Models and Modes of Production

189

During the European Middle Ages, the earth was thought of as the center of the world and Jerusalem as the center of the earth. Jerusalem, both heavenly and earthly, was the dominant model for the Christian settlements of the Middle Ages. Its typical symbolic plan, of Hebrew origin, is a circle divided into four parts by two roads that cross at right angles at a semiotically marked central point (figure 8.5). This spatial organization is closely related to or coincides with the spatial organization of medieval maps of the circular earth (figure 8.6). "Earth" and "world" are Se c of Jerusalem and its model. During the High Middle Ages there appeared in Germany new settlements, the organization and form of which were founded on the Teutonic cosmological model combined with the above Christian model. One of those settlements is Rottweil (figure 8.7). At the center of several German settlements of the High Middle Ages are found large stones, referring to the Se c "omphalos," "Calvary," and "Jerusalem"; they are also strongly linked, connotatively, to the autonomy and the legal jurisdiction of the settlement. Often, these stones are the departure point for four roads that divide the settlement into four regions. These stones are connotatively equivalent to, and also on the level of the pSa associated with or interchangeable with, the perrons, minuscule step pyramids crowned with a column. It is possible that the cosmological connotative code of these settlements with a large stone and/or perron at their center reveals a substratum going back as far as the Megalithic period, during which a cosmology seems to have appeared placing a stone (the cosmic stepped mountain) at the center of a quadripartite earth (Miiller 1961: 5 3 - 1 1 4 , 195-223; Egli 1962: 27, 250; L'Orange 1953: 15). The characteristics, already referred to, of the semiotic urban models of the communal, the Germanic, and the Asiatic modes of production are in general valid for the corresponding models in the feudal mode of production. The dominant codes of the European settlement of the feudal mode of production are the religious and the cosmological codes. There is also probably an anthropomorphic code and for some settlements a zoomorphic code (Lavedan 1954: 33). The data for the German settlements of the High Middle Ages also indicate the existence of a dominant political code. Industrial Mode(s?) of Production. The European Renaissance (generally, fifteenth-sixteenth centuries) was characterized by the ap-

190 Alexandros Ph. Lagopoulos Figure 8.5. Islandic plan of Jerusalem (according to Müller 1961).

pearance of a primitive accumulation of capital. The settlements of nations associated with the Renaissance or showing similar phenomena of development to them were generally considered, on the connotative level, as being only in analogy with the world, and even the analogy w a s of the nature of a general, abstract, and weak allusion. After the Renaissance, the settlement w a s produced by a dominant social minority and addressed itself primarily to visual perception (Choay 1 9 7 0 - 7 1 : 9). In the industrial mode(s?) of production, appearing after the transitional period that extended from the end of the Middle Ages to the eighteenth century, the semiotic urban model governed by connotative codes of a religious nature gradually disappeared, and the dominance and even the existence of a cosmological code ceased. Vestigial traces of this code as it applies to urban or architectural space can be found in

Semiotic Urban Models and Modes of Production

191

Figure 8.6. Map of Beatus (according to Müller 1961).

popular tales and conceptions, as well as in the Church and in initiation societies such as the Free Masons, that is, on the periphery of modernization. New models and new connotative codes characterize the approaches to the settlements of the industrial mode of production. These coexist with the theories focused on urban space of the social scientists—mainly the Marxists, the urban geographers and the human ecologists, and the city planners—aiming at a new approach, the metalinguistic. With the evolution of the industrial mode of production, there is a shift of emphasis from ideological models to scientific urban theories. On the basis of Choay (1965), we can distinguish three main models for urban space. The first two phases of the progressivist model are governed by an ideology of rationalism, hence of functionalism and the rationalization of aesthetics; of efficiency, hence the preoccupation with productivity and health, and the urban constraints; and of modernity and technology. The strong social and political

192

Alexandros Ph. Lagopoulos

Figure 8.7. Rottweil, German settlement of the High Middle Ages (according to Miiller 1961).

interests of the first phase are absorbed by this ideology in the second. Individuals are considered as substitutable and human needs as universal. The resulting urban space is stereotyped, inflexible, functionally segregated, and has imprecise limits. The third phase of this model follows similar principles, which are not unrelated to scientific metalinguistic elements. The two phases of the culturalist model, which is systematically opposed to the progressivist, are governed by an ideology of nostalgia for the past, hence the relation of aesthetics with tradition, and of the predominance of the cultural over the material, which is turned against technology. The strong social and political interests of the first phase weaken in the second. Individuals are considered as unique and emphasis is placed on the urban community. The resulting urban space is specific and its limits are precise. The approaches that succeeded this model cannot in fact be considered as belonging to the semiotics of connotation, but to scientific metalanguage. The naturalist model, which results from the antiurban ideas of the nineteenth century in the U.S.A., is governed by an

Semiotic Urban Models and Modes of Production

193

ideology of nostalgia for nature. The typical representative of it is Wright's Broadacre City. The ideology of this model is founded on individualism and technology, is not focused on efficiency and productivity and, as a consequence of this, includes an abolition of urban constraints. The resulting space is characterized by the elimination of the form of urban space and the continuity of nature, and is simultaneously stereotyped and specific. This space shares characteristics of the space of both progressivist and culturalist models. These three models, which present variants and overlappings, meet in their common requirement to face what is considered as the disorder of the industrial city (see Choay 1965: 7-83). According to the preceding analysis, the dominant codes are the following: vital, historical, aesthetic, naturalist, orderliness, utilitarian (subcode: functional), technological, sanitary, economic, cultural, to which must be added, for the first phase of the progressivist and culturalist models, a socio-political code. Before the Renaissance, the semiotic urban models had their origin, to a greater or lesser degree, in the whole of society. In contrast, the models of the industrial mode of production correspond to the "closed" sub-ideologies of some social groups. Today, they are used by some city planners educated as architects. The models belong to a "logo-technique," a limited language without social impact. The global social structure of precapitalist societies leads to the appearance of a semiotic urban model and permits it to function more or less importantly in the production of space, even large-scale urban or regional space. The transformation to which this model was subject in industrial societies is accompanied by the limitation of its functioning to a small number of small cities and to isolated and limited parts of others. The production of urban space in industrial societies is founded on the extra-semiotic processes (a) and (b) above (figure 8.1) and on ideological elements, among which are ideological positions not in general constituting a semiotic model. The ideological production of real space is characterized by the following elements: (a) in the case of planned intervention, by the urban models, the application of which is, as we saw, very limited, or by ideologi-

194

Alexandras Ph. Lagopoulos

cal positions, dominant among which are the utilitarian and the economic codes, guiding the use of scientific and technical data, and (b) in the case of isolated action, by the ideology of the urban actor, which is subject to the constraints of reality. The constraints are augmented as the income of the actor diminishes. The ideological production of space, on the one hand, is restricted by the scientific and technical data and the existing socio-economic conditions and, on the other, is alienated, since it directly obeys the socio-economic base. Thus, the ideological production of urban space in industrial societies presents a strong desemantization in comparison with that of precapitalist societies (see also Greimas 1974: 13). If we now pass to the semiotic conception of urban space, we observe that a structure exists for every individual that is more or less coherent (Lynch i960: 88-89; and Ledrut 1973: 73-78, 315-73). This structure, in opposition to what is observed in general in precapitalist societies, is not, in general, identical with the urban models of production; nor is it in general accordance with the ideological positions guiding the production of urban space when urban models are not used. Ledrut (1973) demonstrates that the most important dominant connotative code of the structure and models corresponding to the semiotic conception of modern cities is the vital code, and that the aesthetic, ethical, and functional connotative codes are also dominant. He also finds, among the secondary codes, the ludic, economic, historical, social, spatial, temporal, and sanitary codes. There are two general and abstract urban models, which are opposed and complementary, and are clearly related to Choay's typology, a fact showing the existence of a common ideological base between the users of urban space and the intellectuals and specialists concerned with it; however, this fact did not prevent the latter from isolating themselves with their "logo-technique." The dominance of economy leads, in general, to a disequilibrium between the "theoretical" ideology expressed in the rhetoric of urban actors about urban space and their "applied" ideology as it appears in their urban practice, and consequently, to their frustration.

Semiotic Urban Models and Modes of Production

195

CONCLUSIONS The following conclusions were drawn on the basis of the corpus given above.4 The religious plane has a specific function in the global world structure, a function that varies with the mode of production. In societies of the communal and probably the Germanic modes of production, the religious plane seems to be closely linked to the social relations with a dominant function, such as kinship relations. In societies of the Asiatic and feudal modes, it appears to be the religious and political planes, closely linked, that dominate. In the societies of antique and esclavagist modes, dominance lies with the political plane, gradually distancing itself from the religious. Finally, in the early phases of the industrial mode of production, the economic constitutes the dominant plane, while in the recent phase there seems to be a displacement of dominance toward the political plane. As we have seen for societies of the communal, Germanic, and Asiatic modes of production, the settlement as pSa is considered as analogous to but also identical to, identified with, the world (cosmologica! connotative code). In these societies the dominant code for the production of urban space is the cosmological; in the case of the Asiatic mode of production, the royal code is closely associated with the cosmological. The identification between settlement and world appears to be weaker for the feudal mode of production in Europe, or in other words, the analogy appears to be reinforced in comparison to the above modes of production. The cosmological code and perhaps also a political code are semiotically dominant. It is the antique mode of production that historically inaugurates the gradual disappearance of identification in favor of analogy between settlement and world. The political code is semiotically dominant. The esclavagist mode of production seems to present the same characteristics. During the period of transition from feudalism to capitalism, the settlement generally bears only an analogous relation to the world, more abstract than in the societies of antique and esclavagist modes of production. Finally, in the industrial mode of production, the cosmological connotative code in principle disappears. Semiotically dominant, for the produc-

196

Alexandros Ph. Lagopoulos

tion of urban space, are the vital, the utilitarian, the historical, and the aesthetic codes, on the plane of discourse, and the utilitarian and economic codes, on the plane of urban practice. On the basis of the preceding analysis, we conclude that: (a) The dominant connotative codes of the semiotic production of urban space derive from the dominant plane of the social formation. (b) The (strong) identification between settlement and world appears within those modes of production where religion is part of the dominant structure; this identification begins to weaken in favor of an analogy with the appearance of the political plane as the main dominant; and even the analogy disappears when the dominant plane becomes the economic one. If we correlate the general geometrical form (plane of Sad) of the semiotic models of urban space with the modes of production, we can conclude the following: (a) The distribution of the semiotic models among the modes of production does not present any line of evolution; that is, the geometrical forms of the models do not succeed each other in any specific and systematic manner in relation to the evolution of the modes of production. O n the other hand, it appears that the generation of these models is not a function of the transformation of modes of production. However, these two observations cannot be extended to cover the industrial mode of production. (b) It is not possible to correlate a particular semiotic model exclusively with any corresponding mode of production, with the exception of the industrial mode. Thus, the same model is able to function within different modes of production, while different models are able to function within the same mode. (c) Within a particular society, a transformation in the mode of production may or may not be followed by a transformation in the semiotic model. Inversely, a transformation in the semiotic model can take place within one and the same mode of production. There is a persistence of certain general ideological, semiotic, and geometrical complexes through the transformation of modes of production (see also preceding paragraph), a fact that does not deny, however, the determinant role of the latter. In fact, every new mode of production determines nezo functions and new meanings, which integrate the old ones; the geometrical (syntagmatic) plane seems to be the most independent.

Semiotic Urban Models and Modes of Production 197 The general geometric forms of the semiotic urban models of precapitalist societies belong to an extremely limited range among the possible geometric choices. This range is composed of two basic categories of forms: a "centric" category, concentric or radial, generally with circular contour, and an "orthogonal" category (including also the chessboard pattern), which may include a cross, with a rectangle, square, or free polygon as contour. In both categories, a central element dominates. Model types made up of two parts on either side of a diametrical axis ("diametric" type), or of a cross and a circular contour—and the combinations of these types and those in the "centric" category— seem to represent historically the passage from the "centric" to the "orthogonal" category. The following comment can be made in relation to the formal types belonging to the above categories: (a) Though they differ on the level of Sad/ they meet on the level of Sec, for the set of Sec of the cosmological codes corresponding to them signifies "world." In order for this meaning to be realized, that is, for the settlement to signify "world," it has historically been used as the vehicle for the creation of an analogy with the world. This analogy has been created by means of orientation and concentric repetition: in the case of the "centric" category, orientation inward toward the central point having as Sec, "center of the world" and "world," or outward toward all of exterior space—the world—with or without concentric repetition of the contour; in the case of the "orthogonal" category, orientation outward toward the cardinal points of the horizon, the four ends of the world, from a central point, the center of the settlement and the world (see also Eliade 1957: 41-43). (b) If they are to a certain degree semantically equivalent, they nonetheless appear to have different historical origins. The origin of the types of the "centric" category, with the exception of types whose axis or axes are oriented to the cardinal points (a combination with the "diametric" category or the circular type with cross), appears to have preceded agricultural societies, while the archaic type with cross and the other types of the "orthogonal" category seem to be linked to the practice of agriculture, given that this involves the observation of seasonal cycles and related astronomical phenomena (e.g., the movements of the sun, hence the definition of the cardinal points). The most geometrically

198 Alexandros Ph. Lagopoulos evolved type of the "orthogonal" category, the chessboard pattern, is in general an (expected) transformation of the type with cross. (c) These types are characterized by centrality, in view of the dominance of the center, while, on the connotative level, this centrality becomes polarization. (d) They represent, as we have remarked, an extremely limited range among the possible geometric forms, including neither certain elementary geometric forms (e.g., the triangle), nor more complex forms, such as those of the urban models of industrial societies. In precapitalist societies, the transformation of the geometric form of a semiotic urban model can be either diachronic or synchronic. In the second case, it can be doubly synchronic, in which case two or more forms coexist for the same settlement, or synchronic corresponding to a diachronic change: in this case, different forms correspond to different settlement sizes or to successive phases of a single settlement. Some formal transformations are the result of, or simply conform to, the connotative codes of the semiotic model. Others are due to extrasemiotic factors. Although the forms of the semiotic urban models and their transformations are defined by the connotative codes and generally by the global social structure in their relation to their environment, they nonetheless show a relative autonomy. The transformations seem to obey geometric laws of transformation. The forms in industrial societies are probably the object of a different and more complex family of transformations. The Sad of the semiotic urban models, as well as their transcription in cartographic representation, can be classified into three main paradigmatic categories: (a) the category of points: the pSa corresponding to this category may be as varied as the ancient Greek hestia and agora, the perrons of the Middle Ages, and a modern plaza; (b) the category of lines, the principal corresponding pSa of which are the main and secondary roads of a settlement, and its internal and external boundaries; and (c) the category of surfaces, the principal corresponding pSa of which are the zones of the settlement.

Semiotic Urban Models and Modes of Production

199

There is a fourth paradigmatic category, that of volumes, which has appeared principally in recent times. It is used primarily by Utopian architects. The semiotic urban model is governed by a general code, corresponding to the langue of linguistics. This code is actualized in a particular manner in the model and in each specific settlement, the actualization corresponding to the parole of linguistics and constituting a kind of discourse. In precapitalist societies, the semiotic urban model forms a whole obeying syntactic laws of a geometric (plane of the Sad) and probably logico-algebraic (plane of the Sec) nature; the latter seem to be carried by connotative relations governed by the triadic structure + /o/ - or the dualistic structure + / - . In industrial societies, the model also forms a whole governed by syntactic laws of a geometric nature; it has a more flexible geometric form. In the case of both precapitalist and industrial societies, the model is subject to certain transformations, general or specific (e.g., only on the level of the Sa d ), diachronic or synchronic, that take place through autoregulation—when we consider the semiotic system as (relatively) autonomous—and probably obeys laws of transformation concerning the level of the Saj. Thus, the model seems to present the characteristics that Piaget (1968: 6-7) attributes to a structure. This semiotic structure is governed by the global social structure. Both on the connotative and on the denotative levels of the urban model, we can distinguish a set of (sub-) levels corresponding to different ways of dividing the Sa d of a settlement according to different degrees of detail; each level corresponds to an approach to urban space on a different scale. The number of the resulting spatial scales is very limited. A Sad of such a scale is composed of one or, usually, more Sad of the immediately more detailed scale. In precapitalist and industrial societies, the passage from the morphological level of the urban model to the pSa is not, in general, a matter of direct reflection (e.g., the circular contour of the model may become in reality ellipsoidal or a closed free curve, the actual orientation may not accord with theoretical orientation). The passage from the Se of the denotative level of the model to reality frequently presents the same characteristic (e.g., the theoretical number of neigh-

200

Alexandros

Ph.

Lagopoulos

borhoods can be altered). Finally, the projection on its connotative level onto reality is extremely persistent. In the first two cases, the décalage between reality and model is due to social and/or environmental factors.

NOTES 1. On the relationship of explanation and articulation (and comprehension), see Goldmann 1970: 2 0 - 2 1 . 2. For a brief description of the main characteristics of the various modes of production, see Godelier 1974: 60-70. 3. The material presented in this section was partly based on Lagopoulos 1973, 1975, 1978. 4. These conclusions are based on Lagopoulos 1978.

REFERENCES Banu, I. 1974. "La formation sociale 'asiatique' dans la perspective de la philosophie orientale antique." In Sur le "mode de production asiatique" (C.E.R.M.), pp. 285-307. Paris: Editions Sociales. Choay, F. 1965. L'urbanisme: Utopies et réalités. Paris: Seuil. 1970-1971. "Remarques à propos de sémiologie urbaine." L'Architecture d'Aujourd'hui 153:9-10. Egli, E. 1959. Geschichte des Städtebaues, vol. 1: Die alte Welt. Erlenbach-Zurich, Stuttgart: Eugen Rentsch. 1962. Geschichte des Städtebaues, vol. 2: Das Mittelalter. Erlenbach- Zurich, Stuttgart: Eugen Rentsch. Eliade, M. 1957, 1965. Le sacré et le profane. Paris: NRF/Gallimard. Fraser, D. (n.d.). Village Planning in the Primitive World. London: Studio Vista. Godelier, M. 1973. Horizon, trajets Marxistes en anthropologie. Paris: Maspero. 1974. "La notion de 'Mode de Production Asiatique' et les schémas marxistes d'évolution des sociétés." In Sur le "mode de production asiatique" (C.E.R.M.), pp. 47-100. Goldmann, L. 1970. Marxisme et sciences humaines Paris: NRF/Gallimard. Greimas, A. J. 1974. "Pour une sémiotique topologique." Notes Méthodologiques en Architecture et en Urbanisme (Sémiotique de l'espace) 3/4:1-21. Griaule, M. 1966. Dieu d'eau: Entretiens avec Ogotemmêli. Paris: Fayard. Griaule, M., and G. Dieterlen. 1953. "Afrique noir." In Symbolisme cosmique et monuments religieux, pp. 1 0 1 - 6 . Paris: Musée Guimet. 1965. Le renard pâle. Paris: Institut d'Ethnographie, Université de Paris. Lagopoulos, A. Ph. 1973. Structural Urbanism: The Settlement as a System (in Greek). Athens: Technical Chamber of Greece.

Semiotic Urban Models and Modes of Production

201

1975- "Semeiological Urbanism: An Analysis of the Traditional Western Sudanese Settlement." In P. Oliver, ed., Shelter, Sign, and Symbol, pp. 206-18. London: Barrie and Jenkins. 1977. "L'image mentale de l'agglomération." Communications 27: 55-78. 1978. "Analyse sémiotique de l'agglomération européenne précapitaliste." Semiotica 23(1/2): 99-164. Lavedan, P. 1954. Représentation des villes dans l'art du Moyen-Age. Paris: Vanoset, Editions d'Art et d'Histoire. Ledrut, R. 1973. Les images de la ville. Paris: Anthropos. Lefebvre, H. 1971. Au-delà du structuralisme. Paris: Anthropos. Lévêque, P., and P. Vidal-Naquet. 1964. Clisthène l'athénien: Essai sur la représentation de l'espace et du temps dans le pensée politique grecque de la fin du Vie siècle à la mort de Platon. Paris: Les Belles Lettres. L'Orange, H. P. 1953. Studies on the Iconography of Cosmic Kingship in the Ancient World. Oslo: H. Aschehoug. Lynch, K. i960. The Image of the City. Cambridge: MIT Press. Marx, K. 1957. Contribution à la critique de l'économie politique. Paris: Editions Sociales. Müller, W. 1961. Die heilige Stadt. Stuttgart: W. Kohlhammer. Piaget, J. 1968. Le structuralisme (Que Sais-Je?). Paris: PUF. Scheibling, J. 1977 "Débats et combats sur la 'crise' de la géographie." La Pensée 194: 41-56. Travlos, J. 1971. Pictorial Dictionary of Ancient Athens. London: Thames and Hudson. Vernant, J.-P. 1974. Mythe et pensée chez les Grecs. 2 vols. Paris: Maspero.

9 CULTURE, IDEOLOGY, AND THE SIGN OF THE CITY M. Gottdiener

The author's first work in semiotics was a l g j j study of fashion which was influenced heavily by Barthes' analysis of the relation between ideology and material objects. Case study research on the development of metropolitan regions in the United States confronted the author with the critical limitations of mainstream approaches to understanding contemporary urban spatial structure, especially the weaknesses of urban ecology. The author's principal work has been devoted to reconceptualizing urban analysis in light of these limitations (Gottdiener 1985b). The study of the role of ideology in the structuring of urban space according to the socio-semiotic perspective was one logical extension of this effort. This paper examines the limitations of mainstream approaches to signification in urban space. Two perspectives are examined: the culturist school of urban ecology, and the communications model of urban planners. The first view is exposed as fallacious in its basic premises, while the second is simplistic when compared with the socio-semiotic alternative. The author concludes this paper with some remarks on the place of signification in contemporary American cities deriving from his approach called the "social

Culture, Ideology, and the Sign of the City 203 production" of urban space (1985b). This socio-semiotic analysis is also exemplified by reading paper 13 below.

In the introduction to this volume the editors asserted that the sociosemiotic approach is sociologically attuned as well as semiotically cast. The following seeks to develop this argument in more detail. Specifically, if the socio-semiotic approach is the vehicle for the analysis of meaning in the city, then by what conceptual frame can that assertion be understood best? To begin with it is necessary to call attention to more conventional approaches in the literature which seek to analyze the urban environment as a socially shared symbol. Such efforts are few and far between once we discount the inquiry into perceptual imaging started by Lynch. In fact, beginning with the early work of Wohl and Strauss (1958) and Firey (1945), little has been done in the way of analyzing the city as a social symbol until quite recently. To be sure, sociologists have turned their attention to what has come loosely to be called "urban culture," generally associated with the approach of Wirth to lifestyles, but which upon examination is not about cultural processes at all but only their demographic determinants (Fischer 1975, 1976). As Gerald Suttles has remarked, "Practically everyone seems to give local sentiments and culture passing attention, but that is usually the end of it" (1984:283). What seems to be missing is a concern with the way urban artifacts, discourse about the city environment, and locationally oriented ideologies comprise factors in the determination of urban processes and activities. This is precisely the object of analysis chosen by the socio-semiotic approach. Thus, a concern with the social life of signs in urban environments promises to return the study of symbolic processes back to the center of inquiry on culture in the city. In order to bring out in clearer relief the unique contribution which socio-semiotics can make to the study of the city as a symbol, let us pause and glance over some recent work by mainstreamers which purports to improve upon the early efforts of pioneers like Firey and Lynch. Because I mentioned Suttles first, let us begin with

204

M. Gottdiener

him. I will then consider a separate argument recently advanced by Donald Appleyard which tries to rescue the approach of Kevin Lynch. THE FALLACY OF ECOLOGY AND THE ECOLOGICAL FALLACY Suttles (1984) advances a critique of Walter Firey's early casting of the analytical place for sentiments in the equation on urban behavior. In this Parsonian model Firey's notion is intrasubjective. Environmental meanings are not observable directly. Instead, Firey hypothesized that personal beliefs were linked to action through their concordance with overarching values held in common with others in the local area. As such Firey's intrasubjective sentiments can be said to exist whether they are manifested concretely or not and his theory is non-falsifiable. By way of critique Suttles has sought to transfer analysis to objectively accessible elements of culture which are empirically observable, or what he calls "the vast, heritable genome of physical artifacts, slogans, typifications, and catch phrases" (1984:284) associated with urban life. On the surface such a change appears plausible as a means of improving upon the evident limitations of Firey's work. But, we may rightly wonder, after taking his best shot, whether Suttles has provided us with an approach that can replace the one articulated almost forty years ago. In my opinion Suttles has not. Instead his positivist recasting of the problem which Firey addressed leads us away from an important premise of that work and recreates deep-seated errors found in all empiricist thought which seeks to force analysts to confine their attention to mere appearance. Firey's object of analysis was the search for a means of conceptualizing how the meaning of city space becomes an organizing referent for urban behavior. His argument was that, while economic and political factors influenced urban life, socially shared sentiments also played an important and at times determining role. Clearly the specification of such sentiments by some hypothetical intrasubjective model fails to convince us today that the meaning of space is an important determinant of action. However, Firey's original concern remains valid. In fact, much of cognitive geography operates from

Culture, Ideology, and the Sign of the City

205

the very same premise, namely, that urban behavior can be understood through the analysis of environmental meanings subjectively held by individual city inhabitants. Suttles has abandoned this premise and substitutes one with a different emphasis which shifts the source of symbolic meanings. As he asserts, The meaning of these artifacts is embedded less in mass sentiments than in the authoritative knowledge "experts" who interpret them. Their causal importance, in turn, lies less in their capacity to express mass sentiments than in their ability to elicit them. (1984:284)

Furthermore, among those "experts" providing the city with authoritative meanings, producers of written texts such as novelists, essayists, journalists, and poets count most. Suttles seems to like especially Norman Mailer's characterizations of cities. Firey and, later, cognitive geographers wish to call attention to a crucial fact about urban life. Namely, that behavior within it and therefore the understanding of urban processes, such as the determination of land values, is a partial function of commonly held and sustained sentiments. This is not a quality deriving from some discursive interaction between "experts," neither city detractors nor boosters, for example, but an aspect of everyday social interaction involving what Suttles refers to as "the masses" but which w e can more astutely understand as comprised of social groups in a stratified society. To suggest that the source of meaning lies more in the activities of experts and in the elicitation of sentiments, as Suttles does, rather than the use of objects as sign-vehicles for the ideologies of all social groups, is to move us away from the study of the production, circulation, and conception of urban culture as based in the group life of all urban residents. In a passion of positivism we move from the desire to ground cultural analysis onto an objective referent to the assertion that the material culture of the authoritative elite is more significant for behavior than the material culture of the masses because the former determines the meaning of the latter. As Suttles admits, the images of the city which he considers most important may not be familiar to the general citizenry, but that is not significant (p. 302). What does matter is the way in which the typifications of the environment by experts "arm[s] formulaic journalists and

2O6 M. Gottdiener boosters with the instruments of collective mobilization. Jointly these experts provide for much of the fixity of local collective representations and a delayed and selective incorporation of new representations that maintain urban characteriological unity" (p. 302). In short, Suttles' Mass-Elite model of urban culture reproduces arguments voiced long ago by elitist critics of mass culture and is more ancient than Firey's Parsonianism. Clearly the process described above is one of several aspects of urban culture. How then does the socio-semiotic approach differ from this reduction? According to socio-semiotics the symbolic conception of objects is indeed the basis of material urban culture. However, these objects encompass an incredible variety of forms. They include buildings as well as graffiti, or, the sign systems which resemanticize instrumental facades. Even space itself, such as the city street, is an important object of signification. Finally, semiotics also studies discourse both written and verbal as significant symbolic referents in the organization of behavior. Writings about the urban environment include planning commission reports as well as the metaphors of novelists, with the former clearly carrying more behavioral weight. Verbal discourse is intersubjective, hence it can be studied. As Ledrut (1973) has shown, this discourse is a latent artifact of urban behavior and merely requires that a researcher ferret it out and write it down in the technique of field work. Such intersubjective judgments are social constructions and differ from the inaccessible private meanings hypothesized by Firey as lying behind urban behavior. Second, meaning in the city is multi-coded. Each social group possesses its own conception of urban space just as the different interests in the city compete with each other over control of the social surplus. The users of space in a complex, modern society are hardly a homogeneous mass. Even among primitive peoples, as Claude Lévi-Strauss (1963) has pointed out, the differences in environmental conception mirror the principal status cleavage of the tribe. In more modern cities, cognitive geographers have long since established that the environmental image is differentiated according to the social order. Gender, age, and status groups as well as classes hold fundamentally different images of the city and base the organization of daily life on these differentially conceived symbols. It's not

Culture, Ideology, and the Sign of the City

207

that overarching typifications do not exist. Clearly, they do. They do not, however, determine urban behavior, as cognitive researchers have amply shown. Third, given the polysemous or multi-coded nature of urban life, the need of some groups for more stable, uniformly conceived representations requires a management of the clash of oppositional environmental typifications. It is important to ask, on the one hand, what the groups are which possess a need for overarching, historically invariant symbols and, on the other, what the mechanisms are that are utilized to achieve ideological hegemony. In the former case, it is apparent that the real estate industry, homeowner associations, chambers of commerce, and banks are the most active managers of symbolic generalizations. Through the actions of these groups the representational typifications of use values emerging from daily life are converted into ones which are more useful for the exchange value of property and business investment. However, this is by no means an automatic process of sign conversion. It is contentious and contingent depending as it does on the ability of special interests to control the symbolic interpretations of processual outcomes in everyday life. Signs of boosterism which portray a unified image of the community must be superimposed upon a more fundamental socio-semiotic process involving a politics of signs among contending groups within the city and is rarely achieved with any true effect. This contentious process aimed at symbolic hegemony is more significant than the one taking place between separate cities, despite Suttles' contrary suggestion. In order to prove this assertion, it is necessary to digress a bit and discuss the role of signs as symbolic vehicles in the contentious clash between use values and exchange values within urban environments. The limitations of Suttles' approach derive from his need to tailor his remarks to fit the ecological model of city development. According to ecologists, urban processes involve interaction between countless individual producers and consumers whose aggregate effects are felt through the market for urban land. Cities themselves compete with each other for resources. Thus competition between individuals is replicated as competition between urban places. Finally, changes in transportation and communications innovations determine the physical arrangements and intensity of this competition. In recent years

2O8 M. Gottdiener such changes have meant that both consumers of housing and businesses have become more footloose and less tied to any one place. In short, urban places are currently forced to compete with other locations more intensely for the resources provided by revenue sources and employers. Using this model Suttles emphasizes the role of the urban imagery produced by "experts" as a weapon in this world of vigorous competition between separate places under conditions of resource scarcity. A positive and attractive image becomes a tool in the fight to acquire the benefits of growth. Hence the need for an overarching collective representation. This ecological model has not held up at all against the views of its many critics. First, the notion that urban patterns are produced according to neo-classical mechanisms of near perfect competition between large numbers of separate individuals is a patent absurdity. As the French urban sociologists have argued, the behavior of big firms and monopolistic powers are the central determinants of land values (Lojkine 1977; Lipietz 1977; and Scott 1976). Second and most importantly, ecologists act as if the state does not exist. Yet it is clearly a major source of financial support for cities and the principal way in which growth is channeled into the distinctive urban patterns of not only the U.S. (Gottdiener 1985b), but also communist countries (Misztal and Misztal 1984). In a third manner of critique, the technological reductionism of ecologists has been effectively unmasked as a spurious factor in producing spatial organization although it is a means through which more proximate causes operate (Gordon 1977; Gottdiener 1983; Warner 1962; and Schnore 1962). Finally, in addition to the state, recent research has documented the decisive roles that labor force considerations and the domination by capital of the class struggle in a global sourcing strategy play in the location decisions of firms (Frobel et al. 1980; Sassen-Koob 1984; and Storper and Walker 1983). With this formidable critique in hand, what then can be said of the role of urban images in the competition between places for employment and financial resources? First, the urban images which Suttles highlights that are found in the discourse of authorities are largely irrelevant. Perhaps the most important collective representation in the battle between cities for resources is that projected by growth coalitions—the specific combination of public and private interests

Culture, Ideology, and the Sign of the City

209

promoting the development of the city. These are images produced and managed by city interests and for use in specific forums, but hardly left to the whims of journalists or novel writers. Because they are tools for the promotion of regional power blocs such collective representations are not fixed images but vary according to the audience they are intended to reach or the message they are asked to convey. During the 1960s cities did well at the federal level by calling attention to the pathologies of local life. By the 1980s this image had shifted to the more fashionable one connoting a proper business climate and aimed at prospective employers and not the public sector. Consequently, research on collective representations must connect symbols with social interests and the use of images as tools in specific forums. They are mutable and socially grounded rather than freefloating phantasms disembodied from social contexts. Second, most collective images have the same appeal; they are intended to attract resource support for the local area. However, this image is hardly a determining factor in the elicitation of development. In fact, attractive symbols are way down on the list of growth inducers after state intervention, labor force considerations, the needs of corporations as determined by their place in the world system of capital, local tax policies, market or distribution considerations, and public subsidies. Of course, competition exists between places over the attraction of resources but symbolic appeals alone rarely play a determining role in this process. More likely they are conjunctural mediators working together with other, non-symbolic factors. For example, on balance New York is more the recipient of negative feelings than positive ones among media "experts" or influentials, and most Americans would object strongly to living there, yet it is still the wealthiest city in the U.S. If its manufacturing base has declined in recent years, its other sectors have expanded to take up most of the slack just as they have in other cities with mixed media reviews, such as Boston, Houston, and Los Angeles. By the same token the fact that New York symbolizes great wealth and power does not prevent prestige corporations from locating headquarters in relatively less known areas such as suburban Connecticut. Consequently, a variety of factors and trends expose the limited influence of city image, negative as well as positive, in the competition among urban places for new local resources.

210

M. Gottdiener

In addition, because the sign of boosterism is often connected to a specific growth coalition, the interests supporting development are often opposed by alternate interests such as "no-growth" coalitions. In fact, the opposition "pro-" vs. "no-" growth is central to the local politics of most urban areas, with no-growth understood as a coalition advocating more planning or greater self-management of development. Thus, even the most common collective representation for a city may be discovered to typify a special interest or coalition and may be opposed by an alternate image representing an oppositional view of the urban past and future. Beneath the level of appearance, then, a battle of signs often rages which signifies fundamental differences in conceptions of the urban interest. Third, the image of the city is important to land values and local location decisions, but this influence occurs more between the city and its surrounding area, and within city neighborhoods, than between cities themselves. Once busineses and home owners choose to locate within a particular region of the country, the macro factors guiding that decision, such as labor force considerations, are less important and so too is the overarching image of the area. Instead businesses look for tax breaks, land subsidies, industrial parks, and agglomeration incentives such as malls. Homeowners, in turn, are principally concerned with housing values, the quality of local schools, crime, tax rates, and the size of a home for its cost. These are decision factors operating as micro determinants of location within areas. Here the image that a particular neighborhood, section of town, or suburb possesses becomes critical in the competition for resource-producing residents as opposed to resource-draining ones. Once again, it can be emphasized that this image is hardly left in the hands of experts. It is produced and managed by the interest group of local inhabitants in combination with real estate, religious, business, hospital, and local political leaders. Each homeowner, renter, or neighborhood institution plays some role in the casting of the local reputation. The real estate industry alone is responsible for the production of a formidable amount and variety of signifiers which constitute the written artifacts of the regional text along with local histories, promotional brochures, and the like. Added to these signs are the verbal accounts of reputation and even local folklore about places which can be discovered in abundance even for new constructions like shopping malls. Prospective residents will tap into

Culture, Ideology, and the Sign of the City

211

this folklore and, if they are smart, listen to others as well as the advice from their real estate agent. One purpose of collective representations is to attract or repel, but there are others, such as those which connote social functions: entertainment—"downtown"; leisure—"Belle Harbor"; or commerce— "city of industry." The principal aspect of these signs is that they are part of the economic and political group practices which constitute the very core of urban activities. In sum, they are not artifacts removed from the daily life of urban residents which elicit behavior. To believe this is to assert the ecological fallacy. Signs of place are mediations organized at the level of social interaction itself and utilized as tools to facilitate everyday life by helping to organize and direct action. The place of the environmental image in urban activities is uncovered by close examination of the manner by which signs are produced, challenged, and reproduced within cities as much as between them. As cultural vehicles located within a capitalist system of social organization these signs follow a dynamic relation of conversion between symbols of exchange value and use value, a subject which cannot be discussed here (see Gottdiener 1985a). What is important, from a socio-semiotic perspective, is to connect image with its corresponding social group or interest, and to explain the symbolic process as part of social interaction within organized social systems. As stated in the introduction, this social behavior is constituted by non-semiotic processes of economics and politics as well as symbolic ones. Thus, the city is a pseudo-text and signs of place are mediations. They are vehicles for the organization of group action rather than the reified ellicitators of action as ecologists suggest (Suttles 1984:284). If collective representations of spatial environments mediate social action, how then can we specify the manner by which they function along with non-semiotic processes operating within those same places? A second conventional model of urban imaging comes close to answering this question and has been proposed by Appleyard (1979). FROM THE IMAGE OF THE CITY TO THE MEANING OF THE CITY Appleyard, like the ecologists, is a prisoner of a limited conceptual paradigm, in this case it is the outdated "communications" model of

212

M. Gottdiener

urban behavior (Meier 1962; and Webber 1963). Yet, Appleyard's approach is refreshing after a reading of the ecologists. Rather than ignoring the stratified basis of society, he states quite clearly that city signs are symbolic vehicles in the often contentious quest by social groups for identity, status, and power. Furthermore, Appleyard takes hold of the very same basis for meaning which forms the cornerstone of socio-semiotics, namely, the intimate link between the sign and social interest. As he states, "An environment becomes a symbol when it is intended or perceived as a representative of someone or some social group; when social meaning plays an influential role in relation to its other functions" (1979:144). Lastly, Appleyard breaks down the users of signs in a manner similar to sociosemiotics and according to groups of producers and consumers of environmental images. Such assertions have much in common with the approach advocated here. Even the core question posed by Appleyard is quite compatible with the research interests of urban semiotics, as he inquires, "How do environmental actors express the identity, power, and status of their creators and how do they communicate with their audiences?" (1979:148). Yet, important differences remain between our approach and his. On the one hand, Appleyard reduces all symbolic interaction in the city to the referential function of communication because of the limitations of his model. In this approach there is a distinct producer and receiver of messages, and symbols are intentionally produced or perceived. On the other hand, the communicative act is said to be organized around perception, in which case Lynch's analysis of image quality serves as the basis of symbolic action. The socio-semiotic approach takes issue with these premises. Addressing the second one first, we have already discussed the reasons why Lynch's approach is limited and the significant improvements which can be made by substituting a conceptual approach to meaning rather than a perceptual epistemological basis for the city image. From our critique it is not surprising, that, once Appleyard gets down to cases, he emphasizes transportation effects (i.e., movement) as the major source of image perceptual strength (1979:148) (cf. Ledrut's critique in paper 5). As Choay remarks (see paper 7), systems of legibility connected to transportation are merely supple-

Culture, Ideology, and the Sign of the City

213

mentary systems and should not be confused with providing the city with a more fundamental "symbolic weight." Once we focus on conception rather than perception, however, images of place derive their strength and transformative dynamics from a variety of social activities, just as exposure to the discourse of city signs comes from a variety of separate sources or channels of information. What about the first limitation of Appleyard's argument? The communications model is a reductionist view of symbolic interaction. It ignores the complex variety of meaningful acts mediating urban activities. On the one hand, city artifacts are polysémie and possess more than one socially sustained meaning, so that this model's assertion of the unitary sender is quite simply wrong, as both Ledrut (1973) and Choay (1965) also contend (see Ledrut, papers 5, 10; Choay, paper 7). On the other hand, consumers of city signs key into the function of place so that environmental conception is ruled by the poly-functionality of urban practices (see Eco, paper 2; Barthes, paper 3). This convergence of separate conceptions means that consumers of space utilize different meanings for the same location precisely because each location within the city is used for different things by different social groups. At the same time, space itself becomes the object of social interaction over contrasting signifieds. This clash of meanings is often resolved by the atrophy of all signifieds in the sign of place to the benefit of monosemic signifiers of instrumental function. Most often, however, instrumental space becomes re-semanticized through social uses. Spatial systems of signification in Late Capitalist urban environments, therefore, are directed by a dialectic between closed instrumental functions and open social levels of connotation. This dialectic of spatial semiotics corresponds on the material, non-semiotic level to the relation between the use values of social space and the exchange values incorporated in abstract space (see Ledrut, paper 10; Eco, paper 2; Choay, paper 1 1 ; Gottdiener, paper 13). The semiotic implications of poly-functionality for urban places are brought out clearly by Agrest's study of the street (1978) which contrasts with Appleyard's model. The referential function of communication, which is the focus of Appleyard's approach, reduces the street to a simple conveyer of messages. By applying Jakobson's expanded model of meaning, Agrest (1978:218) was able to identify

214

M. Gottdiener

at least six separate symbolic sources including the role of streets in poetic, phatic, and metalingual communication (see also Eco, paper 2). In short, cognitive studies which limit our understanding of the communications process in space to the transfer of messages, i.e., the referential function, also restrict our appreciation for the polysemic nature of city space. As Agrest (1978) astutely points out, however, all functional views of communication, even Jakobson's, are limited. Appleyard's attempt to conflate intention to the three factors of identity, status, and power skirts the important question of the production of meaning in the city and urban space as molded by separate social interests. It is simply not enough to separate social groups into producers and consumers without recognizing the fluid boundaries between the two functional aggregations. Those that produce urban images are not only the elite, that is, architects, businessmen, financiers, they are also the users of space themselves. An inquiry into the production of meaning, therefore, must account for all sources of signification in group practice and must identify not only the alternate cultural codes which intersect in space but their respective means of expression (cf. Certeau 1984). Familiar sign vehicles such as design, advertising, and traffic codes play a role in this process, but so too do laws regulating the symbolic uses of private property, such as anti-graffiti legislation or building codes. Non-elite group producers of meaning, such as teenage breakdancers, bocci-ball players, home builders not imprisoned by dominant design codes, and street corner people also are active in providing sections of the city with distinct images. Consequently, the use of public spaces within the city and its regulation by the dominant city interests is as much a part of the production of meaning as are the activities of those associated with real estate and architectural development. At bottom, then, urban space is not a simple container of social processes. It is the condensation of often contentious group interaction which involves signifying practices as much as non-semiotic processes, such as the class struggle at the place of work. This space has a history. It not only signifies some meaning, but also represents the end result of an economic and political process through which one among many meanings and conflicting uses has acquired hege-

Culture, Ideology, and the Sign of the City

215

mony. As a space it then serves to reproduce that particular interest and use against other contending ones. Oppositional signs exist often only as alternative uses or facades which were repressed during the hegemonic process. Urban space, and, in fact, any space belonging to a stratified society with an oppositional social structure, represents the material manifestation of dominant interests orchestrating social organization along with the traces of historical challenges to this hegemony. The surface naturalness of appearance and its taken-for-granted quality provides false testimony for what is a constant, often contentious process of group struggle over the control of spaces. As Lefebvre observes, One of the most crying paradoxes of abstract space is that it can be simultaneously the birthplace of contradictions, the milieu in which they are worked out and which they tear up, and finally, the instrument which allows their suppression and the substitution of an apparent coherence. All of which confers on space a function previously assumed by ideology. (1974:420)

A final limitation of Appleyard's approach is that it removes us from the urban historical process. In his communication model the contextual nature of urban space is reduced to isolated examples of individual facades. These are architectural statements that people can either accept or reject, such as through graffiti writing or opposition to construction. Such an analysis asserts a synchronic unity which is false. Historical change, the struggle of groups over socio-spatial resources, transformations in the nature of capitalism, changes in the development of the city are all dissipated by this simplistic conflation. As Eco (paper 2) and Barthes (paper 3) suggest above, buildings are not single architectural sculptures free-standing in space, they are parts which relate to a larger whole—the urban context. Thus the fabric of the city and its texture changes as forms of construction and modes of land use alter over time. During the last three decades, in particular, significant phases of change have impacted the city. As a result the design of urban space has shifted and ideological systems of signification have altered along with change as mediators of socio-spatial transformations. For example, Baird (1982) astutely points out that over the past several decades a profound shift has occurred within cities away from single

216 M. Gottdiener plot construction and the nature of urban design suitable to it and toward the building of superblocs. This shift to a larger urban scale has been replicated across the metropolitan region, mixing with it elements of older modes of construction (Gottdiener 1985b). Accompanying this change has been a qualitatively different organization of construction efforts which rely on deep-level changes in Late Capitalism. Mediating the developments of the past decades as ideological vehicles have been spatial systems of signification which have altered the manner by which we think about urban environments and which have reorganized the priorities judged important in the quality of city life. The notion of "community," for example, once important in the 1950s, has been drowned in a sea of fiscal concerns about home costs, financing, tax rates, economic growth, and the decline in the local quality of life. As urban areas have been restructured according to the larger contemporary scale of development under the new sign of revitalization and economic growth, the very face of the city has been altered irrevocably as has the quality of social interaction characteristic of the urban experience.1 In short, the urban image must be read as ideology, as an historical product, as a gesture with a past, as an outcome of a class society propelled by powerful forces of development and change. All analyses at the surface level are limited because they assume that appearance is a simple product of singular interests manifesting themselves in space with little connection to underlying social forces and with no history. On the contrary, the surface calm of the city image belies its constitution as the condensation of the struggle between various organized group expressions about alternate use and design. This sign is produced by economic and political forces, such as property law, as well as cultural expressions. The study of the city image, then, compels us to investigate the struggles for control of space and the manner by which certain ideological representations succeed while alternatives fail to materialize their traces. The image of the city does not float above the oppositional nature of society nor exist alone in the minds of some elite few, it belongs at the level of group interaction and as a banner in the fight between contending social interests, as some vision of what society might be rather than what it has become.

Culture, Ideology, and the Sign of the City

217

NOTE 1. According to the approach of the editors, architectural forms and the experience of the city cannot be analyzed in isolation from exo-semiotic processes. Consequently, I abandon the conventional mode of architectural criticism that depends on the independent interpretations of the critic in favor of one that links deep level societal changes with the production and use of expressive objects in their social contexts (cf. Gottdiener 1985a). Yet, I must caution the reader against using a simplistic Marxian model that would establish some one to one correspondence between modes of production and cultural forms (see Lagopoulos, paper 8). Because urban space is a contingent product of deep level processes, it is not possible to periodize its forms and culture in the same way we periodize changes in the mode (see Gottdiener 1985b, ch. 8). Thus Davis (1985) is correct to reject a critique of Post-Modemist architecture which fails to account for the deep-level processes structuring real estate activity and, therefore, which make changes in modes of architectural practice possible. Analysis of architectural forms should account for all of space, i.e., the contextual space produced by uneven, regional processes of urban growth under capitalism, rather than be restricted to the forms of the built environment themselves.

REFERENCES Agrest, D. 1978. "Toward a Theory of Sense in the Built Environment." In S. Anderson, ed., On Streets. Cambridge: MIT Press. Appleyard, E. 1979. "The Environment as a Social Symbol." American Planning Association Journal 45 (2): 143-53. Baird, G. 1982. "Type and Transformation in North American City Form: A Commentary on the Evolution of Historical Examples in Toronto." Unpublished paper. Certeau, M. de. 1984. The Practice of Everyday Life. Los Angeles: University of California Press. Choay, F. 1965. L'urbanisme: Utopies et réalités. Paris: Seuil. Davis, M. 1985. "Urban Renaissance and the Spirit of Postmodernism." New Left Review 151: 106-14. Firey, W. 1945. "Sentiment and Symbolism as Ecological Variables." American Sociological Review 10: 140-48. Fischer, C. 1975. "Toward a Sub-cultural Theory of Urbanism." American journal of Sociology 80: 1319-41. 1976. The Urban Experience. New York: Harcourt Jovanovich. Frobel, F., J. Heinrichs, and D. Kreye. 1980. The New International Division of Labor. New York: Cambridge University Press. Gordon, D. 1977. "Class Struggle and the Stages of Urban Development." In D. Perry, and A. Watkins, eds., The Rise of the Sunbelt Cities, pp. 55-82. Beverly Hills: SAGE. Gottdiener, M. 1983. "Understanding Metropolitan Déconcentration: A Clash of Paradigms." Social Science Quarterly 64 (2): 227-46.

2i8

M . Gottdiener

1985a- "Hegemony and Mass Culture: A Semiotic Approach." American journal of Sociology 90(5): 979-1001. 1985b. The Social Production of Urban Space. Austin: University of Texas Press. Ledrut, R. 1973. Les images de la ville. Paris: Anthropos. Lefebvre, H. 1974. La production de l'espace. Paris: Anthropos. Lévi-Strauss, C. 1963. "Do Dual Organizations Exist." In Structural Anthropology, pp. 128-60. New York: Basic Books. Lipietz, A. 1977. Le capital et son espace. Paris: Maspero. Lojkine, J. 1977. Le Marxisme, l'état, et la question urbaine. Paris: Centre de Sociologie Urbaine. Meier, R. 1962. A Communications Theory of Urban Growth. Cambridge: MIT Press. Misztal, B., and B. Misztal. 1984. "Urban Social Problems in Poland: The Macro-Social Determinants." Urban Affairs Quarterly 19 (3): 315-28. Sassen-Koob, S. 1984. "The New Labor Demand in Global Cities." In M. Smith, ed., Cities in Transformation. Beverly Hills, Ca.: SAGE. Schnore, L. 1962. "Municipal Annexations and Decentralization: 1950-1960." American Journal of Sociology 67: 315-28. Scott, A. 1976. "Land and Land Rent: An Interpretative Review of the French Literature." In C. Boord, et al., eds., Progress in Geography, pp. 101-46. London: Edward Arnold. Storper, M., and R. Walker. 1983. "The Theory of Labor and the Theory of Location." International Journal of Urban and Regional Research 7 (1): 1 - 4 1 . Suttles, G. 1984. "The Cumulative Texture of Local Urban Culture." American journal of Sociology 90 (2): 283-304. Warner, S. 1962. Street Car Suburbs. Cambridge: Harvard University Press. Webber, M. 1963. "The Urban Place and the Non-Place Urban Realm." In M. Webber, et al., eds., Explorations Into Urban Structure. Philadelphia: University of Pennsylvania Press. Wohl, R., and A. Strauss. 1958. "Symbolic Representation and the Urban Milieu." American Journal of Sociology 63: 523-32.

SECTION B: CASE STUDIES

io THE IMAGES OF THE CITY Raymond Ledrut

Ledrut's The Images of the City, is full of theoretical insights into the conception of the modern city, and is one of the best case studies on the subject in urban semiotics. It is directly complementary to the book by Choay discussed in this volume (see Choay, paper 11). In addition to being a precursor of the approach advocated by the editors, Ledrut's work makes excellent use of empirical techniques of data gathering, including questionnaire analysis on the conception of the city. Because the editors were obliged to present only excerpts from the book, we must alert the reader to the possible distortions of editing. It is necessary, therefore, to introduce these excerpts by presenting them within the context of Ledrut's complete book. This book, which relies heavily on Greimas' Sémantique structurale, is composed of seven parts. It opens with a long and important introduction on the conception of the city (pp. 9-37). The author is not interested in psychological or psychoanalytic conceptions of a particular city, but in the conception of the city as a structured discourse. In the first part of the book (pp. 39-102), the author presents the semantic fields relating to the city: the spatial field, the field of the self, the field of values. He formulates a model of the elementary logico-semantic organizaThis paper consists of translated excerpts from Les images de la ville (Paris: Anthropos, 1973)-

22o

Raymond Ledrut

tion of the urban signifieds. Ledrut presents the lived dimension of the city and its maternal nature, and continues with the symbols of the city, among which are monuments and places. In the second part (pp. 103-54), the author observes that the parts of the city are distinguished according to four factors: geographical-geometric, functional, social, and historical. The standard spatial model is the concentric model, while spatial orientation is achieved through important spaces of the urban form or punctual elements. The author finds that there is a rather strong relation between urban symbolism and the conceptual organization of the city, and that they differ in that the two corresponding spatial images are not identical. The unity of spatial organization and the legibility of the city are secured by the symbolic level. Next, the author studies urban social life, the relation between the use of urban space and symbolism, and the signification of monuments. In the third part (pp. 155-94), the author presents the results of the analysis of the preferences of inhabitants with respect to the city. He identifies a series of classemes of value, observes that there are strong similarities between systems of preferences and the image of the city, and shows that there are great differences between these systems and the symbolic reading of the city. He defines the levels of the image of the preferred city, and closes the third part with a discussion of the correspondence between the preferred city and social groups and classes. This section is based on a survey or questionnaire analysis of a sample of inhabitants in two cities in France. The fourth part (pp. 195-312) rests on the responses by a sample of city dwellers to 11 photographs, and an important part of our text is devoted to this study. Ledrut uses thus the technique of photo-illicitation to complement his questionnaire analysis. The author finds four distinguishing dimensions for the significations of value: ethical, aesthetic, existential, and functional; and four major groups of "objective" traits (caractères): social, practical, recognitional, and formal. He concludes that the reading of the urban elements does not inform us about the elements as they are in reality, but about the urban social experience, which represents a specific organization of reality and transposes a social practice, attainable only through this experience; and that this fact leads to an approach that is complementary to the studies of behavior, the social system and the mode of production. In the fifth part (pp. 313-82), heavily represented by our excerpts, the author makes a synthesis of the results from his two empirical studies, the questionnaire and photo-illicitation research. With the help of a statistical

The Images of the City

221

treatment of the dimensions of the significations of value, he arrives at two opposed models of the conception of the city. Ledrut's general comments on the two models (pp. 338-40) are presented here. Then he shows that the respective discourses on preferences lead to two urban planning models on the part of the persons interviewed: the "functionalist" and the "hedonistic" model. He integrates these models with the models of the conception of the city (from p. 367) in a grand synthesis involving two general models that capture the fundamental distinctions of the empirical data on city images. The first, the abstract model, relates to instrumental, functionalist attitudes, while the second, the concrete model, relates to emotional, personal distinctions. Ledrut finds that the attitudes of the working class are explained best by the abstract model, while those of the middle class are explained best by the concrete model. The author claims that the distance between these two conceptions—partial views of reality belonging to different classes—measures urban alienation.

[pp. 17-19]: Portrait, Figure, Model. The image of the city is not given to us in discourses or statements; it is constructed starting from the statements and from the structure that we can perceive in them. Sometimes this structure is visible without too many difficulties, and this is the case in the true discourse of the "theoriticians" of urban planning, although the readings of it often vary a good deal. The case is not the same with an inhabitant or with a group of inhabitants. Yet the aim of the analysis is the same. What we try to reveal in the work of a professional planner is a more or less coherent urban model. And what do we mean by model in this case? We do not mean a mathematical or logical model of how the city functions, but an "iconic" model, which is in fact simply a rational—i.e., conscious and relatively systematic—image of the urban world. The essentialist character, both descriptive and normative, of these models is evident. If the planner's model is a rational image, the "non-planner's" image can often be considered as a non-rational model, implicit, latent in a sense. The image may thus be the scientific "construct," not of an already existing meta-

222

Raymond Ledrut

linguistic "construct," but of a latent structure of the world of urban significations. Thus the image takes various forms: portrait, figure, model, depending on the level of abstraction, of schematization taken into consideration. The Signifier and the Signified in the Image of the City. The image thus appears as a "symbolic unit." The mode of existence of the image of the city is that of a complex and specific significative whole. The image of the city is not the signified of the word city; it is the "concept" of city which is the signified of the linguistic signifier, the word city. Neither is the image of the city this linguistic signifier, an auditive "image" for instance. It is neither concept nor percept, neither signified nor pure signifier. It is constituted by a certain relation between the level of the signifier and that of the signified. What makes the image of the city a symbol, or rather a symbolic complex and symbolic system (otherwise it is scarcely even image) is precisely the type of relation that exists between the order of the signifier and the order of the signified. The Image of the city is like a "myth" or a "literary work."The image is not the simple expression of a simple, direct, as it were, pure signified; it is a sign which is itself composed of signifiers and signifieds. This means that there are two levels of signification and that above the first level of signification (the denotative level) is superimposed a second level of signification which takes the first one as its plane of expression. Thus, we have a second discourse in which the signs of the immediate discourse become signifiers that refer to another signified. . . . [pp. 21-25]: If we consider the image of the city from a "psychological" point of view, we both perceive and understand that there is no image without an affective resonance. The strength of an image is linked to the extent of these repercussions. There is no image in any case which might be a mere representation. The image is always emotion, whether it impels us toward or away from an object, whether it terrorizes or charms. There is always some fascination in an image: Idol or Imago. Psychoanalysis has recognized this in its very theory of the Imago. For "figures," in fact,, do not belong to the order of objectivity and concepts, even when they are set in the field of

The Images of the City

223

"science" and "technique": the functional image of the city is an image only from the moment when it becomes the locus of a certain "investment" by the ego. Mythical or not, symbolic expressions imply the existential relations which support our relations with reality. The image necessarily includes the value and the values of its object. It is more or less strong, more or less compelling; it cannot be entirely without power over our sensibility, and therefore over human behavior. The image of the city expresses less the city than a global relation of man to city. The "simulated" city is not the city itself as pure object. We are in the Imaginary. There Is No Key to Urban Symbolism. The symbolism of a city, such as it is constituted in the image of the city, makes no reference to a "key" exterior to that image or to those images. What the discourse of the city "wants to say," it "says" itself, even if this only appears on the connotative level. The image itself is the "meaning" of the city and of the discourse of the city. It does not have to be interpreted. It is thus in the discourse itself that the symbolic takes shape. It is all in the game of correspondences, of oppositions, of relations of denotation and connotation, a game which is specific to discourse whether individual or collective. We shall not try to discover another world that the discourse on the city would express in an obscure or dissimulated manner. The city is a symbol, and there is a symbolization of the city, but it is in the image itself, apprehended through and by discourse, that what the city represents for man is revealed and expressed, and that the city and its aspects are manifested in various figures, i.e., symbolized. . . . Sociology of Urban Symbolism. It is not that we would exclude the possibility of explanation, but explanation is not the dissolution of the object. The external symbolic is a destruction of the symbolic fact as such. It always partakes of the spirit that presides over the "dream-book" type of interpretation. Research on the conditions of production of symbols, myths, or images, is quite another matter. If the analysis and description of an image should not exceed the context imposed on them by discourse, the explanation of the formation of these images is obliged to look outside of discourse and image for that which makes them what they are. Here, the context is

224

Raymond Ledrut

history itself and its structures. It is quite obvious that groups (classes or other groups) as well as individuals are subjected to the action of "cultural" conditions, which by way of the media of communication and expression penetrate discourse and "thought." The planner's language and the ideology (or ideologies) that they convey, as well as the "discourse" of urban reality itself, are not completely absent from the expressions of any of us and our "vision" of the city (in the sense that we say "vision of the world"). Other determining factors have to be considered: the obstacles and constraints, or impulsions and impressions resulting from class membership. Like all "visions," the image of the city is a "cultural product." To explain the image is to reveal its process of formation. A reduction is possible here: the one implied by the distinction between fabrication of the image and consumption of the image. When and image is diffused through the "logo-technique" of a dominant group, it acquires a "mythical" character in the sense that Barthes (1957) gave to this word: it depends on the group's ideology and translates a state of dependance on the part of the "consumer" public. It is quite otherwise, however, when this image is different or even opposed to that of the dominant group. One of the essential questions posed by the study of the images of the city is thus the specificity of the various images and of their relations with the various social groups: are the images of the planners and of the group that holds the power of major decisions (financial and political) the same as those of groups without power—i.e., the "public," and in particular, the dominated classes? In the domain of urban planning, the consumer is to a great extent powerless. He is probably also—this is only an hypothesis—less influenced by the ideologies, models, and "myths" of urban planning than the consumer of clothes is by "fashion." This is probably due to the fact that the city inhabitant does not need to be stimulated as "buyer." He is a consumer who does not buy: the urban model is imposed on him. He is thus doubly powerless, but not very manipulated. The manipulated consumer is the buyer of housing and furniture: architecture and design are privileged domains of advertisement. In matters of planning, the most manipulable and sensitive subjects are those who exercise functions in the life of local collectivities without having much power of decision, or party and

The Images of the City

225

syndicate militants who have to take into consideration planning problems in their programs and propaganda. . . . [pp. 27-28]: In the cities, the dialectic between external and internal conditions is the heart of the matter: the adaptation is not automatic, the subject's vital demands may encounter an unfavorable environment which leads to disturbances in adaptation and the phenomena of regression in the relation of the individual to the environment. The image of the city is, like the image of the body, a schema necessary for anchoring ourselves in the world. Perturbations of the body image are inseparable from perturbations of behavior. Legitimately, we can suppose that it is the same with our image of the environment and thus with our image of the city. Yet it is obvious that the notion of environment introduces a restrictive element in the image of the city, that spatial behavior does not exhaust the repertoire of spatial practices, and that adaptation is not all there is to apprehension. The situation of man confronting the city involves other things than schemas of perceptive behavior. It introduces ideology, and, more simply, the consciousness of things. "To inhabit" is not only to have a territory. . . . This does not mean that Lynch's approach is not correct and indispensable. It cannot, however, exclude other approaches which pertain to another level of signification. . . . [pp. 34-37]: After an extensive pre-survey, the survey itself was effected by means of two series of interviews based on two questionnaires given to two different samples. One of these questionnaires is a "photo-questionnaire" composed of eleven photographs representing "parts" of the city submitted for the commentaries and evaluations of the subjects (overall classification of the photos, choice by pairs). Discourse was here stimulated by visual images.' The sample was taken among the inhabitants of Toulouse. The method of the stratified sample was selected as a function of the results of the pre-survey and following the divisions which appeared during the pre-survey: place of habitation (center, periphery, intermediate zone; three blocks of houses were selected in each of the national statistical service's standard areas as a function of geographical differences); sex and occupation

226

Raymond Ledrut

(economically active men and women, inactive women); age (under 35, over 35 years); socio-professional category according to INSEE: managerial staff (cadres), other white-collar employees (employés), industrial workers (ouvriers). In this way 54 terminal groups were obtained. Four subjects by groups were interviewed, for a total of 216 persons. The subjects interviewed were determined, as we can see, in a relatively aleatory fashion taking into account the residence on the basis of the data provided by the statistical service in this matter.2 The second series of interviews were effected using a traditional "verbal questionnaire," including a very large number of "open" questions. With the help of verbal stimulants a quite considerable amount of material was collected and submitted like the previous material to semantic and statistical analysis which will be presented in the context of their application. The sample of subjects interviewed this time was taken in two cities, quite different in size and location, which probably belong to two distinct urban types: Toulouse (a city of 380,340 inhabitants in 1968, with a metropolitan area of 475,887 inhabitants) and Pau (76,227 inhabitants in 1968). The sample, stratified and partly aleatory, included 1 1 2 subjects for Toulouse, and 53 for Pau, i.e., a total of 165 subjects. It included socioprofessional categories not represented in the sample which completed the "photo-questionnaire": merchants and craftsmen in particular. A certain number of urban agents (elected members and employees of the municipality, civil servants of the regional bureau of the Ministry for Equipment, developers, architects . . .) were interviewed at Pau. It is obvious that this sample as well as the previous one has, strictly speaking, no representative value, neither in relation to the "populations" of Toulouse and Pau, nor to any other type of parent population. Given our objectives—which are not "conjunctural" nor "statistical"—this is only a minor inconvenience. As Lewin would say, we are trying to be Galilean! The data of this double survey were subjected to two different and complementary elaborations, themselves linked to distinct types of analysis. In fact, a mode of statistical elaboration was used based on a rather severe categorical reduction of the replies. But, on the other hand, a detailed content analysis of the open questions was effected. This analysis enabled us to discover, within the simplified categories of more specific themes, more expressive variations and also more

The Images of the City

227

detailed matrices of interpretation. The results of these types of analysis have mutually illuminated, and especially corrected, one another, thus authorizing interpretations both better established and more complex. The methods and techniques used will appear more clearly during the discussion of the results of the research. This will be the case in particular for the techniques of content analysis—of a more or less semantic character—and for the statistical techniques, which cannot be presented "in abstract" in a truly interesting manner. . . . [pp. 203-9]: Values and Characteristics. The various orders of "significations" could be presented without order, in so far as nothing in the analysis has introduced any reason for a classification of the secondary semes. However, it may be useful for us to introduce among the orders of significations certain distinctions with no direct bearing on the analysis—if only for the purpose of presenting these various orders or dimensions. The first of these distinctions is the one which differentiates between " v a l u e s " and "characteristics" among the "qualifications." Certainly w e find " v a l u e s " everywhere, because any qualification is the lexematic expression, either positive or negative, of a seme, sometimes with intermediate degrees but always on a scale. Furthermore all apparently purely descriptive features often include a judgment and an element of " v a l u e . " But one could nevertheless say that " v a l u e s " and "characteristics" are distinguished by the mode of their relation to the object and to the subject. The "values" in fact are qualifications attributed to the object represented owing to its conformity (or non-conformity) with the subject's criteria of satisfaction—whatever the particular nature of this satisfaction, and therefore of the order of values. The "characteristics" are on the contrary qualities acknowledged in the object independently of its relation with the subject. . . . The Four Orders of Values: Ethical, Aesthetic, Vital, and Functional. Among the significations of value, w e were able to discover in the corpus the presence of four orders of value. These are the "ethical" values, the "aesthetic" values, the "vital" values, and the "func-

228

Raymond Ledrut

tional" values, with the semes which correspond to them and which articulate the positive and negative values ("functionality" for example). The ethical ( we prefer to avoid the neologism of "ethicity") appears in expressions which refer to the opposition liberty-constraint, either in these lexemes or in other formulations reducible to them. The central idea is that of "autonomy": anything that evokes manipulation of the consumer, "proxemic" pressure (Hall 1969), or any other form of servitude directly or indirectly imposed can be considered as the negative aspect of the ethical. In a certain sense it is the presence of "evil" in the urban universe. On the contrary, the subject who in speaking of a street speaks of the "freedom to live" is referring to the positive aspects of the "ethical," to the autonomy of the agent. In psychological terms, we would speak of the "ethical satisfaction" given by the photo-image in question. This order of value is widely represented in the "corpus": 303 citations. The average number of citations (proportion of the total number of citations to the number of the dimensions selected) being 376, the place of ethical significations and values can be considered as important. 3 In the field of "values," ethical values rank after the vital and aesthetic values, but before the functional values. There is hardly any difference in the role attributed to the ethical values by the various social groups studied: one barely notices that there is a reduced number of citations by the managerial staff (95) as compared to the employees (104) and the workers (104). Yet this difference is not without interest, for it shows, despite the small deviation, a less strong tendency of the managerial staff to read photos in terms of "ethics" and—on the level of "mentalities" and "culture"—a less strong tendency to "feel" social pressure, to "select" and to acknowledge this signification. The differences are still more marked between men and women: the number of citations is 1 1 7 for men, 100 for non-working women and only 86 for working women! Men seem more sensitive to ethical values. On the other hand, there is scarcely any difference according to age. This is not the case for residence, the inhabitants of the intermediate area showing a relatively low number of citations: 86 as against 109 for the inhabitants of the city center and 108 for the peripheral area.

The Images of the City

229

Preponderance of Vital Values. The "vital" values hold first place among the values: 721 citations. They have two opposed and complementary poles: "well-being" and "uneasiness." We find again a semantic articulation already identified in the "comments" on the city previously examined. A photo-image can evoke "happiness," "pleasure of living," "tranquility," or else "sadness," "boredom"; it might seem "dark," " g r a y , " or on the contrary, "light". . . . This order of signification and value has an extremely important place in the "corpus," something which no longer surprises us. All the groups are extremely sensitive to it. Yet there are considerable variations and the variables taken into consideration seem to play a part which is not negligible. The language of the vital values is first and foremost that of the managerial staff; the number of citations is 278 against 222 for the employees and 221 for the workers. They care very much for the "quality" of their living environment. . . . The "aesthetic" values also occupy a relatively important place, though closer to the ethical values than the vital values: 375 citations. They are classified by reduction to the primary axis of "beautiful" and "ugly." Again, this is not a discovery since the aesthetic values have often been discussed in the other sections of this study. The expressions "pretty," "artistic," "elegant," "splendid," "has personality," "marvellous," "well-arranged," "magnificent," "unbeatable," "harmonious," "grandiose" have been reduced to positive values. The wealth of expressions should be noted. The negative aspect of "aestheticity" can be found in the expressions "awful," "no charm," "bastard," "no style," "conventional," "design a bit out of date." Here again the "aesthetic" language is much diversified. Concerning the part played by the socio-professional variables and differences between groups, we might be rather astonished to note that this aesthetic language is especially characteristic of workers (153 citations against 124 for the managerial staff, and 1 1 6 for the employees). The deviation is, however, rather slight. We should also mention here the managerial staff's greater severity regarding the aesthetic value of the urban elements. . . . As far as the "functional" language is concerned, it has decidedly less importance than others: 135 citations only, which no longer surprises us. The positive aspect of the functional is "functionality"

230

Raymond Ledrut

under which we have ranged "the convenient," "the practical," "the useful," "the comfortable," "the functional," "the well conceived." The negative expressions are those of non-functionality or dysfunctionality, a little less diversified in their formulation: "inconvenient," "not well adapted". The workers have a language distinctly more functionalist (54 citations) than the managerial staff (41) and the employees (40). Despite the moderate deviation, we have here a feature which cannot be neglected, for it is not without relation with what could be called the realism of the workers regarding the urban sphere. . . . [pp. 231-32]: Unity and Diversity of the Language on the City. Unity of the Matrix of Interpretation. In the preceding analysis we have been able to observe variations, sometimes quite significant, in the language used about the city. These variations have an indisputable interest on a sociological level. An accentuation of this or that order of signification depending on social class, residence, or sex is clearly evident, and this informs us about the changes that may appear in the modes of apprehension of reality. Certain significations and certain values "affect" one group more strongly than another; certain conditions of life accentuate the importance of a certain "characteristic" of the urban elements. However, on the level of the orders of significations (dimensions) and of the set of the semes involved in the reading of the urban elements, we are struck by the unity of the matrix used. In spite of the use of different filters and lenses which permit a valid differentiation of the sample—by class, sex, etc.—we have seen that the same orders of significations (or classemes) were always used and that there were only variations in intensity, sometimes negligible, in the use of these various dimensions. We can thus say that there is indeed a single language of the city for the subjects of the sample and the images which are presented to them. . . . We can also note that the verbal questionnaire—which is independent of the restrictions imposed by the photoimages—reveal to analysis a language which is not fundamentally different. The "vital" values, the "aesthetic" values, the "ethical" values, and the "functional" values are definitely stable semantic axes of the current discourse on our cities.

The Images of the City

231

Unity in the Scale of Significations. This unity of language, however, might not exclude, on another level, large variations on the scale of importance of the significations—a hierarchy which starts from the most widespread (general) to the least cited (particular). Yet this is not the case at the level of groups, the social groups for instance. . . . [pp. 242-46]: If we take into consideration the variable "social groups," we see that the managerial staff accentuate significations related to vital values more strongly than the workers and the employees, while the workers are more sensitive than the other groups to functional values and aesthetic values, and as sensitive as the employees to ethical values; the position of the employee being rather poorly differentiated. Therefore, if we examine only the values, we here have a characteristic difference in the degree of sensitization in relation to values. The workers are a little more "functionalist," the managerial staff "vitalist"; the former are more realistic, as already mentioned, more technical, and more sensitive to inconveniences; the others are more concerned with pleasure and less with "technical" problems. One must add that the workers are also more aware than the managerial staff of aspects of constraint (or of liberty), and speak more frequently in aesthetic terms; they more easily than the managerial staff say: "it's beautiful" or "it's ugly." As far as the other significations are concerned, they also reveal a strong opposition between the workers and the managerial staff. The latter accentuate significations related to "ludicity" (which probably links happiness to the beautiful), the "order" of the things of the city (which are also related to aesthetic significations and "sociability"). The workers are more sensitive than the managerial staff of the "utilitarian," to "urban planning practice," and to the "spatiality" of urban elements. Workers and managerial staff are both more concerned than the employees about décor: it is their only common point of emphasis. The opposition is thus very strong. We shall see its meaning a little better further on, after the study of the relations between orders of significations. One can already note that the "realism" of the workers is very distinctly developed: the utilitarianism and the interest in a more evident social practice are opposed to the greater

232

Raymond Ledrut

importance that the managerial staff gives to the "ludic." The order of things is a more global and more abstract signification than their extension. Another equally important fact is the attention given by the managerial staff to the "forms of sociability" in an urban environment. Urban social life seems somewhat "underestimated" by the workers. As for the employees, they are particularly more sensitive than the others to the "social appearance" of places (the "rich" and the "poor"), to the mark of the social groups, to significations related to "economicity," to hygiene, to historicity, and naturality. This is what distinguishes them from both the managerial staff and the workers. However, they resemble the workers in their interest in the utilitarian, and the managerial staff in their attention to the utilization of time (distinction between the regular and the irregular). Consequently they seem more "socialized," very sensitive to urban "sociality" (appearance and marking) and, on the other hand, interested in external characteristics of urban forms (clean-dirty, presence-absence of nature). If we want to establish a synoptic table of the significations which are ranked highest in relation to the others by the three main social groups we have distinguished, we can note first of all that the managerial staff is characterized by the accentuation of the vital values, ludicity and the forms of sociability and, to a lesser extent, by order and utilization of time. There seems to emerge here an apprehension of the city of an existential type, where sociability as the spectacle of things plays a leading role. It is the city of a slightly passive, we might say "hedonistic" enjoyment, the city of pleasure received. Let us repeat that it is only an emphasis. The industrial workers attach relatively more importance to the functional values and the aesthetic values, to planning aspects and spatial dimensions. Their view is more "realistic" than that of other groups, as already stated. The workers' city is the one which is convenient and inconvenient; it is also that of social practice: the city on which one must act by destroying, renovating, or preserving. The workers, like the employees, and unlike the managerial staff, are sensitive to the constraining or liberating characteristics of the city. They are not in a hedonistic world, but in a world of alienation and practice. As for the employees, they distinguish themselves by their sensi-

The Images of the City

233

tivity to social appearance, to the marking of social groups, to history, to nature, as well as to the economic and hygienic aspects of the things of the city. The employees' city is close to a concrete and complex model, in which the social, the historic, the natural, the hygienic, the economic are mixed. It is neither the city of pleasure nor that of practice.4 What is dominant in it is appearance and, especially, social appearance and marking, the signs of prestige given to the urban elements by the dominant class ("the middle class," the "rich"). The dependence of the employees in regard to these values and signs is apparent here. The beautiful may well be confused with the old, and this with the rich and the middle class. [PP. 338-40]: The study of factors and groupings had already allowed us to apprehend some of these relations. They are of very great importance, in so far as they reveal that there is not only distinction, but opposition, between the eudaemonic register and the functional register. We are dealing not simply with two levels of discourse but also with two systems in partial opposition. The opposition of structures on the same level does not exclude a more limited opposition between levels. We are gradually moving toward the acknowledgment of implicit urban models more or less irreducible at present. In our civilization and our urban experience, the modern cannot be invested by the values which at present take refuge in the old, particularly "the positive vital values" (the nice, the pleasant, etc.) and to a lessser degree "the positive aesthetic values" (the beautiful, the pretty, etc.). On the contrary, the practical (or convenient) is linked quite naturally to the modern and not to the old. And this modern is associated more often with the negative values of the vital than with the positive values. We meet here again the unsolved problem of our urban planning—reality or theory: the reconciliation of the modern and the vital. The modern and modernism have not yet succeeded in presenting themselves and in establishing themselves as beautiful and especially as euphoric or eudaemonic. The pleasure of the modern is more intellectual than emotional. The beautiful and the practical are not associated in our experience. Thus, we can distinguish and oppose two positive structures (which do not proceed from the negative values of the aesthetic, vital, ethical, or functional order).

234

Raymond Ledrut

The first may be called the structure of the functional, where the practical, modern and vast are associated; the second, which links the pleasant, the free, and the beautiful with the old and the "natural" can be called, as we have already called it, hedonistic, eudaemonic, or simply vital. The urban planning ideology that corresponds to the first is both "functionalist" and "modernist" ( or "progressivist"); the second is at the same time "naturalist" and "culturalist." The essential cleavage depends on the role of modernity, but also on the fact that the practical is more often disagreeable than the non-practical (87 percent of the citations against 73 percent). . . . Structures of the Discourse on Preferences. The analysis of the criteria of preference in regard to cities, which we carried out on the basis of the replies to the "verbal questionnaire," has enabled us to reveal a certain number of significations and, in particular, significations of values implied in the preferences (these significations are almost the same as those of the discourse on urban elements) 5 . . . . [PP. 356-59]: The "Concrete" Model. It seems that there exists a system of preferences, of a vital and personal character, that we could call "concrete" insofar as it is based less on concepts and "objective" judgments than on the affective, emotional relations of the subjects with the city they like best. The city is then not an object preferred because of its conformity to general principles and qualities objectively appreciable and determinable. It is a being with whom we have "intimate" relations, interior and not exterior. This system of preferences is thus not a set of abstract determinations which would enable us to pronounce judgment on any city that we might examine. It is a "practico-emotional" structure, in which the concrete relations of the subject to the thing are much more important than the objective characteristics of this thing. "The model" or "the image" of the city that satisfies us will thus be defined, not in terms of urban planning in the strict sense of the word—in terms of spatial configuration—but rather in terms of relations, where the emotional and effective content counts more than the developed form. . . .

The Images of the City

235

The "Abstract" Model. On the other hand we can conceive of an "abstract" model which leaves the subject out of the city, outside of it. The subject is in a sense "free" or "uprooted." Although attached to other things, things more general and abstract that can be found in various places, he remains unattached in regard to concrete places, to the places of his life insofar as they have particular "faces." It is not that he does not enjoy himself or find some advantage in certain places and in their fréquentation, but he is a bit like the "tourists," the lovers of "beautiful scenery," the "utilitarians" who calculate their choice, or the "wanderers" who find amusement in the successive places of their existence. The abstract model would therefore be an objective system in which spaces, forms, and spatial relations dominate, where exteriority prevails over the inferiority of objects as well as that of the object and the subjects. This model would thus be more a "planning" model in the strict sense than the other. It refers to an urban planning conceived as a simple arrangement of spatial relations between volumes and outlines, "social" relations being subordinated to them. It is the most understandable model for the "planner"—the "ideal type" of the "urban development specialist"—it speaks his language. It certainly is not the most widespread among the "inhabitants" and is seldom found in a pure form. The love for a place of residence is still very strong and is opposed to this model. Thus the "concrete" city or reality is opposed to the "abstract" city represented as a pleasant décor, as a set of commodities, or even sometimes as both of them. Here everything is based on the opposition of feeling and calculation; we could even risk saying of "passion" and "consumption." [PP. 367-73]: Urban Models and Structures of the Urban Experience. We can now relate what has been extracted from the discourse on the city (replies to the "verbal questionnaire") with the structures we believe we have recognized in the discourse on the urban elements (comments on the "photo-images"). In comparing the urban models to the structures of the discourse on the urban elements, we notice very quickly a close link, on the one hand, between the concrete model

236

Raymond Ledrut

and the hedonistic or eudaemonic structure (language of the vital values and of the ethical values with the significations attached to them) and, on the other hand, between the abstract model and the "functionalist" structure. In the "concrete" model, as in the hedonistic structure, the principle of unity is an "ethico-vital" order; it is the emotional link of the person to the city which is in question; it is his comfort (pleasure and absence of constraints) which is involved. . . . The two models are not in themselves exclusive. It is nonetheless true that in the present conditions of urban life there is a separation, a gap—as we have already remarked—between the vital and the functional, the pleasure principle and the reality principle. We do not have a synthetic model, harmonious and holistic—the functional has dried up, shrunk and isolated itself, while the vital has been rejected, shut u p in imaginary paradises. This is a fact, not a necessity. We find here again a question which inevitable appears when we examine the multiplicity of urban planning theories since the nineteenth century: why are most of these theories partial? Why do they emphasize one point, one value, one order of signification in relation to which their view of the city, their planning theory, is constructed? Françoise Choay, with some exaggeration sometimes, has brought to light some characteristic accentuations. Her classification of the "utopias" is perhaps a little too a priori and it would be preferable to rely on the categories of the urban experience to classify planning theories. The designations do not seem to fit well into a set of oppositions which would have a solid structural value: progressivism is opposed to conservatism and not to naturalism (which is, however, opposed to culturalism). Nevertheless, the acknowledgement of the three directions (progressivism, culturalism, naturalism), which are not opposed, correspond rather well to these accentuations of values, which are not in contradiction, as we have just seen. Progressivism—which it would perhaps be preferable to call "functionalism" in the wide sense of the word—gives first rank to the functional values and results more or less in what we have called the "abstract" model. We find here again, without difficulty, the practical, the clean, the orderly. The natural is subordinate to it. This is indeed a coherent theory—latent in our urban experience—but partial! . . . Opposing functionalism, we shall find not a precise theory, linked

The Images of the City

237

to action and attempting to be effective, that is, "technical," but on the contrary "a model" which, while it is very concrete by its relation to life and to the personal relationships, remains very undetermined. For, in fact, it is linked only to the individual use of the city and not to an urban planning practice. We can call this latent model "hedonistic," for it places the ludic in the forefront. It is the model of free pleasure where the requirements of urban practice are conjured away. How could it be otherwise? In the same way that through a multiple and profound dissociation the urban agents cannot but be pushed towards functionalism and technicism, the "ordinary inhabitant," entirely powerless in regard to urban development and totally marginalized, cannot but withdraw into purely affective relations with the city which tend to become more and more infantile and phantasmal. The two models and their specific accentuation are the result of a single procedure of increasing separation of two dimensions of the urban universe, and of a reciprocity of action which involves an increasing specification of each model. When something which could be unified and form a complex and complete whole is divided and dissociated, each separate part not only accentuates its characteristics but also loses its virtues and presents itself much more in its negativity than its positivity; that is, the loss of its complementary part, which has become its contrary, liberates it and allows it to run riot. The isolated imagination becomes delirium: reason by itself, calculation and rationalization. It is a law of dialectics. Thus, when the pleasure principle and the reality principle—enjoyment and action—are deeply dissociated, pleasure becomes passivity, and in a certain way unreal, and reality for its part becomes object and instrument. The attraction to lived experience is changed into hedonism, and realism into functionalism. Who would deny that thus pleasure as well as practice, the felt as well as the useful, destroy themselves—contradict themselves? The two models described thus seem to be in fact the fundamental models, distinct and linked together, which in our civilization really and necessarily divide the field of urban planning projects. They result from a historical development and appear among the theoreticians of planning as well as the inhabitants and the agents of urban development. The other theories, Utopias, or systems of attitudes do

238 Raymond Ledrut not possess the historical "truth" of these models. Aestheticism or formalism, culturalism, naturalism are ambiguous wholes. Attempts at aestheticizing, or culturalist and naturalist rationalizing belong to ideologists who are attached either to formal values—often the case with architects, or with "artists"—or to social values, usually of the passéist kind (nostalgia for traditional local communities or for archaic societies closely linked to nature)—this is the case of "idealistic" intellectuals. As for the city inhabitants, even if they belong to one of the two models, they reveal a complementary synchretism, explicable by the insufficiency of the principal models. If we examine their desires we very soon discover their attachment to an urban décor which is inscribed in history by some feature, but also to "nature," to the natural elements (sun, water, trees, etc.) and to their presence in urban monuments (gardens, quaysides, etc.). This double demand ("culturalist" and "naturalist") is very widespread. These requirements, which are not contradictory in themselves, are very difficult to conciliate nowadays. "The city in the country" constitutes neither a model nor a project. It is as yet only a dream. [pp. 379-8i]: Models and Social Classes. The social class whose position is most clearly characterized and which seems to be decidedly most attached to the concrete model is that of the employees.6 The employees are—together with the middle managerial staff—the group with the highest proportion of references to the "personal links" with the city of their preference. On the other hand, they show little tendency to be innovators or progressivists: they constitute the group which shows the smallest proportion of subjects stating that "there are many things that should be changed" in the city where they live; they show little interest in the creation of entirely new things; they are—with the workers—the group least disposed to desire the disappearance of certain things in their city. The employees also in their comments show that they are not much for state intervention and planning: they are one of the groups where declarations in favor of free urban development are proportionally the most numerous— after the class of craftsmen and merchants and well ahead of the managerial staff. It is also the group that proportionally gives the

The Images of the City

239

highest number of citations where it is considered that the principal function of the city is to assure "tranquility," "peace." The employees are a class whose position is well-marked in relation to the two proposed models. This is not the case for certain classes w h o s e situation is much more ambiguous, sometimes contradictory, and w h o seem to live this contradiction. This is the case in particular of the managerial staff (superior and middle) w h o manifest a rather strong desire for change but w h o are among the least numerous in desiring certain destructions. On the other hand, while they have a greater inclination for intervention and planning, they are often linked personally to their city of preference, and seem rather strongly attached to tranquility (second position after the employees). The workers, on the contrary, despite some internal oppositions, seem relatively closer to the "abstract" or functional model. They make much fewer references to "personal links" (like the craftsmen and merchants); they are among those that speak the least about "tranquility." This is the class that most often mentions hygiene—a functional value even if it is "symbolic" in its w a y . They constitute— with the craftsmen and merchants—the group where the declarations that "there are many things that should be changed" in their city are proportionally most frequent. Progressive and very strongly disposed to see the disappearance of whatever is inconvenient, though they have not made the "abstract" model entirely their o w n , they have a more functionalist and technical—a more "practical" view than most other groups. The employees are for their part more dreamy and sentimental. This is confirmed by the analysis of the replies to the "photo-questionnaire."

NOTES 1. The idea is Christian R o y ' s . 2. Once again I wish to thank m y friend Louis Amiel, currently Regional Director of the national statistical service, for his cooperation in this study, as in m a n y others. 3. We can use the absolute number of citations, since the number of photos is a l w a y s the s a m e and the g r o u p s compared h a v e the s a m e number of m e m b e r s ,

240

Raymond Ledrut

give or take one unit. Also, the total number of citations varies very little from one group to another, whether relative to each variable or absolutely. No group is more talkative than another. It was thus pointless to relate the citations for each order of signification to the total number of citations made by the group in question. 4. Here we find again the dream and the city of dreams which we have discussed above. The "dream" is distinguished equally from "pleasure" and from "practical realism." 5. Question: Which is the city that you like best? Why do you prefer this city? 6. All comments in this paragraph are based on the responses to the verbal questionnaire.

REFERENCES Barthes, Roland. 1957. Mythologies. Paris: Seuil. Hall, Edward T. 1969. La dimension cachée (French translation of The Hidden Dimension). Paris: Seuil.

11 U R B A N I S M IN QUESTION Françoise Choay

As we have already seen, the book from which these selections are taken, L'urbanisme: Utopies et réalités, can be considered as a landmark in urban semiotics. The typology of urban discourses presented in it was later enlarged by the author and we shall begin with this enlarged typology. The author differentiates between the discourses linked to the production of urban space and the "commentary" discourses referring to existing space (conception, advertising, historico-aesthetic approaches). The first are divided, as we have seen (our comments on Choay, paper 7), into inaugurative and connotative discourses. The inaugurative discourses finally, are divided into static "utopian" and dynamic "semiogenetic" discourses. The subject of Choay's book is the Utopian discourses on urban space, in Europe, and the United States, during the nineteenth century—the stage of "pre-urbanism" (Ruskin, Owen)—and the twentieth century—the stage of "urbanism" (Gamier, Le Corbusier, Howard). These Utopian discourses are understood best as critiques of the industrial city. The book includes two parts. In the first part (pp. 7-83), the author analyzes the models underlying urban discourses, and in the second part (pp. 84-435) she presents excerpts from the body of work of the authors she has used. This paper consists of translated excerpts from L'urbanisme: Utopies et réalités. Une anthologie (Paris: Seuil, 1965).

242

Françoise Choay

It is only the first part of the book that is represented in the following excerpt. Choay discusses initially the models corresponding to pre-urbanism (pp. 10-29), and then passes to their twentieth-century descendants (pp. 3°~35)• Continuing, she discusses (pp. 53-73) what she considers as a second degree urban critique, which, according to her, is effected by two different approaches, "technotopia" and "anthropopolis," corresponding to the Utopian urban models of "progressivism" and "culturalism." She observes in her conclusion (pp. 74-83) that these trends reject the previous models which are dominated by the imaginary, but that, even if undoubtedly metalinguistic, they cannot escape entirely from fantasy.

[p. 10]: PRE-URBANISM Genesis: The Critique of the Industrial City. In order to outline the conditions under which, in the nineteenth century, the problems of urban development appear, we need to briefly recall certain facts. . . . [pp. 1 5 - 1 6 ] : The Two Models. That which is felt as disorder calls for its antithesis, order. Thus we find opposed to this pseudo-disorder of the industrial city, proposals for urban arrangements freely constructed by a process of reflection developed in the imagination. Since it cannot give practical shape to its questioning of society, this reflection situates itself in the dimension of Utopia; there it orients itself along the two basic directions of time, past and future, and thus assumes the features of nostalgia or progressivism. From a set of political and social philosophies (Owen, Fourier, Considérant, Proudhon, Ruskin, Morris) or of genuine Utopias1 (Cabet, Richardson, Morris), we see emerging, with a greater or lesser luxury of detail, two types of spatial projections, of images of the future city, which we will hereafter refer to as "models." By this term we intend to emphasize both the exemplary value of the constructs proposed and their quality of reproducibility. Any structuralist resonance should be elimi-

Urbanism in Question

243

nated from the use of this word: these models of "pre-urbanism" are not abstract structures, but on the contrary they are monolithic images, indissociable from the sum of their details. . . . [p. 26]: Critique Without a Model: Marx and Engels. Contrary to other political thinkers of the nineteenth century, and in spite of their borrowings from the Utopian socialists, Marx, and more explicitly Engels, criticized the large contemporary industrial cities without having recourse to the myth of disorder or proposing its counterpart, the model of the future city. . . . [PP. 28-54]: The Anti-Urbanism of America. The majority of writers who criticized the large industrial city in the Europe of the nineteenth century were nonetheless marked by a long urban tradition; throughout history, the European cities appeared to them as the cradle of the forces which transform society. The opposite takes place in the United States, where the heroic age of the pioneers is connected with the image of a virgin nature. Thus, even before the first effects of the industrial revolution are felt, a nostalgia for nature inspires in this country a violent anti-urban trend. The attack is merciless but does not rise to any alternative model. An anti-urban tradition thus begins with Thomas Jefferson, continues with Ralph Waldo Emerson, Thoreau, Henry Adams, Henry James, and closes paradoxically with the greatest architect of the Chicago School, Louis Sullivan. The works of M. and L. White notably have analyzed the stages of this trend, in relation to which the bards of the American city, from Walt Whitman to William James, represent merely a few voices lost in "the city wilderness," completely submerged by the "anti-urban roar in the national literary pantheon" (White and White i960: 215). The large city is thus criticized successively from a series of different angles; in the name of democracy and political empiricism by Jefferson; in the name of a metaphysics of nature by Emerson and mainly by Thoreau; 2 finally, as a function of a simple analysis of human relations, by the great novelists. All these writers, in unison,

244

Frattfoise Choay

naively place their hopes in the restoration of a kind of rural state which they think is compatible, with a few reservations, with the economic development of industrial society and which alone will ensure the safeguarding of liberty, the blossoming of personality and true community. American anti-urbanism does not share the tendency of the trends of thought examined above; it was never established as a method. It should, however, be mentioned here because of its influence on American planning in the twentieth century. URBANISM Urbanism differs from pre-urbanism on two important points. Instead of being the work of generalists (historians, economists, or politicians) it is in its two forms, theoretical and applied, a domain of specialists, most generally of architects. "The urban planner is nothing but an architect" affirms Le Corbusier. Thus urbanism ceases to be part of a global vision of society. While pre-urbanism was linked to political options throughout its history,3 urbanism is removed from the sphere of politics. This transformation of urbanism can be explained by the evolution of industrial society in capitalist countries. After the militant, heroic phase of the nineteenth century, capitalist societies became liberalized and their ruling classes borrowed certain ideas and proposals of the socialist thinkers of the nineteenth century, cutting them off from their origin. Moreover, these ideas were to be put into practice. Instead of being confined to Utopia, urbanism was going to give its technicians a practical task. However, urbanism does not completely escape the imaginary dimension. The first urban planners had a limited hold on reality: at times they had to face unfavorable financial conditions, at times they confronted the omnipotence of economic and administrative structures inherited from the nineteenth century. When this happened, their polemical and educational purpose in turn manifested itself in a movement towards Utopia. This is why, in spite of the differences pointed out above, and even though we cannot talk of a consciously assumed ideological

Urbatiism in Question

245

continuity between pre-urbanism and urbanism, the latter also makes the imaginary play a role in its method. We will rediscover there, in modernized form, the two models of pre-urbanism. A New Version of the Progressivist Model. The new version of the progressivist model finds a first expression in La cité industrielle by the architect Tony Gamier. . . . From 1928 on, the organ of diffusion of the progressivist model is an international movement, the group of C.I.A.M.; 4 in 1933, this group formulates its proposals as doctrine under the title of The Athens Charter. This charter is, thus, the common property of the progressivist urban planners; its contents are repeated in their numerous individual works. We have, however, borrowed most of the following citations from Le Corbusier: an exceptional journalistic talent (maintained by the necessity of unceasing polemics against the conservatism of the French public) inspired in Le Corbusier the most striking images and formulas for forty-five years. 5 The key concept underlying progressivist urbanism is the concept of modernity. "A great era is beginning, a new spirit exists," declares Le Corbusier in the review L'esprit nouveau, which he had just founded together with Ozenfant in 1919. He sees this modernity at work essentially in two areas: industry and avant-garde art (in this case cubism and the movements deriving from it). Just as in progressivist pre-urbanism, we find at the basis of progressivist urbanism a conception of the industrial era as a radical historical rupture. But the interest of urban planners had been displaced from economic and social structures toward aesthetic and technical structures. The large city of the twentieth century is an anachronism because it is not truly the contemporary of either the automobile or Mondrian's paintings: this is the historical scandal which they will denounce and try to suppress. . . . Thus, industry and art have in common the universality of their aim, and their twin expansion on a worldwide scale confirms for the progressivist planners the pre-urbanist conception of the universal man: identical in all latitudes and every culture, man for le Corbusier is defined "by the sum of the psycho-physiological constants identified and inventoried by the appropriate experts (biologists, doctors, physicists and chemists, sociologists and poets)." 6

246

Françoise Choay

This image of the universal man inspires the Athens Charter, which analyzes universal human needs classified under the four great functions: living, working, transportation, recreation of body and spirit. This is the base which should allow one to determine a priori, with complete certitude, what Gropius calls "the ideal type of the human establishment". . . . The plan of the progressivist city is linked no more to the constraints of cultural tradition than it is to the site; the only thing it wants to be is the expression of a creative liberty of reason, placed in the service of efficiency and of aesthetics. These are the two imperatives which confer on the space of the progressivist model its particular characteristics. The desire for efficiency is first of all manifested in the importance given to the question of health and hygiene. The obsession with hygiene focuses around the notions of sun and greenery. It is linked to the contemporary advances of medicine and physiology, to the practical applications deriving from them, 7 as well as to the new role ascribed after World War I to the care of the body and exposure to sunlight. These objectives will lead the progressivist planners to explode the old closed space in order to de-detisify it, in order to isolate, in sunlight and greenery, buildings which are no longer attached one to the other, but have become autonomous "units." The major consequence is the abolition of the street, stigmatized as a barbarian trace, a revolting anachronism. At the same time, most planners will extol high-rise construction, in order to substitute a limited number of units or vertical pseudo-cities for the continuity of the old low buildings. In terms of Gestalt psychology, we observe an inversion of the terms of figure and ground; instead of stretches of free space playing the role of figures on the constructed background of the city, space becomes the backround, the environment in which the new settlement is deployed. This new background is to a great extent invested with greenery. "The city will gradually be transformed into a park," anticipates Le Corbusier (1946:86), and Gropius adds (1935:100): "The goal of urban planning should be to create closer and closer contact between the city and the country." Thus one is led to the concepts of the vertical "garden city" of Le Corbusier and the urbs in horta of Hilberseimer. This exploded space is nonetheless governed by a rigorous order,

Urbanism in Question

247

corresponding to a new level of efficiency, that of productive activity. In effect, the industrial city is also industrious, meaning for progressivist urbanism "a working tool." The city must be "classified," analyzed, every function in it must occupy a specialized area, so that the city can fulfil this function of being a tool. Following Tony Gamier, progressivist planners carefully separate working zones from living zones, and living zones from civic centers or areas of recreation. Each of these categories is in turn divided into subcategories equally classified and ordered. Each type of work, administrative, industrial, or commercial, is assigned its label. Down to "coffee-houses, restaurants, shops . . . remaining vestiges of the present street," everything must "be given form or set in order, in a condition of full efficiency. Concentrated places for strolling and socializing" (Le Corbusier 1946:74). Transportation, in turn, is conceived as a separate function which, paradoxically, is treated independently from the constructed whole to which it belongs; there is a "mutual independence of constructed volumes and traffic routes," says Le Corbusier, and he adds "highways will pass through in transit and following the most direct, most simplified route entirely linked to the terrain . . . but perfectly independent of the edifices or buildings which happen to be in greater or lesser proximity." 8 The street is thus not only abolished in the name of hygiene, but also because it is "the symbol in our era of the chaos of transportation." Orderly transportation, actually, often runs the risk of becoming an unconditional surrender to the power of the automobile, of which it has been said, not without justification, that in the end it was the only factor determining the main lines of a great number of plans. Tool-city, the progressivist model, is also spectacle-city. Aesthetics is an imperative as important as efficiency for these planners-architects who in the European tradition have received primarily an artist's education. But, faithful to their modernism, they reject every sentimentality concerning the aesthetic contribution of the past. Of the old cities to be reconstructed, they retain only the outlines, practicing a surgical planning which is equally satisfactory for the exigencies of profit. "The more Haussmann cut, the more money he made," remarks Le Corbusier (19238:255). Le Corbusier also, in his plan for Paris, would raze without hesitation all the "picturesque" old neighborhoods (a passeistic attribute, proscribed in the progres-

248

Frangoise Choay

sivist settlement), retaining only certain major edifices (Notre-Dame, Sainte Chapelle, the Invalides) that he promoted to the dignity of symbols and to a museological function. The planner "composes" his future city on the drawing board, as he would compose a painting. Following the principles of cubism, and still more those of purism and of Stijl, he eliminates every incidental detail in favor of simple forms, reduced to essentials, where the eye cannot stumble against any particularity; it is in a sense a question of constructing the a priori framework for any possible social behavior.9 The composition returns to the theme of explosion: it is organized around multiple centers of vision, in a process which evokes that of synthetic cubism. Each one of these dissociated focal points is ordered according to the principles of a simple geometry, which also characterizes the composition of the schools related to cubism. "Geometry," said Apollinaire, "is for the plastic arts what grammar is for the art of writing." However—as Kahnweiler and Raynal have emphasized—the geometry of the cubists was a matter of instinct, bearing little relation to mathematics. On the contrary, for most of the progressivist planners, like Le Corbusier and his disciples, geometry becomes the point of intersection of beauty and truth: art is governed by a mathematical logic. "Geometry is the basis. . . . The whole contemporary era is thus eminently one of geometry; it orients its dream toward the joys of geometry. Modern art and thought, after a century of analysis, search beyond the accidental fact, and geometry leads them to a mathematical order" (Le Corbusier 19238:35). And yet we must not be trapped in the mirage of words. The geometry ordering the progressivist model is quite elementary. It consists essentially of arranging cubic or rectangular elements according to straight lines intersecting at right angles: the right angle is the golden rule which determines relations between buildings and between buildings and transportation routes. Le Corbusier affirms: "Culture is an orthogonal state of mind" (1923a: 35). Finally, there is a rigorous correspondence between the exploded but orderly space of the object-city and the dissociated but geometrically composed space of the spectacle-city. . . . To the extent that the progressivist model, in opposition to the culturalist model, privileges the typical individual rather than the

Urbanism in Question

249

typical community, it is not surprising that their most advanced research concentrated on housing. The first works of the C.I.A.M. were centered around housing. The Athens Charter bears witness to this. Sert, in his work summarizing the Athens Charter, entitles one chapter: "Dwelling, the First Urban 'Function.' " In a general way, two types of housing are envisioned in parallel, exactly as in the age of Fourier and Proudhon. On the one hand, w e find the law house, individual or reserved for a small number of families: this solution is mostly studied by the Anglo-Saxons, the Dutch, and certain members of the Bauhaus. On the other hand, we find proposed the Gigantic collective building, which corresponds better to the ideal of a modernist society. Remarkable prototypes were perfected at the Bauhaus and by certain Soviet avant-garde architects, such as Ol and Ginsburg, during the 1920s. Le Corbusier would later conceive the most elaborate model: the unit of habitation or the radiant city, realized for the first time in Marseilles, 10 and then repeated in Nantes, Briey, Berlin. The radiant city explicitly resumes the Fourierist conception of the phalanstère. Constructed for sheltering the same number of families (1,500 to 2,000 persons), offering the same collective services and the same "organs," in particular "the street gallery," the "unit" is a version of a modernized phalanstère and marked by the progress of technology: the invention of reinforced concrete and of elevators makes it possible to replace horizontality by the verticality of a building seventeen storeys high. But the cell or family lodging, which Fourier's system deliberately left undetermined ("one finds a suitable dwelling according to one's fortune and preference"), becomes on the contrary, for Le Corbusier, a standardized apartment with classified functions in a minimal space, unalterable. The occupant has no choice but to submit to the circulation pattern and the way of living implied by these lodgings, which the architect has deduced to be the best ones possible. The material order which we have just defined by its projection in space equally contributes to the creation of a particular mental climate. In so far as it was conceived as an artistic expression of modernity, it immediately evokes a manifesto atmosphere. Rupture with the past is undertaken in an aggressive, provocative manner, new values (mechanization, standardization, rigor, geometricity) af-

250

Frangoise Choay

firmed in an avant-garde style, in a certain way exposed to the public, whose adherence one tries to acquire by giving an impression of futurism. The ambition of the project, its historic quality, create a feeling of exaltation. But the non-conformism of the progressivist urbanists is threatened by a new conformism. Their intransigence, their polemical refusal to allow for the negative aspects of the human experience by eliminating all the elements capable of interfering with the theoretical arrangement of a plan, runs the risk of stiffening into academicism. In other respects, the ruling climate of the progressivist settlement is not truly urban. Such a statement may seem paradoxical if we recall the cities of many millions of inhabitants proposed by Hilberseimer or Le Corbusier. But it is significant that one of the words most frequently used by the latter is "unit." He even specifies that the "tools of urban planning take the form of units" ( of housing, of transportation, etc.). This terminology neatly betrays the atomization, the fragmentation of this settlement which groups, within the greenery, rows of skyscrapers or little vertical cities. Finally, the settlements of progressivist urbanism are places of constraint." Here there is still another key word: efficiency. This value justifies the rigid determination of the way of life. The inscription, irremediably fixed, of every single human activity in spatial terms, symbolizes the reifying role of this planning for which one could not find a better image than the one given by Le Corbusier himself: "Nothing is contradictory any more . . . everything is in its place, properly arranged in order and hierarchy" (1946: 11). And, as a matter of fact, once the human individual has been defined in terms of physical development, functioning, productivity, universal standard needs, what room is left for the infinite and undetermined field of the creation of values and of potential desires? . . . A New Version of the Culturalist Model. The culturalist model takes a truly planning form very early, before the progressivist model, before even the creation of the term "urban planning". . . . The ideological principles of this model are comparable to those of its precursor. The whole (urban settlement) takes precedence over the parts (individuals), and the cultural concept of town over the material notion of the city. But while the socialist Ebenezer Howard

Urbanism in Question

251

was, like all pre-urbanists, prompted primarily by political and social considerations, the visions of Unwin (1909) and Sitte (1889) are depoliticized—in favor, especially in the case of Sitte, of an aesthetic approach, which all the resources of archaeology and the imaginary museum of town planning are marshalled to support. "It is only by studying the works of our predecessors that we can reform the commonplace plans of our great cities," writes Sitte (1889: 118). Thus the space of the culturalist model is opposed point by point to that of the progressivist model. Precise limits are assigned to the city. The metropolis of the industrial era horrifies Howard, who fixes the number of the inhabitants of his town at thirty thousand or fifty-eight thousand. 12 His city is circumscribed in a precise manner with a green belt destined to prevent it from coalescing with other settlements. A garden city cannot expand in space; it can only multiply in the manner of live cells, the surplus population going off to found a new center, at a sufficient distance, which will also be surrounded by greenery. Every city occupies space in a particular and differentiated way; this is the result of the role that the culturalists ascribe to individuality. In the search for differentiation, Howard stresses particularly sociological factors; the population must be balanced among different age groups and among all sectors of the economy. Sitte, for his part (followed faithfully by Unwin as concerns the organization of the central nucleus of the garden cities), is concerned exclusively with the means of ensuring individuality and variety in the interior space of the city. He makes use of analysis of the city of the past (from antiquity to the fifteenth century): there he tirelessly studies the layout of the road network, the arrangement and size of squares in relation to the streets leading up to them, to the building that surrounds them, to the monuments that adorn them. The Viennese master even identifies, in the greatest possible number of cases, the location and dimensions of the vanishing points. If the study stops at the Italian Renaissance it is because urban planning at that time is already (unfortunately, in Sitte's view) making use of the drawing board to achieve effects of perspective. From the multiplicity of plans and analyses, Sitte draws a definition of a model spatial order. Instead of the abstract, exploded space in which, in the progressivist model, the unit-forms of building are

252 Françoise Choay outlined, Sitte proposes a concrete space, defined in the continuity of a background of buildings. Even in the case of monuments, one should react against "the modern disease of isolation" (Sitte 1889: 39). For typological analysis, Sitte substitutes relational analysis; the street is a fundamental organ, the guiding forms are no longer those of buildings but those of places of passage and of meeting, which means streets and squares; and greenery itself, practically eliminated from the urban center by Sitte, is carefully shaped when, occasionally, it appears in certain residential neighborhoods. This space is closed and intimate; for the basic characteristic of ancient cities consists in the limitation of space and of impressions. . . . The ideal street must form a closed entity. The more limited the impressions there, the more perfect will be the picture. W e feel more comfortable if our g a z e cannot lose itself in infinity. (Sitte 1889: 137)

This space should, in addition, be unpredictable and diversified, and for this reason it must refuse to be subordinated to any principles of symmetry, must follow the natural sinuosities of the landscape, the angles of the sun, must adapt to the dominant winds, or to the greatest existential comfort of the user. The mental climate of this model is reassuring, simultaneously comfortable and stimulating; it is favorable to intensity and the multiplication of interpersonal relations, even if, in the case of Sitte, one defers resolutely to pure aesthetics, understood in the same vitalist sense as in Ruskin and Morris. But the promoters of this model, even if they are essentially attached to history, disregard the historical originality of the present and the specificity of its problems. Giedion is not mistaken when he accuses Sitte of wanting, in the middle of the twentieth century, to return to the "medieval city," and when he calls him a "troubadour"; Le Corbusier will observe, more cuttingly: "The cult of the donkeypath has just been created. The movement originated in Germany, the result of a work by Camillo Sitte" (1923a: 9). In effect, the Viennese planner is so obsessed by the aesthetic problems and the forms of the past that he disregards completely the evolution of working conditions, as well as of traffic problems. Unwin himself is well aware of the contradiction and as a good empiricist tries to reconcile the

Urbanism in Question

253

culturalist model with the exigencies of the present. In spite of his efforts, particularly those concerning public transport, he does not a l w a y s succeed. In the case of garden cities, the requisite control of urban expansion and the strict limits imposed on it are not easily compatible with the necessities of modern economic development. T h e fact is that this model is decidedly nostalgic. To completely grasp the nature of this nostalgia, w e will refer to a series of German authors more or less contemporary with the first planners; the vision of the nineteenth-century historians is in these works deepened, completed by certain later acquisitions, and occasionally illuminated with some Hegelian-Marxist concepts. Thus, in spite of the divergence of their positions and their preoccupations (in which philosop h y , history of culture, and political economy respectively play the dominant role), minds as diverse as Max Weber, Sombart,' 3 or Spengler present us with a very similar picture of the pre-industrial European city; for all three of them it is an exceptional place and moment w h e r e , thanks to the particular climate of the urban community, the individual w a s able to reach fulfillment and culture could develop. O n the last p a g e of his introductory remarks to Weber's collection The City, Martindale gives a g o o d summary of this vision and the nostalgic echoes that it still evokes today: Max Weber's theory of the city, thus, leads to a rather interesting conclusion. . . . The modern city is losing its external and formal structure. Internally it is in a state of decay while the new community represented by the nation everywhere grows at its expense. The age of the city seems to be at an end. (Weber 1958: 62) T w o critical conclusions should be d r a w n from this desire to recreate a dead past which is ultimately the ideological motivation of culturalist urbanism. O n one level—methodological and speculative—this uncritical valorization of the past leads to a reification of time, which is treated like space, as if it were reversible. O n e thus arrives by a different route at the same result as in progressivist urbanism. To the progressivist Utopian impulse is opposed a nostalgic utopianism, and to the religion of functionalism, the cult of ancestral values, w h o s e m o d e s of functioning have been revealed by history and archaeology. If w e place ourselves on a second level of criticism, that of the

254

Françoise Choay

unconscious, even culturalist urbanism itself betrays certain neurotic tendencies. In place of the progressivist recourse to the father image, we have this time a frank regression. And the quasi-ritual repetition of ancient modes of conduct betrays a lack of adaptation, a flight from a confrontation with an unbearable present. In its extreme form, this attitude would culminate in the loss of the reality function, compensated for by a kind of magic behavior of compulsory character. A New Model: Naturalism. The ideas of the American anti-urban current crystallize, in the twentieth century, in a new model. Too radically Utopian to lend itself to realization, but nonetheless destined to mark the thinking of a section of American sociologists and urban planners, the model was developed under the name of Broadacre City by the great American architect Frank Lloyd Wright. . . . The ideological principles on which he founds Broadacre are those of a faithful disciple of Emerson. The big industrial city is accused of alienating the individual by the artificial. Only contact with nature can restore man to himself and permit a harmonious development of the whole person. . . . Starting from these premises, Wright proposes a solution for which he always retained the name city, although it eliminates not only megalopolis but the idea of the city in general. Nature here again becomes a continuous environment, in which all the urban functions are dispersed and isolated under the form of limited units. Housing is individual: not apartments, but private houses with at least four acres of property each, land which the proprietor uses for agriculture (the privileged activity of the civilization of leisure, according to Wright) and for different leisure activities. Work is sometimes attached to housing (studios, laboratories, and individual offices), sometimes incorporated in little specialized centers: industrial or commercial units are each time reduced to the minimum viable size, destined for a minimum of persons. The same is true for hospitals and cultural establishments, the large number of which compensate for their dispersion and their generally reduced scale. All these cells (individual and social) are linked and related to each other by an abundant network of land and air routes: isolation has meaning only if it can be broken at any moment. The American architect thus

Urbatiism in Question

255

imagined an acentric system, composed of pointlike elements inserted in a rich transportation network. Broadacre is the model of a random section of a uniform tissue which can be expanded to cover the whole planet with more continuity than the progressivist model. Wright proposed to make a trial of it first in a limited region of the United States; but for him this was a universal solution, destined for world-wide application. Space in this naturalist model is complex; certain of its characteristics relate it to the progressivist model, others to the culturalist model. It is at the same time open and closed, universal and particular. It is a modern space which proffers itself generously to the liberty of man. The great city engineering works (highways, bridges, landing fields) which constitute the transportation network give Broadacre a cosmic dimension: everyone is linked to the totality of space, which lies open to his investigation in all directions. Broadacre's relationship to modern technology is even more decisive than that of the progressivist model: it is the automobile, the airplane, the parkway, television, the most advanced techniques of transport and communication which give this dispersed mode of settlement its meaning. . . . It is also clear that the garden cities of the culturalist vision present a certain number of points in common with the progressivist models. It is not by mere chance that for a large number of American critics, garden city and radiant city are seen as similar. Ebenezer Howard never ceased assigning an important place to hygiene. 14 And his city plan, with its six concentric boulevards and its well-defined neighborhoods, recalls the precision of Fourier's illustrations. Nevertheless, Howard's garden city clearly belongs to the culturalist model in the pre-eminence given to communal values and to human relations, and by the urban Malthusianism which is its result. . . . We could, finally, try to relate to the naturalist model certain propositions by Buckminster Fuller (1963) or Henry Ford (My Life and Work); these propositions are as acentric as Broadacre and place the same emphasis on the role of transportation routes. . . . These three models (progressivist, culturalist, naturalist) have not had the same repercussions in practice. The study of the concrete applications of urban planning reveals, as one would imagine, the great numerical superiority of progressivist settlements. The natural-

256

Françoise Choay

ist model found only very partial expression, and mostly in the United States, in suburban forms. The culturalist model continues to inspire the construction of new towns in England; elsewhere it has occasioned only limited experiments (certain reconstructions and some tourist resorts). . . . A SECOND-LEVEL CRITIQUE: URBANISM IN QUESTION The response to the urban problems posed by industrial society does not end either in the models of urbanism or in the concrete applications inspired by them. These models (originating as criticism) and their applications have in their turn provoked a new criticism, a criticism of the second degree. The movement started in the decade of 1910-1920, but its real rise began after World War II, connected to the growing practical application of urban planning.15 This criticism, which is still theoretical, remains vague. It is, however, oriented in two major directions, corresponding to the dichotomy (progressivism-culturalism) that we have detected and revealed as already existing in the age of pre-urbanism. . . .

NOTES 1 . The nineteenth century was the golden age of Utopias. Among the most interesting, we do not cite Looking Backward (1888) by Edward Bellamy, or Freiland: Em soziales Zukunftsbild (Vienna, 1890) by Theodor Hertzka: both are too exclusively centered on the legal and economic issue to be included here. On the problem of Utopia, cf. Hertzler 1923; Ruyer 1950; and Riesman 1947. 2. Cf. Emerson, Nature (1836); and Thoreau, Walden (1854). 3. We are thus in disagreement with Benevolo (1963), who dates the "de-politicization" of thought concerning urban planning from 1848. In a general way, the approach of the pre-urbanists is more interested in a theory of social relations than in politics as such. But this holistic view of the city persists until the beginnings of the twentieth century. William Morris is an excellent example. The technical plans that attract Benevolo's attention from 1848 on are only a—particularly spectacular, due to the industrial revolution—case of an applied practice which has always existed; he tends to confuse urban planning and civil engineering. 4. The group of C I A M (Congrès Internationaux d'Architecture Moderne) included not only Europeans such as V. Bourgeois, Gropius, Hilberseimer, Le Corbusier, Rietveld, Sert, Van Eesteren, but also representatives of the United States (Neutra, Wiener), of Brazil (Costa), of Japan (Sakakura), etc. Though at first interested

Urbanism in Question

5.

6. 7.

8. 9.

10. 11.

12.

13. 14.

15.

257

in the housing problem, the CIAM placed planning in the forefront of their concerns after the congress of 1930, at which date Van Eesteren, who was then head of the Town Planning Department of the city of Amsterdam, became president. The architects of CIAM in 1933 drew up the Athens Charter or Town Planning Charter during their fourth congress, which took the form of a mediterranean cruise toward Greece and Athens. The principles then defined were later presented in two works destined for the non-specialist public: La charte d'Athènes: L'urbanisme des CIAM by Le Corbusier (Paris: Pion, 1943), and Can Our Cities Survive? by J. L. Sert, vice-president of CIAM (Harvard University Press, 1944). For forty-five years, we detect no development or change in the urban planning theories of Le Corbusier. Today his ideas appear outdated in certain countries, but not in France, where, essentially, the conditions of construction and the mentality of the public had not changed from 1918 to the 1950s. Le Corbusier 1946: 38. Cf. "All men have the same organism, the same functions. All men have the same needs" (Le Corbusier 1923b: 108). Cf. Rey and Pidoux, "Une révolution dans l'art de bâtir: L'orientation solaire des habitations" (paper presented at the Pasteur Institute Congress on Hygiene, 1921). These authors extol "sunlight, supreme element of life" and propose a "rigorous solution of the problem of sunlight in housing," which they return to later in Rey and Pidoux La science du plan des villes (1928). Le Corbusier 1946: 27, 77. Cf. the opposite thesis in the Buchanan Report (1963). Kahnweiler has quite perspicaciously compared the cubist method to Husserlian philosophy (which was actually unknown to the cubists); cf. Kahnweiler 1946: 267. The first stone was laid in 1947 and the building was completed in 1952. The constraining character of Le Corbusier's cities has been particularly well demonstrated by Mumford. Cf. notably in Mumford (1964) the essay entitled "The Marseilles Folly:" "In short, this plan, with its arbitrary dimensions, the way in which it deprives its occupants of every possibility of privacy, its failure in the use of natural light, offers a perfect example of the procrustean conditions which are beginning to govern modern architecture. As that ancient Greek innkeeper, the architect of the radiant city has recourse to violence in order to bend human beings to the inflexible dimensions of his monumental edifice" (p. 77; our italics). The maximum population figure that Howard assigned his cities is 30,000 inhabitants plus 2,000 farmers. These cities (by definition isolated from each other by green belts) could possibly be grouped around the periphery of a central city (at a distance of three to twenty miles) whose population should not exceed 58,000 inhabitants. Cf. especially Sombart 1902-27, 2: 2, 1907. Cf. his paper at the 1910 First International Conference on Town Planning in London: "What are the basic requirements of a dwelling? They are above all sufficient space, light, and air. We have shown that it is possible, scientifically and systematically, to attract the industries of the overpopulated centers to specific spots, having at their disposal, for maximum efficiency, water, light, and power, and where the population can be lodged in appropriate, inexpensive houses surrounded by gardens, in proximity to their work and recreations, in such a manner that the infant mortality rate would be ncsjiigher than 31.7 per thousand, compared to 107 per thousand in London." In effect, the second-degree critique has developed parallel with and in propor-

258

Françoise Choay hon to the importance of applied urban planning. Thus it has developed much more rapidly in the United States and England than in France.

REFERENCES Benevolo, N. 1963. Le origini dell'urbanistica moderna. Bari: Laterza. Buchanan Report. 1963. Traffic in Towns: A Study of the Long Term Problems of Traffic in Urban Areas. London: Her Majesty's Stationery Office. Fuller, Buckminster. 1963. Nine Chains to the Moon. Carbondale: Southern Illinois University Press. Gropius, Walter. 1935. The New Architecture and the Bauhaus. 3d ed. London: Faber and Faber. Hertzler, J. O. (1923). The History of Utopian Thought. London: Macmillan. Kahnweiler, D. 1946. Juan Cris. Paris: Gallimard. Le Corbusier. 1923a. Urbanisme. Paris: Crès. 1923b, 1958. Vers une architecture. Paris: Vincent Fréal. 1946, 1963. Manière de penser l'urbanisme. Paris: Gonthier. (First published in L'Architecture d'Aujourd'hui, 1946.) Mumford, Lewis. 1964. The Highway and the City. London: Harcourt, Brace and World. Riesman, O. 1947. "Some Observations on Community Planning." Yale Law journal (December). Ruyer, R. 1950. L'utopie des utopies. Paris: PUF. Sitte, Camillo. 1889, 1918. Der Städte-bau nach seinem künstlerischen Grundsätzen. Vienna: Carl Graeser. (Page references are to the 1918 French edition.) Sombart, W. 1902-1927, 1928. Der moderne Kapitalismus. Munich: Duncker und Humboldt. 1907- "Der Begreif der Stadt und das Wesen der Städtebildung." Brauns Archiv 4. Unwin, Raymond. 1909. Town Planning in Practice: An Introduction to the Art of Designing Cities and Suburbs. London: Unwin. Weber, Max. 1958. The City. Trans, and ed. by Don Martindale and Gertrud Neuwirth. New York: Free Press. London: Macmillan. White, Morton, and Lucia White, i960, 1962. "The American Intellectual Versus the American City." In Lloyd Rodwin, ed.. The Future Metropolis, pp. 214-32. London: Constable.

12 SEMIOLOGICAL URBANISM: AN ANALYSIS OF THE TRADITIONAL WESTERN SUDANESE SETTLEMENT Alexandros Ph. Lagopoulos

This article belongs to the author's first period and is complementary to his 19-72 article "Semeiological Analysis of the Traditional African Settlement" (EkisticsJ. Both articles are based on his unpublished dissertation L'influence des conceptions cosmiques sur l'urbanisme africain traditionnel (1970).

Lagopoulos presents briefly his views on "semiological urbanism," which are influenced by European structural linguistics, Barthes, and Eco. He sees semiological urbanism as a branch of spatial analysis, whose specificity, compared to urban semiotics, derives from its closeness to spatial analysis. The author differentiates the semiotic from the non-semiotic, the semiotic conception from the production of urban space, and the theoretical mental image from the actual, perceptual image. The author then discusses the settlement models of a series of African tribes and ends with a number of conclusions, which extrapolate from the body of work studied to the whole of First published in P. Oliver, ed., Shelter, Sign, and Symbol (London: Barrie and Jenkins, 1975).

260

Alexartdros Ph. Lagopoulos

the "non-Western" (i.e., precapitalist) settlements and which include comparisons with the "Western" (i.e., modern capitalist) settlements.

THE SUDAN AS A CULTURAL UNIT Ecologically, the Sudan consists of two almost parallel main biomes: arid grassland or semi-desert and tropical savannah and grassland.' Within this zone Baumann identifies, from west to east, eight cultural areas: the Western Atlantic, that of the Upper Niger, the Voltaic, the Eastern Atlantic, the Semi-Bantu, the Central Sudanese, the bipartite Eastern Sudanese, and the pastoral area of north-east Africa. Paulme bands together the first four of these and part of the next two into one area, the Western African area, thus defining an area corresponding to the Western Sudanese area of Herskovits. Both define a cultural area, which Herskovits calls "Eastern Sudan," lying east of Lake Chad and, according to Paulme, extending to the Red Sea. Horowitz believes that the Eastern Sudan is culturally heterogeneous, but forms a unity by reason of the observed economic and cultural interdependence, and that the whole of the Sudan—at least its central region—can be regarded as a unity,2 a view to which we subscribe. In the following study we will deal with the western part of the Sudan's cultural area. BASIC SEMIOLOGICAL CONCEPTS Semiology is the science of the study of signs; it rests on structural linguistics. The linguistic sign consists of a concept and an optical or acoustical image of a psychological nature. The concept is the signified and the image is the signifier. Certain basic semiological concepts will be dealt with below. Social language (langue) and speech (parole). According to Saussure, language (langage) presents a social aspect, the social language, which constitutes a system of conventional signs, and an individual aspect, speech, which consists of individual combinations of those signs, obeying the rules of the social language. The two aspects of the language are linked dialectically. Extended speech, the combina-

Semiological Urbanism

261

tion that is of the larger linguistic elements termed sentences, is called discourse and is no longer bound by linguistic but by logical restrictions. Hjelmslev distinguishes three levels of social language: schema, norme, and usage. The opposition "social language— speech" is closely related to the opposition "code—message" of information theory and corresponds to the oppositions "structure— event" and "system—procedure." According to Barthes, the opposition "social language—speech" is a fairly general one as regards semiological systems. One difference between linguistic social language and most semiological social languages is that the former is elaborated by the mass of speakers, whereas the signs of the latter are relatively arbitrary and become carriers of a "logo-technique." Signifier (signifiant-Sa) and signified (signifié-Se). According to Hjelmslev, the level of the signifiers and that of the signified of linguistics present two aspects: form, which can be described systematically by linguistics, and substance, the description of which lies outside the field of linguistics. This verification can be generalized with all semiological systems. The semiological sign frequently possesses a signifiées substance which carries the substantive signifier and is called the pre-signifier {pré-signifiant). Signs of this sort belong to non-isologue semiological systems. A pre-signifier is presented by semiological systems where the substance of the signifier of the signs is a useful object. Barthes calls the signs of these systems functions-signs (fonctions-signes). The effect of the act which connects the signifier and the signified, entailing the creation of a sign as an absolute entity, is called signification. The relative position of a sign inside a system of signs establishes its value. According to Barthes, the relation between signifier and signified is exact and unmotivated as concerns sign, while it is inadequate and analogical as concerns symbol. Their analogic relation in the case of a symbol also lends it expression beyond signification. The linguistic sign (morpheme) is one of the two unitary elements of linguistics; it is accompanied by meaning (unité significative) and belongs to the first articulation of language. A second unitary linguistic element is the phoneme, a distinctive unit (unité distinctive) "devoid" of meaning, which belongs to the second articulation of language. Human language is characterised by the existence of those two articulations.

262

Alexandros Ph. Lagopoulos

Syntagm (syntagme) and paradigm (paradigme). Syntagm (metonymy) is a linear combination of signs governed by syntactic rules. Paradigm (metaphor), in linguistics, is the set of the signs which are related by sound or meaning or both. Barthes observes that there can exist erratic semiological systems, in which inert pre-signifier material will bear at certain places discontinuous and even disconnected signs; in the interspaces between the signs there are temporarily inexistent syntagms. In the opinion of Saussure, there are only differences in social language. The differences between linguistic signs become the reason for their classification into three basic dual paradigmatic categories, according to Cantineau, in accordance with which the relations between the similar and the dissimilar element of two signs are adopted as criteria. Particular interest is presented by privative oppositions, in which, according to Martinet, the signifier of one sign is detected by the existence of a signifier (marque) which is lacking in the signifier of another sign. The one signifier is marked (marqué) while the other is unmarked (non-marqué). The latter is called zero degree of the opposition and does not entail absence of signification, but an absence which possesses signification. Denotation (dénotation) and connotation (connotation). A system is called denotative when its signs by themselves are regarded, and connotative when its signifiers (connotators) are signs of another system, which is denotative with respect to the first one. It is possible for a number of signs of the denotative system to form only one signifier of the connotative system. The form of the signifiers of the connotative system constitutes rhetoric and that of the signifieds ideology. The level of connotation is the symbolic level of a denotative system of signs, and it is there that the aesthetic message must be sought. Language-object (langage-object) and meta-language (méta-langage). A system is called meta-language if its signified are signs of another system, called language-object.3 INTRODUCTION TO SEMIOLOGICAL URBANISM The physical space (the artificial environment of ecology) is the result of social action. As with every social phenomenon, it becomes

Semiological Urbanism

263

the carrier of a meaning to the group or groups which have created it and, generally, to every individual who is familiar with it directly or indirectly, whether intimately or superficially. In principle, the approach to it on this level falls within the scope of semiology, in order to form ultimately, in our opinion, a branch of urbanism (semiological urbanism), when immediately related to the urbanistic approach. In the present study, a semiological analysis of the settlement of the western part of the cultural area of the Sudan, as defined in the first paragraph, will be attempted. There are two inverse processes which are explored by semiological urbanism. The first presupposes the existence of a physical space and the activities pursued within it, the perception of which leads to the emergence of a set of signs, related to signifieds of denotation and connotation. The object of its semiological study is the mental morphology (according to Boudon) of the physical space and the creation of an interpretative structure-model. The urbanistic sign is a function-sign, because it presents a presignifier of a useful character. According to Eco, the signified of denotation of the sign of architecture and urbanism is the function performed by the

pre-signifier: the signified indicates the primary function (funzione prima) of the pre-signifier viewed as an element of physical space. This thesis relates to space production, but in so far as the mental image is concerned the signified of denotation does not belong solely to the urbanistic functional level, or to the functional level generally, but may also belong to other levels, such as the social level generally, the economic level, the chronological level, and others. A pre-signifier may relate, either by itself or with other pre-signifiers, to a series of signifieds of connotation belonging to the same sign of connotation, in regard to the fact that both the signifier as well as the signified of denotation carried by it form a new signifier, or a part of a new signifier, of a series of signifieds. T h e latter

mainly indicate the social role of the pre-signifier (funzioni seconde, in regard to space production). The semiology of the mental morphology of the settlement may comprise t w o d i f f e r e n t levels. The first relates to its theoretical mental image and the second to its actual perceptual image. This distinction d o e s

not lead to the radical separation of the two levels for an individual, because the continuous passage from the one to the other is possible for him.

264 Alexandros Ph. Lagopoulos The second process explored by semiological urbanism relates to the physical morphology (according to Boudon) of space, from the scope of its production with the aid of its creators' cultural representations (representations' structure-model). The semiological study of this process can adopt two avenues of approach, one of which explores the mechanisms of the creation of the physical space. This creation sets out from a set of values (ideological level), which, if we isolate the functional level generally, define a set of functions. The latter, which belong to the signifiers of the signified values, are translated into forms. These forms, in turn, become signifiers of the signified functions. One part of urbanistic, and architectural, design falls under semiology.4 The following symbols will be used in the semiological analysis of the western Sudanese settlement below: (d) = at the level of denotation; (c) = at the level of connotation. SEMIOLOGICAL ANALYSIS OF THE WESTERN SUDANESE SETTLEMENT South Mali: Dogon. The shape of the settlement, according to the Dogon (Dagon, Dagom, Tombo, Habbe-sing, Kado), must be elliptic (figure 12.1). The main square, tay, according to this model, is located to the north and comprises the council house and the oval altar anakazu dummo; the latter is quadrangular in plan and its corners are oriented toward the cardinal points of the compass. The signified (c) of the square, which plays the role of pre-signifier, is the sky (alagala) and the primordial field. This field is held to be divided into 60 parts, in the image of the earth and of the world; into 64 parts; or into 6,400 (80 by 80 cubits). The signified (c) of the council house is a head, while those of the altar are the egg of the creator Amma, the center of the world, and the point of departure of the (second and third world) creation. North of the square there is the forge, the signified (c) of which is a mythical forge. At the center of the settlement there is a set of stones, the signified (c) of which is the female genital organ. Respect for women leads to the construction of the foundation altar, having as its signified (c) the male genital organ, outside the settlement, while normally it ought to have stood beside the set of stones already referred to. Two other elements of the settlement have as their signified (c) one hand each,

Semiological Urbanism 265 Figure 1 2 . 1 . Elliptic model of the Dogon settlement (according to Griaule 1966). ^ ^ Forge

, Council -Assembly ¿place U U J t shelter Head

'

\

/

\

/ gl^BS \ •

Chest

Women's i house |

J

Pjlfc

o\

Hand I Oil s t o n e d 1FemahsaiKg ^

| |

Jo

/ I

Village «Her

• Hand '

7

Masculine sex r

%

\

I

V

Altars Feet

/