IRENE PORTIS WINNER AND THOMAS G. WINNER THE SEMIOTICS OF CULTURAL TEXTS I. INTRODUCTION

Transcription

1 Semiotica 18:2, pp , Mouton Publishers, IRENE PORTIS WINNER AND THOMAS G. WINNER THE SEMIOTICS OF CULTURAL TEXTS I. INTRODUCTION The search for unifying approaches which interrelate, but do not reduce, the complexities and levels of the various phenomena which compose the vast domain called culture has motivated much of the research in the area of semiotics of culture. This timely and ambitious topic, which by its nature is fraught with difficulties and complexities of interpretation, is the theme of various recent studies, one of the most important of which is entitled Structures of Texts and Semiotics of Culture, edited by Jan van der Eng and Mojmir Grygar (1973). This work, composed of various articles assembled at the occasion of the Seventh International Congress of Slavists, opens with the "Theses on the Semiotic Study of Culture" drawn up by five Russian scholars who have led semiotic studies in that nation dedicating themselves to its pursuit from the most concrete level of specific applications, from verbal and nonverbal behavior (including mythology, folk art, high arts, the cinema, and the most various cultural systems) to the most abstract considerations of theory, methodology and metasystems encompassing theories of signs, texts, and communication in general. The growing interest in the West in this broad subject is evidenced by the number of theoretical studies as well as collections of articles that have been translated into English in the last five years, and works written in English originally, that assess the interdisciplinary potential of semiotics. (Greimas, Jakobson et al, eds., 1970; Kristeva et αϊ, eds., 1970; Eimermacher, 1971a; 1971b; Meletinskij and Segal, 1971; Rey-Debove, ed., 1973; LP. Winner, 1973; T.G. Winner, 1973a; 1973b; Segal, 1974; Sebeok, 1974; 1975; Sebeok, ed., 1975; Eco, 1975; 1976; Jakobson, 1952; 1960; 1965; 1970a; 1975a.) The semiotic view of culture assumes the multiplicity, diversity, stratification and intercorrelation of sign systems which are investigated on various levels from that of technology to social, economic, and expressive behavior

2 102 to ideologies. Indeed it encompasses all communicative behavior that is cultural (meaningful, shared, organized, and dynamic). Following this approach, the synchronic and diachronic aspects of semiotic systems are viewed to be inseparably related and to be appropriate subjects for investigation. While the science of signs has an ancient history, going back to the Greeks, and its formal extension from linguistics to culture in general has long been implied and was specifically called for by Saussure in the early twentieth century (1966: 16, 17), over the last decades the work of scholars in the most varied disciplines, in Eastern and Western Europe and in the United States, who have considered signs from many approaches, have laid the basis for significant progress toward the construction of a broad theoretical and unifying point of view. But most fundamental to modern semiotics were the theories of the Prague Linguistic Circle and the related early Russian structuralists, as they evolved under the leadership of Roman Jakobson and Jan Mukarovsky, departing from, and extending, Saussurian insights, leading to the extremely fruitful application of semiotics to aesthetic and other cultural systems. A pioneering work in this direction, of over thirty years ago, was Bogatyrev's study of folk costumes of Moravian Slovakia (1971). By the 1940s Jakobson brought the semiotics of Peirce to bear upon the developing semiotic point of view, thereby fundamentally broadening approaches to typologies, as well as to the dynamics of sign systems, particularly in the area of pragmatics (cf. Peirce, ). Moreover, the wartime contact between Jakobson and Levi- Strauss stimulated both these seminal thinkers, as is evidenced by their fundamental postwar studies in various aspects of cultural semiotics, demonstrating important mutual influences. As the collection of primary consideration here illustrates, extremely significant work in semiotics is being carried out in Eastern Europe, principally in the Soviet Union, where the group of scholars concentrated in the University of Tartu and the Institute of Slavic Studies at the Soviet Academy of Sciences in Moscow are collaborating in their studies of the theory and methods of a semiotics of culture. Starting from the semiotics of the various art systems, the Moscow-Tartu group has devoted increasing attention to the semiotics of other cultural systems and their mutual translatability. Some of the findings of the "Summer Schools on Sign Systems", from 1964 to 1972 under the leadership of Jurij Lotman at the University of Tartu, have appeared under the titles Researches on Sign Systems (Trudy po znakovym sisteman) and Semeiotike. There is also an English language summary of the Summer School Session of 1970 (O.G. Revzina, 1972). Furthermore a number of the findings have been published in English in Semiotica and in Soviet Studies in Literature (Fall, 1974 and Baran, ed., 1975). Finally, an important work by Lotman in translation has appeared

3 103 (Lotman, 1976). An eminently welcome, and fundamentally important, addition to the growing number of original texts of this group becoming available in the West is The Structure of Texts and the Semiotics of Culture, edited by Jan van der Eng and Mojmir Grygar, particularly since it opens with an English translation of the "Theses", a compact summary of the basic principles of semiotics advanced by the Tartu-Moscow group. The Russian version of the "Theses" was published simultaneously in a work edited by Mayenowa (1973). The telescoped conclusions of the "Theses", encompassing the considerable work in this field (we mention only a few of the many works) by E.G. Lotman (1970b; 1973; 1974a); Pjatigorskij (1962; 1974); Uspenskij (1972), etc., Ivanov (1970; 1971) and Toporov (1971), introduces the individual articles in Russian, English and.french which discuss particular aspects of semiotics. The approach to semiotics of culture by this group leads toward a conjunction of the linguistic and aesthetic theories of the Prague school and the Russian structuralists, the traditional anthropological view of culture as patterned, communicated, learned behavior composing an inherited tradition, and the more recent view of culture as information. In 1970, Lotman described culture as a "semiotic mechanism for the output (yyrabotkd) and storage of information" (1970b:2) and "a historically evolved bundle (pucok) of semiotic systems (languages) which can be composed into a single hierarchy (supralanguage) which can also be a symbiosis of independent systems" (1970b:8). The analogy is to the memory of mankind or of some narrower collective (nation, class, etc.), memory implying the capacity of systems for storage and accumulation of information. Furthermore, the attempt is to prescribe culture 'types' as specific 'languages' (1970b:12). The broadening of the linguistic concept of text is fundamental to these departures, since it is the basic significant unit of cultural semiotic systems. In an early statement by Pjatigorskij (1962), a text is defined as a variety of signals composing a delimited and autonomous whole. Such a communication is characterized in three spheres: (a) In the syntactic sphere it must be spatially (optically acoustically, or in some other fashion) fixed so that it is intuitively felt as distinct from a nontext. (b) In the pragmatic sphere, its spatial fixation is not accidental, but the necessary means of conscious transmission of communication by its author or other individuals. Thus the text has an inner structure. (c) In the sphere of semantics a text must be understandable, i.e., it must not contain unsurmountable difficulties hindering its comprehension (Pjatigorskij, 1962:79). Since Lotman held that all cultural semiotic systems were to be seen as secondary modelling systems shaped 'along the lines' ('po tipu') of language

4 104 (Lotman, 1970a:16) the linguistic concept of texts began to be applied by analogy to all cultural behavior. However, since many nonlinguistic communications seem to depart radically from the structure of language, the unqualified concept of secondary modelling systems has been questioned, modified and altered by the members of the Lotman group as the "Theses" suggest, and as subsequent publications of this group clearly indicate, which will be commented upon later in the discussion. II. THE THESES We now turn to a discussion of the "Theses" which are widely encompassing, penetrating and cryptic. Inevitably this broad statement touches on more areas than can be considered in one essay, and leaves uncertainties requiring interpretation, only some of which can be reviewed here. We divide our discussion of the "Theses" into three sections: 1. a preliminary delimitation of the domain of the "semiotics of culture"; 2. a consideration of the concept of 'text'; and its bearing upon some theoretical controversies in the West relating to statics and dynamics as they are held to apply to the abstract and the concrete aspects of cultural systems; 3. a discussion of the concept of culture, understood as operating on three levels. 1. Semiotics of culture The "Theses" open with the following statement: In the study of culture the initial premise is that all human activity concerned with the processing, exchange, and storage of information possesses a certain unity. Individual: sign systems, though they presuppose immanently organized structures, function only in unity, supported by one another. None of the sign systems possesses a mechanism which would enable it to function culturally in isolation. Hence it follows that, together with an approach which permits us to construct a series of relatively autonomous sciences of the semiotic cycle, we shall admit another approach, according to which all of them examine particular aspects of the semiotics of culture, of the study of the functional correlation of different sign systems (1.0.0).* * In the discussion of the "Theses", numerals in parentheses will refer to the outline of the "Theses" and not to pages.

5 105 Among the various assumptions underlying and extending this view of culture, we point to five groups. (a) Culture is a unifying concept transcending local boundaries. Thus a radically diverse, emic approach is not the final one. Deeper analysis will lead to more basic components which, on higher levels, are transformed as they are affected by varying contexts. The search for specifics and correlated universals is sought in three interconnected realms which may somewhat loosely be described as: syntactics (inner structural organization, relations within and between structures), semantics (manifested for example by binary oppositions such as culture/nature, upper/lower etc., and by fundamental symbols), and pragmatics (for example the oppositions in point of view of inner and outer interpreters, the creation and perception of sign systems, and the dynamics of sign systems). (b) The concept of culture bears a direct relation to our knowledge of mental processes which integrate the diverse facts perceived into communicative models. "If we regard the collective as a more complexly organized individual, culture may be understood by analogy with the individual mechanism for the storage and processing of information" (6.0.0). (c) Natural language is the primary and universal model, in relation to which all other communicative modelling systems are perceived as "secondary modelling systems", the arrangement being hierarchical.... particular importance is attached to questions of the hierarchical structure of the languages of culture... (1.0.0) (underlining supplied). As a system of systems based in the final analysis on a natural language (this is implied in the term 'secondary modelling systems', which are contrasted with the 'primary system', that is to say the natural language [I.P.W. T.G.W.] ), culture may be regarded as a hierarchy of semiotic systems correlated in pairs, the correlation between them being to a considerable extent realized through correlation with the system of the natural language (6.1.3) (underlining supplied). However, as will be noted later, the phrase "to a considerable extent" reflects an area of unclarity or diversity of interpretation, suggesting that other "primary models" beyond the linguistic ones may also be considered. (d) Culture is a dynamic system that must be understood both from the inner and the outer points of view. From the inner point of view, culture is linked to its opposition, non-culture. This opposition is also conceived of as information and organization versus entropy and chaos, as culture versus

6 106 nature, and, in more formal terms, as inclusion versus exclusion. From the inner point of view, culture does not need its outer, chaotic, agent, but can be understood immanently (1,1.0, 1.1.1, 1.1.2) and may appear as an immobile, synchronically balanced mechanism (1.3.1). "The opposition 'culture/extracultural space' is the minimal unit mechanism of culture on any given level... in the center there is situated a certain normal 'we' " (1.2.4). As opposed to the above consideration of the inner view, from the outer point of view, culture and non-culture appear as two interrelated spheres. "The mechanism of culture is a system which transforms the outer sphere into the inner one, disorganization into organization... entropy into information". Culture moves from one sphere to another. It needs its outer sphere which it continually both destroys and creates (1.2.0). From the outer point of view, that which from the inner point of view would be seen as chaos may appear to be a different form of organization (1.2.2). And from the outer point of view culture will not represent an "immobile, synchronically balanced mechanism, but a dichotomous system the 'work' of which will be realized as the aggression of regularity against the sphere of the unregulated and, in the opposite direction, an intrusion of the unregulated into the sphere of organization". The incorporation into the cultural sphere of outside texts sometimes acts as a powerful stimulant for cultural development (1.3.1). Thus culture is constructed not only of a hierarchy of semiotic systems but also of a multi-layered arrangement of the extracultural sphere surrounding it. The problem of the limits of a particular cultural system leads to the next assumptions, regarding specific culture types, culture boundaries and comparisons of cultures (2.0.0). (e) The inner structure, or the composition and correlation of particular semiotic subsystems, determines the type of culture. However, the boundaries of a culture, or of cultures, may only be determined in terms of specific contexts (genetic, areal, and others). Thus, from the changing points of view of narrower or broader contexts, several cultures may form smaller or greater functional or structural unities which are correspondingly contrasted to the outer sphere of culture. A comparative approach requires the definition of inner paradigms of culture (2.1.0). The complex question of cultural boundaries and culture types receives only brief direct consideration in the "Theses", but many aspects of this complex problem are implied by other writings of the authors. Elsewhere Uspenskij has emphasized the general semiotic importance of the problem of frames, or borders, of an artistic text. Considering Lotman's view that all cultures may be thought of as having a beginning and an end, Uspenskij notes that there are cultures where the beginning is marked, others where the end is marked, and finally there are cyclical culture systems, all of which

7 107 help shape the general world view of the whole culture, as well as particular cultural texts. For example, he cites special procedures for entering Russian Orthodox church indicating "the border between the special semiotic world of the church and the daily world" (Uspenskij, 1972: 14-15). Similarly Leach (1966:134) has commented upon a typical characteristic of primitive culture, the Durkheimian ritualization of boundaries surrounding the sacred and the profane, that delimit units of both cultural space and time. The general phenomenon of the perception of boundaries in archaic societies is understood by Lotman and Uspenskij as an aspect of "mythological consciousness" (which is discussed in our Section 2.4 and elsewhere), where the perception of space is limited and bounded even though myth may deal with cosmic dimensions. For example, a frequent theme in a mythological text is the hero's ability to cross closed space into the outer unbounded world (Lotman and Uspenskij, 1975:24). The problem of cultural typologies, as briefly discussed in the "Theses", requires considerable clarification, particularly since the problem of "culture" versus "cultures" is never clearly delineated. The authors move from discussions of "culture" to phrases such as "a culture", "cultures", "type of culture" and "culture complexes". Throughout the "Theses" various classificatory criteria are suggested for grouping cultures into types, or elements into complexes, which may be placed into two sets, synchronic and diachronic: (a) Those of a synchronic order include: similarities of inner structure (2.0.0); areal similarities (2.1.0); common types of texts (as discrete vs. nondiscrete); orientations of cultures towards the speaker (3.2.1) or toward writing, oral speech, the word, and the picture (9.0.3); and the degree of unity of the entire system and awareness of this unity (9.1.5). (b) Those of a diachronic order include: genetic criteria (2.1.0); types of texts which may represent stages of development (3.2.1); and evolutionary criteria where, for example, more advanced systems are those in which group behavior is regulated by the memory of real history, as opposed to concepts of cyclical change (6.0.2). (f) The last assumption we mention, underlying and extending the semiotic concept of culture, appears to be the crucial one, that is the notion that "texts" are the primary elements or basic significant units of culture (3.0.0). The implication of this key concept underlies the entire argument of the "Theses" and it is to this problem that we direct our remarks in the following section. 2. The concept of texts As noted earlier, in the early sixties A.M. Pjatigorskij applied the concept of "text" to verbal communications of all kinds. The expansion of the term

8 108 "text" to culture in general, including nonverbal systems, followed from the assumptions that the linguistic system is the most fundamental of all semiotic systems and that all other systems are "secondary modelling systems". Yet it is also clear that the notion of texts may be interpreted as serving to free semiotics from the dominance of the linguistic model over all others. In this sense the term "text" must cover various subtypes with radically differing structures in spite of certain very general invariants. How the concepts of the "Theses" address themselves to this urgent question is of particular interest. Some lack of consistency seems to reflect the reevaluations being undertaken by various scholars. We divide our discussion of the concept "text" into four main areas: 2.1 the relation of text to culture; 2.2 typologies of texts and their general characteristics; 2.3 reconstruction of texts; 2.4 the relation of varying interpretations of the concept of text in East and West to the general controversy being argued in the West over 'explanation' as opposed to 'description', or 'statics' as opposed to 'dynamics'. 2.1 The relation of text to culture From the point of view of texts, culture is understood on three levels, namely "as a hierarchy of semiotic systems composed of texts, as the sum of the texts and the set of functions correlated with them, or as a certain mechanism which generates these texts" (6.0.0). The implications of these views for definitions of culture will be discussed in section 3. We confine ourselves here to a consideration of the problem of what is meant by a "culture text". Texts in general are understood to be particular kinds of messages or groups of messages. But it is not always clear whether, in specific instances in the "Theses", the terms "text" and "message" refer only to verbal communications or whether they also refer to nonverbal communications, and 'whether they imply only "cultural" texts and messages or also other texts and messages. It would have been preferable if these distinctions had been more clearly made and if those texts and messages discussed in sections 3.1.0, and of the "Theses" which are not considered to be "cultural" would have been unmistakably differentiated terminologically from "cultural" texts and messages. We feel that in most cases the terms 'message' and 'text', when used without qualifications, denote both verbal and nonverbal communications that meet certain cultural criteria, thus claiming cultural status. For example, it is stated that the concept of culture text is applied not only to verbal messages, "but also to any carrier of integral ('textual') meaning" including ceremonies, works of art, as well as 'genres' such as 'prayer', law', 'novel', etc. (3.1.0). Accordingly, cultural texts, composed of cultural messages relate to all kinds of communication applicable

9 109 to any semiotic system (4.1.0). The various criteria required for a text to be a cultural one appear to include the following: the carrying of integral meaning, the fulfilment of a common function (3.1.0); the ascription of value and the preservation of text (4.0.0); the organization of the texts into some kind of genre or type (3.1.0); and the construction of texts according to definite generative rules (4.0.1). In considering this last point it should be noted that the phrase "generative rules" appears in the original Russian but is lost in an incorrect translation. The following phrase is correctly translated from the Russian edition, "By text (here "text" appears to mean cultural text. I.P.W., T.G.W.) we imply only a message which is constructed within the given culture according to definite generative rules" (underlining supplied), (po opredelennym porozdajuscim pravilam) (Mayenowa, ed., 1973: 18). However, in the English edition this phrase reads, "By text we imply only a message which performs within the given culture a textual function" (Van der Eng and Grygar, eds., 1973:12). Only with the correct translation in mind can the next statement be understood, which holds that a message may be a text in one culture but not necessarily in another (4.0.1). Presumably what is meant is that a message is a cultural text in a particular culture, but following the analogy of the linguistic concept of grammaticalness a cultural text must adhere to certain systematic rules or lose its "cultural status". But all these points relating to the cultural status of messages and texts remain extremely general and need to be extended and clarified. Based upon the above view of texts, texts are compared to culture in the following three ways: (a) If we distinguish between the position of the investigator of culture and that of its carrier, texts like cultures can only be understood from two points of view, the inner and the outer. From the inner point of view, the text has integral meaning, while from the outer point of view, from that of the investigator of culture, it is the carrier of integral function, (b) Like culture, texts can only be determined in terms of specific contexts. Thus, for example, the same "message" may appear as part of a text, or a set of texts, depending upon the framework utilized (3.0.0). (c) Text is opposed to nontext or antitext as culture is opposed to nonculture or anticulture. Thus while those texts which the culture does not preserve are nontexts, those texts which the culture destroys are antitexts (4.0.0). 2.2 Typologies of texts and their general characteristics Discrete versus nondiscrete texts. One of the most important problems discussed in the "Theses" is the question of typologies of texts, distinguished by the nature of their internal structure. All texts appear to be made up of signs (with certain possible exceptions to be noted) but the nature of these signs differ. In a discussion presenting some disturbing

10 110 terminological unclarities, there are noted the following oppositions: On the one hand there is the primary text composed of an integral sign which "represents a whole and is segmented not into separate signs but into distinctive features" (3.2.1), (which is variously referred to as the nondiscrete text, and the continuous text). Its characteristics are typically, but not exclusively, exemplified by the visual text. On the other hand there is the secondary text derived from a sequence of individual signs (variously called the discrete text or the discontinuous text) the characteristics of which are typically, but not exclusively, exemplified by the verbal text. (The use of the terms 'primary' and 'secondary' texts are not to be confused with 'primary' and 'secondary' modelling systems. "Primary" in relation to texts appears to mean unsegmented into separate signs and does not relate to the characteristics of the natural language.) Although it is stated that the primary text is segmented into distinctive features, nevertheless it appears that some primary texts are not even characterized by this form of segmentation, a problem which needs clarification. Thus it is held that the primary text is demonstrated by two contrasting examples with certain basic similarities. (a) The first example is composed of the most modern communication systems such as television and cinema as well as early systems of painting, sculpture and dancing. These are primary texts where the sign appears only as a secondary notion definable in terms of the text. The implication appears to be that in such continuous texts the discrete is deemphasized even at the level of distinctive features. (However, as is noted subsequently, Ivanov [1973a:122, ] points to just such systems and particularly to the cinema as being based upon discontinuity and the discrete.) Following the argument of the "Theses", it is held that in television the basic unit is the elementary life situation which, before televising, is irresolvable into smaller elements. Furthermore, it is held that segmentation of continuous texts, such as painting or the cinema, into discrete distinctive features such as upper/lower, left/right, dark/light, black/white may in fact represent archaizing tendencies which "impose" on the continuous texts binary symbolic classifications of the mythological or ritual type (although the reservation is made that such features may be archetypal even during the creation and perception of continuous texts) (3.2.1). The apparent conflict between these two statements is not resolved. (b) The second example relating to primary texts is composed of systems of mathematical logic and formal grammars, in which the "language is understood as a certain set of texts". The text may be represented by a chain of discrete symbols such as elements of an initial alphabet of a set. In other words, in the second example orientation remains toward the discrete "which was characteristic of the linguistics of the first half of the century",

11 Ill as opposed to contemporary semiotic theory which is concerned with "the continuous (indiscrete) text as primary datum" (3.2.1). (Yet, as is discussed later, Ivanov [1973a] points to the contemporariness in semiotic theory of the orientation toward the discrete, and not toward the continuous, as the primary datum of perception.) The opposition between primary and secondary texts receives various modifications. Thus both tendencies may be present in one text. For example, there are discrete signs within a continuous text, such as a film or a painting, but they are reinterpreted in the context of the whole. Often such discrete signs are present as 'raw material' from earlier systems (examples are the Cross, the World Tree, and others) (3.2.1). It is suggested that predominance of discrete or nondiscrete texts in particular cultures may be associated with particular cultural stages of development, or secondly that the two types of texts may coexist. (Here again the conflict between these two statements should be resolved.) "The tension between them [discrete and nondiscrete, I.P.W., T.G.W.] (for example between verbal and visual texts) constitutes one of the most permanent mechanisms of culture" (3.2.1). The problems presented by the opposition of discrete and nondiscrete texts, woven throughout the "Theses" and in other publications by these authors, hardly appear to be resolved. Among the various issues, we note two implied by the above discussion: one relates to the presence of symbols within the primary or nondiscrete text, and the other relates to the resolution of the dichotomy between discrete and nondiscrete structures. Turning first to the question of the symbol, we have noted that the "Theses" hold that a primary text is typically composed of an integral sign, although the sign may be a secondary notion (3.1.2), while later the "Theses" suggest that a continuous text, such as early Slavic depictions in vegetative designs, may be based on one symbol, an 'extrasystemic sign' composing the entire code (5.2.2). However, elsewhere Lotman has taken this argument farther, suggesting that some continuous texts (using examples from the arts) may not even be reducible to one symbol since the text in its entirety is the bearer of the message. "And if we introduce discreteness into it, singling out symbol-like structural elements, this must be regarded as the result of the habit of seeing in verbal contact the basic or even sole form of communicative contact and seeing the depictive and verbal texts as similar" (Lotman, 1974b:77). Thus the question of whether a continuous text necessarily implies a sign or symbol-like structure of some kind does not seem to be resolved. Secondly, various resolutions to the apparent dichotomy between discrete and nondiscrete have been suggested. In a continuing attempt to distinguish and to interrelate discrete and nondiscrete structures, Lotman

12 112 has discussed the reconciliation of spatial (nondiscrete) and temporal (discrete) textual systems within the framework of two forms of narration within the arts (1974b). The effort to overcome this dichotomy has earlier roots, as Ivanov has shown, quoting Florenskij who wrote that "there is no necessary boundary between the visual arts which are quite erroneously supposed to be the art of space, and music in its various forms, which is said to be pure time" (Ivanov, 1973a:135, quoting Florenskij, n.d., para 71). As Lotman asserts, Levi-Strauss has also offered resolutions of spatial and temporal texts, showing that the "elaboration of myth and of the work of music over time is a mechanism to overcome the irreversibly linear direction of real time" (1974b:76). Similarly, Jakobson has discussed the interrelation of continuity (or space) and discontinuity (or time) in the arts (1964; 1967), making it clear that continuity (simultaneity) is a part of the socalled discrete, sequential arts. In music, simultaneity is represented by the chord, and in the verbal arts by metaphor and all types of parallelisms, though discrete qualities dominate in these two arts. On the other hand, while the visual arts are usually considered as continuous and nondiscrete, they also partake of discrete elements, though in a subsidiary role. In Lotman's attempted synthesis of these two types of texts, the two forms of narration distinguished are the following: one is associated with discrete verbal texts, built upon the addition of new elements such as words and symbols, which are joined into chains. The second form of narration is associated with nondiscrete primary texts, frequently, though apparently not necessarily, of the iconic type, and is identified with transformation which is based upon the internal reorganization of elements. The kaleidoscope, where the moving of colored pieces of glass form an infinite variety of symmetrical figures, models such a narration (1974b:76, 77). Just as discrete and nondiscrete tendencies may be a part of one text, as Lotman points out, these two models of narration are not mutually exclusive. "Various aspects of the two possible semantic models of narration are realized differently in each of the real forms of artistic narrative (1974b: 79)." Both the literary and iconic arts strive to gain freedom relative to the dominant principles of their respective narrational types. Thus "verbal narration becomes a revolutionary element in inherently iconic narration and vice versa" (1974b:80). And here we might extend the application of the view of the dynamic interrelation of these two forms of narration (and their underlying principles) beyond the arts to communicative aspects of other sign systems, which it seems is clearly implied. How distinct Lotman's two models are is not always clear, however, and the difficulties are increased by the differing use of the important terms "discrete" and "continuous" by Ivanov for whom discreteness is the primary form of perception which underlies continuity. In his stimulating essay on

13 113 time, Ivanov discusses the problem of inner psychological time, pointing out that the basic problem for the contemporary artist has become the exploration of the psychological reality of discrete forms (1973a:134). The modern view of time, where immediate perception is no longer held to correspond to physical phenomena (1973a:130) and events no longer necessarily correspond to time chronology (1973a:133), is shown to parallel archaic concepts of time. For example, in archaic cultures "the past is usually conceived of as a repetition of several qualitatively different cycles or else is assigned to a particular mythological tinie to which events not only of myths but also of epos are assigned" (1973a:117). Folk genres add fantasy time to mythological and epical time. All of these categories of time are separated from the time of the narrator by a certain interval, and cannot be measured, all of which is opposed to historical time of the West European type based on development, direction in time, and accumulation of information of the past which functions to correct programs of behavior (1973a: 121). Ivanov points out (1973a:134) that Florenskij has shown in 1922 that, reacting to nineteenth century evolutionism and the associated concepts of continuous development and dissolution of form, the twentieth century preoccupation has been with the exploration of the discontinuous character of phenomena as well as with the discontinuous nature of its form. This interest has, however, been united with that of the temporal nature of any process of perception, and most particularly of aesthetic perception (1973a:135). Correspondingly, modern approaches to time have taken up again the question of pre-eternal time, of deformations of time, shifting of time back and forth, stopping of time, and dismembering it (1973a:112, 123, ). Indeed, Florenskij, in his Analiz prostramtvennosti, has shown that in twentieth-century culture the role of the discrete means that the organization of time is always achieved by dismemberment or discontinuity (Florenskij, n.d., paragraph 72, in Ivanov, 1973a:136). An artistically organized, book is "the simplest example of the unity achieved during the visual perception of discontinuous images" (Florenskij: 'n.d., paragraph 73. cited in Ivanov, 1973a:138). Other examples are the Chinese scroll, frescoes and avantgarde arts. "The presence of the temporal and discrete structure not only in the sequence of images, but also in each individual image", underlies modern art styles such as Cubism and the cinema, where the discrete principle of construction is brought into the foreground. All of these tendencies in twentieth-century art lead toward "revelations of discontinuous elements in temporal structures" (1973a:141). As Ivanov poses the contemporary argument, the questions are: is the cinema, where the impression of continuity arises from a succession of discrete forms, a model of our perception of the world, thus making inner psychological time discrete? Or is inner

14 114 psychological time continuous? (1973a: ). Ivanov gives greatest importance to the underlying discontinuity of time which he traces not only to research on biological factors marking time intervals, but also to the limitations of time imposed on all living organisms which face death, the latter also accounting for the belief in eternity or mythological time (1973a: 146). Thus where the "Theses" seem to be discussing surface impressions of continuity (an example being the cinema), as well as surface impressions of discontinuity (examples being the verbal message), and the resolution of these differences, Ivanov seems to be discussing the underlying discrete basis of psychological perception which is being explored by dismemberment of surface impressions, an example being the cinema. It is clearly important that the varying uses of these terms be clarified. Levi-Strauss has taken the argument of the discreteness of inner perception as opposed to the apparent flux of empirical reality considerably farther, suggesting that "what the brain perceives in language is not by the nature of sounds, but of distinctive features", and "Current research on the visual system suggests a similar conclusion. The eye does not merely photograph the outside world. It rather encodes its formal characteristics". Thus the discrete emic level of structural binary oppositions is the real one and the etic level remains the artifact" (1973:20-21) Other typological criteria for texts. Other characteristics which determine type of texts, in addition to the factors relating to discreteness and nondiscreteness, include the orientation of texts, and the dynamics of texts. Thus texts and cultures as a whole may be oriented towards the sender or towards the teceiver (3.2.2). In cultures oriented towards the receiver, the most highly valued texts are held to be widely intelligible ones (prose, documentary film), and truth value is important. On the other hand in cultures oriented towards the sender, the most highly valued texts are closed and unintelligible ones. In such esoteric cultures, prophetic and priestly texts, glossolalia, and poetry are valued. Whether such a dichotomy distinguishes culture types in general, or simply opposing trends within the same culture, needs to be clarified. In relation to these distinctions, the age of enlightenment and the later nineteenth-century world are held to be examples of the first type whereas the romantic age, and the first two decades of the twentieth century, represent the second type. Types of texts are also distinguished by their relative age, a complex and relative question since "... in the real existence of cultures, texts, transmitted by a given tradition or introduced from the outside always functions side by side with new texts". Since different levels of culture change at different speeds the synchronic state of culture may include its diachrony

15 115 (4.1.2). The whole problem of the close relation of synchrony to diachrony is expanded upon in the sections of the "Theses" (5.0.0 and 5.0.1) which deal with the problem of changes in texts. It is stated that "the place of the text in textual space is defined as the sum total of potential texts" (5.0.0). In the process of culture change, the same work may sometimes gain, and sometimes lose the ability to appear as a text. It is concluded that a broad typological approach to texts removes the absoluteness of the opposition of synchrony and diachrony (5.1.0). (The resolution of this dichotomy follows from the early "Theses" of Jakobson and Tynjanovin 1928.) Furthermore, as Ivanov notes, Bakhtin's early concept of genre memory, acted to eliminate the opposition between historical and synchronic poetics (Ivanov, 1975: 19 l;bakhtin, 1963:142). 2.3 Reconstruction and recoding of texts To a certain extent the sections on the reconstruction and recoding of texts may be considered to be the heart of the "Theses" since they involve all aspects of texts: semantics, syntactics and pragmatics. The problem of the historical reconstruction of texts is extended from linguistics to the entire field of reconstruction of cultural traditions (folklore, mythology, ritual, music, dress, ornament, lifestyle, and others [5.2.0]). Reconstruction involves the sender's intention, the text itself, and the receiver's interpretation. It may be assumed that the statement, "To a certain extent, every reading of a poetic manuscript is a reconstruction of the creative process and a successive removal of superimposed levels" (5.2.1) may be at least partially applied to all cultural texts pointing to the broad area of pragmatics that is being increasingly investigated and reevaluated today. (Cf. T.G. Winner, 1973b and 1976a, forthcoming.) The question is raised of a general theory of reconstruction where each textual level requires different procedures. The highest level, the general intention of the text, which is purely semantic, is in the final analysis the language of certain universal notions, a problem which has been extensively dealt with by Toporov (see, for example, his treatment of paleolithic symbols discussed subsequently [1975: ]). The greatest success in reconstruction, it is held, is that of the two extremes, signified and signifier, since they correspond most closely to textual reality, whereas intermediate levels are more closely correlated with the metalinguistic system assumed in description (5.2.1). In other words, it appears that the segmentation of the text, if it can be segmented, remains the most complex problem underlying the need to reevaluate the internal structure of textual systems (already alluded to in the consideration of typologies of texts). The general diagram of recoding of texts by levels reproduced in the "Theses" (5.2.2) applies to linguistic texts, but appears to refer by analogy to nonlinguistic texts as well.

16 116 Nevertheless, the use of specifically linguistic terms would seem to limit its applicability. Five levels are traced from the phoneme to syllables to words to sentences to major semantic blocks and finally to the highest level, the general intention of the text. The problems that this diagram suggests lead to the question whether we can, as the "Theses" imply, successively reduce a text from the general intention to lower levels through a process of recoding (5.2.1, 5.2.2). As Ivanov has noted elsewhere, "the possibility of separating various stages in the process of synthesizing an artistic text... (is) problematical since in it the surface structure, which is determined by formal restrictions, may affect the deep image structure" (Ivanov, 1975: 188), a point also brought out in the later section of the "Theses", to be discussed subsequently. The problem of the difficulty of separating levels in interpretation of texts has been discussed by Western semioticians. For example Damisch, in discussing the semiotics of art (1975), and following Panofsky and Uspenskij, discusses a primary level of articulation corresponding to a phonological level in natural language, that is the level of figuration corresponding to the most general processes which may be nonsignificant in themselves and which allow the artist to represent relations of time and space. Over and above this level, according to Damisch, are "superimposed semantic (figurative), grammatical (ideographic), even idiomatic (symbolic) levels. This cleavage is modelled on the 'linguistic blueprint' ". But the entire problem of the coexistence of different levels of organization for the same figurative material cannot be resolved, according to Damisch, until there is a revision of the concept of the split between the image, or signifier, conceived as a prop for a signified which is injected from the outside, and the signified itself. The issue is the rejection of the radical distinction between the "aspect of the image which belongs to the order of perception, and that which has properly semiotic dimensions" (1975:34). In this connection, Damisch invokes Peirce's hypoicon, a kind of icon which is not necessarily a sign, and which Damisch suggests applies to the images of art (1975:35). In a similar direction Eco has underlined the importance of a concept of "sign-function" which "has brought into crisis the category of 'sign' and the naive view of the signifier/signified relationship.... Expression and content are not the precise qualities of mutually exclusive semiotic objects; they are the correlation of function which any object or event can assume in the general relationship of meaning" (1975:15). The problem of the relativity of interpreting the semantic level of texts, since interpretation depends upon changing temporal and spatial dimensions and on reconstructions and recodings, is briefly raised in the "Theses". For example, there is the case of a text where an "extrasystemic sign" or single symbol may form the entire code as instanced by the general symbol of a

17 117 vegetative design (composed of suns, birds and horses) forming a single text exemplified in Proto-Slavic materials. However, subsequently, in individual Slavic traditions, this symbol may become separated into individual elements becoming decorative designs, incorporated into artifacts, embroidery and clothes, on carvings, ritual goods (5.2.2). Another extremely informative example, a most archaic wide-spread single text, the "world tree "»confirmed in Slavic tradition by charms, riddles, song, tales, and traced to early Indo- European, and Eurasian shamanistic traditions (5.2.2), has been written about extensively by Toporov and others, dramatically illustrating the relativity of sign systems to their contexts. Toporov has traced the text of the world tree to early potential or marked elements or symbols in paleolithic art which later became organized into semiotic systems, one of which was a broad class of texts of the world tree. This image is composed of universal concepts modelling the world of human societies traced to both old and new world cultures in the era before the rise of high archaic civilizations, and still persisting in Siberia. The image of the world tree, held to contribute fundamentally to the later anthropomorphic art of ancient high cultures, is described as a "three-term vertically projected system" with birds at the top associated with the sky, hooved animals at the middle associated with the earth, and snakes and fish at the bottom associated with the subterranean world. "The world tree or its alloelements (mountains, temple, pillar, anthropomorphic beings, etc.) became means for expressing the idea of vertical structure" (Toporov, 1975:133). The horizontal composition is expressed in the central composition of the world tree and two adjacent positions (to the right and left) occupied by hooved animals or human beings. Depth is suggested by circular motion (1975:134). Finally, the world tree expresses the seven coordinates (top, center, bottom, north, east, south arid west), thus representing cosmological space (1975:135). Toporov concludes that the world tree system is opposed to that of paleolithic painting where syntactic and semantic organization within individual pictures is lacking or weak, being compensated for by the contextual organization of the totality of pictures determined by the structure of the underground sanctuary. In contrast, after the Paleolithic the world tree becomes an "ideally organized system that models both the static and the dynamic aspects of existence, while the elements outside lose all traces of organization and are surrendered to the power of chaos" (1975:136). Common factors remaining between the paleolithic and all subsequent periods include the capacity of humans to translate the "environment into symbols and build from them a world of symbols parallel to the material world" (1975:140). Underlining the universal significance of the semantic level of such fundamental texts as that of the world tree, and the varying transformations and interpretation resulting from differing contexts, the "Theses" point out

18 118 that where early linguistic texts describing the world tree are lost, a semantic reconstruction remains possible relying upon the typological similarities of cultural complexes using a set of basic semantic oppositions such as those reconstructed from Protoslavic (fortune/misfortune, life/death, sun/moon, and land/sea) (5.2.3). These similar semantic structures (here called somewhat unclearly cultural complexes) may also, according to the "Theses", be associated with similar social systems in abroad sense and are manifested, for example, in the shape of settlements, or houses, in marriage rules and in kinship terms. Such types of correlations follow from the study of Slavic antiquities (and presumably culture in general) as a single semiotic whole, and of the later transformation and differentiation of traditions (5.2.3). This view bears comparison with that of Alan Lomax's cantometrics which attempts to see underlying and pervasive relations between musical behavior and social and economic factors (1968; 1976). Whether Lomax's cross-cultural method based on comparison of fixed and quantitative factors can lead to the determination of deep structural-semantic oppositions, however, may be questioned. Yet the aim was stated, early in the project, as the establishment of "a scientific musicology that could speak with some precision about formative emotional attitudes pervading culture and operating through history" (Lomax, 1962:425). 2.4 Explanation versus description of texts The problems discussed in the preceding section (the reconstruction of texts from the highest semantic level of general intention to the lowest and most specific level, the question of the possibilities and justifications for such reconstructions, decodings and reductions, as well as the relation of levels to discrete and nondiscrete structures) all suggest a general and wide controversy. As phrased by Ricoeur (1970; 1971) two trends of interpretation oppose each other, namely description, hermeneutics, or close reading of texts held to take place at the most concrete level, on the one hand, and the deciphering, decoding or explanation of texts held to apply to the most abstract level, on the other hand. Ricoeur identifies the first approach, that of description, with dynamics and the second, that of explanation, with statics. Static explanation is attributed to 'structuralism' which Ricoeur holds has not achieved a necessary synthesis between the concrete and the abstract levels of reality Terminological differences. Terminological differences appear to have contributed to some of the divergences in conceptualization between Ricoeur, Geertz, Turner, and other Western scholars on the one hand, and those structuralists and semioticians in both East and West who have departed more directly from the traditions laid down both by the Prague

19 119 Circle and by early Russian structuralists, on the other. However, an essential assumption uniting the varying structural semiotic concepts that have emanated from Prague and Moscow has been the reconciliation of statics and dynamics, suggesting a more serious conflict with those who view structuralist theories as necessarily static. A clear terminological confusion lies in the differing uses of the term 'symbol' and its varying references to 'mythological' and 'nonmythological' consciousness. Within the "Theses" we find the expression "general symbol" and "extrasystemic sign" which may apply to mythological texts (5.2.2). Toporov has written of primitive 'symbols', the first evidences of which are marked characteristics in paleolithic expressions, and which later developed into various archaic symbolic systems such as the world tree (1975). Meletinskij, following Levi-Strauss, has also pointed to metaphorical and symbolic qualities in myth: "Some myths prove to be metaphorical transformation of others... revealing symbolic meaning" (1975:62). Of course the association of symbols with primitive systems is commonly made by Western scholars (Turner, 1964; 1967; 1969; 1974; Douglas, 1970;Evans-Pritchard, 1973; and Kluckhohn, 1968; etc.). Yet two of the authors of the "Theses", Lotman and Uspenskij (1974), have pointed out elsewhere that in their particular understanding of mythological 'symbol', a mythological text is not symbolic from the inner point of view. Reminiscent of Malinowski, who wrote in 1926 that in primitive man's ideas, tales and myths "there is but little room for symbolism " (1948:97) and "studied alive myth... is not symbolic, but a direct expression of its subject matter" (1948:101), Lotman and Uspenskij write that primitive myth is perceived as symbolic only when translated into the categories of nonmythological consciousness, when it is interpreted as composed of iconic or quasi-iconic signs (1975:28). They see poetic language as the antipode of mythological languages, a kind of middle stage between mythological, logical and scientific perceptions. In a distinction only partially reminiscent of Levi-Strauss, who does not deny metaphorical thinking to mythology, nor science to primitive thought ("It is as if the necessary connections which are the object of science, neolithic or modern could be arrived at by two different routes, one very close to, and the other remote from, sensible intuition" [Levi-Strauss, 1966:15]), Lotman and Uspenskij describe mythological consciousness in the following way: A kind of perception where objects are seen as singular and integral wholes, where signs are not ascribed but recognized, being synonomous with proper nouns, where the word and the denotate are identified, and where the text is isomorphic with the described word, where understanding is based on recognition, identification and transformation, and the entire system is monolinguistic. This type of consciousness breaks down when texts are not isomorphic with the described world, when they are poly linguistic, when

20 120 mythological texts are reinterpreted as metaphorical, and when synonomy is developed at the expense of periphrastic expressions, all of which lay the conditions for growth in " 'flexibility of language' " (Lotman and Uspenskij, 1975:34). Lotman and Uspenskij concede in the end that mythological and nonmythological thinking have been a part of all culture from the moment it made its appearance and that "one can only speak of the dominance of certain cultural models or of a subjective orientation of them toward the culture as a whole" (1975:36), but the significance of this assertion, implying the co-presence of these forms of cognition in all culture is unclear in the context of the following statements: Mythological consciousness is characterized by a specific kind of semiosis which amounts to a process of nomination (p. 20). A purely mythological state must remain undocumented probably because it pertains to too early a stage of culture development and since even if it were present it would be impossible to make contact with it (unless both were always co-present) (p. 26). Nevertheless the picture drawn (of mythological consciousness) is confirmed by numerous examples from archaic texts (p. 34). And finally: "We must emphasize the fundamental difference between myth and metaphor; although the latter is a natural translation of the former into the normal forms of our thinking. In fact, in a rigorously mythological text metaphor, as such, is strictly speaking impossible" (pp ). Yet it is held that symbols may be directly generated by mythological consciousness, but such a symbol appears to be highly limited, corresponding to a concrete element of the text and does not "go beyond the framework of mythological consciousness" (p. 28). Lotman and Uspenskij's conclusions seem to strengthen the original dichotomy between mythological and nonmythological thought, since they write that while mythological thought established the idea of isomorphism, at the same time being based solely on identification, analogies and equivalences, it did not give logical-syllogistic thought a chance to develop. Thus the specificity of mythological thought lies in the fact that identification of isomorphic units occurs at the level of the objects themselves, not at the level of their names, and mythological identification assumes transformation of an object which occurs in concrete space and time, while logical thought operates with words possessing relative independence outside time and space (1975:36). Various aspects of the treatment by Lotman and Uspenskij, which asserts the nonsymbolic nature of myth and three stages of perception from mythological to artistic to scientific, recalls the position of Cassirer and his followers. Reacting against the traditional technique of allegorical interpretation of myth which prevailed from the classical period until modern times, and which Cassirer held was only partially improved by the view that myth,

21 121 although a fiction, is an unconscious one since the primitive mind is not aware of its own creation, Cassirer, following Malinowski, has asserted the literalness of myth. According to Cassirer, in mythical imagination there is always an act of belief in the reality of the object of myth, while as Kant showed aesthetic thought is indifferent to the existence or nonexistence of its object. Finally science, like myth, is also in quest of reality but its method differs fundamentally (Cassirer, 1945:99, 101). For Cassirer, recalling Levy- Bruhl (1910; 1926), myth is a different mode of perception where, in opposition to the scientific, emotion pervades all phenomena (1945:102). Similarly Susanne Langer has written of the subjectivity of mythological consciousness. While the subjective model of reality is really a "great metaphor", apparently this is only the outer view, since in mythic consciousness "the symbol and its meaning are not separable", "the great metaphor is identified with its meaning", and "mythic thinking determines the form of language" (Langer, 1953:189). That this view, which accords little significance to metaphor and symbols in mythological thought, is not consistently shared by all members of the Tartu-Moscow group is evidenced by Meletinskij's writings. For example, he writes in approval of Levi-Strauss that "some myths prove to be metaphorical (more rarely - metonymic) transformations of others, transmitting the same 'message' through different 'codes'; transformation of mythological texts becomes a means of revealing symbolic... meaning" (1975:63). Nor would it seem that many contemporary Western anthropologists would accept the full implications of the dichotomy posed by Lotman and Uspenskij. Thus Turner writes that in thinking of other cultures, one must "pick ones root metaphors carefully" but this form of thought is not barred from primitive societies. In fact Turner finds metaphor, as described by Nisbit, at the basis of Ndembu definition of ritual symbols. "Metaphor is, at its simplest, a way of proceeding from the known to the unknown. It is a way of cognition in which the identifying qualities of one thing are transferred in an instantaneous, almost unconscious, flash of insight to some other thing that is, by remoteness of complexity, unknown to us" (Turner, 1974:25 quoting Nisbit, 1969:4). Similarly Evans-Pritchard vividly describes Nuer symbolism. For example a Nuer, when leaving his tribal territory, may mix in a liquid, earth from his natal territory and drink it. "Since we at once perceive the meaning of the symbolism of the ritual action, we may suppose that Nuer also perceive its logical fitness to its purpose: and, indeed, it is often certain that they do so, for if asked to explain what they are doing they interpret the symbol of a rite in terms of its purpose. The symbolism is manifest to them as it is to us". A deeper symbolic meaning is also present, however, which is not obvious or explicit either to the culture bearers or to the outside interpreter. Symbolic interpretation at this deeper level, though difficult, is also warranted (Evans-Pritchard, 1973:93).

22 122 It would be helpful if the various approaches to symbols and to mythological and non-mythological thinking by members of the Tartu-Moscow group were clarified. We know that the Peirceian trichotomy is not used and that symbols are understood as iconic (Lotman and Uspenskij, 1975:42, footnote 19). But symbols are not consistently defined, and whether they are, or are not, a part of mythological consciousness is hardly agreed upon. We need more clear differentiations between the various views of the authors of the "Theses" and between this group as a whole and the approaches of others. There is the broadest linguistic interpretation of the 'symbol' by Cassirer (See Ricoeur, 1970:11, footnote 5 referring to Hamburg, 1956:59, citing Cassirer, 1975:3, 93), the somewhat more restricted use by Ricoeur (for Ricoeur symbols denote linguistic signs that encompass double or multiple meanings that require interpretation, 1970:9); the extension of the concept to nonverbal systems by semioticians and many anthropologists in West and East (for example the whole area called 'symbolic anthropology'); and finally there is the more specific use of the term symbol (understood to include more limited aspects of verbal and nonverbal behavior) by Jakobson and other semioticians, as well as Levi-Strauss, who have followed, extended and modified the Peirceian trichotomy of index, icon and symbol. However, deeper differences can be traced to divergent interpretations of structures and their textual manifestation, three aspects of which we now briefly discuss, namely 2.4.2: the hermeneutic critique; 2.4.3: statics and dynamics in relation to myth and ritual; and 2.4.4: the relation of phenomenology to structural semiotics The hermeneutic critique. This view is conveyed by the comments of Geertz, who defines his concept of culture as a 'semiotic' one (1973:5) and who, going back to Ricoeur, holds that the notion of text, freed from the notion of scripture and writing, allows us to see culture as an "assemblage of texts" (1972:26). However, Geertz holds that "the idea remains theoretically underdeveloped: and the more profound corollary, as far as anthropology is concerned, that cultural forms can be treated as texts, as imaginative works built out of social materials, has yet to be systematically exploited". For Geertz the problem is that interpretation of culture or penetration of a text to discover its social semantics, which he calls "thick description", is in itself opposed to deciphering a code or ordering a system (1972:26). The latter is held to be reductionist and meaningless. Reminiscent of Saussure's dichotomy, to Geertz the level of action, which he terms symbolic because it is meaningful and which may be composed of

23 123 complex, irregular and inexplicit conceptual structures (or parole), and the level of structure (or langue) are forever separated; but the distance between them is greater than it was for Saussure who saw these levels as interrelated even if in a relatively static sense, while for Geertz, "Once human behavior is seen as symbolic action... the question as to whether culture is patterned conduct... loses sense" (Geertz, 1973:10). Somewhat less dichotomously, Ricoeur sees hermeneutics as "the theory of rules which presides over an exegesis, that is over the interpretation of a particular text or group of signs that may be viewed as a text". Such an exegesis is extended "to all signs bearing analogy to a text" (or all signs that are symbols, encompassing double meanings of expression, and thereby requiring interpretation) (Ricoeur, 1970:8). However, Ricoeur holds that there is no general hermeneutics, "no universal canon for exegesis, but only disparate and proposed theories concerning the rules of interpretation" (1970:26-27). As he states, "according to one pole, hermeneutics is understood as the manifestation and restoration of a meaning addressed in the manner of a message... ; according to the other pole, it is understood as a demystification, as a reduction of illusion" (1970:27). "A general theory of interpretation would thus have to account not only for the opposition between the two interpretations of interpretation, the one as recollection of meaning, the other as reduction of illusions, but also for the division and scattering of these two great 'schools' of interpretation into 'theories' that differ from one another" (1970:32). The former approach places confidence in the object that will address itself to the subject (1970:29-39) while the latter approach doubts the existence of the object, contests its primacy, tears off its masks and reduces its disguises (1970:30, 32). According to Ricoeur, at issue is the fate of the "mythopoetic core of imagination" which is sacrificed by the school of demystification (1970:35). A general hermeneutics reconciling the above conflicts does not yet lie within our scope and would require a most critical resolution which would return to the listening of "the fullness of language". This does not however mean a "return to the dense enigma of initial, immediate speech, but to speech that has been instructed by the whole process of meaning" (1970:496). Yet Ricoeur's theory of interpretation, or depth semantics that he has called "appropriation", bears marked analogies to some tenets of the semiotics of culture we have been discussing. Such an approach involves selfinterpretation on the part of the receiving subject, encouragement of reflection, overcoming of cultural distance and estrangement in regard to cultural values, and actualization of semantic dimensions (Ricoeur, 1971:145). Here Ricoeur invokes Peirce's view of the relation of sign to object that does not exclude the role of the interpretant. The sign is never closed "in the sense that there is always another interpretant" (1971:149). Only when structural

24 124 analysis can contribute to rendering manifest such a depth semantics, as for example the "live-semantics" of a myth, will it become possible "to integrate the opposed attitudes of explanation and understanding within the unique concrete act of reading' 1 (1971: ). It seems clear that Ricoeur's conflict has received various resolutions by the Prague structuralists, beginning as early as 1928 when Jakobson and Tynjanov first set about reconciling the dichotomy between synchrony and diachrony. As Matejka has written, Jakobson has always been interested "not only in the disclosure of antinomies but also in the forces which bind together antithetical properties". Thus he comments upon Jakobson's philosophical bent, the overcoming of "the radical dualism, which dogmatically insists on the isolation of opposites rather than on their synthesis" (Matejka, 1975:114). Throughout his lifetime Jakobson has advanced and considered the implications of "the inseparable tie between... substance and form, sound and meaning" (Matejka, 1975:93), and, we might add, between Ricoeur's "restoration of meaning" and reduction of illusion, or "understanding" and "explanation". Furthermore, the aesthetic function elaborated by both Jakobson and Mukafovsky, and which Mukafovsky has shown to be at least potentially present in all communicative behavior, entails a process of actualization of semantic dimensions ever present, but no longer consciously perceived by the receiver. Moreover, as Mukafovsky has shown, while structures are autonomous, they are not closed, and particularly the aesthetic text is polysemantic, its many-levelled meanings being affected both by the intention of the encoder and that of the decoder (Mukafovsky, 1934a; 1936; 1937a; 1937b; l.p. Winner, 1973; T.G. Winner, 1973a; 1973b; 1976a forthcoming; 1976b forthcoming). All these concepts have been fundamental to the later developments in Eastern European semiotics and seem to bear affinities to Ricoeur's "appropriation" which he considers to be beyond the present confines of "structural analysis" Statics and dynamics in myth and in ritual. The position that the structural approach to myth is necessarily static appears to be another variant of Ricoeur's theme of the two as yet irreconcilable schools of interpretation. For example, Levi-Strauss is frequently placed in the camp of reductionist and static explanation, a point of considerable importance in considering the position of the Tartu and Moscow semioticians, since many of their studies of myth have been strongly influenced by the approach and methodology of Levi-Strauss. In an expression of the above assessment of Levi-Strauss, Geertz concedes that Levi-Strauss treats symbolic forms as texts, but holds that Levi-Strauss does not interpret these texts, since he understands them entirely in terms of their inner-structure, independent of any meaning and context (Geertz, 1972:36, footnote 38). Yet on the open-

25 125 ing pages of Le cru et le cult, Levi-Strauss underlines the relationship between the concrete context and abstract structure. Le but de ce livre est de montrer comment des categories empiriques... definissables avec precision par la seule observation ethnographique et chaque fois en se pla9ant au point de vue d'une culture particuliere, peuvent neanmoins servir d'outils conceptuels pour degager des notions abstraites et les enchainer en propositions. L'hypothese initiale requiert done qu'on situe d'emblee au niveau le plus concret, c'est-ä-dire au sein d'une population... (1964:9) (underlining supplied). Meletinskij has attempted to show the dynamic potentials of this view. He notes that Only Levi-Strauss was able actually to describe mythological thought in its aspect of generation of semiotic modelling systems and... (to show) the intellectual capacity of myth to classify and analyze, while simultaneously explaining those of its specific features that bring it close to art: thought at the sensual level: thought attaining its objectives by indirect paths ('bricolage') and using a kaleidoscopic rearrangement of an available set of elements (1975:62). From this point of view Meletinskij has considered the cycle of paleo-asiatic raven myths in relation to their variations and permutations and in relation to their function of mediations between binary oppositions (Meletinskij, 1973), as is noted later. Also similar to Levi-Strauss' analysis is Ivanov's treatment of Gogol's Vij (1971; 1973b), who mediates between oppositions such as visible/invisible, which Ivanov compares to the oral art of the most varied cultures, East Slavic (e.g. the Russian baba-jagd), Abkhasian, Khett, Scandinavian (the Edda), Egyptian, Chinese, American Indian, etc. The question of the potentially dynamic function of myth and ritual bears upon the closely related issue of the degree of flexibility as opposed to rigidity in "mythological cultures". (Note: The general question of the rigidity of archaic cultures is discussed in section 3.1 of this essay. Here we consider only those aspects of this issue that bear specifically upon the question of myth and ritual.) A part of French anthropological studies since Durkheim, this area of investigation has reached its most complex development in Levi-Strauss' monumental analyses (1964; 1967; 1968; 1971) which have brought into fruitful contact the French structural tra-

26 126 ditions and those emanating from the Prague school. While the influence of Levi-Strauss' studies upon Russian semiotic approaches to myth has clearly been profound, as Ivanov has shown (1975), structural approaches to myth and ritual were a part of Russian studies since Bakhtin's works of the 1920s. Bakhtin's analyses, which contribute a dynamic approach, may be contrasted to the more inflexible view implied by Lotman's dichotomy between mythological nonsymbolic cultures and historical cultures discussed earlier, and to some of the later remarks in the "Theses" emphasizing rigidity of archaic cultures. According to Bakhtin, the mediating and dynamic function of carnival culture, a system widespread in folk culture but originating historically in ritual performance, raises to the forefront the whole question of structural approaches to areas of freedom and flexibility in archaic cultures. Thus Bakhtin distinguished between unofficial consciousness as opposed to official consciousness. Unofficial folk language, or carnival language, evidenced at the holiday, carnival and market situations. "... is a means of connecting the lower strata of inner speech with a broad social milieu (in other words a means of translating individual-biological into the social)" (Ivanov, 1975: , citing Bakhtin). According to Ivanov, the series of binary oppositions found by Bakhtin in the Rabelaisian world (Bakhtin, 1965) correspond to those which Levi-Strauss finds in myths (Ivanov, 1975:212). Bakhtin showed " 'the carnival image seeks to embrace and combine within itself both poles of the formation or both terms of the antithesis: birth and death, youth and age, top and bottom, face and backside, praise and abuse' " (Ivanov, 1975:213, citing Bakhtin, 1966:238). Furthermore, Ivanov holds that Bakhtin's inversions and reversals also foresaw Leach's later study of the carnival (Ivanov, 1975:213). Similarly to Bakhtin, Leach connects the problem of social role reversals at rites of passage, festivals, and carnivals, with the concept of discontinuities of time, "probably the most elementary and primitive of all ways of regarding time" (Leach, 1966:134), when time is seen as oscillation or repetition of sequences of polar opposites (life-death, night-day, summer winter, the sacred the profane) (1966:126). The "thing" that oscillates would need to change into its opposite since "at the end point of any oscillation everything goes into reverse" (1966:129). Turner has also demonstrated the dynamic relation of rites of passage and ritual to extra-ritual processes. As he noted, Van Gennep (1909) showed that in all ritualized movement there was at least a moment when those being moved in accordance with a cultural script were liberated from normative demands, when they were, indeed, betwixt and between... In this gap between ordered worlds almost anything may happen.

27 127 Thus, In this interim of 'liminality', the possibility exists of standing aside not only from one's own social position but from all such positions and of formulating a potentially unlimited series of alternative social arrangements. without liminality program might indeed determine performance. But given liminality... multiple programs can be generated (Turner, 1974:13-14). The quintessence of the mythological and ritual dynamics of primitive role reversal (one manifestation of which is Bakhtin's comic double) (Ivanov, 1973; 1975:216) is, according to Meletinskij, the widely occurring mythological, syncretic figure, the culture hero and trickster (Meletinskij, 1973 and 1975) (an ambiguous figure described by Western anthropologists since Boas, notably Radin, 1956 and others). According to Meletinskij, the essential fact is "the duality of the mythological figure who is at the same time both the serious creator of the organized world order (natural and social) and a jester-rogue constantly bringing chaos into the very organization he created, violating the taboo..." (1975:94). Trickster plots (like Turner's liminality) are seen by Meletinskij as a safety valve in a strictly regulated society, accounting for the emphasis on erotic and anal details, ridicule of obligatory rituals and even the element of social criticism (1975: 94). In Meletinskij's analyses (1973:152), the Raven of Paleo-Asiatic and northwestern American Indian mythology (a parallel to other American heroes and tricksters) exemplifies the above concepts, since the Raven combines both the culture hero and shaman, as well as the mythological trickster and mischief monger. In these capacities according to Meletinskij, following Levi-Strauss, this figure functions as a progressive mediator of general categories including life/death, carnivore/herbivore, above/below, nature/culture, human/beast, male/female, etc., and more specifically, land/sea, salt water/non-salt water, cleverness/stupidity, winter/summer, etc. (Meletinskij, 1973:114). Bluffs and tricks exhibit Raven's shamanlike skill, yet at the same time they seem to caricature and parody genuine shamanlike activities (1973:121). Since the Raven's action is set in a mythological age before creation was finished and before social norms were stabilized, considerable freedom is allowed. Social norms are flouted, the sacred may be profaned and carnivalized and a respite is provided to the ordered society (1973: ). Recently the Japanese structural anthropologist, Yamaguchi, has also assessed the dynamism of the trickster complex which he sees as a synthesis of cognitive and aesthetic functions. Creatively utilizing some of the aesthetic theories of the Prague Circle, Yamaguchi depicts the Jukun trickster cycle in northeastern Nigeria as not only a logical instrument which me-

28 128 diätes oppositions, but also as an aesthetic device which "actualizes" reality. Through introducing chaos into the world, the trickster creates a kind of marginal reality, on the basis of which order is then recreated. In Yamaguchi's view, the trickster becomes a synthesis of a kind between phenomenological and structural senses of reality (Yamaguchi, 1973) Phenomenology and structural semiotics. Yamaguchi's conclusions point to a fundamental historical and theoretical issue that can only be briefly discussed here, namely the relation of the structural semiotic movement emanating from the Russian structuralists and the Prague Circle to the phenomenology of Husserl and his followers, a problem to which Holenstein (1975a, see also 1975b) has addressed himself in his "Jakobson phenomenologue?" Holenstein questions the traditional notion that structuralism and phenomenology are two distinct, and mutually exclusive, intellectual currents, pointing out that Husserl lectured at the Prague Linguistic Circle in 1936, and that the two movements proceeded in parallel fashion in three directions: in considering the role of the subject in the constitution of language; in considering the role of language in the constitution of the world, and in considering the conception of the phenomenology of the theory of relations (1975a:30). Indeed Jakobson, according to Holenstein, went beyond phenomenology. In relation to the first direction, the role of the subject in the constitution of language, Holenstein points out that both phonemics and poetics, the first areas of interest to the Prague Circle, are defined by the aid of reference to the subject who, far from being eliminated, is an important correlate of the system, both as an encoder and as a decoder. Furthermore, Jakobson stresses, more than Husserl, the view that all perception is a systematic classification of content within a system of values or categories (1975a:32). Taking the linguistic example, Holenstein reminds us that individual creativity and freedom, though close to zero at the most discrete level of phonemes, becomes almost unlimited at the most general level of the text, where the obligatory linguistic rules are almost entirely lacking (1975a:33). The second parallel area noted by Holenstein, the role of language in the constitution of the world, is both a phenomenological presupposition and underlies Jakobson's system. This is demonstrated, in Jakobson's analysis, by the child's developing ability to use subject and predicate, thereby gradually liberating himself from the concrete situation, and increasingly employing imagination to obstruct and destroy the world and to plan (Holenstein, 1975a:34). Corresponding to this view, it would seem to us that mythology may be understood as a means of liberation from the concrete situation through its function of classification by means of mediation of oppositions, thereby successively resolving fundamental conflicts.

29 129 Finally, the third area common to phenomenology and structuralism is Husserl's theory of fundamental relations that unite all matter, namely reversible reciprocity and irreversible unilaterality. Jakobson has shown the existence of such fundamental relationships in his system of linguistics, particularly the reciprocal one of oppositions. As Holenstein points out, phenomenological analysis is distinguished from a purely logical one of formal exclusion, since phenomenology studies data not only as they are " 'in themselves' " (" 'elles-memes' "), but equally as they are in the consciousness " 'for us' " (" 'pour nous' ") (1975a:35). From the phenomenological point of view, the opposition is shown to be one of both exclusion and inclusion, since the excluded member is necessarily co-present in the consciousness, one implying the other. Jakobson is quoted as citing Pos: L'opposition n'est pas un fait isole: c'est un principe de structure. Elle re unit toujours deux choses distinctes, mais qui sont liees de teile fa9on, que la pensee ne puisse poser 1'une sans poser l'autre. L'unite des opposes est toujours formee par un concept, qui, implicitement, contient les opposes en lui et se divise en opposition explicite quand il est applique ä la realite concrete... (Jakobson, 1949 in 197la, 423, as cited in Holenstein, 1975a:35). Holenstein points out that the phenomenological view not only implies that the excluded aspect of opposition is equally present in the consciousness, but also relies on material or substantive traits directing us to their content. Similarly, Jakobson's use of the relation marked/unmarked is held to be a specific form of the relation presence/absence which is equally oriented toward material or substantive traits (Holenstein, 1975a:35). According to Jakobson, the two members of the relation are not equivalent and the opposition marked/unmarked is a secondary phenomenon, derived only from the content. Thus only a content analysis makes it possible to explain why one of a pair of oppositions is marked (more informative, indicating the presence of a property) and the other unmarked (either saying nothing about the presence of the property or signalling its absence) (Holenstein, 1975a:36). Holenstein concludes that the position that structuralism is a movement unilaterally oriented towards form, neglecting matter, overlooks the fact that Jakobsonian structuralism holds that all forms depend upon matter, and because of this fundamental condition the determinants of content enter equally into the system (1975a:37). This conclusion is also relevant, it would seem, when considering Levi-Straussian structuralism, which although emphasizing formal elements, does not overlook content.

30 130 In a recent statement on the relation of structuralism to teleology, Jakobson has underlined some of these points. He has stressed the clear and necessary relation between the abstract and concrete in the system of that branch of structuralism which developed from his own early efforts and from that of other scholars united in the Prague Linguistic Circle. In opposition to the antiteleological orientation of Saussurian linguistics, Jakobson's approach from the beginning was teleological (Jakobson, 1975b:50-52). As he states, from the time when he formulated the theory of the bundle of distinctive features in 1931, following up the implication of earlier insights, the concept of opposition, which he began thinking about in 1920 after reading Saussure, received considerable elaboration. Ces elements oppositifs sont reellement pergus par les sujets parlants et on peut demontrer les correlats physiques et moteurs des oppositions en question. Mefions-nous des modeles abstraits en dehors de la realite perceptive. C'est une autre question que de savoir si ces rapports sont con$us par nous de fa9on consciente ou subliminale; en tout cas le metalangue les met en relief. Si nous reconnaissons ces rapports malgre toutes les distortions possibles, c'est parce qu'ils existent et restent valides: on peut deflnir en quoi consiste l'invariance du rapport. L'idee d'une in variance topologique est incontestablement realiste. Les deux elements qui s'opposent Tun ä l'autre ne sont jamais equipollents: Tun d'eux, hierarchiquement superieur, fait contrepoids au partenaire non marque. C'est un point essentiel de la linguistique structural teile que je la deflnis ä la suite de Troubetzkoy (Jakobson, 1975b:52). If we follow the implications of this point of view, it is clear that binary oppositions are not empty abstractions, and which of a pair is marked or unmarked depends on how the elements are used in the cultural context by subjects. These several areas of parallel development between phenomenology and structural semiotics are reflected in the writings of various members of the Tartu-Moscow group and of other semioticians, particularly in respect to the emphasis on the role of the subject, as the following remarks demonstrate. Thus Pjatigorskij has underlined, that the concept or sign is inseparable from the subject who uses things as signs by virtue of his ability to imagine, choose and project. In Pjatigorskij's words: "Things usable by living beings as signs, objectively present the possibilities of such usages as a result of the fact that they possess the qualities of duality, position and projection" (1974: ).

31 131 These qualities are the " 'pure possibilities' " of sign-ness that are converted by subjects "into sign reality in acts of communication and auto-communication" (1974:187). Lotman has also held that structural research should not describe the language or code of a system to the detriment of parole, including the phenomena that are "random" from the point of view of language (Lotman, 1974a:304). These phenomena are created by non-identical individuals or subjects who purposely multiply the mechanisms that impede message transmission. Thus those exchanging information use not one common code, but rather two different ones that to some extent intersect, and the problem of the communicative act is one of translation. "Non-understanding, incomplete understanding, are not side products of the exchange of information, but belong to its very essence". At the cultural level, there is a constant and purposeful multiplying of the mechanisms that impede message transmission (1974a:302). This process presupposes the variety of the individuals composing a collective, which "excludes their absolute intersubstitution". All of which points to the problem of what is the origin of the contents of texts and the mechanism by which new communications and texts arise (1974a: 303). Lotman concludes that it is imperative that structural research takes full cognizance of such "random" phenomena that "are the working mechanisms in the semiotic structure of culture and a way to describe them must be sought. Hence the most pressing demand is for research into the semiotic nature of untranslatability and into different kinds of culturally created "noise", and also into the degree of intersection of various codes active in a single cultural system" (1973a:304). Taking a somewhat more extreme position than that of the Russian semioticians, but showing important similarities, nevertheless, Kristeva writes that semiotics must identify both the systematic constraint within each signifying practice as well as those aspects within the practice that fall outside of the system and characterize the specificity of the practice as such (Kristeva, 1975:48-49). Emphasizing the role of the speaking subject, as distinguished from the abstract transcendental ego, she wishes to make explicit the signifying practice where practice means "the acceptance of a symbolic law together with the transgression of that law for the purpose of renovating it" (1975:51), Thus: "The moment of transgression is the key moment of practice", since practice means "transgression of systematicity, i.e. a transgression of the unity proper to the transcendental ego" (1975:51). Semiotics, Kristeva concludes, must recognize the heterogeneity of the signifying process, and thus must not neglect the "remainder" or the "waste" which, from the point of view of the speaking subject, sets the semiotic process in action (1975:52). It seems clear from the preceding discussion that a semiotic approach to

32 132 culture demands and suggests syntheses of the most concrete and dynamic, as well as the most abstract and formal levels, and that the statements in the "Theses" demonstrate significant advances in this direction, although important areas of unclarity remain. Final resolutions can hardly be reached at this stage in semiotic research. Among the various issues that remain problematical are the questions of the degree of systematization of semiotic systems, and their adherence to one (linguistic), or more, models. These points will be expanded upon in the following discussion of definitions of levels of culture advanced in the final sections of the "Theses". 3. Definitions of Culture on Three Levels The various approaches to cultural texts advanced in the "Theses", relating to the semiotics of culture, where the fundamental unit is a culture text, lead to three definitions of culture (quoted earlier), which we now explore (6.0.0 to 9.1.0). On the culturological level, culture is "a hierarchy of particular semiotic systems". Since semiotic systems are composed of texts, moving one level down, culture is "the sum of the texts and the functions correlated with them". Thirdly, reducing texts to their origin, culture becomes a certain mechanism which generates the text (6.0.0). 3.1 Culture as memory We consider first the third definition which rests on an analogy of culture to the individual. "If we regard the collective as a more complexly organized individual, culture may be understood by analogy with the individual mechanism of memory as a certain collective mechanism for the storage and processing of information". The analogy of culture with memory "does not contradict the dynamism of culture: being in principle the fixation of past experience, it may also appear as a program and as instructions for the creation of new texts" (6.0.0). Participants then create texts, and texts contain the memory of participants (6.0.1). Similarly, Evans-Pritchard, in rejecting ahistoricism, stated that a people's traditional history forms part of the thought of living man, hence part of their observable social life (Evans- Pritchard, 1961:6). Carrying farther the implications of the analogy of culture to memory, the "Theses" point out that through the centuries texts transmit certain structures of personality and types of behavior and may appear as a condensed program of the whole culture, reminding us of Geertz's "thick description" of the Balinese cock fight (Geertz, 1972) which, in spite of Geertz's disavowal of systematization, seems to lay bare structures that compose such a "condensed text". That such a condensed program would necessitate uniformity of culture is countered by the existence of the

33 133 phenomenon of polyculturality. Thus the assimilation of texts from other cultures may result in polyculturality, providing the possibility of choice on the part of participants between alternative cultural styles (6.0.1). (However, the whole question of pluralism and heterogeneity within culture is an area rather weakly developed in the "Theses" and needs considerable expansion). The comparison of culture to memory, or to a collective mechanism for the storage of information, is the basis of an evolutionary view that contrasts archaic cultures to more advanced systems, invoking the dichotomy discussed earlier between mythological and non-mythological cultures (see our Section 2.4.3). In archaic systems, verbal texts are based on rigid schemes, and literature is reduced through ritual formulae to embodiment of mythological plots handed down from generation to generation. On the social level, such sign systems may be synchronized with "rigidly determined systems of relationships, in which all possibilities are covered by rules correlated with the mythological past and with cyclical ritual" (6.0.2). In more advanced systems, memory is regulated by real history and the concept of development or direction of time is then associated with accumulation and processing of information which is gradually used to correct programs of behavior. Such an evolutionary position recalls earlier formulations including the Durkheimian oppositions between mechanical and organic solidarity, and the sacred and profane, and Hertz's seminal early exposition of the evolution of dual symbolism. Speaking of the most primitive societies (the "mythological" cultures of the "Theses"), where tribes are divided into equal moities, one sacred and one profane, Hertz writes that, "the evolution of society replaces this reversible dualism with a rigid hierarchical structure: instead of separate and equivalent clans there appear classes or castes of which one at the summit is essentially sacred..., while another at the bottom is profane..." (Hertz, 1911, 1973:8). A more recent expression of these'thoughts is of course Levi-Strauss's distinction between mechanical and statistical models (1963:286). Taking up the Levi-Strau$sian position (see Le'vi-Strauss, 1963:29), Ivanov (1973a) distinguishes two types of archaic systems only briefly alluded to in the "Theses" (6.0.2). Treating marriage systems as systems of information relating to the number of choices available, Ivanov points to the following contrast: (a) The dual principle, the basis for four marriage classes, exemplified by such societies as the Australian Aranda, and the Pueblos and also reconstructed for the ancient Chinese, is the most ancient one, associated with the perception of only two kinds of time, cyclical and ritual on the one hand, and mythological on the other, historical time being absent, (b) On the other hand there is the less rigid Omaha-Crow system (from

34 134 which the Roman is held to have originally developed) where the criteria for choosing marriage partners depend on memory of data of a statistical nature (the partners belong to classes carrying the greatest amount of information, i.e. "marriage partners on whom there is no information stored in the memory concerning marital ties already established between marriage classes" (Ivanov, 1973a:120)). It seems doubtful, however, that Levi-Strauss meant to suggest the mutual exclusiveness or strict evolutionary sequence of archaic mechanical and advanced statistical systems such as appears to be proposed in the "Theses" and by Ivanov. As Levi-Strauss cautioned "... the same phenomena may admit of different models, some mechanical and some statistical, according to the way in which they are grouped together and with other phenomena" (Levi-Strauss, 1963:284). Moreover, it would seem that the extreme position taken in the "Theses", that in the most ancient sign systems, on the level of social interpretation, all possibilities are covered by rules correlated with the mythological past and cyclical ritual, is unclear and may conflict with the implications of other statements by the signers of the "Theses" and other members of the Tartu-Moscow group. It would be necessary first of all to define the meaning of "social interpretation". Secondly it would seem that if culture can be interpreted as composed of sign systems which function in some kind of unity where each system supports the other (1.0.0), the above assertions concerning the rigidity of the most archaic social systems on the level of social interpretation, would imply the rigidity of the entire culture. Yet as noted earlier, Pjatigorskij's view of semiotics would seem to imply a less rigid approach even to archaic cultures, since it assumes that the creation of signs, which is not limited to advanced cultures, depends on several operations including that of freedom of choice of the subject (Pjatigorskij, 1974:187). Furthermore, Lotman has isolated as particularly fundamental the problem of constructing a communication model that can account for the origin of the content of messages and the mechanism by which new texts are created (Lotman, 1974a:304). Again this dynamic process can hardly be limited to "historical societies". And the dynamic role of the primitive trickster culture hero has been shown to oppose the regularities of rules in archaic cultures (Meletinskij, 1972; 1973), which is traced back to the earlier inspirations of Bakhtin. Finally, in the "Theses" itself the dynamic relation of rigid rules to variety and flexibility in poetic forms is explicated, which we discuss subsequently. Turning now to the concept of development in time characteristic of advanced culture, as discussed in the "Theses", we note that this characteristic must also be modified, again weakening the dichotomy posed between archaic and advanced cultures. It is necessary to take account of secondary processes within the historical cultures that may take the direction of decreasing complexity or entropy, resulting in "simplification of structures".

35 135 Then "movement is not in the direction of increasing the amount of information but in the direction of increasing the amount of entropy" (6.0.2). This remark is strengthened in the opening of the "Theses" where it is stated that "in culture itself entropy increases at the expense of maximum organization" (1.2.0). Such a negative trend in advanced societies is emphasized by Levi-Strauss: "Thus it is that civilization... can be described as an extraordinarily complex mechanism, which we might be tempted to see as offering an opportunity for survival to the human world, if its function were not to produce what physicists call entropy, that is inertia" (Levi-Strauss, 1974: 413). Levi-Strauss's "cold societies" (the archaic rigid societies of the "Theses") create minimum disorder, while "hot societies" (the historical societies of the "Theses") use up and destroy energy (Charbonnier, 1969: 32-33). Similarly, Ivanov states (1973a: ) that the application to human history of the principle of increase in complexity of structures assumes the still largely hypothetical idea of the replacement of the biosphere (the sphere of living matter) by the noosphere (the sphere of the activity of reason). As Ivanov remarks, not only has the store of information about the past not been effectively used for the extrapolation of the future, but the difficulties increase with the growing volume of information which cannot be physically processed (1973a:113). Thus it seems that "systems transmitting information in society seem to be governed by the same laws of development in time, i.e., the increase of entropy, as other physical systems". Further intensifying this dilemma is the growing standardization of mass culture and even the standardization of the increasingly taboo deviations from mass culture. Thus we cannot assume that "the accumulation of information, in and of itself, signifies the Overcoming of entropy' " (1973a:115). 3.2 Culture as a hierarchy ofsemiotic systems We turn now to the first definition of culture on the highest cultural level, where culture is seen as composed of a hierarchy ofsemiotic systems. A minimal mechanism constituting a culture would require a pair of correlative systems. "The text in a natural language and the picture demonstrate the most usual system of two languages constituting the mechanism of culture". However, "the pursuit of heterogeneity of language is a characteristic feature of culture" (6.1.0). Thus it would seem that a minimal pair does not really characterize culture, which is typically composed of numerous correlated pairs of semiotic systems, hierarchically organized, "the correlation between them being to a considerable extent realized through correlation with the system of the natural language" (the primary system, while other systems are conceived of as secondary modelling systems) (6.1.3). We note several problems in the above statements. For example, is

36 136 natural language alone a sufficient premise for the construction of culture even when the reservation is made that "a natural language is not a strictly logical realization of a single structural principle" (6.1.4)? Furthermore, we receive only limited help in solving the question of the meaning of "secondary modelling systems" by the following comment (omitted from the translation into English in Section 6.1.5). Under secondary modelling systems we understand such semiotic systems, with the aid of which models of the world or its fragments are constructed. These systems are secondary in relation to the primary system of natural language, over which they are built - directly (the supralinguistic system of literature), or in the shape parallel to it (music, painting) (6.1.0 original Russian version in Mayenowa, ed., 1973). As has been suggested, statements made by some of the signers of the "Theses" do not appear to have consistently upheld the primacy of the linguistic model. Lotman (1974b:80) has identified two types of narration possible in primary semiotic systems and it is not clear that both are fudy derived from an underlying linguistic model. Thus transformational narration (texts elaborated in space and frequently visual) and linear narration (texts linked with time and typically verbal) are interrelated, yet distinct (see section 2.2 of this discussion). Furthermore myth partakes to a large extent of the transformational form of narration having parallels to music as Levi-Strauss has shown. If we follow Lotman's argument, here, it would seem to contradict the view that music and painting may be seen as taking a shape which is merely "parallel" to the primary system of natural language as is stated in the above untranslated section of the "Theses". Rather, the contrasts between the two forms of narration would appear to reflect a more radical divergence. Moreover, in the "Theses" themselves some statements seem to imply that secondary modelling systems are not necessarily secondary to primary ones. Instead, two equivalent systems seem to exist in a state of mutual dependence. For example, it is stated that the tension between discrete (or verbal) texts and nondiscrete (or visual or iconic) texts "constitutes one of the most permanent mechanisms of culture as a whole" (3.2.1). And later it is stated that the opposition between discrete and nondiscrete semiotic models or texts plays a special role in the system of culture generating semiotic oppositions since each type of text needs the other "in order to form the mechanism of culture". But each type also needs to be different "according to the principle of semiosis, that is to say, on the one hand equivalent, and on the other hand not entirely mutually convertible" (6.2.0).

37 137 Similarly, Ivanov has held that one of the basic tasks for contemporary semiotics "remains the development of a general semiotic set of concepts and correspondences suited to the description of various sign systems, including those structured quite differently from the natural language" (1975: ). Kristeva has taken a more critical approach to secondary modelling systems than the Tartu-Moscow group, holding that if semiotics limits itself to the linguistic model alone, or even to the general principle of "systematicity", it cannot satisfactorily deal with signifying practices in the areas of transgression and pleasure, in the domain of art, ritual and myths (Kristeva, 1975:48). In considering these various theoretical divergences and critiques, it should also be noted that how we conceive of a cultural system, whether it is based on one primary modelling system or more than one, and to what extent we consider Lotman's "noise", Kristeva's "waste", and other such concepts, are themselves factors affecting the resulting perception. Indeed this point is implied in the "Theses" where it is held that "the degree of awareness of the entire system of systems within a culture varies, which by itself may be regarded as one of the criteria of a typological evaluation of a given culture" (6.1.5). For example, while medieval culture was traditionally conceived of as uniform, Bakhtin's discovery of the layer of unofficial carnival phenomena implies a less uniform organization of medieval culture than originally thought and other examples are given. Thus the demand made in the "Theses" for a broader investigation of "the interrelations between natural languages and secondary (superlinguistic) semiotic modelling systems" (6.1.5) seems to be warranted and may even be too narrow a conclusion since a broader framework is needed. 3.3 Culture as the sum of texts Turning now to implications of the middle level definition of culture as the sum of the texts and the functions correlated with it (7.0.0), we suggest that the term "sum" is misleading and refer to Lotman's statement elsewhere which holds that the multiple texts of a given culture are not additive, but are variants of an invariant text of a given culture which ultimately can be reconstructed. From the point of view of the particular culture, such a "culture text" is its most abstract model of reality or "world image" (Lotman, 1973:549). In considering the interrelation of texts within a culture, a fundamental problem is the question of equivalence, yet not identity, of the structures and functions of specific texts, and consequently the question of translation, which never can be complete. As was pointed out in the preceding section, the antithesis between discrete and nondiscrete texts as understood in the

38 138 "Theses" means that they are not entirely mutually convertible (6.2.0). Moreover, the systems themselves that generate texts retain their autonomy "no matter how extensive may be the identity of the text they generate" (7.0.0). Various examples are offered of transmissions of texts through different channels, which may involve related languages or different ones, and related cultural traditions or different ones. Transmissions across cultures and languages require broad understanding of the entire culture including " 'sublinguistic systems' of custom, life style, and technology... which are not constructed on the basis of signs and texts of a natural language and cannot be transposed in them" (7.0.1). Only in subsequent stages can sublinguistic systems be detected in "secondary supralinguistic systems" (7.0.1). Here the problems of what is meant by "sublinguistic systems", and how they are distinguished from "secondary supralinguistic systems", need to be clarified. One may ask why, for example, is life style "sublinguistic", while art is a "secondary modelling system". A particularly complex problem is that of the functioning of a text on different cultural levels, which is opposed to the concept of immanence of structures. As the "Theses" point out, in corrected translation, "The view according to which cultural functioning is not achieved within the framework of any one semiotic system (let alone within a level of the system) implies that in order to describe the life of a text in a system of culture... it does not suffice to describe the immanent organization of separate levels" (8.0.0). It is necessary to study the relations between structures at different levels. Structural isomorphism, and appearances of intermediate levels, allow the passing from one level to another and of recoding. Saussure 's Anagrams are referred to as an early insight into the relations of levels which are not immanent (8.0.0). To expand beyond the "Theses", it would seem that Saussure's remarks suggest the infinity of possibilities emanating from the interrelation of structures. As Saussure wrote, "The anagram may develop either from the word actually present in the text, or from one not mentioned at all but suggested naturally by the context" (1907:68). And, "One ends up by wondering if all the anagrams in the world could not be found in three lines of any author" (1907:68-69). Transcoding from one level to another may require the rewriting of rules. As one switches to a lower level, a rule is rewritten to show the expansion of "one symbol" into a whole.text and vice versa. Referring to context bound rules primarily, it is held that the order of rules for synchronic synthesis of the text may coincide with diachronic development (8.0.1). These observations offer a critique which may be applied to cross-cultural studies carried out in the United States where the "factor" retains its independence at various levels and thus can be added, subtracted etc., rather than being absorbed into higher levels or expanded at lower ones.

39 139 The possibilities of receding from level to level are qualified by the following reservations. The dynamic relationship between levels makes problematical the possibility of experimentally dividing the different stages in the process of synthesizing a text since "its surface structure, which is defined by formal limitations, may influence the deep figurative structure" (8.0.2). (See our Section 2.3 on reconstruction of texts.) Underlining this point is the dialectical relation of rules and variety in poetics, where it is held that the greater the limitations imposed on the form, the greater the increase in flexibility of poetic language (8.0.2). As expanded elsewhere by Ivanov "with an increase in the number of formal restrictions as in terza rima, an increase in the number of synonomous expressions is imperative, and that is achievable thanks to metaphorical figurative word usages, unusual word combinations and so forth" (1975:188). The relevance of this insight for semiotics of culture in general needs to be explored. For example does this view imply that certain types of flexibilities necessarily are a part of the more regulated cyclical cultures organized by mechanical rules, as opposed to cultures governed by statistical rules, and if so how would the relationship between linguistic and other levels be affected? As Ivanov has noted, Bakhtin showed that each semiotic system has its own "language", one example of which was carnival language which "expressed the unitary (but complex) carnival perception of the world" that cannot be translated fully into verbal language but rather into the language of artistic images (1975: 189). We refer here to what appears to be a particularly pertinent example of the above dilemmas, namely Turnbull's analysis of the music of one of the world's most primitive groups, the BaMbuti Pygmies of the Ituri Forest, whose vocal music gives the impression of remarkable independence for individual performers, including great flexibility and wide improvisation as well as virtuosity. At the same time the music also shows a highly developed harmonic, rhythmic and technical complexity governed by strict rules. Yet, as Turnbull tells us, in spite of impressions of complexity in some respect the musicianship of these people is simple "because given a dominant mode, whatever notes are sung at the same time result in the same chord. There can be no disharmony. The genius of the BaMbuti music lies in the fact that given this very limitation they can do so much with it". Thus, "the thing to look for is the ability of the BaMbuti to improvise within this self-imposed limitation". Similar parallels between wide variations and strict rules, noted in harmony, exist in the areas of melody and rhythm. Thus highly individualized melodic forms are combined with the insistence on descending melodic patterns, and varied and individualized rhythmic patterns must all compose a four or eight beat sometimes with syncopation. An important area of widest freedom from rules becomes the structure of the language itself,

40 140 on the phonemic, morphemic and syntactical levels, since songs may have a general meaning, or no words or very simple words, and vowels and consonants are changed "mercilessly" to please the ear. "The sound itself is what is important, and the Pygmy listens intently to his own music, even as he is making it" (Turnbull, 1958:1-2). Apparently in Lotman's terms, the primarily transformational form of narration demonstrated by music is prior to the linear linguistic one in this particular musical system, and in Ivanov's terms the rigidity of the rules governing the musical forms on the surface structure is associated with an increase in flexibility at deeper levels. 4. Concluding remarks Considering cultures as a single semiotic whole, the "Theses" distinguishes two mutually opposed mechanisms: on the one hand the tendency toward diversity, toward the increase in differently organized semiotic languages (the polyglotism of culture), and on the other hand the tendency towards uniformity (the attempt of a culture to interpret itself or other cultures as uniform and rigidly organized). The striving towards organization is contrasted to the "need of relatively amorphous formations which only resemble structure" (9.0.0). For example highly regular artificial sign systems which act as a model of the organization of culture, such as ranks and badges in Russian society under Peter the Great, are contrasted to the "motley irregularity of the real life of those times" (9.0.0). Extending this concept, it is held that the idea of regularity, which grammars of natural languages may embody, as well as other regulating texts such as instructions and directions, may represent a systematized myth created by the culture about itself (9.0.0). The implications of this statement for the real, as opposed to the possibly mythological, role of the regulating forces needs to be investigated. As Ivanov has commented, the notion of subjectivity and variability of real life was stressed by Bakhtin who wrote " 'the constitutive factor' for a linguistic form as a sign is not at all its signal of self-identity but its specific variability, and for the understanding of the linguistic form the constitutive factor is not recognition Of the same thing' but understanding in the strict sense i.e., orientation in the given context and the given situation, orientation in formation and not Orientation' in some immobile existence" (Ivanov, 1975:196, quoting Volosinov, 1972:83). Bakhtin's approach emphasized not only the point of view of the listener but most importantly the point of view of the speaker, and their interrelations, a view which was further developed by Jakobson and forms the basis of his communication model applicable to linguistic communications, poetic texts, and in fact

41 141 communication in general (Ivanov, 1975: ). Furthermore, as has been noted, the aesthetics of Mukafovsky assumes the violation of norms as essential to the aesthetic function, which is posited as universal in human behavior. We have also noted Lotman's view of the limitations of message transmission due to the individual variations affecting codes (1974a: ) and we might add Kristeva's stress on semiotic practices (1975). These notions of variability in real life have their counterparts in the linguistc concepts of Chomsky (infinite variations on the level of performance) and in anthropological concepts such as Turner's liminality, Geertz' thick description, etc., a whole area of inquiry that needs much further investigation. More attention in the "Theses" is devoted to mechanisms which impart regularity. These include the regulations mentioned above, various artificial languages (9.0.1), myths of a culture about itself (9.0.2), and orientations of the culture towards a particular semiotic system (as for example towards writing, or toward mathematics in the Age of Reason) where the structural principles of such a system "penetrate the other structures and the culture as a whole" (9.0.3). (The opposing tendencies in culture, while not neglected, need to be investigated further.) The emphasis on general orientations imparting regularity are reminiscent of Ruth Benedict's relativistic dominant drives and overall configurations (1934). Relativistic tendencies suggested themselves from the first discussion of the inner and outer approach to the conclusion of the "Theses" which asks (9.1.0) whether modern structural semiotics studies, such as this one, represent not only a metatext (an instrument for the study of culture in order to cognize itself) but also an object of the culture, indeed a text of the culture of itself (9.1.0), recalling Levi-Strauss's view of history as a particular myth about.itself. Thus, in the last dialectical twist, we are presented with a text on two levels at once, and a text perceived from two points of view, the inner and the outer. To decide to what extent this text (the "Theses") is to be understood as a text of Slavic culture as opposed to a metatext would be too ambitious a task. However, since all metatexts must be to some extent particular cultural texts, claims of universality must be independently assessed, regardless of their origin in a specific cultural milieu. The compactness of the stimulating composite statements which compose the "Theses" belies the richness and heuristic insights which lead us on to the most various and inexhaustible paths taken by these ambitious researchers in their explorations of the semiotics of culture THE ESSAYS The "Theses" are followed by eleven papers from Slavic and Western

42 142 scholars, unified by their semiotic interests, each one contributing specialized concepts and applications of some of the questions touched upon in the "Theses". While they variously relate to, develop, alter, or expand upon broad concepts and implications of the material contained in the "Theses", not all of the contributions were written by members of the Tartu-Moscow group; and thus not all the contributions were written from a position of prior knowledge of the "Theses". It is thus difficult to group the essays into a productive taxonomy directly related to the various topics of the "Theses", because some do not lend themselves to such an organization, although all of them touch on the materials of the "Theses" at least tangentially. For these reasons, we shall take the essays one by one, rather than superimposing upon them an organization which would be partially artificial. 1. The first article, Jan Van Der Eng's "The Device: The Central Factor of the Semantic Construction of the Narrative Text" ("Priem: central'nyj faktor semantifceskogo postroenija povestvovatel'nogo teksta") (29-58) reevaluates and redefines, in semiotic terms, several of the concepts utilized by the Russian Formalists as they relate to specific features of the semantic construction of prose texts. Van Der Eng discusses the linear-temporal aspect of the decoding process as applied to verbal art, where the total meaning of the work cannot be perceived until the entire work has been decoded, an approach developed earlier by Mukafovsky (1940:22; 1946: 112). Van Der Eng sees "the artistic activity as a pragmatic problem in semiotic terms. It is a collaborative process between the author and the receiver, and their successive generations, which uncover hidden and dynamic patterns of the work. Thus the essay addresses several topics of the "Theses": that of the linearity and sequentiality of the discrete (verbal) text, the complex problem of cultural encoding and decoding as specifically applied to texts of the verbal arts, and that of contextuality. Following Sklovskij's and Mukafovsky's earlier premises, Van Der Eng sees the motifs of a literary text as acting through the entire text in progressive and regressive directions, taking on ever-new meaning in the process of decoding, setting up ever-new expectations (Lotman's "forks"), as the receiver (decoder) and the author "collaborate" to bring them to life. Literary texts, as all cultural texts, are not context-free, but are related, in complex ways, to other texts belonging to the same system (e.g. other literary texts), or belonging to other systems (other culture texts) which raises many questions including that of inter-systemic and inter-level translation or conversion discussed in the "Theses". As had Mukafovsky and Jakobson before him, Lotman has stressed the relational quality of the "device" within the context of other systems, introducing the concept of "minus-device" (Lotman, 1970a:121), i.e. the absence of a normative, and therefore expected, device. This absence becomes in itself a marked feature

43 143 (for example, free verse in a poetic tradition where rhyme is, or was, a norm, is a minus-device, as opposed to mere lack of rhyme in pre-rhymic poetic traditions [e.g. Greek poetry, the Russian by liny]). Van Der Eng's "functional device" expresses the growing trend to expand such concepts as that of the device beyond their early connotation. Thus the functional device becomes the central factor of the semantic unification and dynamic structuration, of the narrative work since it establishes semantic ties between the various elements of the fabula, thus providing a focus for the structuration and unification of the work of art (Mukafovsky's "semantic gesture") realized through the psychological intention of the decoder. 2. Mojmir Grygar's "Cubism and the Poetry of the Russian and Czech Avantgarde" ("Kubizm i poezija russkogo i cesskogo avangarda") (59-102) applies a semiotic view of culture to the study of the interrelations between different art forms within the same culture and time, eschewing the traditional causalistic and genetic approach and the tracing of influences, typically a part of traditional comparative studies. The semiotics of culture, as developed in the "Theses" and earlier, directs us to the investigation of inner-structural and typological similarities of strategic levels of cultural phenomena and systems. Such similarities follow from the assumption that all such levels partake directly or indirectly in social communication and transmission of information within an autonomous, but not immanent, structural whole, namely a culture. Drawing on Jakobson's insights into the interrelation of the various art media (Jakobson, 1919; 1935; 1964; 1970b), and Mukafovsky's earlier fruitful approach to inter-art comparisons, Grygar discusses the work of the leading theoretician of Czech surrealism, Karel Teige. While, as noted earlier, Roman Jakobson, in earlier discussions (1964; 1967), had distinguished the temporal arts (verbal arts and music) from the spatial (visual) arts, he pointed out that both auditory and visual signs occur in space and time, the temporal dimension being dominant in auditory signs, while the spatial dimension takes priority for visual signs. A fundamental problem that inter-art translation must consider, that has preoccupied Mukarovsky and Jakobson, and which the "Theses" take up, is that of the varying principles of semiosis of the different arts, which necessarily qualifies the isomorphism of structures and texts belonging to different artistic systems, even if functioning within the same culture. Rather than searching for causal or chronological relationships, diffusion and borrowing, a semiotic study of inter-art relationships attempts to isolate inner-structural and typological similarities and specific differences between subsystems. Thus, the interrelations between such autonomous systems as painting and literature are seen as permutations, rather than "borrowing" or mechanical transfers of external impulses. From these points of view, Grygar attempts to analyze structural interrelationships between painting and poetry of the

44 144 Czech and Russian avantgarde of the 1920s and 1930s, and similarities as well as divergences between parallel literary movements of both countries. As Jakobson has already pointed out (1970a), the novelty of the avantgarde of the 1920s and 1930s lies in a fundamental shift of semiosis to an "introversive" semiosis, a dimension implied in Peirce's triad, where nonfigurative art is a kind of self icon. When no object exists in reality, or when the grasping of the relation between the sign and an objective denotatum is extremely difficult, the inner form of the art work becomes its own object. According to Jakobson, signs based on "introversive" semiosis (such as music [except for program music] and glossolalic poetry) possess a minimal referential component. Such an orientation describes those autolelic artistic schools of the early twentieth century, such as the Symbolists and post- Symbolists in the verbal arts and the post-impressionists in the visual arts. The rapid development of these trends, and their widespread reception, Grygar reminds us, was directly related to the chaotic condition of European culture at the threshold of the present century, when scientific advances, new philosophical directions, and the political upheavals after the first World War all contributed to the further disintegration of already questioned traditional codes, norms and restraints, and to the weakening of traditional hierarchies. (How all these philosophical and cultural trends affect modern art, particularly in terms of views of time, is the focus of Ivanov's article on time, discussed earlier as well as below.) The revolutions of avantgarde art led not only to the above-mentioned tendency towards introversive semiosis, but also to the dissolution, or relativizing, of the traditional boundaries existing between the different arts and between art itself and its periphery. Consequently, numerous new combinations were the foci of experimentation. The cubist painters dramatized a combination of the visual arts and language, by incorporating verbal material into their paintings in the shape of collages, and Kandinsky wrote poetry in which the visual elements of shapes and colors played a dominant role. The futurist poets expressed symbiotic relationships between poetry and the visual arts by emphasis upon graphic, optic, and palpable aspects of the poetic text. Thus poems were created in special graphic shapes, or were even printed on specially textured paper. Xlebnikov linked the individual sounds of language to certain geometric and kinetic concepts, recalling Malevio' view of minimal spatial units which appear to be the basis of combination of all "suprematist" art (cf. Malevic, 1975; Vallier, 1975:12). The attempt to break traditional inter-art boundaries, and those between art and non-art, thereby breaking out of the closed sphere of artistic forms and norms, has its counterpart, as Grygar noted, in other areas of social consciousness (e.g. Freud, Bergson, etc.), and neatly corresponds to similar trends in the sciences, as well as in the social and humane sciences.

45 The interrelation of art and other aspects of cultural systems continues to be the focus of the outstanding contribution of V.V. Ivanov's "The Category of Time in the Arts and the Culture of the Twentieth Century" (Kategorija vremeni v ikusstve i v kul'ture XX veka)( ), where new views of time, as they organize new semiotic dimensions of the arts, are the dominant theme. Parallels are drawn between contemporary perceptions of time in the sciences and philosophy, and in the avantgar de arts where the category of time becomes important not only as a thematic component, but also as a basic structural principle. Theories advanced by A. Eddington ("time arrow") are compared to those of the Russian philosopher P.A. Florenskij, who, in 1927, had depicted the essence of culture as ektropy (logos), which he opposed to entropy (chaos), an opposition which is discussed in the "Theses". Social information transmission systems, a part of culture, are - as Ivanov notes in large measure subject to the same laws of the growth of entropy in time as are physical systems, as is clearly evidenced in natural languages, where the increasing difficulty of digesting the growing amount of information, and the growing standardization of mass culture, create redundancy and entropy. Ivanov depicts in greater detail the fundamental opposition which the "Theses" poses, that of the contrast between archaic and later cultures. In Ivanov's archaic cultures the past is mythologized, and present time is cyclical, while in modern cultures (where behavior is regulated by memory), the past is "real" and historical time and present time is linear. Earlier, Jakobson and Bogatyrev had contrasted the stability and normativeness of folk art to the dynamism of "high" art, which they interpreted as a result of the censorship of the collective (Jakobson, Bogatyrev, 1929). Similarly, Ivanov's archaic systems are controlled by rules which exhaust all variant possibilities. Mythological themes are inherited and incarnated by utilizing fixed ritual formulae, whereas in the more evolved cultures, the underlying principle of the arts is the struggle against entropy, and hence the search for the statistically least frequent devices and themes, which consequently bear a maximum of information. However, anthropological findings have increasingly modified the view of the static quality of archaic art, as is shown by some Russian semioticians (Meletinskij, Toporov, etc. and by many other Eastern European and Western scholars). Thus in primitive and folk arts, spaces of indeterminacy also exist, and are also filled by innotative changes and variations within the limits of the specific structures of archaic art systems. (Cf. I. Winner 1973:17-18 and our sections 2.3, 2.4.3, 2.4.4, and 3.3 for a discussion of the implications of this theoretical problem.) More attention is paid to the avantgarde arts of today which share a preoccupation with the concept of time with such philosophers as Heidegger and Sartre, and such literary theoreticians as George Poulet. Thus the verbal

46 146 arts are recast by Proust, Joyce, Wells, Faulkner, Gide, Remizov, and many others, whose time is no longer one-directional, but rather a relative force which can move forward and backward, or which can project different time slices simultaneously (as for example Joyce's stream of consciousness and Faulkner's construction in his The Sound and the Fury). According to Ivanov, as we have noted earlier, the art which by its very nature and the technique of its production offers greatest possibilities for exploration of time structures is the cinema, where synchrony and diachrony are inextricably interwoven and superimposed upon each other. The motion we see on the screen, which appears to be a dynamic propulsion of time, is in reality a composite of hundreds of thousands of static pictures following upon each other in too close a succession to be detected by the eye which is "deceived" into seeing a sequence of dynamic events. Modern film makers, such as Eisenstein, Fellini, Bergman, Resnais, have used the film for an exploration of the problem of time. For example Resnais' Last Year in Marienbad is based on the structural principle of the superimposition of two discrete time periods into one consciousness. Cubist painters have also constructed new timespace relations by sacrificing traditional principles of perspective and presenting spatial relationships, normally perceived in linear time, simultaneously. 4. Fortunately, there is a second article by Ivanov, offering vivid applications of particular views, entitled "The Category of 'Seen' and 'Unseen' in the Text: Once More About the East Slavic Folklore Parallels to Gogol's 'Vif " (Kategorija 'vidimogo' i 'nevidimogo' v tekste: E e raz o vostoönoslavjanskix fol'klornyx paralleljax k gogolevskomu 'Viju' ) (151-76). Parallels are found between the motifs of East Slavic folklore and Gogol's "Vij", as well as between folk archetypes and "high" literature. Developing a theme of an earlier article (Ivanov, 1971), where the opposition seeing/not seeing was demonstrated to be a transformation of the opposition visible/ invisible, examples are given from Slavic and other folklores and confirmed by old Icelandic data. The semiotic relation of these oppositions is specified in Slavic grammatical categories conveying unwitnessed action; for example in the Bulgarian and Macedonian "evidential" forms. According to Ivanov, the ritual role of eye-closing by the Thompson Indians reported by Levi- Strauss seems to have similar grammatical function (Levi-Strauss, 1971:357). 5. Jurij Levin's "The Lyric from the Communicative Point of View" ("Lirika s kommuniktivnoj toski zrenija") (177-97) attempts to formulate a taxonomy of lyrical poetry oriented upon the actors, both within and outside of the text. Poetry is approached from three points of view: (1) the inner view, relating to the personages active within the texts themselves (the hero, the "I"), (2) the external view, focusing on the concrete sender and the concrete receiver of the communicating text, and (3) the inter-

47 147 mediate view, concerned with the implicit author (Mukafovsky's author's personality; see Mukafovsky, :228). The paper represents an attempt at a new taxonomy of lyrical poetry from this point of view. 6. Maria Renata Mayenowa's "An Analysis of Some Visual Signs, Suggestions for Discussion" ( ) describes non-verbal sign structures needed for a methodology which may be applied to texts which combine verbal and non-verbal character. Four categories of signs are distinguished: 1) diagrams, limited to relation models requiring the aid of symbolic signs (captions); 2) maps and hybrid sign systems, composed of iconic signs (the map as a model of spatial relationships); indexical signs (the longitudinal and latitudinal lines on the map which mark relations to the earth's poles); linguistic and visual signs which are symbolic (proper names of rivers, towns, countries, etc.); and partially conventional signs indicating water, lowlands, mountains, etc. by various shadings of blue, green, brown, respectively); 3) paintings without verbal texts which are iconic-metaphorical (Mayenowa's example being a drawing of the sixteenth-century Iconologia of Cesare Ripa, "the Academician"); and 4) paintings without symbolic sign elements, or iconic signs of the classical type, which lack narratives, since temporal or causal links are not shown in the painting itself. Reflecting the implications of secondary modelling systems, Mayenowa holds that a purely iconic message cannot be fully decoded without verbal narration, which elucidates cultural norms (e.g. the traditional division of the Renaissance painting into an upper and a lower half, the upper half representing landscape and the lower, human affairs, and the "action" of the painting). 7. In "Literature as Information. Some Notes on Lotman's book Struktura xudozestvennogo teksia" ( ), Jan M. Meijer finds Lotman's approach lacking in convincingness. Criticized are Lotman's view of art as information, the concept of secondary modelling systems built "like a language" (po tipu jazyka), and Lotman's use of the concept of text. Some of Meijer's criticisms have been answered by Lotman himself in later works, where the concept of secondary modelling systems has been greatly delimited, allowing nondiscrete texts their own autonomy. Other criticisms seem exaggerated within the context of the particular work of Lotman examined, as well as within the context of the tradition of the semiotics of poetics. Meijer holds that Lotman's analysis overlooks the importance and lack of specificity of endings in artistic, as opposed to non-artistic, verbal communication (216). But the literature of structural and semiotic poetics, stemming from Mukafovsky and Jakobson, as well as many works of Lotman, underlines the significance of both beginnings and endings of artistic works (e.g. Lotman, 1966). Meijer also rejects Lotman's assertion that the specific feature of the artistic sjuzet is the co-presence of several meanings at the same time for every one of its elements. But here it would seem that

48 148 Lotman is simply following, and developing further, Mukafovsky's view of the polysemantic character of the work of art as a sign, which is the result of the weakening of the relationship of the sign to a denotatum. Meijer criticizes Lotman for invoking the intention (he calls it "will") of both author and reader in the equation of art. He argues that "if a kitsch text is both encoded and decoded as art", that does not make it art. This statement seems to reflect a misreading of Lotman, since he has not attempted to establish external criteria for the ontology of art. Rather Lotman defends the autonomy of art and its irreducibility, as well as its interrelationships with other systems, a view posited earlier by Jakobson and Tynjanov (1928). As Mukafovsky held (1937b:223), what he called "external stimuli" enter the artistic system through the intermediacy of the individual, through both the encoder and successive generations of decoders. There is projected into the work of art the psychological energy, or intentionality (zamernosf) (Mukafovsky, 1943) of both encoder and decoders, binding the components of work into a semantic cohesiveness. Encoding and decoding are both subjective and conditioned by cultural norms (cf. T. Winner, 1973b; 1976 forthcoming). Thus it is that constructions first intended as utilitarian objects are "read" today as art (e.g. archaeological finds of pots in a museum). This is the kind of mutability Lotman is describing. The philosophical issue of the ontology of art is not argued in Lotman's work. 8. In his second article, "Verbal Art As Interference Between a Cognitive and an Aesthetic Structure" ( ), Meijer sees the structure of the work of literature as a form of interference between a cognitive and an aesthetic structure. The argument holds that, since the material of verbal art is language the function of which is denotive, a work of literature must create tension between cognitive and aesthetic elements. The two structures are said to interfere with each other, just as in physics when waves meet, the ensuing interference is either supportive or counteractive. Metaphorically then, a literary structure is composed of two "wavelengths" (319), the cognitive and the aesthetic. But it seems that Meijer's "aesthetic" component is, in fact, merely form, while the "cognitive element" becomes meaning, a separation which would not be acceptable to the signers of the "Theses". This argument was traced earlier in this discussion where Ricoeur and Geertz relegated structuralism to formalism and where Ricoeur called for a synthesis of form and content which, in fact, has already been to a considerable extent achieved. For example, Meijer cites the line: "Eeny, meeny, miny mo", which he holds demonstrates merely the aesthetic principles of parallelism (320). But the semantic level of this line does exist when we return the phrase to its full context, when it becomes a part of a two-line counting rhyme, where each "lexical" unit would stand for a numeral, "eeny" would stand for the numeral one, "miny" for two, etc.

49 149 The seemingly meaningless first line precedes a second rhyming phrase "catch a tiger by the toe". While it is true that "the main carrier of meaning in art is the cognitive structure" (332), it need not follow that artistic tension is produced by the clash (interference) of cognitive-aesthetic principles, since every element serves both purposes, creating the polysemic, ambiguous and flexible character peculiar to art. 9. Complementing Ivanov's treatment of "seen/unseen", is V.N. Toporov's "On the Structure of Dostoevsky's Novel in Connection with Archaic Schemes of Mythological Thinking (Crime and Punishment)" (0 strukture romana Dostoevskogo v svjazi s arxaicnymi sxemami mifologiceskogo myslenija) (Prestuplenie i Nakazanie) ( ). This work attempts to demonstrate the underlying "deep structural" elements in Dostoevsky's novel. Toporov sees Dostoevsky's novel exceptionally related, through exact correspondences, to the texts of archaic, mythological cultures, where a crisis situation, frequently of cosmological proportions, requires a solution under the threat that the organized, predictable, cosmic principle (the "seen") threatens to become transformed into chaos or entropy (the "unseen"). Following Propp, the solution of the task is seen as a test, or some form such as a duel, won by physical means or by greater wits. As Bakhtin has shown, the carnival also expresses this mythological drama which is reflected as well in some literary texts, including those of Dostoevsky. For Dostoevsky, such schemes are shorthand plot mechanisms, where the space of the adventure novel is significantly widened, permitting great independence of the action of the hero, as he is confronted with the elements within the plot. The novel space is split into a great number of interpenetrating, but semi-independent, units (temporal, spatial, causal, etc.), the transitions between which are characterized by maximal entropy, that is minimal predictability. In Dostoevsky's idiom, time and space are not limited to framing, but become mythic. Thus they are active determinants of the hero's behavior (for example, the sunset becomes a direct influence upon the hero). They also actively participate in creating an atmosphere (for example, the sunset creates a mood of indeflniteness and the fantastic, and space creates a feeling of chaos in the image of the city of St. Petersburg). The basic oppositions in the structure are spatial oppositions, central or inner/peripheral or outer, and a series of graduated oppositions lying between inner/outer which are paralleled by the psychological-ethical contrast of minimal freedom/maximum freedom, where "inner" comes to mean minimal freedom, and "outer" means maximal freedom. The oppositions are actualized by the movements of the hero, which are in turn associated with the hero's internal moral changes, as in mythic texts. Both elements of the pair of oppositions have linguistic peculiarities of description which become so-to-speak iconic of novelistic space. According to Toporov,

50 150 analogies of Dostoevsky's novels to mythic texts also extend to the structure of abundant loci comunes, which include a great number of semantically (and frequently symbolically) marked pieces of texts, appearing and reappearing in various contexts, in the same work or in several novels (repetitions, doublets, situational rhymes, etc.). Names in Dostoevsky's novels also demonstrate a technique peculiar to myth, that is a carnivalesque bricolage, where proper nouns and namings are difficult to distinguish, i.e. where the names are semantically active (e.g. Zametov: zametil, etc.). Other elements are the peculiar use of significant numbers. 10. In S. Zolkiewski's essay, "Des problemes de la litterature populaire" ( ) he attempts to arrive at a cultural-semiotic and functional delimitation of popular (ludic) literature, or para-literature within the framework of an over-all classification system of culture texts, where different types of texts of a culture can be confronted and classed as subtexts of an overarching cultural system of science. 11. The final essay, Krystyna Pomorska's "On the Segmentation of Narrative Prose" ("0 clenenii povestvovatel'noj prozy") ( ) attempts a structural definition of the difficult area of segmentation of prose texts, using three works, Maria Dabrowska's novel "Days and Nights", Pushkin's "Voyage to Arzrum", and Pasternak's "Oxrannaja gramota". The criteria are the selection and organization of the material in terms of the paradigmatic and the syntagmatic axes. 12. The above essays are united by a common theme, namely the semiotics of aesthetic texts seen as cultural texts interpretable on various levels of culture. Some attempt to account for the translatability of cultural material from one level to another including translations from myth to verbal "high" art (Ivanov, Toporov), translations between various art systems (Grygar and Zolkiewski) and translations between various sign systems (Mayenowa). Some authors attempt to provide answers to specific problems of structural semiotic analysis of the verbal arts (Meijer, Van Der Eng, Pomorska), and some of the studies offer specific applications of individual points of the "Theses" (Ivanov, Toporov, Mayenowa, Zolkiewski, Grygar). If the heuristic potential of the "Theses" is to be realized even in part, they should stimulate a continuing series of theoretical, methodological and critical essays as well as the widest possible applications. Individual aspects need to be approached variedly and both assumptions and concrete materials relating to the arts and mythology, as well as to all other cultural systems, should be evaluated and extended. Only then can the fruitful initial expectations of the "Theses" be fulfilled.

51 151 REFERENCES Bakhtin, Mixaü (see Volosinov, V.N., 1972) 1963 Problemy poetiki Dostoevskogo (Moscow) Tvorcestvo Fransua Rable i narodnaja kuvtura srednevekov'ja i Renesansa (The Work of Frangois Rabelais and Medieval and Renaissance Folk Culture) (Moscow). Baran, Henryk 1975 Structuralism. Soviet Studies in Literature. Spring-Summer 1975, IX, 2-3 (New York: International Arts and Sciences Press). Benedict, Ruth 1959 [1934] Patterns of Culture (Boston: Houghton Mifflin). Bogatyrev, Petr G Functions of Folk Costume in Moravian Slovakia (The Hague: Mouton). (Original title: Funkcie kroja na moravskom Slovensku) (Turciansky sv. Martin: Matica Slovenska, 1937). Cassirer, Ernst 1954 [1944] An Essay on Man (New York: Doubleday) The Philosophy of Symbolic Forms (New Haven: Yale University Press). Charbonnier, G Conversations with Claude Levi-Strauss (London: Cape) (Original title: Entretiens avec Claude L vi Strauss [Paris: Plon]). Damisch, Hubert 1975 "Semiotics and Iconography". In The Tell-Tale Sign, Sebeok, ed., 1975: Douglas, Mary 1970 Natural Symbols (New York: Pantheon). Eco, Umberto 1975 "Looking for a Logic of Culture". In Sebeok, ed., 1975: A Theory of Semiotics (Bloomington: Indiana University Press). Eimermacher, Karl 197la Teksty sovetskogo literaturovedceskogo strukturalizma (Texts of Soviet Literary Structuralism) (München: Fink). 197Ib "Entwicklung, Character und Probleme des sowjetischen Strukturalismus in der Literaturwissenschaft" (Development, Character and Problems of Soviet Structuralism in Literary Scholarship). In Eimermacher, 1971 a: Evans-Pritchard, E.E Anthropology and History (New York: Humanities Press) "Nuer Spear Symbolism", Right and Left, Essays on Dual Symbolic Classification, Rodney Needham, ed. (Chicago: University of Chicago Press), (Reprinted horn Anthropological Quarterly 1,1053). Florenskij,P.P. n.d. Analiz prostranstvennosti v xudozestvenno-izobrazitel'nyx proizvedenijax (Analysis of spatiality in works of visual art), cited in Ivanov, 1973a, islo kak forma (Number as Form), unpublished ms. in possession of V.V. Ivanov. Cited in Ivanov, 1973a. Geertz, Clifford 1972 "Deep Play: Notes on the Balinese Cockfight", Daedalus (Winter), The Interpretation of Cultures (New York: Basic Books). Greimas, A.J., R. Jakobson, and M.R. Mayenowa, eds Sign. Language. Culture. (The Hague: Mouton). Hamburg, C Symbol and Reality (The Hague: Nijhoff).

52 152 Hertz, Robert 1909 "The Pre-eminence of the Right and Left Hand: A Study of Religious Polarity". In Right and Left, Essays on Dual Symbolic Classification. Rodney Needham, ed. (Chicago: University of Chicago Press), Holenstein, Elmar 1975a "Jakobson phenomenologue?" L 'Arc 60, bRoman Jakobsons ph menologischer Strukturalismus (Frankfurt a/main: Suhrkamp Taschenbuch). Ivanov, V.V "K semioticeskomu analizu mifa i rituala (na belorusskom materiale)" (Concerning the Semiotic Analysis of Myth and Ritual [based on Belo-russian Material]) in Greimas, Jakobson, and Mayenowa, eds. 1970, "Ob odnoj parallel! k gogolevskomu 'Viju' " (Concerning One Parallel to Gogol's 'Vij'), Trudy po znakovym sistemam V (= Ucennye zapiski tartuskogo gosudarstvennogo universiteta vyp. 248) (Tartu), a "Kategorija vremeni v iskusstve u kufture XX veka" in Van Der Eng and Grygar, eds. 1973, (English version: "The Category of Time in Twentieth Century Art and Culture", Semiotica 8/1:1-45). 1973b "Kategorija 4 vidimogo' i 4 nevidimogo' v tekste: esce raz ο vostocnoslavjanskix paralleljax k gogolevskomu 'Viju' " (The Category of 'seen' and 'unseen' in the Text: Once more About the East Slavic Parallels to Gogol's 'Vij'). In Van Der Eng and Grygar, eds. 1973, "The Significance of M.M. Bakhtin's Ideas on Sign, Utterance, and Dialogue for Modern Semiotics". In Baran, ed., 1975, Original title: "Znacenie idej M.M. Baxtina ο znake, vyskazyvanii i dialoge dlja sovremmenoj semiotiki. Trudy po znakovym sistemam VI (= Ucennye zapiski tartuskogo gosudarstvennogo universiteta vyp. 308) (Tartu). Jakobson, Roman 1919 "Futurizm" (Futurism). Iskusstvo (Moscow), August "Randbemerkungen zur Prosa des Dichters Pasternak" (Marginal Observation concerning the Prose of the Poet Pasternak), Slavische Rundschau (Prague) VII, "Results of a Joint Conference of Anthropologists and Linguists", Supplement of International Journal of American Linguistics XIX, 2, April. See Jakobson 1971a, "Closing Statement: Linguistics and Poetics", Style in Language, Thomas Sebeok, ed. (Cambridge: MIT Press), "On Visual and Auditory Signs", Phonetica XI, "Quest for the Essence of Language", Diogenes 51: "About the Relation between Auditory and Visual Signs. Models for the Perception of Speech and Visual Form: Proceedings of a Symposium, Wathen-Dunn, ed. See Jakobson, 1971b: "Language in Relation to Other Communication Systems", see Jakobson, 1971b: a "Language in Relation to Other Communication Systems", Linguaggi nella societa e nella tecnica (Milano: Edizione di communita). 1970b "On the Verbal Art of William Blake and Other Painters", Linguistic Inquiry 1,1: a Selected Writings I (The Hague: Mouton). 197lbSelected Writings II (The Hague: Mouton). 1975a "Coup d'oeil sur de developpement de la semiotique" (= Studies in Semiotics No. 3) (Bloomington: Indiana University Publications). 1975b "Structuralisme et teleologie", L 'Arc (Paris), 60: Jakobson, R., and P. Bogatyrev 1929 "Die Folklore als eine besondere Art des Schaffens" (Folklore as a Special Kind of Creation). Donum Natalicum Schrijnen (Nijmegen-Utrecht),

53 153 Jakobson, R., and J. Tynjanov 1928 "Problemy izucenija literatury i jazyka" (Problems of the Study of Literature and Language) Novyj Lef 12: English translation in: From Marx to Levi-Strauss, Richard and Fernanda de George, eds. (New York: Doubleday) 1972: Kluckhohn, Clyde 1968 "Myths and Rituals: A General Theory", Studies on Mythology, A. George, ed. (Homewood: Dorsey) (Reprinted from the Harvard Theological Review, Vol. 35,1942). Kristeva, J., J. Rey-Debove, and Donna J. Umiker, eds Essays in Semiotics (The Hague: Mouton). Kristeva, Julia 1975 'The System and the Speaking Subject", The Teil-Tale Sign, Thomas Sebeok, ed. (Lisse: Peter de Ridder Press), Langer, Susanne K Feeling and Form (New York: Scribners). Leach, E.R [1961] Rethinking Anthropology (= London School of Economic Monographs on Social Anthropology, No. 22) (New York: Humanities Press). Levi-Strauss, Claude 1963 Structural Anthropology (New York: Basic Books) Mythologiques (Le cru et le cuit) (Paris: Plon) The Savage Mind (Chicago: University of Chicago Press). Translated from La pensee sauvage (Paris: Plon) Mythologiques (Du miel aux cendres) (Paris: Plon) Mythologiques (Origine des manieres de table) (Paris: Pion) Mythologiques (L 'Homme nu) (Paris: Plon) "Structuralism and Ecology", Social Science Information:!'-32. First published in Barnard Alumnae, Spring Tristes Tropiques (New York: Atheneum). Translated from the French (Paris: Plon, 1955). Levy-Bruhl, Lucien 1910 Les fonctions mentales dans les societes inferieures (Paris) La men tali t e prim itive (Paris). Lomax, Alan 1962 "Song Structure and Social Structure", Ethnology 1: Folk Song and Style in Culture (Washington D.C.: American Association for the Advancement of Science) "The Coevolution of Expressive, Productive, and Cultural Systems". Report at the Annual Meeting of the American Association for the Advancement of Science. Boston, February 22. Lotman, Jurij M "O modelirujuscem znacenij ponjatii 'konca' i 'nacala' v xudozestvennyx tekstax" (On the modelling meaning of the concepts 'beginning' and 'end' in artistic texts). Tezisy dokladov v torof letnef skoly po vtoricnym modelirujuscim sistemam (16-26 avgusta 1966 g) (Tartu): a Struktura xudozestvennogo teksta (The Structure of the Artistic Text) (Moscow: Iskusstvo). Reproduced in Brown University Slavic Reprints IX. l91qbmaterialy k kursu teorii literatury. vyp I. Tipologija kuvtury (Materials for a Course on Literary Theory. I. Typology of Cultures). Tartu "O metajazyke tipologiceskix opisanii kul'tury" (On the Metalanguage of typological descriptions of Cultures) in J. Rey-Debove, ed. 1973: a"Sign Mechanism of Culture", Semiotica 12:4, (First published in Semeiotike: sbornik state/ po vtoricnym modeliruju$ im sistemam [Tartu 1973: ].) 1974b "Observations on the Structure of the Narrative Text" Soviet Studies in