% FOUR Performance Codes and Theatrical Conventions Bottom: Some man or other must present Wall; and let him have some plaster, or some loam, or some rough-cast about him, to signify wall. Shakespeare, A Midsummer Night's Dream VVho is so stupid as not to know the difference between being and pretending? Niccolo Barbieri, La supplica discorso famigliare a quelli che trattano de' comici 4.1. THE CONCEPT OF "CODE" IN RELATION TO THEATRICAL PERFORMANCE In the last chapter I defined the theatrical structure of performance as the single combination of (more or less) multiple codes, which means that these codes can also be found in other combinations in other texts (whether related to performance or not). I will now attempt a closer examination of the ''simple'' components of the textual structure of performance that constitute the /performance codes/. For the moment I will designate all of them as such, and will describe their characteristics and modes of operation. First, however, I must pause to ask if it is appropriate to use the term /codes/ in relation to theatrical performance, and if so, in what sense, i.e., according to which of the many meanings commonly attributed to the term. ■: The notion of the code has been extended to a very general application by such scholars as Prieto and Eco in a way that is very useful for the purpose of the present mquiry into the basic conditions of coding. Accordingto the definitions offered by these writers, the only prejudicial property of a code (i.e., what a code must possess in order to qualify as such) is to be made up of two systems ("universes of discourse" for Prieto, and "s-codes" for Eco), one being present, assumed as the level of expression, and the other being absent as the level of content (for Prieto, these are the "semantic field" and the "noetic field," respectively). In fact, according to Prieto, Eco, and others, a code simply represents the rule (or set of rules) that links the elements in a (conveying) system ill I —*~ Bit H eti§s mmm: WSSmmmm^ mm^ -v-' mmm. I mm miM. mm mm. 98 The Semiotics of Performance to the elements in another (conveyed) system. No other more restrictive condition is necessary for the existence of a code: neither double articulation (characteristic of the linguistic code, as we already know) nor even the bi-univocal correspondence between expression-units and content-units that is proper to codes in the strict sense, such as the Morse alphabet.1 With the help of this generalization, a preliminary definition of /performance code/ can be stated in the following terms {for the time being I will continue to use the expression in the broad sense to refer to any convention, rule, or organizing principle manifest in the performance text): a /performance code/ is the convention in performance which permits the association of particular contents with particular elements in one or more systems of expression. Thus when speaking of the gestural code in a given performance (Ryszard Cieslak's performance in Grotowski's The Constant Prince, for example) I mean the rule(s) in that performance2 capable of assigning one or more meanings to one or more expressions isolated from the continuum of physical movements. In other words, a gestural code is the convention (or set of conventions) on the basis of which the gestures in a performance signify and communicate their meaning. Naturally, in performances that do not conform to the canons of representational theater, the performance codes function above all as rules of self-signification through which the expressive elements of the performance text refer back to (and signify) themselves—at least in the first instance, not taking into account eventual, perhaps inevitable, metaphoric and connotative processes. Furthermore, except in the case of powerfully coded genres (such as Indian classical dance, Noh theater, and even Sicilian puppet theater),3 the performance codes constitute at most a particularly obvious example of the "systems of vague correlation" mentioned by Eco (1976) which provide for the correlation between a single expression and several contents and the precision of very broad contextual and circumstantial selections. In other words, we are dealing with weak or very weak codes, like those that are proper to (audio)visual commurrication in general. According to Eco, these codes are "imprecise, changeable, and poorly defined, and their acknowledged variables predominate considerably over their pertinent traits" (1975: 280).4 The characteristics of "vagueness" and "weakness" obviously depend on co-textual and contextual properties of the performance text, some of which have already been described and some of which are still to be examined: heterogeneity of expressive media, multiplicity of codes, multidimerisionality, non-persistence, and (as always happens, by definition, in the case of texts of an aesthetic type) the presence of conventions and set codes, instituted ex novo.5 4.2. DECODING, COMPREHENSION, INTERPRETATION Before continuing this discussion, I will make two comments on the definition of the "performance code" formulated above. My first obser- Performance Codes and Theatrical Conventions 99 vaion concerns the (already mentioned) necessity of distinguishing between the codes of the sender (meaning the codes of textual production) and the codes of.the addressee (meaning the codes of textual reception and interpretation). It is well known in fact how these two lists of codes, otherwise open, are very far from being equal to each other, contrary to the claim of the communications model popularized by information theorists. In short, according to Eco: "the competence of the addressee is not necessarily the competence of the sender" (1979: 53).6 The possibility and degree of noncoincidence between the two series of codes (and the two relative competencies) obviously increases in the case of very "complex" texts such as performance texts, which, in addition, are directed at a collective addressee which is in general highly differentiated from a sociocultural standpoint. In particular, the degree of awareness of the class of performance codes that I will later call /theatrical conventions/ strongly con- ■ ; ditions audience reception, causing it to fluctuate between various levels of "naive" fruition and "educated" fruition.7 Finally, it is also appropriate to distinguish between the analysis of the performance text "at the point of departure/' meaning from the point of view of the sender (the codes, competencies, intentions, and presuppositions of the sender)8 and the analysis of the same text "at the point of arrival," meaning how it is 1 received and interpreted (by the audience, by critics, and by semiotic theorists). Space is thus created for at least two different kinds of semiotics of the performance text: a semiotics of production and a semiotics of reception. It is hardly necessary to mention that we are dealing with two perspectives that have been largely ignored up to the present, given their contextual-! pragmatic character, in favor of structuralist approaches focusing especially, if not exclusively, on co-textual aspects, i.e., on the internal rules of texts.9 As for my .second observation, I now wish to restate the comments made at the end of the last chapter regarding specialized reception in particular. I am referring to the fact that the spectator's reception of the performance text is an interpretive activity much greater than a simple act of decoding. As we shall see below, it requires competencies of a contextual, intertextual, and encyclopedic order, involving pragmatic as well as syntactic and semantic problems. The performance text, like any other text, consists not only of code material but also of material to be freely interpreted and material for the inferential process.10 Therefore, it demands of the addressee an active textual cooperation, requiring a competence that goes very far beyond a simple "knowledge of the codes." This means that it is now necessary to link a semiotics of theatrical codes to a larger model of the >■ textual analysis of performance, a model that attempts to account for all the ■ semantic-pragmatic factors involved, in accordance with the insights of-: ered by the most recent research in the field of textual semiotics. As for my own view, I fully agree with Eco's proposal that we must establish a relationship of interdependence rather than opposition between textual semiotics and code semiotics: 100 The Semiotics of Performance In this way we postulate a semantic description in terms of the structure of the code whichis constructed for the purposes of the comprehension of texts, and postulate at the same time a theory of the text that does not deny but rather encapsulates (through the notion of encyclopedia or thesaurus and that of the frame) the results of a wider analysis of the components. ... In this way . . . the theory of codes and the theory of text appear closely interrelated. In a semantics concerned with textual actualizations the sememe must appear as a virtual text, and a text is nothing other than the expansion of a sememe.31 (1975: 23) 4.3. CLASSIFICATION ACCORDING TO CODES AND CLASSIFICATION ACCORDING TO EXPRESSIVE MATERIAL Two fundamental problems emerge in analyzing the codes of a textual structure as multi-coded and as heterogeneous in its expressive materials as the performance text: distmguishing and enumerating these codes on the one hand, and classifying them on the other. I will begin with the first problem. As we have seen in 2.5., the performance text is doubly heterogeneous; that is, it has multiple codes and is created from multiple expressive media (in addition to being multidimensional). There is currently a rather widespread tendency in theater semiotics to create a two-way correspondence between these two kinds of heterogeneity (i.e., heterogeneity of the expressive materials used by the performance text in question, and heterogeneity with respect to the codes that can be found in these through the process of analysis). For example, references are constantly made to the ''linguistic code/' "gestural code," "scenographic codes," "accessories code," and so forth, as though each "level" of expression within the performance text were ruled by only one specific code. I have previously designated these levels as partial codes.12 For several reasons I cannot accept the two-way symbiotic function often established between the level of the expressive components and the forxnal-systemic level of the performance text—i.e., between the modes of expression and the codes— and on the basis of which the codes would necessarily amount to the same number as the expressive media. Some of my objections are specific, while others are related to the general characteristics of the code as such. (1) A single expressive material can be created from multiple codes in different ways. This is certainly what happens in the case of theatrical performance. In fact, as we have already seen, the performance text is constructed through the association of complex, autonomous expressive media, which are physically and even historically distinct.. These expressive materials constitute actual texts in themselves, produced by the horizontal and vertical coexistence of various codes and subcodes. Take, for example, the verbal Performance Codes and Theatrical Conventions 101 text sets, music, or dance. None of these "texts" can, as such, be reduced to a single, homogeneous code scheme (see Ertel 1977: 139). (2) Conversely, there are codes that draw on several elements of different signifying media for their expression plane.13 We can locate what Buyssens defines as heterogeneous seme in this category (1943). Buyssens offers a specific example of heterogeneous semes in the theatrical context, the seme of "audience reactions to a theatrical work." This seme includes (a) phonic or even phonemic signifiers such as /bravo!/, (b) auditory, non-phonic, and labial signifiers such as whistles, and (c) auditory, non-phonic, and manual signifiers such as applause (Metz 1971: ET 29). (3) Finally, there are- cases of "one and the same code being manifested in several arts or language systems," that is, codes that may appear "fundamentally" unchanged in different expressive media (Metz: ET 216-17).14 Metz demonstrates this by developing Buyssens's observations on heterogeneous sememes, and formulating the notion of "characteristics that are pertinent to given codes" in a post-Hjelmslevian direction (2i8ff.). According to Metz, because a given code can be transferred (unchanged in its internal correlational structure, or, to use HjelrnsleVs term, in its form) from one language to another, it is necessary (and sufficient) that in the course of this transfer the sigrufying material maintains in an unaltered state some physical traits that are pertinent in relation to the code considered. Metz offers the chiaroscuro code as a hypothetical example. Since this is an "mtrinsically visual" code, the chiaroscuro effect may appear fundamentally unaltered in "languages" (texts) whose expressive materials all possess a visual feature: painting, photography, cinema, and so forth. Conversely, however, it is incdrrect to speak of a "single code with multiple manifestations." We are dealing instead with "codical transpositions" involving various portions of the code, but never the code in its entirety. This is the case with literary "chiaroscuro" in contrast with painterly or photographic chiaroscuro.15 On the basis of (i), (2), and (3) we can argue that, generally speaking, a two-way correspondence between signifying materials and codes in texts does not necessarily exist, and that this is even more true for the performance text because of the reasons given above. Therefore the analysis of the performance text from the perspective of its expressive materials and an analysis conducted from the systemic-functional perspective (i.e., the codical perspective) will lead to different classifications of the textual unit in the theatrical event. In the first case, it leads to a classification by expressive materials (of slight interest to semiotic analysis) and, in the second case, to a classification by codes (the only direction of real interest to the semioti-cian).16 This tendency to confuse the standpoint of the codes with the standpoint of the expressive media (or the deliberate attempt to make the different classifications produced by these coincide with each other) is responsible for many of the errors and blind alleys that can be noted in the 102 The Semiotics of Performance scholarship dedicated to identifying the performance codes. A particularly interesting case in point is provided in Kowzan's classification of thirteen ''sign systems" (1975: i82ff.).17 Because of the confusion mentioned above, these "systems" sometimes turn out to be simple classes of "typical signs" (Barthes 1964: ET 47), which have the same expressive media but are differentiated or are capable of being differentiated on the functional level. At other times, however, the "signs systems" are classes of heterogeneous "signs" having the same function: involving mime, accessories, costumes, sets, etc.18 The many inaccuracies and incongruities that Ruffiiu has sharply criticized in scholarly attempts to identify the theatrical codes (1974a: 72ft.) derive from the effort to make the two classifications coincide, a tendency that continually confuses the two perspectives (or moves from one to the other without distmgmshing between them). Some examples are: the system that Kowzan defines as facial mime represents a class of elements that share the same signifying medium (the expressive continuum of the human face) and are thus available to various types of simultaneous sign-correlation. At the same time, however, this system seems to be presented as a full<-fledged code (we could call it the mime code) capable even of forming different materials from the continuum of the human face—for example, those proper to various types of mask. Kowzan's remarks on the mask are especially revealing: "In my opinion, the mask refers to the system of makeup, even if it belongs to costume from the perspective of expressive material, and to mime from the functional perspective" (191-92, emphasis added). It is clear from this quotation that Kowzan's taxonomy vacillates continuously between the material and functional viewpoint, and that it offers an inaccurate concept of the code. According to Kowzan a code can be identified simply on the basis of its content-plane (on the basis of the meanings that it is specifically delegated to furnish: this, I think, is what Kowzan means by the term /function/) independently of the expression-plane, without taking into account the properties that the content-planes have to create logical relationships of inclusion, intersection, and even identity with each other (Prieto 1966). Obviously Kowzan is not the only writer to commit this type of error. It also appears, for example, in the otherwise valuable and ground-breaking work of the Prague school of theater theorists (whose writings are now anthologized in Matejka and Titunik, eds., 1976, and critically assessed in Slawinska 1978). It reappears more recently in Steen Jansen's critique of the multi-coded model of performance, casting doubt on the very possibility of coding in the expressive components of theater (see Jansen 1977, partially repeated in 1978b). In fact, the most important arguments adopted by Jansen against the possibility of accepting the hypothesis of theatrical codes is that theater does not offer "classes of elements" (in practice: systems of expression) with a constant function that are capable of carrying the same type of meanings from one performance to another. These objections can only be agreed upon if they are intended simply to disagree (as I am dis- >l Performance Codes and Theatrical Conventions 103 agreeing here) with the one-to-one correspondence that often tends to be ■ created between the material components of expression and the codes. ,;: They are objections that can be seriously considered when they intend to call attention to the fact that it is performance itself that furnishes many of the rules necessary for its own interpretation. But if Jansen's comments also aim to present themselves as prejudicial in a general way to the pos-tulation of a code—as indeed they seem to—then I believe that they are unfounded because: (a) a code cannot be identified on the basis of the content-plane alone, (b) the content-planes of the codes are not exclusive of each other (in other words, different codes may carry common messages), and (c) it is in fact inherent in the very definition of the code (see 4.1.) that an expression can carry several contents, incurring, within the code, different sign-functions, according to the contexts and the circumstances (Pri-eto 47-64; Eco 1975: 74; 1976: 33). To conclude this section, I must repeat that for all the reasons expressed above, it is not possible to make a single, specific code correspond to every textual "level" of the performance text. Systems of expression within performance and performance codes must be subjected to different classifica-i tions. Obviously, once this extremely important point has been clarified, we can continue to speak of verbal, paralinguistic, gestural, scenographic, and other codes to designate the code-relationships that utilize as their plane of expression the signifying components that share the same designation. We must be careful, however, to take into account what has already been mentioned regarding the absence of a two-way correspondence between the latter and the former. 4.4. THEATRICAL AND NONTHEATRICAL MEANINGS In addition to the argument just discussed, Jansen made a second objection, linked to the first, against the concept of theatrical codes. This objection, though unacceptable in its conclusions, is more interesting than the first, since it throws light on the crux of a real problem. According to Jansen, the expressive elements of a theatrical occurrence do not usually convey in the performance the same meanings they convey outside performance ("outside" meaning other theatrical works and other artistic practices, or even daily life), not to mention the fact that they are also altered in form to a greater or lesser degree. More precisely, it is only within a given performance, and not on the basis of a preexisting and external code, Jansen claims, that a certain element becomes associated with a certain function, that it acquires one or more meanings. Among the many examples he lists is the scene in a production of Lorenzaccio (1953) where Gdrard Philippe indicated (signified) a change of location with a simple movement of his head, and the scene in Gaston Baty's production of Hamlet (1928) where a single point of light on a black backdrop designated the ghost of Claudius Blllti 104 The Semiotics op Performance (at least from the moment when the spoken lines were associated with it). Jansen writes: It geems difficult or even impossible to affirm that prior to Gaston Baty's staging this point of light constituted an element linked to a given class of functions, among which was included the function of signaling "the apparition of a (certain type of) character." This is the necessary precondition to provide us with an example of a system, and even more so, of a code, and of one of its manifestations in performance, which would be linear and distinct from a manifestation of other systems (and codes).19 The positive aspect of these arguments lies in the fact that they draw attention to the difference between the theatrical meaning and extra-theatrical meaning of the expressive components of theater, in other words, of the specific and particular modalities (this formulation will suffice for the moment) according to which nonspecific, vaguely specific, or very specific systems of expression are used in a theatrical performance. I would argue that it is possible to take Jansen's objections into account, without repudiating—as he would tend to—the theoretical model I am attempting to construct, and with it the basic possibility of a textual analysis of performance. This model includes the concepts of code and of multi-coded text as two of its fundamental elements. To counter Jansen's argument I will draw on the idea of "double con-textualization" that Ruffini proposed some years ago (1974a, 1974b), although it presents an excessively simplified understanding of the production and stabilization of meaning in performance, and Jansen was correct perhaps in this aspect of his appraisal. This idea of "double contextualiza-tibn" must be integrated with a classification of performance codes, which does not appear in Ruffini's work, and which nevertheless seems extremely useful in enabling the analyst to proceed beyond the impasse criticized by Jansen. In this way, the performance text begins to delineate itself more clearly as an aesthetic text, characterized by all the modes of sign-production proper to the latter, and its analysis (or reading) becomes a complex set of different inferential and extra-codical modalities. 4.5. PERFORMANCE CODES (IN THE STRICT SENSE) I now wish to argue for the division of the codes of the performance text (until this point simply designated by the generic term /performance codes/) into two large classes: performance codes (in the strict sense) and theatrical conventions. From now on I will designate as /performance codes/ the codes that are not specific20 to theater, since they are also found in other artistic practices and in daily life, and which are used in theater according to more or less particular, characteristic modalities (or without appreciable modifications). In effect, we are dealing essentially with the rules of signification Performance Codes and Theatrical Conventions 105 for the textual units of performance. Obviously it is very difficult to enumerate these even in an approximate way, but this is not of great importance to the present argument. Nevertheless, apart from the codes that Metz (1971: ET 248) calls "codes of content" (i.e., ideological, axiolog-ical, and epistemological codes, and the like, all of which display minimal specificity or a complete absence of it),21 also included in this class are linguistic, paralinguistic, kinesic, proxemic, perceptive, rhetorical, narrative, and iconographic codes, as well as the code of iconic recognition and naming (see Ertel 1977: 138; Ubersfeld 1977: 30-31, 40; Pavis 1976; 1978: 54).22 As is well known, all the semiotic codes are by definition cultural in the general sense. Yet many of the codes just mentioned (for example, the linguistic, kinesic, and perceptive codes, and the code of iconic recognition and naming) are also cultural in the strict sense since they are created over a long period of unconscious absorption and deep assimilation within a given culture to the point that they sometimes seem innate, not learned. They appear "natural," or "naturalized," according to Ertel's description {138).23 In this sense, these codes are distinguished from the explicitly instituted and consciously acquired conventions which I will discuss in the next section.24 All performance codes are nevertheless part of the general text of a given period which, as we shall see, functions as the motivating level of contemporary theatrical conventions, whether these are "particular" or "distinctive." Nevertheless, in order to forestall facile objections, I must immediately point out that, as a rule, the performance text does not use these codes exactly as provided by the general text, i.e., it does not simply limit itself to "quoting" them, to "placing them on stage" (to use Ruffini's expression) except perhaps in extreme cases (sub judice),25 such as those mentioned by Ruffini: the "ceremonial" and "a certain type of Renaissance comedy," and the like. In an attempt to arrive at a definition, we might say that /performance codes/ result from the more or less particular usage, in performance, of nonspecific cultural codes, which will now be designated as /extra-theatrical/. In effect, we are dealing simultaneously with specific and nonspecific codes. The distance separating a given extra-theatrical (for example, a code of ordinary gestures, or even a kinesic code in painting or in literature) and the "corresponding" performance code (the actors' gestures within a specific theatrical event) will obviously vary a great deal according to genre, author, historical period, and cultural or geographical context. Staying with the example of the gestural code, this distance will be predictably slight, almost nonexistent, in the case of an actor's gesture produced by the almost exact imitation of an ordinary, everyday gesture and the use of the same, appropriate form of expression (examples can be found in traditional pantomime of the late nineteenth century, or in the productions of naturalist theater during the same period: from . Antoine to the early work of Stanislavski). The distance is very great, however, in the case of a gesture produced with extremely antinaturalistic techniques and 106 The Semiotics of Performance supported by code relations that are instituted ex novo (for example, the "abstract" mime of Etienne Decroux, or certain types of "physical" theater in the 1960s). We could attempt to create typologies of the performance code from this perspective (using the graduated concept of specificity), but the resulting list would very probably resemble a continuum of infinite nuances. In order to understand how, and to what degree, an extra-theatrical code is transformed on stage into a performance code (whether general, particular, or distinctive), it is essential to draw upon a second class of codes: theatrical conventions. 4.6. THEATRICAL CONVENTIONS Theatrical conventions are technical, specialized codes that found and regulate the theatrical use of the codes that I have designated above as /performance codes/. These include such conventions as those of architecture or stage design, as well as the proxemic, kinesic, and delivery patterns that are proper to an actor, a specific period, or an entire tradition of theater. The conventions assist the performance texfs "legibility," and hence the proper functioning of theatrical communication—providing the (competent) addressee with enough information to discern the underlying performance codes, and supplying points of reference and inter-textual patterns to the inferential process of interpretive cooperation. In effect, we are dealing with agreements that flow between the audience and the performance (or rather, the participants in the performance). More precisely, in Lewis's terms (1969), we can conceptualize theatrical conventions as "regulators of behavior" produced by systems of expectation, to which those participating in theatrical interaction almost always tacitly conform— generally speaking and under normal conditions26—m order to find a "balance of coordination'' for their problem of coordination, i.e., the problem of making sense of the performance text, of comprehending it.27 The classification of theatrical conventions that I am about to propose focuses on a single dUstmguishing aspect, namely the degree to which they can be described as general, which is particularly crucial for the purposes of my discussion. But theatrical conventions (like all others) are differentiated from each other in many other ways: (a) in the explicit or implied quality of their stipulations (for example the conventions created by classical Western and non-Western poetics are explicit: from Aristotle's Poetics to Boileau's Art poitique, from the treatises on Noh drama by Zearni Motokyio to the Natya-Sastra in India); (b) in the degree to which they are prescriptive or constructive: generally speaking, explicit conventions acquire the power of institutional rules and norms; and (c) in their varying degrees of social acceptability (what Lewis has termed "degrees of conventionality"):28 later I will note the difference that exists in this regard between both distinctive and particular conventions on the one hand, and Performance Codes and Theatrical Conventions 107 general conventions on the other; (d) in the type of awareness possessed by spectators who conform to these conventions. Lewis points out that "our awareness of our conventions . . . can be quite slight. . . . (1) It can be a purely potential awareness. ... It can be an inescapably nonverbal awareness" (76). This last problem is particularly interesting in relation to theatrical conventions, partly because it also involves the same elements mentioned in relation to the other issues. To begin, I disagree with Ertel's claim that theatrical conventions always require special learning and a conscious act of decoding {138). Indeed this often happens, especially when the audience member has to deal with conventions foreign to his culture (for example, the conventions of Asian theater for a Western spectator).29 But in other cases theatrical conventions are "unconscious [or] too automatic to be noticed: the laws of perspective, euphony, the ideological marks that govern the mise-en-scene" (Pavis 1980a: 65), just like many of the performance codes mentioned above.30 This tacit and often unconscious awareness constitutes the "nontheoretical competence" that I referred to in 3.4. (in contrast with the critic's "theoretical" or explicit, rational competence).51 In addition to the various types of awareness of theatrical conventions, we must also distinguish between awareness (whether theoretical or nontheoretical) of conventions and their usage. I will devote greater attention to this custinction in chapters 6 and 7, defining an active competence (constituted by the maximum of awareness plus use: this is the competence of the sender) and a passive competence (which is nontheoretical for the audience member and theoretical for the critical analyst).32 Before moving on to the classification I promised to provide, a brief terminological parenthesis is required in order to explain my preference of the term /theatrical convention/ over other possibilities (performance convention, theatrical code, etc.). The use of /theatrical/ instead of /performance/ as a qualifier is meant to underline the conventions' greater specificity (within the limits about to be explained) in comparison with the codes in the strict sense that I have just examined. We are in fact dealing with the technical "rules" of theatrical performance. As for the noun, I prefer the more generic term /convention/ to /code/ in order to stress the asymmetrical position that these institutions of theatrical performance (conventions) occupy with respect to the codes whose use they regulate, a position that is simultaneously both fundamental and transversal. Above all, I prefer the term /conventions/ because we are not dealing with codes in the proper sense, but with systems, or s-codes (Eco's "codes as systems" [1975: 56]), like all artistic institutions and the rules of genre in particular.33 It should be clear that the difference between performance code and theatrical convention concerns the type of coding, and not the degree of coding, as Pavis seems to think in juxtaposing codes and conventions in terms of strong codes and weak or vague codes (idSoa: 93). While performance codes are correlative, therefore codes in the proper 108 The Semiotics of Performance sense, theatrical conventions are institutionalized sets of systematic rules and norms (as strong or as weak as the correlative codes). Having said this, I would argue that the crucial problem here is not to ascribe the theatrical convention to the code or system,34 but to provide an adequate clarification of the type of action that the conventions exert on performance codes in the strict sense, for the purpose of the discursive functiorung of the performance text. I will attempt to offer such clarification in the pages that follow. Finally, I will elaborate on my classification of theatrical conventions according to general, particular, and distinctive types.35 4.6.1. General Conventions General conventions are preliminary and basic to theatrical fiction. According to these conventions the stage represents but is not the world, the actor acts but is not the character, the time-space of the performance text statement is cHstinguished from the time-space of the performance text utterance, and so forth.36 This first type of convention defines both in specific and general terms the particular communicative situation known as "theatrical representation," "a performance that refers in its entirety and/or in its constitutive elements to a signified other than itself" (Molinari and Ottolenghi 1979: 8). Such "rules" may be violated and are systematically violated, for example, by the historical avant-garde movements. But, as the Russian formalists have shown, it is this very transgression that reveals and confirms their existence. Nevertheless, the fact that they are not known to some spectators has given rise to rather farfetched incidents which have been passed down in an abundance of traditional anecdotes: an audience member waits for the traitor at the stage door intent on punishing him for his deeds; or a man tries to climb on stage to save the heroine from danger; and so on.37 As will become clear at a later point, in terms of textual pragmatics these naive and inappropriate reactions can be attributed to deficiencies in encyclopedic competence, and above all to a deficiency in intertextual competence. In a good encyclopedia the sememe "theatrical performance" should normally register a distmguishing trait such as /make-believe/, /untrue/, and the like. These deficiencies, in turn, provoke an erroneous attribution of frames (ordinary "scenarios" are applied instead of intertextual "scenarios") and the wrong selection of a world of reference (the world of real personal experience instead of the world made possible by the theatrical fab-ula)?8 We could designate as general competence the competence that is (particularly if not exclusively) constitutive of an awareness of the general conventions of theater. It is thus distinguished from particular competence or genre competence which involves the particular conventions that will be defined in 4.6.2. Both are part of a broader receptive competence proper to theater. Nevertheless, while particular competence makes us capable of distinguishing between various classes of theatrical performance, and of judging the appropriateness of the occurrence ■ to the type—as we shall see— general competence is what enables us to recognize a theatrical perfor- Performance Codes and Theatrical Conventions 109 mance as such, and to distinguish it from other artistic (macro)genres and from ordinary life through the correct deciphering of the co-textual and contextual signals of the (macro)genre.39 Obviously, we find a very different situation with theatrical phenomena produced outside these general conventions, i.e., phenomena common to cultures and civilizations—such as Asian culture—to which such conventions are in fact foreign and in which the conventions are positioned in a very different way. I must therefore limit the claim that I made above regarding the general applicability of what I described as /general/ conventions (which are in reality ^Hflsz-general) as well as what I said regarding their specificity. Beginning with the issue of specificity, I would argue that as s-codes of the performance text, "general" conventions are without a doubt more specific than performance codes and in two ways: because they concern a smaller number of phenomena, and because, within the phenomena where they are manifest, they take on an essential character inherent in their very definition.40 Theatrical representation could perhaps dispense with almost all of the performance codes (as in Grotowskfs "poor theater") but it would not even exist without the conventions that create it as such (e.g., fiction, and so forth).41 Nevertheless, these conventions are not "specific to the greatest degree" if by this we mean the conventions proper to a single language (following Metz), and hence, in the present case, proper only to theatrical performance (in the definition given in 2.2.). Without needing to delve too deeply into the matter, I believe it can easily be argued that the conventions enumerated above (and others on the same level) are common at least to all the "arts of visual representation," mduding dnema, television, and painting, as well as theater.42 Similarly, not too much needs to be said on the "general applicability" of these conventions which should be more accurately understood as a "quasi-general applicability." The heterogeneity of theater and the irreducible multiplicity of its manifestations make it impossible to attempt any definition of a prescriptive type, and undermine any attempt to single out theatrical codes that are foundational, constitutive, or generally true, of the kind proposed by Metz in his analysis of cinema. In reality, the conventions that I referred to as "general" are fully valid only in the case of representational theater (and therefore for the majority of theater in the tradition of Western high culture) but much less so in the case of theatrical phenomena situated partly or completely outside the canons of dramatic illusionism or fiction, or phenomena that reformulate them in a radical way. I am thinking of the experience of contemporary avant-garde theater, from "happenings," to street theater, to performance art, but also traditional genres like the circus, ballet, sporting events/ ceremonies, and festivals. Not to mention the theatrical traditions of Asia, where mimesis and make-believe are established as a rule according to modalities that are very different from the realist-illusionist modalities of the Western theatrical tradition of Aristotelian origin.43 110 The Semiotics of Performance 4,6.2. Particular Conventions These are conventions, rules, and styles proper to an artist (playwright, director, actor), a genre, a school, a movement, a historical period, or a cultural-geographical region.44 Having noted the quasi-general applicability of "general" conventions, it should be clear that the difference between general and particular conventions is not qualitative but merely quantitative. "Particular" conventions only involve a very small number of performance texts. Some examples of particular conventions randomly chosen from various theatrical periods are as follows: the three classical unities, the deus ex machina, and the ekklyklema in Greek theater; the assigned locations of medieval drama; the rules of perspective in Renaissance stage scenery; the decor simultanee in classical French theater; the "verbal stage sets" of Elizabethan theater; the typology of set characters and situations in the Commedia dell'Arte; the "fourth wall" in late riineteenth-century drama; Brecht's Ver-fremdungsejfekt; and the techniques of physical provocation in the "theater of involvement" during the 1960s. But there are obviously many more if we look at dance, mime, opera, the circus, spectator sports, or Asian theatrical traditions. Many, conventions of the dramatic text also belong to this category; not those that concern it as a literary text (stylistic, rhetorical, and narrative conventions, and so on) but only those that belong to its "specific" destination as performance, such as conventions of dialogue (invectives, sticho-mythia, and so on), monologues and asides, "the play within the play," conventions of stage directions, and the dramatic treatment of time (see Burns 1972; Gulli Pugliatti 1979; Elam 1980).45 As we shall see, conventions can be "particular" to a greater or lesser degree, and coded to a greater or lesser degree. Yet they are usually more rigid in their canons and more long-lasting (I am tlrinking in particular of the conventions of historical periods, genre, and cultural-geographical areas) than conventions that are proper to other lands of artistic texts. Undoubtedly, this circumstance can be related to the fact that theater is "a product necessarily intended for coEective use" and hence strongly connected, at least by tradition, to its social consumption. Paola Gulli Pugliatti has offered a very interesting comment oh the principle articulated by Eco according to which /acceptance of a rule/ always leads to "conformism" (Eco 1976: 32): It is particularly true in the case of the dramatic text that if one can make the claim that nonconformity to the established institutional norms can be aesthetically meaningful, one certainly cannot claim that conformity in itself is no-naesthetic. . .\. Indeed in many cases one could say that the dramatic text bases its aesthetic value on conforming to such rules (which is not "conformism" but a something that emerges as proof of the success of the assumption of simulation). (Gulli Pugliatti ibid.: 146-47) Performance Codes and Theatrical Conventions 111 i;/ This is also valid, mutatis mutandis, with reference to the performance text, V as will be clear from what follows. Focusing on the pragmatic viewpoint as 1 have tried to do through- ■ out this book, I believe it might be useful to consider the relationship between particular conventions and comprehension (what I have to say about this can also be applied at least in part to the other two types of conventions). First, the audience's degree of awareness of particular conventions greatly conditions the level of their comprehension and fruition of a performance. Like all the rules of genre and other conventions that are at the basis of various artistic practices, these too are over-coded rules, which provide the receiver, if he knows them, i.e., if he possesses adequate particular competence (or genre competence), precise intertex- :■ tual information on the performance text in question, facilitate comprehen-sion, and create certain expectations, which can of course also be frustrated. For example, an awareness of the rules of "tragedy" as a genre ': 1 enables us to know from the beginning that the hero must die. The types of costumes and masks in classical comedy allowed the audience member in ancient times to recognize the characters immediately and to make predictions about the plot. In medieval theater the audience already knew ^ the "content" of the mystery plays and passion plays as well as their inevitable epilogue. The same happened in the case of the Commedia 'i MU'Arte, where characters and situations never varied. I have already men- ; honed the disadvantages that can derive from the addressee's lack of awareness of general conventions. To give an example regarding the case , of a particular artisf s conventions, it is easy to imagine that a knowledge of the staging principles and the ideology of Brechf s dialectic-epic theater would be of considerable advantage to a spectator attending a production ! by the Berliner Ensemble at the Theater am Schiffbauerdamm in the early : 1950s.46 Genre as a socially, historically institutionalized theatrical type is , only one of the classes of theatrical events to which particular conventions apply, as I have already mentioned. Other classes are made up of the performances of a particular director (or actor), of an artistic "school," ■ of a given era, or of the same cultural-geographical region. There is no doubt, however, that genre (as it is traditionally understood) represents the most important of these for reasons that concern the receptive process and which I will examine in greater depth in chapter 7. Consequently, , the rules of genre can be considered the principal example of particular con-% vehtions. It is hardly necessary to state that not all theatrical events behave in the same way in relation to the rules of genre. As is well known, in many performances (and more frequently in recent times) the relationship takes the form of an infraction or violation and the conventions that dominate are those that are created ex novo. I will describe these below as /distinctive conventions/. It is more interesting to examine how reception operates in 112 The Semiotics op Performance the case of highly standardized performance texts, that is, performances that conform more or less completely to their genre and which are largely predictable on several levels: in the material that is represented, the expressive modalities brought into play, the acting style, and the like. Pavis has observed how in these so-called performances: Everything seems arranged in advance: the natural signs of the actor seem tamed and brought under constant check. Where does the creativity of such a performance lie? It lies in the syntax of signs and not in their semantics. It hardly matters what the sign says, since once it is known it remains the same in all instances (thus, in Chinese theater, white is the sign of an intellectual). The only necessary focus of our attention is the concatenation of signs: what follows a given gesture and why? When the code is completely known, we can concentrate on linking together the message. (1976: 130, emphasis added) These are very interesting comments, although, at least in principle, I do not share the negative evaluation that Pavis attributes to this kind of theater. According to Pavis, in "genre performances" our attention shifts from the fabuia and the plot to the expressive strategies used to convey them, from the written text to the "game" on the stage; in short, from the what to the how, from semantics to syntax. The individual moments are predictable, taken for granted. The creativity and the possibilities for variation in a performance lie in the links, in the shifts from one moment to the next, and it is on these that the attention of the competent receiver must focus.47 An extreme case of genre theater, heavily standardized both in its narrative structures and in its style of theatrical execution, can be found in nmeteenm-century opera. In one of his famous quips, George Bernard Shaw described this genre as "the story of a tenor and a soprano who want to go to bed with each other, and a baritone who wants to prevent them."48 Because of the simplicity and stereotyped quality of its plots, the predictable developments and results, the lack of narrative variations, the recogmzability of the characters, everything in melodrama conspires to activate in the addressee precise expectations and desires of easy satisfaction: The man sitting in the audience receives the first, decisive information from the musicians: the voices constitute in themselves a peremptory system of signals. If a tenor comes forward to the front of the stage, sings his aria, and utters an invitation to the audience and the supporting cast "to hear the torments of his heart," the spectator has the immediate assurance that this is the hero, that the actions that will be attempted against him will all have a negative connotation, that the soprano will love him with all her might, and that the bass and the baritone will be numbered among his enemies, at least at the start. (Lavagetto 1979:166). On the specific issue of the competence and receptive modalities of the nineteenth-century opera's addressee, Lavagetto develops Lotman's notion of the "structure of expectation,"49 asking himself how many levels 73 Performance Codes and Theatrical Conventions 113 this is composed of, and how it functions in the specific case of operatic librettos: when we speak of a structure of expectation we must refer to a stratified structure in which different levels are superimposed and integrated with each other. ... (a) Expectation is organized on the basis of a series of data culturally held by the addressee. History, mythology, an awareness of previous literary versions, and the like allow a forecast of the plot just slightly overshadowed by the (also foreseeable) possibility of eventual, premeditated, more or less consistent, infractions. . . . (b) Expectation is set up by the genre. . . . (c) Expectation is set up by the text, as the result of the consecutive triggering of classemes that have not yet been fully saturated. ... In the case of librettos the problem can be reformulated in the following terms: the poet tells a tale, basing it (a) on a cultural code, which, theoretically, the spectator may or may not know, but which in fact forms part of the sphere of control of the habitual addressees of opera; (b) on a genre code which the spectator knows because it conforms to the structure of his historically and socially determined desires, and because it follows a series of territorially fixed rules; (c) on the dissemination of a series of taxonomies of classemes within the text which heighten the possibility of formulating precise predictions. (Lavagetto 176-78, emphasis added) If I have quoted at great length, it is because nineteenth-century opera represents such an exemplary case of a "genre performance" that is highly predictable and standardized. Opera—like many other genres, from ;J phenomena of the past, such as the Commedia dell Arte, to genres that % are still current, such as the circus or puppetry—has an abundance of explicit textual signals that permit the pragmatic operations of textual rec-j ognition and the identification of genre, thus stimulating expectations V ; and predictive hypotheses in the receiver. But it is important to bear in j> ■ mind that particular conventions are always inevitably at work in every theatrical event, even if they are not always signalled in such an abundant and explicit way as in the "strong" genres that I have just mentioned. In Western culture, especially since the end of the nineteenth century with ;.: the rise of the historical avant-garde movements and modern stage directing, there are performances that are so interwoven with the nonspoken, c": with "empty spaces," as to make the spectator's "inferential walks''50 much ■I more adventurous, and his abductive guesses much more risky and often k{ frustrated. Even if a performance text is not a "genre performance" in the narrow |f: sense of the word as used above, it always refers to a greater or lesser ' degree (positively or negatively) to a genre or a theatrical type, and is comprehensible only on the basis of a rather vast and well-ordered inter-textual backdrop.51 In the final chapter of this volume I will have the opportunity to analyze in depth the function and the characteristic of "genre labels" (title, and the like), "paratexts" (theater programs, playbills, ■ newspaper reviews, and the like) and other co-textual and contextual signals which prepare the performance text to be recognized, and which 114 The Semiotics ojf Performance define its pragmatic appropriateness, functioning as genuine instructions for use. For the moment I will limit myself to stating that the recognition of genre (in the broader sense) is an operation of primary importance for the comprehension of a theatrical event.52 If it is indeed true, as the most recent theories of reading seem inclined to maintain, "that textual comprehension is largely ruled by the application of pertinent scenarios" (Eco 1979: 8i),53 it is also true that genre (textual type) guides the selection of interpretive frames, and hence that only its correct recognition allows us to apply the pertinent scenarios (whether common or intertextual) and to choose accurately the "possible world of reference." It is here that the cowboy in the audience at the Grand Guignol failed (for lack of general competence, as well as genre competence) when he shot the "killer,"' having decided to pay him back; or, while attending an opera performance, when he wondered how people managed to sing when they found themselves in moods or situations hardly conducive to singing. We could say therefore that every performance text hypothesizes a Model Spectator provided with the genre competence necessary to chose adequate interpretive schemes and correct possible worlds of reference.54 Naturally, as we will see in chapter 7, the Model Spectator is quite a different entity than real spectators, i.e., actual audiences to whom the performance text can be communicated. I have already noted in this regard that the usual noncoincidence between sender codes and receiver codes means that numerous analyses and readings are possible for a given performance text: in each case, many more than those "foreseen" by the text and more or less expUcitly "inscribed" in it. But the Model Spectator is also able to infer correctly the conventions that the text itself institutes ex novo: I am alluding to the /distinctive conventions/ which in the West began to dominate rnan's distinction) when faced with an avant-garde performance piece or a 116 The Semiotics ojf Performance Japanese Noh play on the one hand, and with a traditional staging by a company such as a Teatro stabile on the other.56 It is equally imperative to draw a clear separation between artist's conventions and those of genre or historical period.57 The former are classified among particular conventions since they are hyper-coded rules that are applied to many works. When witnessed at the moment of, their inception, these are, however, distinctive conventions. With a slightly more careful analysis, it becomes clear that for various reasons artist's conventions cannot be placed on the same level as other particular conventions, whether of genre, school, or historical period. They bear a greater resemblance to distinctive conventions, both in the (rarely high) degree of their social institutionalization and acceptance, and especially in their subversive position vis-a-vis the other particular conventions, and sometimes the general conventions as well. There is a wealth of examples. Take, for instance, the case of Etienne Decroux, whose new grammar of body mime (developed during the 1930s) breaks with and radically reformulates the conventions of pantomime, forcing itself beyond the canons of representation. We might also recall the operatic productions of Ronconi (from his 1970 Carmen to the Tetralogy of 1979-1981) which, with the help of stage designer Pier Luigi Pizzi, literally revolutionized the staging of melodramatic opera, stripping it of nmeteenm-century stereotypes still evident in stage design, gestures, and mime. Another observation concerns the status and function of particular conventions other than artists' conventions. It is indeed well known that these too may have a subversive affect on general conventions, and this happens more and more frequently from the second half of the nineteenth century onward with the blossoming of the schools, styles, and movements of the theatrical avant-garde. There is nothing strange about this, if we consider for a moment that particular conventions are almost always the result of the transformation of more or less solidly institutionalized "inventions." This transformation of distinctive conventions into particular ones (to be more precise: the origin of the latter in the former) offers an example of the general process illustrated by Eco (1975:339): idiolect of a work idiolect of a body of works -> idiolect of a historical period or trend. From the perspective of the modes of sign production, the same process corresponds to the catachresis of inventions into "over-coded stylizations" (De Marinis 1979a: 18-20). 4.7. THE PERFORMANCE TEXT AS AN EXAMPLE OF INVENTION 4.7.1. Transformation and Institution of the Code In 4.5. I stated that the nontheatrical cultural codes become performance codes of the performance text through transformations of varying degrees of intensity. Having introduced the category of theatrical conventions, we are now better able to understand the terms of this process. The particular 72 Performance Codes and Theatrical Conventions 117 and specific use of nontheatrical cultural codes that is made by theatrical performance (and on the basis of which the expressive unities of a code are variously modified in performance, and hence correlated to unities of content that are different from those which would convey them in nontheatrical situations) can be seen, I believe, as the result of an action exerted on them by theatrical conventions, in the three ways distinguished above. This action can be appropriately defined in terms of the transformation or institution of codes (Eco 215, 32s).56 To be more precise, general and particular conventions carry out transformations of codes, producing, respectively, general and particular performance codes. Distinctive conventions activate instead the institution of codes, giving rise to specific performance codes. We thus have the following taxonomy: /general genre conventions conventions -— particular <^7^ \ artists' conventions distinctive codes- general particular distinctive genre codes artists' codes From the point of view of function and origin, this model must be reformulated for the moment in the following way: theatrical conventions extra-theatrical cultural codes general