Semiotics and the Limits of Architecture

Diana Agrest and Mario Gandelsonas

Part I: The State of the Art.

INTRODUCTION

In the past fifteen years interest in the semiotics of architecture has grown in various parts of the world, as is shown by the publication of the Italian and English bibliographies prepared for the First International Congress of Semiotics in Milan (June 1974) and published in the journal Versus 8/9.^1. To this impressive number of works already published in English and Italian, an equally large Spanish bibliography will very soon be added. This great quantity of work is, however, not just a simple indication of the vitality of the field, but rather the result of a quite complex situation.

On the one hand, a great number of these works are part of developments in the behavioral sciences (psychology, sociology, etc.). This produces a situation in which the study of communication problems (transmission of meaning by means of sign systems) and moreover of signification (structures of such systems) is approached from presemiotic perspectives which have since been criticized and superseded by semiotics. On the other hand, in the proper semiotic works the dispersion of themes of analysis instead of the determination of the object of study, the disarticulation between empirical analysis and abstract generalizations, the quantitative expansion rather than the accumulative development of knowledge all become quite obvious.

Very few works have understood the sense and importance of the Saussurian gesture of the definition of the object of study^2. and developed their analysis in reference to a theoretical object previously defined. Very few are the works which may be considered theoretical developments in a strict sense, that is, hypothetic-deductive constructions, descriptions, or classifications which are explicitly inscribed within a given theoretical context, such as the work of M. Bense^3. or E. Garroni.^4. Finally, there are few works which propose a conceptual development of the problems stated in previous works; most works transpose models developed in another field (linguistic or semiotic) and cannot be considered the product of a consistent development of a more specifically architectural problematic. This situation is a product, in our view, of two major types of problems which a semiotics of architecture has to confront: those which have their origin in architecture and those which originate in semiotics itself.

TWO OBSTACLES: ARCHITECTURE AND SEMIOTICS

From the decade of the sixties architecture has been going through one of its periodic crises, due this time to several reasons: a) the failure of the massive interventions of post World War II and the corresponding collapse of the functionalist doctrine, the basic ideology of modern architecture; b) the increasing gap between the accelerated process of industrialization which affects the whole of the economic production and the craftslike character of the architectonic production; c) the expansion of the field of intervention, i.e., industrial design, architecture, urban design; d) the ever-growing distance between the figurative vocabulary and syntax created by modern architecture (simultaneous with cubism and neoplasticism) and the languages generated by mass media, in particular advertising, television, and film; e) the antihistoricist position of the modern movement and the fast consumption of formal vocabularies.

As a result of this crisis, discussions take place about the dubious architectural character of some realms of production. An example of this is the problem of housing which in certain contexts appears as “that” which opposes architecture—architecture versus housing—the latter representing the “death” of architecture. Parallel to this, the concern for the problem of industrialization displaces the architectural creativity from the design of buildings to the design of construction systems, of assembling, “montage,” etc. The expansion of the field of intervention produces an hypertrophy of “architectonicity”; from the design of objects to the design of buildings and urban spaces, everything is considered architecture. Finally, the crisis and its results are manifested at the level of language in the form of eclecticism, a confusion of languages characterized by the use of figures and syntaxes belonging to heterogeneous and sometimes contradictory codes.

This set of circumstances acts as a first obstacle to the development of a semiotics of architecture, an obstacle which makes the identification of an object of analysis a very difficult task.

The development of semiotics has also gone through a succession of crises to the point of negating the validity of its own demarche.^5. The semiotic approach, which at this point is just beginning to expand, cannot be seen as a peaceful “factory” of models but as a field where many times antagonistic fights between different trends develop. We will mention some of the events that have been influential in the development of semiotics of architecture, a) The acceleration of semiotic analysis which took place in the 60s in France owed a great deal to the structuralist perspective, and adopted a theoretical and methodological outlook from the confluence of models drawn from structural linguistics and anthropology.^6. b) Strong criticism is directed from philosophy and epistemology centered on the empirical nature of the approach and on the problem of the absence of two notions: the subject and history.^7. The diffusion of theoretical hypothetic-deductive models from generative transformational linguistics reinforces the challenge to the empirical structural models, c) The theories of enunciation, of speech acts,^8. and the reaction against Chomsky’s model influence the displacement of the semiotic problematic towards a pragmatics which introduces the problem of subjective and sociological determinations in language.^9. The analysis of text, on the one hand, opposes to Chomsky’s Cartesian subject a more complex notion of subject drawn from psychoanalysis,^10. and, on the other hand, poses the problem of the articulation of the theory of meaning with history through the analysis of ideology considered as production of meaning.^11.

It is important to note that this schematic description of the sequence of events is applicable in particular to the semiotics of literature where since the Russian Formalism of the 1920s most work has been done. This is a field which received from the theory of literature and poetics an already structured theory, which had inventoried codes and specific systems with a high degree of sophistication. In the case of the semiotics of literature, we should seek the reason for these developments not so much in external factors (such as the rapid succession of various linguistic trends) as in the limitations of different analytical approaches which were made apparent by the result of such analysis.

This is not the case in the semiotic approach to architecture where, although a similar process takes place, most of the analyses are mechanical transcriptions of models which do not go beyond the general level. The changes in approach do not reflect a need or development provoked by the process of research itself but are instead a mere reflection of developments in other semiotics. This situation, which affects the possibility of producing specific models, acts as a second obstacle for the development of a Semiotic approach to architecture.

Nevertheless, it is possible to distinguish a number of works where a body of notions suggests—whether in an explicit or implicit way—the possible boundaries and the internal organization of the field of semiotics of architecture.

SEMIOTICS AND ARCHITECTURE: THE STATE OF THE ART

A panoramic view of these developments can be presented, classifying the works according to: a) their relationship with various approaches in linguistic theories which had a strong influence in the production of semiotic models; b) their relationship with different general semiotic perspectives, in particular those posited by the founders of this approach, F. de Saussure and C. S. Peirce.

The first attempts to develop a semiotic approach to architecture from a structuralist perspective center their effort in the definition of systems and the units of which they are composed, following the model of the sign.

From the impulse given to this approach by the works of C. Lévi-Strauss and R. Barthes Umberto Eco produced one of the first attempts to establish the basis for a semiotics of architecture.^12. This work develops in the field of architecture (defined in a wide sense which includes any man-made object with the exclusion of esthetic objects) the semiotic notion of sign, the function as first meaning of the form, the notion of connotation, the non-functional secondary meanings, the notion of code, the rules that establish equivalences and oppositions between forms and functions, and the diachronic transformation of primary and secondary meanings. It has also the virtue of reformulating previous theoretical attempts—such as communicational theories—and of introducing models from other disciplines, such as the anthropological analysis of space, which are then displaced and reelaborated within a problematic of signification.

Parallel to this work, other analyses are developed by R. de Fusco and M. L. Scalvini^13. which focus especially on architectural material in a more restrained way, that of the buildings designed according to a strict system of rules in different historical periods. In contrast with Eco’s position the two aspects of the sign, signifier and signified, are considered at the level of configuration, as architectural shapes. The external configuration is considered as signifier which refers to an internal organization which becomes the signified. This particular definition of the architectural sign started a parochial controversy which involved E. Garroni, who has tried to show that these apparently antagonistic views do not exclude each other.^14.

The model of the sign has also been applied in the field of urbanism, an extension of the semiotic approach which is of major importance since the practice of architecture in our culture has been, since the Renaissance, associated with problems derived from the urban context. F. Choay’s early works propose a typology of interpretative models of the city based on the analysis of texts.^15. She has further proposed the analysis of the city as a system of signification in terms of syntagmatic and paradigmatic relationships which define types of urban configurations.^16.

More recently, a more careful structural approach to architecture may be seen in the recognition and distinction among the different media (matters) in which architecture is manifested.^17. As a consequence there is a development of more specific theoretical models for their analysis.

The complexity of architecture—its heterogeneous nature—is revealed by the diversity of forms manifested by “architecture” in the various products of architectural practice, i.e., written, drawn, and/or built architecture. Traditionally the building has been considered as the “natural product” of architectural practice, while texts or drawings were seen as neutral means to achieve that product. However, not only is the main part of the creative process developed in the text and specifically in the drawing, but also on many occasions the process is interrupted at some of these levels without reaching the point of an effective realization of the building. The complex character of architecture has undoubtedly influenced the present tendencies to analyze these levels separately.

The development of techniques for the analysis of discourse has contributed to the accentuation of this emphasis on the analysis of texts. In particular M. Gandelsonas’ work has centered on the analysis of a great number of theories developed since the Renaissance.^18. One of the most interesting characteristics of this material is that these theories are rewritings of a text, that of Vitruvius, written during the second century A.D. and resuscitated during the Renaissance, a fact which gives a certain unity to the material and facilitates the analytical work. These analyses, by artificially separating the normative texts from the other phases of the productive process—drawings and buildings—accentuate their differences.

The analysis of drawing in architecture as a system of representation and particularly as an instrument for creation has only recently begun and there is still much to do in this respect.^19. It is unquestionable that the creation of the system plan-elevation-section-perspective has contributed in great measure to the signifying complexity of architecture, both classic and modern.^20. This system of transformation of tri-dimensional space into two dimensions allows the play of symbolic relationships among the representational elements not completely noticeable in the building without the help of a mental reconstruction of plans or sections.^21.

The most complex problem has up to now been presented by those cases in which the analyst is confronted with a building without the possibility of access to a text or a two-dimensional representation. In such a case, the analyst confronts a situation which poses in a crude way the basic problems of a semiotics of architecture, that of the need for producing some notational system which would allow the description of the building by reducing the infinite figures that space and movement suggest, and also to find or generate a text which allows access to functional, figurative, or symbolic codes.^22.

The analyses developed by anthropologists like Lévi-Strauss^23. and Bourdieu^24. have not made this problem explicit. These analyses consist of the reconstruction of codes and rules based on a discourse and a graphic description provided by an informant, which are not considered as possible determinants of the results of the analysis but rather as neutral means containing information for the research. In more complex situations, such as the analysis of urban environments (historical or modern), the problem of description becomes insoluble for the structural approach; consequently the results are rather poor and schematic. This situation has forced the analyst to look for other modeling alternatives to which we will refer in the last section of this paper.

Finally, it would seem that the problem of the manifestation of architectural semiosis in different matters promotes not just different levels in the productive process (the process of design) but different languages; and what has been lost in this analytical schism is the problem of the relation between languages within the process of design, that is in the process of transformation from writing to drawing, from drawing to the building. This collision of languages is a determinant factor of the architectural semiosis; and in this sense the analysis of this problem might contribute to the construction of theories which describe the generation of meaning and forms in architecture and the “translation” of ideology into architectural formal codes and texts.^25.

At this point we would like to suggest some of the possible reasons for the almost exclusive development of structural models and the limited number of works based on generative models.

The production of architecture—as in the other artistic practices—is manifested in “works”; it is not the result of an infinite productivity controlled by more or less explicit rules as in the case of language. This fact makes it very difficult to apply the notion of grammaticality to an architectural “work.” In this case the analyst could propose, at the most, criteria of acceptability which should be decided from the consideration of a set of notions which would include in particular gender, reader, or use. This might be one of the main reasons why most semiotic analyses are carried out from the structural perspective.

One of the first works within the generative approach is “Le statut de l’objet” by P. Boudon.^26. Boudon attempts to establish the basis for a very general theory of the production of objects elaborated on the basis of notions originated in anthropology, generative transformational linguistics, and philosophy of language. In later works Boudon has conducted his research with an anthropological emphasis manifested in his “Semiotique des lieux” or in his analysis of popular Arabic architecture.^27. The transposition of Chomsky’s model has also been applied to the theorization of the process of design in terms of its operations as in Eisenman’s work,^28. and to explain popular^29. and non-Western architectures.

There is a series of works of architectural semiotics which have been produced as part of a more general semiotic theory. Within this particular type of analysis, i.e., the structural approach, in addition to U. Eco we will mention L. Prieto’s and E. Garroni’s developments.

The interest of Prieto’s work,^30. which takes E. Buyssens’ notions as basic references for his developments, resides in the fact that his theory, in which objects and language occupy privileged positions based on the logic of classes, has given origin to developments which present interest at a taxonomic level such as in the case of the works of C. Jannello^31. and in the recent development by H. Pinon.^32.

Among the theories developed from general semiotics, one of the most interesting is that proposed by the Italian semiotician E. Garroni^33. who has not only indicated the validity, limitations, and complementarity of Eco’s and de Fusco’s approaches, but has also classified other approaches originating in architectural theories, such as typological analyses. In addition, he has proposed other theoretical alternatives for the analysis of architecture in terms of meaning, based on geometrical models and on Saumjan's generative and transformational grammar. The interest of these works, developed as part of a general semiotic theory, resides not only in their development of notions at a high level of abstraction but also in their potential production of newer and more powerful models. This might be the result of both their involvement with comparative analyses within a wider range of signifying systems and their distance to the more specific characteristics of such systems. Nevertheless, this distance is responsible for their limitations, which derive from their detachment with respect to the practice of architecture and their limited knowledge of indigenous theories—a detachment which makes difficult the testing of models and therefore the proposition of more “concrete theories,” that is, theories that might produce more specific knowledge about architecture.

SEMIOTICS AND THE LIMITS OF ARCHITECTURE

The general aim of the semiotic approach to architecture is the production of knowledge of architecture seen as a system of signs, as a systematic and specific organization of forms and meanings. Similar objectives have been postulated by other theoretical and critical approaches to architecture. The basic difference between semiotics and these other approaches lies in the former's more specific and powerful models and therefore in its potential for a substantial contribution to the production of knowledge.

The problem is that most of the semiotic production has not fulfilled this potential role. In our opinion, the fulfillment of this role depends on a more critical position with respect to the following attitudes and notions:

a) The mechanical transference of models elaborated in other fields. Semiotics has been just one of the latest models in a long list of technical or scientific models used to develop new pseudo-theoretical approaches to architecture.

b) The two forms in which the traditional approaches to the problem of the relations between architecture and culture and society are manifested: first, architecture seen as an autonomous discipline with respect to other semiotic systems and second, architecture seen as a direct “representation” of political and economic instances.

c) The normative definition of architecture which separates a few buildings designed according to institutionalized rules from the rest of the man-made physical environment which is not given any consideration. U. Eco’s definition of architecture involving the whole of the man-made environment has been postulated, precisely, as a critique of the traditional definition.^34. Nevertheless, it merely reflects a new ideological attitude developed in particular by the approach known as “Design Methods,” where design is considered as a universal process which underlies every manmade object.

Our own work has been developed as a critique to these historically determined ideological notions. We consider that, inasmuch as they both acknowledge and distort pertinent questions, these notions should be taken as basic material for a theoretical analysis which in turn should displace them in order to develop a new and more productive theoretical problematic.

Instead of applying in a mechanical way linguistic or semiotic models, our analyses are based on, first, what C. Metz calls “indigenous” theories,^35. that is, theories developed through centuries by historians and critics of architecture; and second, on semiotic theoretical procedures developed for the analysis of other complex semiotic practices.

C. Lévi-Strauss has given two reasons why one cannot dispense with studying cultural “home-made” models. “First, these models might prove to be accurate or, at least, to provide some insight into the structure of the phenomena; after all each culture has its own theoreticians whose contributions deserve the same attention as that which the anthropologist gives to colleagues. And second, even if the models are biased or erroneous, the very and type of error are a part of the facts under study and probably rank among the most significant ones.”^36.

An example of the efficacy of “indigenous” models coming from the theory and criticism of architecture is the work of Gian Carlo Argan on the notion of type.^37. This notion of type refers to a formal structure which implies the possibility of infinite formal variations. These types are basically configurational structures rather than functional ones. Argan relates this notion of type to the process of design in terms of two instances: on the one hand, what he calls the moment of typology, an instance associated to the configurational structures established in the past; on the other hand, the moment of invention, the instance of response to the demands of the contingent historical situation through criticism or subversion of past solutions sedimented and synthesized in the codification of the type.

In the elaboration of this notion Argan analyzes the same problem developed by C. Metz in relation to film semiotics, a problem that transcends this practice to become a fundamental question for semiotics of artistic facts: the consideration of a semiotic fact in terms of the notion of language —as a combinatory of codes—or as text produced following the rules established by codes and at the same time subverting them, a text written with and against language.

Architectural theories might not only provide “good models,” as in the case of Argan’s theory of type. Their close relationship with the practice of architecture might provide to semiotics a closer and more direct view of the problems that architects have tried or are trying to solve.

One of the central problems being discussed in the fields of history and criticism of architecture is the relation between architecture and the city. Of course, this is not a new problem, since architectural theories have always provided models and norms for the design and interpretation of urban spaces. These different models—the Renaissance city seen as a building-monument, the 19th-century garden city, the 20th-century futuristic city, or the recent “instant” city—have something in common: their simplistic, systematic nature. However, the city itself—and in particular the contemporary city—has developed in a manner which makes the application of these traditional or modern models difficult if not impossible. Therefore, the development of new interpretative theories and models for the understanding of the symbolic nature of new urban environments is not just a pertinent theoretical problem but also a practical demand.

We will mention two different analyses of this problem which might be of interest for a semiotic approach to architecture. First, Manfredo Tafuri’s analysis of Piranesi’s “Campus Martius”^38. —an anticipatory vision of the modern city—where architecture is seen as a systematic game with architectural rules, developed between the limits of redundancy and chaos, between the order of norms and the disorder and irrationality characteristic of the modern city. Second, Colin Rowe's thesis suggesting the use of “Villa Hadriana” as a model for the understanding of the meaning of modern urban spaces in terms of contextual relationships.^39. Villa Hadriana is an “architectural collage” in which fragments of architectures belonging to different periods and cultures are juxtaposed by means of operations in which the classical procedures (proportion, symmetry, balance) play a minor role and the relationships of juxtaposition and contiguity assume the major one.

The interest of these analyses is that they take as a reference complex graphic and/or built projects very different from those used as a reference in the traditional theories. However, these projects still imply an architectural or urban logic, inasmuch as the models derived from them allow the understanding of the traditional models of the environment as partial or simplified versions of the new models.

Of the different semiotic models developed for the analysis of complex practices, the models from both semiotics of culture and semiotics of film seem to be the most interesting ones to compare with architecture.

Architecture has been defined many times as the physical manifestation of a culture. This notion is implied since Vitruvius, when he prescribes that the architect’s knowledge should be so vast that no aspect of culture is unknown to him. We have approached this question by placing less emphasis on the prescribed relationship between architecture and culture and more on the description of similarities between the analysis of architecture and the analysis of culture.

The analyst of architecture and the analyst of culture are confronted with a set of common problems, among which we mention the complexity conferred on it by its heterogeneous nature (a conglomerate of semiotic systems, of languages, of codes and texts), the articulation and transference of forms among the various systems involved, and the transformations which are produced either in the interior of each system simply by its coexistence with others, or in the movements of migration of forms between systems.

Among the most interesting approaches to the problem of culture from a semiotic standpoint is that of J. Lotman, who considers culture as significative information which can be seen either as a hierarchy of codes or as invested in texts which enter the public domain of society.^40.

This dual possibility of analysis, which has also been postulated and developed by Metz for the analysis of movies, has provided us a conceptual framework for the analysis of architecture in terms of codes belonging to different semiotic systems and in terms of the particular mode of articulation of these various codes in the architectural or urban texts.

The analysis of film is particularly interesting to compare with the analysis of architecture because both confront similar problems.^41. First is the interaction and articulation of different languages, a fact which is not immediately obvious in architecture where the various languages appear in sequence within the process of design and construction (and not juxtaposed as in the case of movies) and therefore the complexity does not show in the final product, the building. Second and related to this complexity is the notion of heterogeneity which applies to the different languages structured on the basis of different matters of expression (linguistic, two or three-dimensional scopic matters). Here, we should distinguish between internal and external heterogeneity: the former applies to the design process which has been analyzed in terms of codic differences or in terms of intertextual relationships between written, graphic, or built texts; the latter applies to the contextual relationships and to the process of interpretation of the building or its “reading.” Third, the notion of specificity, which, as in semiotics of movies, should not be just applied to codes but also to the articulation of codes belonging to different semiotic systems. Furthermore, this notion seen from a diachronic point of view allows us to describe changes of the structure of articulation of codes through time and therefore to build a more dynamic model which accounts for the historical dimension of the production of meaning.

The notion of ideology, through the articulation of the theory of meaning with a theory of society, allows us to discuss the questions of autonomy of architecture and its relationships to other social instances overcoming banal and mechanical approaches.^42.

A strong criticism has been addressed to the so-called anti-historicist position of the semiotic approach to architecture in general and to the structuralist perspective in particular. It is important to notice that this criticism comes from the fields of architectural history and criticism, and that it refers to the emphasis on analyses of systematic characteristics of architecture that ignore the historical articulation between architecture and its social context.

In our opinion, structural and taxonomic procedures in the semiotic analysis of architecture are justified when they are just a stage of a more complex analysis which accounts for the articulation between symbolic, economic, and political instances. However, the historian’s criticism is fair when addressed to those cases where the structural analysis (whether systematic or processual) of architecture is developed without making this wider theoretical framework explicit.

Architecture, more than any other practice analyzed by semiotics— perhaps with the exception of film—requires an enormous economical apparatus, and at the same time is submitted to tremendous political pressure. These economic and political forces influence in a very definite way both the structure and the production of meaning in architecture. Therefore, the artificial separation of the symbolic aspects of architecture from the economic and political determinations, which might be a productive procedure within a first analytical stage, should be followed by the development of models that account for more complex determinations.

The interest of the new approaches to the notion of ideology resides in the fact that they seem to offer the possibility of building such models, inasmuch as they try to interrelate the social theory of ideology and semiotic, the theory of the different systems of signs.^43. This notion of ideology concerns the different systems and processes of investment of meaning in different matters. These systems of operations define processes of production of meaning, the products of which are the different social texts, among them the architectural texts in a broad sense. However, ideology is not a property immanent to texts but rather a relationship between textual and extra-textual instances. The pertinency of an ideological analysis cannot be defined without considering the linkage between products (texts) and conditions of production (extra-textual instances, i.e., economic-political and/or subjective determinations). Therefore, in architecture, the ideological functions that characterize architectural notions—technical, cognitive, “masking,”etc.—should be explained on the basis of analyses of the relationships among architectural text and extra-textual circumstances. An example of this kind of analysis can be seen in M. Gandelsonas’ critique of the ideological notion of building, conceived as an unique and simple product of the design process—a notion related to the ideology of artistic creation—which excludes the notion of complex language and intermediary texts.^44.

The aim of semiotics is neither the acceptance of the ideological definition of architecture nor the proposition of new definitions, but rather displacement of the boundaries established by the ideological definitions. Within the practice of architecture, a first displacement of the definition of building as product of architecture, and therefore as object of analysis, has enabled us to show the importance of the analysis of texts and drawings as well as buildings.^45. With respect to the relation between architecture and the non-designed environment the displacement from the exclusive or inclusive definition of architecture has enabled us to show the need for considering their differences in terms of distinct semiotic systems and processes.^46.

The displacement of the limits established by traditional or modern definitions of architecture could be seen in our own work as a change of focus which is produced in two different directions: as a close-up that allows one to see a more complex internal structure within the process of design; or as a panoramic view that allows one to incorporate a wider physical and semiotic context in the process of “reading” the built environment.

Analysis of the intratext determined by the different languages that take part in the process of design allows the analysis of the specific characteristics of texts, the operations of meaning between texts, and their relationships with the intertext. First, each of the texts produced in the process of design and construction must be considered as a mixed text. The written texts are complemented with figures, the drawing with diagrams, written texts, and symbols; the built text has at least a double nature expressed in the opposition of two-dimensional figures versus volume and/or space. Second, in the relation between texts, there is a transmission, reduction, or expansion of meanings; for instance, the written text reduces the meanings opened by the graphic text, the built text acts as a condensed written text, etc. Third, the architectural complex intratext relates in a more direct manner with the intertext like the single texts of other practices, i.e., literature. The intervention of the intertext as a generator in architecture and the great number of “quotations” from texts external to architecture are some examples of this particular relationship which is caused by the impurity of the internal mixed texts which provide an easier access to external figures and signifiers.^47.

The analysis of the non-designed environment allows the theoretical development of the notion of productive reading, which is not the reproduction of a unique or complete sense of the built environment but rather a way of entering into the sequential mechanisms which are part of the production of that meaning, a meaning that society as a whole has put in the built world and not merely that imposed by the architect. A further important difference between reading in design and productive reading in the non-designed environment is the position of “direct reading” required by the latter. Instead of reading by following a previously written text, the reading of the non-designed environment starts from a first mark, not only towards an architectural text but towards other texts in culture, putting into play a force analogous to that of the mechanisms of the unconscious.

In the second part of this paper the semiotic problem of the relation between the designed and the non-designed environment will be developed. The focus on this specific theoretical problem will not exclude consideration of the other problems mentioned above—the relation between semiotics of architecture and other semiotic fields and the relation between theories of meaning and theory of ideology—but rather it will enable one to understand some of their interrelationships.

Part II: Design versus Non-Design.

INTRODUCTION

The specific relationship of architecture to ideology has been generally excluded from consideration in traditional architectural criticism. Concerned only to relate architecture formally, or internally, to itself, or at best to relate architecture externally to society in general, criticism has failed to truly incorporate the cultural problematic of architecture into its domain of concern. When the cultural dimension has been introduced, it has more often been as a simple explanation of architecture as “reflecting” a particular culture—the notion of style as the expression of the spirit of the age—than as a problem to be confronted independently from a consistent theoretical standpoint.

Practicing architects and critics of architecture have repeatedly emphasized the need to relate architecture to its social or cultural context. Positions have been developed around such concepts as “contextualism” and “ugly and ordinary” by writers like Colin Rowe and Denise Scott Brown and Robert Venturi. Rowe, for example, speaks of an architectural contextualism that situates the object of design or analysis in its physical-historical surroundings in terms of formal elements and relations; Venturi and Scott Brown speak of the need to recognize mass culture as the necessary cultural product of our time and as a new source of inspiration for designers. However, rather than attempting to appeal to the notion of collage—a familiar architectural strategy in periods of transition—or to the simulation of the objects of mass culture, this analysis will attempt to investigate the mechanisms of the built environment at this specific historical moment.

We wish to explore here these “external” or cultural relations of architecture—that is, between architecture and its social context—by means of a theoretical model that posits two distinct forms of cultural, or symbolic, production. The first, which we will call design, is that mode by which architecture relates to cultural systems outside itself; it is a normative process and embraces not only architectural but also urban design. The second, which is more properly called non-design, describes the way in which different cultural systems interrelate and give form to the built world; it is not a direct product of any institutionalized design practice but rather the result of a general process of culture.

In thus examining the mechanisms which relate architecture to culture— the processes by which meaning is produced, not only within architecture or design, but also in the domain of non-design—we are, of course, analyzing ideology itself. For ideology is no more than the social production of meaning. Thus, all cultural production, such as architecture, when articulated at the economic and political levels, manifests the ways by which ideology is produced as a part of a given social structure.^48.

In this sense, it is unnecessary to compare one type of architecture to any other type of architecture—as in the accepted mode of “formal,” internal criticism—or to compare it to society in general. Rather, one must oppose the notion of architecture as design to the notion of a radically different kind of symbolic configuration— non-design. This opposition allows analysis of the built environmental in terms of the relationship between different cultural systems. Design and non-design, in fact, can be seen as two modes of social discourse; and to consider them in this way opens up the question of what might be called the “active relationship” between design, as one cultural system, and other cultural systems.

DESIGN AND CULTURE

Design, considered as both a practice and a product, is in effect a closed system—not only in relation to culture as a whole, but also in relation to other cultural systems such as literature, film, painting, philosophy, physics, geometry, etc. Properly defined, it is reductive, condensing and crystallizing general cultural notions within its own distinct parameters. Within the limits of this system, however, design constitutes a set of practices—architecture, urban design, and industrial design—unified with respect to certain normative theories. That is, it possesses specific characteristics that distinguish it from all other cultural practices and that establish a boundary between what is design and what is not. This boundary produces a kind of closure that acts to preserve and separate the ideological identity of design. This closure, however, does not preclude a certain level of permeability toward other cultural systems—a permeability which nevertheless is controlled and regulated in a precise way.

Culture, on the other hand, is understood to be a system of social codes that permit information to enter the public domain by means of appropriate signs. As a whole, culture can be seen as a hierarchy of these codes, manifested through various texts.^49.

The relationship between design and culture may, then, be stated as the mode by which design is articulated (as one cultural system) in relation to other cultural systems (at the level of codes). The transformations in these articulations are historically determined, and they display themselves as changes in the structures of meaning. Thus, the development of specific forms of articulation between design and other cultural systems can be seen as a dynamic process, the study of which opens up the problem of the production of meaning.

The relationship between design and other cultural systems is heightened and intensified at certain moments in this process, and its precise articulations become clearer. In architecture, this occurs when new economic, technical, functional, or symbolic problems force the production of new formal repertories, or the expansion and transformation of existing vocabularies.

Thus, during the French Enlightenment, elementary geometrical figures (the sphere, the pyramid, the cube, etc.) were introduced as the primary constituents of a new formal vocabulary by the “revolutionary” architects Boullée and Ledoux. For Ledoux these forms expressed the new notions of the sublime, while for Boullée they represented the universe and its scientific explanation developed in the context of profound social and political change.^50.

This recognition of articulations between design and other cultural systems also implies the recognition of differences between them—differences which may be understood through the notion of specificity. ^51. This is a notion which permits the clarification of codes according to their relation to design or to other cultural systems.

Three types of codes regulate the interpretation and production of texts in design. First, there are those codes which may be seen as exclusive to design, such as codes establishing relationships between plans and elevations or plans and cross-sections. Second, there are those codes which are shared by various cultural systems, among which design is included (i.e., spatial, iconic). Third, there are those which, while they are crucial to one cultural system (such as rhythm to music), participate—albeit transformed—in another (such as architecture) by virtue of a shared characteristic, i.e., in the case of rhythm, the temporality of the sequence, audial in one case and visual in the other.^52. In a decreasing order of specificity, the first type of codes is specific to design, the second has a multiple specificity, and the third is non-specific.

The specificity of a signifying system is not, however, defined solely by the specificity of its codes, but also by the form in which those codes are articulated; that is to say, the combination of codes may be specific, although the codes themselves may or may not be specific to the system in question.^53. Examples of specific code articulation in architecture are found in classical theories of harmony that utilize the articulation of musical codes and arithmetical proportional series for the invention of specific architectural codes, which are then used to determine the proportions of and relationships between the different elements of a building.

Specificity manages to maintain the limits of architecture despite the apparent changes that occur under the pressures of history, technology, social action, or symbolic change. On the one hand, the most specific codes remain within the system of architecture; on the other hand, the less specific codes link design with other systems through the opening and closing of its limits. This mechanism allows for the articulation of design with some systems and not with others, a process which operates according to the “internal” determinations of design—that is, according to the rules of architectural language, to the logic of the configuration, and to the meaning proper to the “text” of design.^54.

The Mannerist inversion of the established architectural rules—by which each element is used in contradiction to what should be its prevailing ideological function—is an excellent example of such internal determination, in which the inversions so weaken the limits of architecture as to allow an opening to codes external to it; thus the “painterly” architecture of the sixteenth century in Italy.^55.

This process of articulation might, however, take place according to “external” determinations—to the forces of economics, politics, or other ideologies foreign to design. The influence of hermetic thought on the design of the Escorial Palace, for example, demonstrates the role of such external factors in architecture. Both the plan and the general configuration seem to have been derived from mystical or hermetic geometric regulating lines, based partly on parallel developments in quantitative mathematics, and partly on chapters eliminated from Renaissance editions of Vitruvius,^56. but not, as might be assumed, directly from classical architectural theory. Magic codes were thus substitutes for the Albertian geometric codes. Geometry, while represented by similar figures, was imbued with an entirely different meaning. At the same time, these geometric magic codes remained distinctly separate from other magic codes, such as those based on verbal or gestural practices, which never entered in their physical-spatial implications into architecture.

The concept of the closing and opening of limits introduces the notion of an ideological fdtering in the production of design, which takes place by means of certain processes of symbolization. In this case an equivalence, or exchange, of sense is produced by restricting the access of certain codes and figures from other systems into architecture.

The notions of metaphor and metonymy allow for a more systematic analysis of this symbolic functioning. These should be considered as the mechanisms of opening and closure, ultimately revealing the way in which design maintains its limits in relation to culture and acts as a filter in relation to meaning.^57.

Metaphor and metonymy are, of course, notions that have been used principally in the analysis of discourse and text. Since in this context we are analyzing the production of meaning and not its structure, the reference in general will be to metaphoric or métonymie operations rather than to these figures as they applied to classical rhetoric.^58.

These tropes or rhetorical figures represent the most condensed expression of two basic kinds of relationship in discourse: the relation of similarity, which underlies the metaphor, and the relation of contiguity, which determines the metonymy. Each may exist in the relationship between the figure and the content or in the relation between figure and figure.

The development of any discourse (not necessarily a spoken one, and in this case the architectural discourse) may develop along two semantic-syntactic lines: one theme in the expression or content may lead to another either by means of similarity or by means of continguity.^59. The most appropriate term for the former relation is “metaphoric,” while the latter might be termed “metonymie.”^60.

In its relationship to other cultural systems, which is a necessary condition for the regeneration of sense, architecture takes part in a game of substitutions which, thought of in terms of metaphoric or metonymie operations, explains, at the most specific level of form, the translation from extra-architectural to intra-architectural systems in a recoding which, by means of reducing meanings, maintains the limits of architecture.

The well-known nautical metaphor in Le Corbusier’s Villa Savoye exemplifies this functioning. Here, two different signifying systems are related: dwelling and ocean liner. The necessary condition for this relationship is provided by the existence of an element common to both, in this case the window. Through a metaphoric operation, a figurative substitution of the signifying element common to both systems is produced (dwelling/window— liner/window), carrying and transferring codes from one system (liner) to the other (house). The new form is thus loaded with the new meanings required to translate into figures the proposed new architectural ideology.

The similarity of functions—in this case, both liner and house are forms of habitation—makes the metaphor possible.

To these metaphoric transpositions other metonymic operations are added—for example, the promenade architecturale—which also carry further meanings related to the liner.

At an urban scale, where the system of architectural design coexists with many others almost by definition, the role of the metaphor as a filtering device becomes particularly evident, especially in the functional approach to urban design.

At the moment when urbanism was constituted as an institutionalized practice in the first decade of this century, urban formal codes were developed on the basis of the prevailing architectural codification. From the set of possible systems that give meaning to form, the functional approach was emphasized almost exclusively. Le Corbusier may serve once more to exemplify the type of functionalism that is at work in a filtering operation in the substitutive relation between architecture and other systems.

In Le Corbusier’s texts Vers une Architecture (1923) and Urbanisme (1925), these metaphoric operations function clearly as a mechanism for contact between different cultural systems and, on other levels, as a means to architectural recodification.^61. At the building scale, Le Corbusier establishes a connection between architectural systems and other systems, such as technology, tourism, sports, and geometry. This connection is established through a metaphor based on similarity of function.^62. Geometry, for example, had acted as an internal code for formal control from the classical period of Greek architecture. It had not, however, functioned as the provider of the formal vocabulary itself, geometric regulating lines being the “invisible” elements in the construction. For Le Corbusier, however, geometry became not only an instrument of formal control, but also the provider of the formal vocabulary itself in two and three dimensions. The instrument (tool) for representation, that is, drawing, became first the project itself, and then the construction, without alteration.

At the urban scale, Le Corbusier’s metaphoric operation establishes a relation between geometry as a signifying system and the city by means of the common element of “order,” which is manifested as a “grid”; a system of equivalences is established between the geometric grid with its connoted codes and the city grid with the set of values ascribed to it by Le Corbusier.

Thus, in Urbanisme, the existing city is seen as equivalent to disorder, chaos, illness, and irrationality. On the other hand, the grid, the geometric order, is seen as equivalent to order, health, beauty, reason, modernity, and progress. “Geometry is the foundation. ... It is also the material basis on which we build those symbols which represent to us perfection and the divine. .., ”^63.

In the plans for the Ville Contemporaine, and later for the Ville Radieuse, Le Corbusier establishes the equivalence between those two systems by means of the common element of grid-order. The appropriate connoted codes of the geometric grid are transferred through a figurative substitution to the city plan and become the codes of the city itself.

It can be seen, in this case, that while there is an initial opening of the system, its closure is produced by means of a metaphorical equivalence by which the means of representation are imposed as ideological filters in order to develop an architectural recodification. In this substitution, meanings are limited and filtered by a system (geometry) which, while it may not be specific to architecture, will, in its recoding, become specific to urban design. This is made possible by the fact that a system such as geometry may participate in a double “game”: symbolic at a formal-cultural level, and instrumental, or representative, at the level of the specific practice where physical configuration becomes the device that allows for translation and recoding.

The relationship between geometry as a symbolic system on the one hand, and as a basic organizational system on the other, is not, of course, a new problem and may be found at other points in the history of architecture. In the work of Piranesi, for example, the figurative and the geometric coexist, juxtaposed in a clear dialectical relationship. The rear of the altar of S. Maria del Priorato, for example, crudely displays the set of geometric volumes which serve as its support, while the face presents itself as almost pure allegory. The architectural contradiction between geometry and symbolism is here critically posed.^64.

When Boullée and Ledoux adopted geometry in itself as a formal system, the sacred symbology was substituted for a more secular symbology—that of man. In Le Corbusier, however, there is no longer a separation between the geometric and the symbolic; rather geometry itself represents the symbolic aspect of form, and carries with it an entire set of implicit values.

With the waning of the enthusiasm for functionalism in the late 1940s, a series of works appeared which, conscious of the cultural reductivism of the heroic period, were explicitly concerned with the cultural rather than the functional aspects of design. This cultural concern was demonstrated by an intention to make explicit the articulation between architecture and other cultural systems.^65. The work of the active members of Team 10 (Alison and Peter Smithson) reintroduce culture in this sense, and again new openings and closures are produced by means of metaphoric operations: openings to incorporate “the culture”; closures to preserve the specificity of the system.

However, while in Le Corbusier the metaphor was reductive in terms of the possible inclusion of other cultural systems—a product of the exclusive nature of geometry and its concomitant modernism—the intention of Team 10 was to establish relations between architecture and other systems. “Our hierarchy of associations,” they stated, “is woven into a modulated continuum representing the true complexity of human associations. ... We must evolve an architecture from the fabric of life itself, an equivalent of the complexity of our way of thought, of our passion for the natural world and our belief in the ability of man.”^66.

This criticism addresses itself precisely to the functionalist reductivism of the 1920s and to its elimination of cultural aspects, here described as “human associations” and “the fabric of life itself.” These aspects were considered as an intrinsic aspect of architecture by Team 10.

Once more, metaphor is being used as the substitutive operation to incorporate “vital” aspects into design. Two types of metaphor are used. The one, which accounts for urban form in general, resembles Le Corbusier’s use of geometry at an urban scale. The other, which accounts for the realization of ideas at a building scale, is itself conceived as a fundamental element of urban design.

The first metaphoric operation links two systems through the common element “life,” and thus relates the city to nature (a tree). Hence the plans for Golden Lane. The city is overlaid with the attributes of a tree and given qualities of growth, organicity, movement; at the level of form, the city is understood as a tree possessing a stem, branches, and leaves.

The second type of metaphoric operation articulates the relationship between design and life at the scale of the building and operates on the basis of a common function: circulation of people (street). In the proposal for Sheffield, the corridor is transformed through substitution into a street, carrying with it the urban codes which, when transferred to the building, give it “life.”

Despite the explicit intent of Team 10 to open the system of architecture to culture, however, the result does not, in the end, differ much from the reductive system they criticize. The type of substitution utilized—the recodification of architecture by means of yet another formal analogy—is fundamentally similar to that effected by Le Corbusier. The process by which the Smithsons assimilate “life” to design is described exclusively in socio-cultural terms, even though “nature” is invoked, while the form adopted is taken directly from nature, that is, from organic, physical life. The other systems to which architecture is supposed to be actively linked (in this case, life or nature) are, in this way, filtered and reduced through the metaphor of one system, that of architectural forms. Thus, there is little real difference between the street in the air and the open corridor; the symbolic functioning which would make an architecture “out of life itself ” is in fact absent. We may now see that metaphoric operations, rather than functioning to open the design system beyond its limits, in fact operate as filtering mechanisms which precisely define those limits.

It is paradoxical that the metaphor which allows for the interrelation of different codes is here used as a closing mechanism. Design is once again a sieve which allows the passage of certain meanings and not others, while the metaphor, which is used as a translating device from other codes to architecture, provides a mechanism by which ideology operates through design. In the infinite field of signifying possibilities, the metaphor defines, by a complex process of selection, the field of “the possible,” thus consolidating itself in different regions by means of a language or languages.

DESIGN/NON-DESIGN

There is, however, another possible way of stating the relationship between design and culture. Rather than seeing systems of culture from a point of view that imposes a hierarchical relationship in which architecture or design is dominant, we may posit a notion of the “non-designed” built environment—“social texts,” as it were, produced by a given culture.

The act of placing design (that is, both architecture and urban design) in relation to the rest of the built environment—the non-designed environment—immediately changes the level at which the problem is formulated. While in the work of Team 10 the problem is stated as internal to a single cultural system (architecture or urban design)—the relating of architecture to the city in such a way that the former acquires the “life” of the latter, here the signifying function of design is considered to relate to and, in relating, to oppose the rest of the built environment. It is regarded as a problem internal to culture, and thus to an entire set of cultural systems.

In these terms, architecture is no longer either implicitly or explicitly seen as the dominant system, but simply one of many cultural systems, each of which, including architecture, may be closed or “designed.” But it is the entire set of different cultural systems configurating the built environment, which we call non-design.

In the world of non-design, that no-man’s-land of the symbolic, and the scene of social struggle, internal analysis of single systems is revealed as inadequate and impossible to apply. Here there is no unique producer, no subject, nor is there an established rhetorical system within a defined institutional framework. Instead there is a complex system of intertextual relationships.

The opposition between design and non-design is fundamentally defined by three questions: first, the problem of institutionality; second, the problem of limits and specificity; and third, the problem of the subject. While the first establishes the relationship between design and non-design, the second establishes their respective types of articulation within culture (ideology), and the third establishes the processes of symbolization.

Design may be defined as a social practice that functions by a set of socially sanctioned rules and norms—whether implicit or explicit—and therefore is constituted as an institution. Its institutional character is manifested in the normative writings and written texts of architecture, which fix its meaning and, therefore, its reading. These texts insure the recording of the codes of design and guarantee their performance as filters and preservers of unity. They assure the homogeneity and closure of the system and of the ideological role it plays. The absence of a normative written discourse in non-design, on the other hand, precludes defining it as an institution and makes possible the inscription of sense in a free and highly undetermined way; we are here presented with an aleatory play of meaning. Thus, while design maintains its limits and its specificity, these defining aspects are lost in the semiotically heterogeneous text of non-design.^67.

Non-design is the articulation—as an explicit form—between different cultural systems. This phenomenon may be approached in two ways: as empirical fact—the actual existence of such systems found, for example, in the street, where architecture, painting, music, gestures, advertising, etc. coexist—and as a set of related codes. In the first instance, at the level of “texts,” each system remains closed in itself, presenting juxtaposed manifestations rather than their relationships. At the level of codes, on the other hand, it is possible to discern the mode of articulation between the various systems and, in this way, to define the cultural and ideological overdetermination of the built environment, or rather the process by which culture is woven into it.^68. The predisposition of non-design to openness implies permeable limits and an always fluctuating or changing specificity.

Finally, if design is the production of an historically determined individual subject, which marks the work, non-design is the product of a social subject, the same subject which produces ideology. It manifests itself in the delirious, the carnivalesque, the oneiric, which are by and large excluded or repressed in design.

To study the reality of non-design and its symbolic production in relation to culture, it is necessary to perform an operation of “cutting”—“cutting” and not “deciphering,” for while deciphering operates on “secret” marks and the possibility for discovering their full depth of meaning, cutting operates on a space of interrelations,^69. empty of meaning, in which codes substitute, exchange, replace, and represent each other, and in which history is seen as the form of a particular mode of symbolizing, determined by the double value of use and exchange of objects, and as a symbolic modus operandi which may be understood within the same logic of symbolic production and which is performed by the same social subject of ideology and the unconscious.^70.

The moment one object may be substituted for another beyond its “functional” use-value, it has a value added to it which is the value of exchange, and this value is nothing but symbolic. Our world of symbolic performances is comprised of a chain of such exchanges in meaning; that is how we operate within the realm of ideology. Non-design leaves this ideology in a “free-state,” while design hides it.

The mode of analysis for these two phenomena of design and non-design (at least from the first moment that the difference between them is recognized) must therefore vary.

READING. MISE-EN-SÉQUENCE

As a complex social text, a semiotically heterogeneous object in which many different signifying matters and codes intervene, non-design has a disposition to be open to a situation which we will call here a mise-en-séquence.

We propose here for non-design a productive reading, not as the reproduction of a unique or final sense, but as a way of retracing the mechanisms by which that sense was produced.^71. Productive reading corresponds to the expansive potential of non-design and permits access to the functioning of meaning as an intersection of codes. The object of analysis is not the “content,” but the conditions of a content, not the “full” sense of design but, on the contrary, the “empty” sense which informs all works.^72. Instead of reading by following a previously written text, the reading starts from a “signifier of departure,” not only toward an architectural text but toward other texts in culture, putting into play a force analogous to that of the unconscious, which also has the capacity to traverse and articulate different codes.

The metaphoric operation participates asymmetrically in both readings, design and non-design. While in design the metaphor is not only the point of departure but also the final point of the reading, in non-design the metaphoric and métonymie operations function similarly to dreams, as chains which permit access to meanings that have been repressed, thus acting as expansive forces. This expansive mechanism may be seen to be a device used for the purpose of criticism in the work of Piranesi. His opposition to the typological obsession of his time is an indication of his perception of the crisis of architecture and the consequent need for change and transformation. His Campo Marzio is a true architectural “explosion” that anticipates the destiny of our Western cities.^73. Piranesi's “explosive” vision comprises not just the architectural system per se but rather a system of relationships, of contiguity and substitution.

Non-design may also be seen as an explosive transformation of design. This kind of explosion implies in some way the dissolution of the limits of architecture, of the ideological limits which enclose different architectural practices.

In front of two drawings of Piranesi’s Carceri, one of the Carcere Oscura of 1793 from the series of the Opere Varie and the other on the Carceri Oscure from the Ivenzioni, the Russian filmmaker Eisenstein makes a reading which may be considered as an example of this type of analysis. Eisenstein applies a cinematographic reading to the first prison, his reading producing displacements with respect to the limits imposed by pictorial and architectural codes, thereby making it “explode” in a kind of cinematographic sequence.^74. This is the starting point of a reading that travels across literary, political, musical, and historical codes, multiplying in this way perceptions which are potential in the Piranesian work. A proof of this potential lies in Eisenstein’s reading of Piranesi's second engraving, done eighteen years later, in which Eisenstein finds that the second is actually an explosion of the first prison, done by Piranesi himself.^75. It should be noted that Eisenstein is here dealing with a closed cultural system, such as architecture or painting. What Eisenstein takes, however, is not just any closed work from these fields but rather the work of someone like Piranesi, who poses the problem of the explosion in form (or form as explosion) in his Carceri, or in his Campo Marzio, which is a delirium of typological chaining. Although this Piranesian strategy touches problems specific to architecture, it also comes very close to the problem of the explosion of sense in architecture, to the problem of meaning as signifying chaining. In creating this extreme situation, Piranesi is implicitly assessing the problem of the limits of architecture as a “language,” that is, as a closed system.

“One evening, half asleep on a banquette in a bar, just for fun I tried to enumerate all the languages within earshot: music, conversations, the sounds of chairs, glasses, a whole stereophony of which a square in Tangiers (as described by Severo Sarduy) is the exemplary site. That too spoke within me, and this so-called ’interior’ speech was very like the noise of the square, like that amassing of minor voices coming to me from the outside: I myself was a public square, a sook; through me passed words, tiny syntagms, bits of formulae, and no sentence formed, as though that were the law of such a language. This speech, at once very cultural and very savage, was above all lexical, sporadic; it set up in me, through its apparent flow, a definitive discontinuity: this non-sentence was in no way something that could not have acceded to the sentence, that might have been before the sentence; it was: what is eternally, splendidly, outside the sentence.”^76.

The built environment as the object of reading is not “seen” as a closed, simple unity but as a set of fragments, “units of readings.” Each of these units may be replaced by others; each part may be taken for the whole. The dimension of the built environment, empirically determined, depends upon the density of meanings, the “semantic volume.”

Since these fragments appear as an articulation of different texts belonging to various cultural systems—e.g., film, art, literature—it is possible to read them by starting from any of these systems, and not necessarily from design.

Certain types of configurations, like public places (streets, plazas, cafes, airports), are ideal “fragments of readings,” not only for their “semantic volume,” but also for the complexity they reveal as to the signifying mechanisms in non-design. They may be characterized as signifying “nodes,” where multiple codes and physical matter are articulated, where design and non-design overlap, and where history and the present are juxtaposed.^77.

The reading that can be produced by these places is not a linear discourse but an infinite and spatialized text in which those levels of reading, organized along various codes, such as theater, film, fashion, politics, gesture, are combined and articulated. The reading example we choose to present below is in itself metaphorical. It is the metaphor of architecture as theater. It is not a specific detailed analysis, but rather it exemplifies the mechanisms of chains and shifters.

A metaphor begins to function by articulating the referential codes in relation to other codes by means of replacing the referential codes in the signifier of departure with another code. In this way, a chain linking the codes is developed. Once the intersemiotic metaphor, such as that between architecture and theater, is produced and a possible level of reading is established, the chain of signifiers along the codes and subcodes of that cultural system is organized by “natural association”—that is, metonymically.

Signifiers appear and disappear, sliding through other texts in a play that moves along the codes of, for instance, the theater (i.e., scenic, gestural, decorative, acting, textual, verbal, etc.) in an intertextual network. This play continues until some signifier becomes another departure signifier, opening the network toward new chains through what we have called the mise-en-séquence, thus starting other readings from other cultural systems like film, fashion, etc. These signifiers which open to other systems may be called shifters. ^78.

SHIFTERS

Such a reading presents a symbolic structure of a “decondensed” kind. Here, by decondensation we refer to an operation which is the reverse of that in the elaboration of dreams. Condensation and displacement are the two basic operations in the work of elaboration of dreams. By them, the passage is produced from the latent level to the manifest level of the dream. These two operations of condensation and displacement are two ways of displacing meanings, or of overdetermining, or giving more than one meaning to, some elements; they are produced precisely by means of the two operations already discussed, namely metaphor and metonymy. The metaphor corresponds to condensation, and metonymy to displacement.^79. In this way, it is possible to see the relationship between ideology (cultural codes) and subject (of ideology and of the unconscious) in the logic of symbolic production in the environment as determined by a particular mode of production.

Some signifier fragments function as “condensers” from which decondensation is possible through a network of meanings. These will be called “shifters.” A set of readings could be regarded as a musical staff in which various signifiers are situated in a polyphonic organization with each voice at a different level of reading. Certain of these signifiers organize several different readings and allow for the intercrossing of codes and for the shifting from one to the next. These are the shifters; they are part of a process of exchange of codes. They are the conditions of the probability of producing different readings; they are structures of transition, the organizers of symbolic space. These connective, condensing structures are the key to the understanding of the complexity of the built environment as an infinite text. They are not concerned with signification but with the linking of signifiers. They are the key to an intertext where meanings are displaced, thereby forming a network in which the subject of the reading, the laws of the unconscious, and the historico-cultural determinants are articulated. The importance of this notion of shifter is that it accounts for the process of configuration and for the dynamic aspect of a configuration, rather than for objects and functions. It accounts for the symbolic aspect of exchange. It provides an insight into the problem of the mode of operation of ideology within the built world. It allows us to enter into a mechanism of production of sense that corresponds to an ideology of exchange.

If the system of architecture and of design, even when we play with it, is always closed within a game of commentaries of language—a meta-lingual game—it is interesting to speculate on the outcome of a similar “game” of non-design, a game of the built world. For non-design is a non-language, and by comparison with a language, it is madness since it is outside language, and thus outside society. This non-language, this non-sense constitutes an explosion of the established language in relation to a sense already established (by conventions and repressive rules). It is symbolic of the built world outside the rules of design and their internal “linguistic” games. It permits us finally to understand another logic which informs the significance of building.