TWO
The Performance Text
The theater is the only place on earth where a gesture, once performed, is never enacted a second time.
Antonín Artaud, Theater and Its Double
We can thus define theater as "what happens between the actor and the audience." AH the rest is supplementary—necessary, perhaps, but supplementary.
Jerzy Grotowski, Toward a Poor Theater
2.1. PERFORMANCE AS TEXT
1 will now definitively shift my attention away from the dramatic text, and turn to the other pole of the transcoding of the theatrical event, i.e., to performance itself, which I have already frequently invoked as /performance text/. To speak of a /performance text/ means to presume that a theatrical performance can be considered a text, even if an extreme example of tex-
. tuaiity. This also implies that we conceptualize the semiotics of theater in terms of textual analysis. The textual approach to performance is linked to the increasingly generalized conception of the "text" in semiotic theory over the past few years. The term has now taken on a much broader meaning than allowed by its traditional linguistic and literary application, or even its current usage in textual linguistics. From a semiotic standpoint, the term /text/ designates not only coherent and complete series of linguistic statements, whether oral or written, but also every unit of discourse, whether verbal, nonverbal, or mixed, that results from the coexistence of several codes (and other factors, too, as we shall see) and possesses the constitutive prerequisites of completeness and coherence.1 According to this understanding of textuality, an image, or group of images, is, or can
pel, a text. A sculpture, a film, a musical passage, or a sequence of sound effects constitute texts also, or rather, they can be considered as such (Metz 2:9,71: ET 87). Clearly, therefore, even the units of theatrical production
, Jjchowri as performances can be considered as texts, and can thus become the object of textual analysis, provided that they possess the minimal prerequisites for consideration as texts.
Nil
II ■
III
-a. '"^
: ■■v^v•jv^-rv^■!v^-.;j;,;.1^^.
"lil!
48
The Semiotics of Performance
To anticipate some of my observations in this and in the following chapters, let me say that by /performance text/ is meant a theatrical performance, considered as an unordered (though complete and coherent) ensemble 0) textual units (expressions), of various length, which invoke different codes, dissimilar to each other and often unspecific (or at least not always specific), through which communicative strategies are played out, also depending on the context o their production and reception.2
It should be clear that the concepts of /theatrical performance/ anc /performance text/ do not completely coincide. /Theatrical performance/ in volves theater as a material object, the phenomenal field that is immedi ately available to perception and to an analytical approach. /Performance text/ refers instead to a theoretical object (or an "object of thought," as Prieh defined it in 1975), or to the theatrical event considered according to semi otic-textual pertinence, assumed and "constructed" as a performance tex within the paradigms of textual semiotics. We could therefore say that thi performance text is a theoretical model of the observable performance phe nomenon, to be assumed as an explanatory principle of the functioning c performance as a phenomenon of signification and communication. Wha is therefore at stake here is the theoretical model of an aspect of the perfoi mance-object: its textual aspect.3 The following chapters have the goal of de fining and structuring as precisely as possible the textual aspect of the vas category of concrete facts that constitute theatrical performances. As w shall see, however, not all theatrical performances constitute eo ipso perfoi mance texts. Similarly, not all verbal discourses are considered by textual lir guistics to constitute texts. Having thus attempted an initial narrowing c the range of objects on which the textual-analytical approach can be pe: formed, I will proceed to investigate the properties that are constitutive t performance textuality (its quidditas). It is perhaps appropriate to anticipal the main conclusions I will draw: in the long run it is reception (both gei eral and specialized)4 that determines the coherence and completeness of theatrical event, and thus qualifies or disqualifies it as a performance tex Hence any performance can be considered a performance text when the ii terpretive cooperation of the addressee desires (and is able) to "construe! it as such. According to Prieto, "the point of view that gives rise to the pe tinence of the way a material object is conceived is never imposed by tl object itself" (1975:131).
2.2. THEATRICAL PERFORMANCE: A DEFINITION
Before continuing, some clarifications must be made concerning the cat gory or class of material objects known as theatrical performances and rJ criteria according to which the category can be defined.
To begin, iTvould argue that it is not only representational theater that b longs to this class, but also all theatrical phenomena where the so-calli presentational aspect variously prevails over the representational aspe
The Performance Text
49
. where turning inward (self-reflexivity or self-referentiality) prevails over iurriing outward; where production (of meaning, reality, etc.) prevails over reproduction. To eliminate the trait /representation/ from a definition of theatrical performances as a class involves broadening the class sufficiently to include traditional and contemporary phenomena which our current sensibility as audience members or theater specialists tends to include more and more frequently in the domain of "theater"; celebrations, ceremonies, rituals (all the "social dramas" described by Duvignaud [1965]); contemporary avant-garde phenomena such as "happenings" and performance art; but also traditional genres such as dance and ballet where, as we all know, presentational aspects and self-reflexive abstraction take precedence over representation and the referential function. Even military parades and athletic events belong to this category. Indeed they have been linked to performance since the Renaissance; one has only to remember the case of the eighteenth-century Opera-torneo.5 It is obvious that we are dealing with a held that is far broader and more varied than the category consisting exclusively of traditional stagings of dramatic texts, to which some scholars still restrict the class of theatrical performances.6
By "representational theater" and "presentational theater" I obviously intend to designate two extremes, two polarities between which, as Valeria Ottolenghi writes:
It is thus possible to map out the single performance events ... in relation to the degree that self-displaying elements can be identified within them . signs that refer to something other than themselves. (1979: 29)
or
Mentioning in particular the actor's performance, Ottolenghi goes on to state-that it is possible to distinguish:
different modes of self-presentation: from total display of the self (public confession) to the presentation of one's social or professional role (the dancer or the boxer), right up to the complete representation of someone or something other than oneself (a disguise, a character). (31-32)
This means that while it is very difficult, if not impossible, to find representational performances where there is a complete absence of some presentational and self-reflexive element (given the indispensable fact that a staged event has a real, concrete existence, and that the actor is "truly" there, in the here and now), it seems equally difficult to imagine performances of a presentational type that are completely lacking in representational and symbolic components. As I will demonstrate in chapter 4, the production conventions within which the performances in question are located are extremely important. In representational theater, the mise-en-scene functions on the whole as a transparent semiotic system of renvoi, and as a fictional event, by virtue of the conventional canons that ground it, though it can also obviously contain presentational elements^real actions, concrete objects, and so on—which by themselves, at least to'begin with, lack a symbolic aspect. The exact opposite occurs with theatrical events that
I       .   ■  ■ . '
■
MAP':
.....
",.,1. ::«Í«Í
50
The Semiotics of Performance
can be placed mostly or entirely outside the canons of representation. In these cases, the underlying production conventions cause the performance to present itself generally as a self-reflexive and nonfictional entity, although it can contain fictional elements that point beyond it. Here I am thinking of events such as ''happenings'' or performance art, or Grotow-ski's "para-theatrical acts"; but the same holds true for traditional genres such as ballet, the circus, vaudeville, and the like.7
It therefore seems problematical, for different reasons, to use the concept of representation or its opposite to articulate an adequate definition, neither too vague nor too narrow, of the field of theatrical performance. If, on the other hand, we are to transcend the level of purely tautological deflations,8 as seems necessary, there is no other option than to adopt the cornmunicative-productive perspective, meaning the pragmatic one.
I will begin with Kowzan's definition of the "art of performance":
[It is] the art form whose product is communicated through time and space, meaning that it requires time and space in order to be communicated.. . . This is what the circumscribed definition of performance implies: a work of art which must, by necessity, be communicated in time and space. (1975: 24, emphasis added)
Obviously, the "autonomous and well-differentiated domain" that this definition circumscribes with respect to the other domains of art is extremely vast. It quite legitimately includes many very disparate forms oi performance in addition to theatrical productions in the narrow sense. It is particularly problematic that this definition does not distinguish between theatrical phenomena and other phenomena such as films or television programs, which are certainly performances, but which our "native" competence as theatergoers enables us to recognize intuitively as nontheatrical.
In order to overcome this lack of precision, I will articulate two bask requirements for theatrical communication (in which all the "performance arts" classified by Kowzan are brought together) by developing and honing Kowzan's own hypothesis. These are:
1. The real physical co-presence of sender and addressee (the latter is collective, as a rule).9
2. Simultaneity of production and communication.'10
Let us say therefore that the class /theatrical performance/ is constituted by all performance phenomena that possess the two properties just articulated.11 We thus arrive at the following definition: theatrical performances art performance phenomena communicated to a collective addressee, the audienct (physically present at the reception), at the very moment of their production (transmission). By /production/ is of course meant the production of the theatrica performance in its entirety and as performance (meaning, as we will see later, the integration of various partial texts into a single, coherent performance text), and not the production of individual parts of it (individual partia texts, in fact) which can precede, and often do precede, theatrical communication: for example the dramatic text, if one exists, or the staging, the co&
The Performance Text
51
tumes, the music. To insist on the condition of simultaneity between production and communication (between transmission and reception) means, I believe, to emphasize the discrete and specific aspect of theater—that every theatrical performance (every single theatrical occurrence) constitutes an unrepeatable, unique event, an ephemeral production that is different each time in spite of all attempts at standardization (rehearsals, director's notebooks, Modellbuch, repertory theater) and recording (descriptions, graphic transcriptions, audiovisual playbacks). Theater, in short, always involves event, as well as code and structure, as we shall see.
2.2.1. Technical Reproducibility, Repeatability, Duplicatability
At this point, it might be useful to articulate some further comments and distinctions in order to avoid possible confusion. In the first place, I would like to emphasize that I understand /unrepeatability/ in the strongest sense; that is, as non-total duplicatability. This does not exclude partial duplicatability. Theatrical performance is therefore unrepeatable, insofar as it is not entirely, reproducible (as is the case with film, painting, or novels), but it is nevertheless usually" possible to replicate it in part. It is no coincidence that in Italian the term /replica/ is used to indicate each single theatrical performance of a play.13 As for the level of repeatability, this can obviously vary according to each genre, tending to increase in the case of theatrical genres that are highly coded, such as ballet, circus, mime, classical Indian dance, Noh drama, and even Italian director's theater (in the style of the featro Stabile, for example).14
,: i But what are the reasons for this unrepeatability (i.e., "the non-total reproducibility") of theatrical performance? I must return to what I said earlier in chapter i regarding the total irreversibility of the process leading from the dramatic text (or script) to the performance and from there to its eventual graphic transcription. In fact, the non-repeatability of the theatrical performance can be explained in the first place (and not only in the first case, as we shall see) by the lack of a notational system that could unequivocally record the essential characteristics of a given occurrence in order to provide the opportunity of re-delivering it (repeating or reproducing it) as can be done with a musical work, thanks to the existence of sheet music.
At a basic level, therefore, the repeatability of a work seems to depend on whether or not it is suited to transcription into notational language. At least up to now this has not been possible for theatrical performance, since, as we have seen, theater lacks a close equivalent to sheet music, None of the "records" produced before or after a theatrical performance (from the dramatic text to the technical script, from the graphic transcription of a performance to audiovisual recording)15 fully satisfies the requirements of no-tationality. These documents cannot therefore be favorably compared to sheet music. As Goodman has pointed out (1968), these "transcriptions" do not succeed in unambiguously determining the class of occurrences that constitute the exclusive and equivalent examples of a performance text, nor
52
The Semiotics of Performance
can these transcriptions themselves be determining in an unambiguous way based on a given theatrical occurrence.16 All this is nevertheless usually possible with a musical score. It can therefore be stated that a musical work is in general repeatable, as are literary and architectural works.17
But could it also be said that a musical work, a concerto, for example, is reproducible in its entirety {or, rather, is capable of precise duplication) in the same way as a film, a photograph, or a literary text? Certainly not.18 And the reasons for this fact allow us to discover another level of motives that explain and give rise to the unrepeatability of theatrical performance. Indeed, while admitting that the elaboration of a system of complex notation could allow us one day to record and to re-execute a given theatrical occurrence in its totality, determining in an unequivocal way the specific way in which the work must be executed,19 we would still have to deal with a very different type of reproduction from what is found in cinema, photography, and television (and also from what occurs in painting or sculpture, which can by now be reproduced with the benefit of electronic instruments).20 While in these cases we can speak of materially identical copies (up to a certain point, and in different ways according to the techniques of duplication), in the case of theater, however (and also in the case of musical concerts), we are still faced with renderings that bear an imperfect resemblance to the original, despite the use of a hypothetical system of notation.21 This depends (and thus we arrive at the second reason for the unrepeatability of theatrical performance) on the "concrete characteristics of theater" (real-life actors, etc.): a medium that is not technically reproducible, unlike film, television, painting, and literary texts.22 In sum, it is this very material reproducibility, achievable through different procedures (manual, mechanical, electronic, and so on) which accounts for the possibility of furnishing almost perfect duplicates of works which belong to the autographic arts, i.e., lacking notational systems, and therefore lacking "renderings," as indeed in the case for works of sculpture, painting, cinema, and television. Although the technical reproducibility (or complete duplicatabiliry) of a work is independent of notationality, it nevertheless appears to be closely linked to the work's duration and material persistence, as the examples already given demonstrate. A painting, a sculpture, and even a film or a photograph are examples of texts that last through time and are always materially available. It can thus be stated, in the final analysis, that the non-reproducibility of theatrical performance depends on the ephemeral and transitory quality of its presence.23
It is opportune at this point to distinguish between works that are re-peatable, because they lend themselves to notation, and works that are technically reproducible (whether possessing the capacity for notation or not), and to introduce a third level beyond the two already examined:
1. Partial duplicatability (e.g., theatrical performance).
2. Repeatability (e.g., a musical concert).
3. Technical reproducibility (e.g., film).
The Performance Text
53
I must make one final observation on this matter. As already stated, even in the case of technically reproducible (or duplicatable) works, one can never achieve perfect copies, that is, "absolute doubles." And this is on the level of the utterance itself, of the co-text. If we now consider the problem from the point of view of utterance and reception, the concept of total reproducibility is subject to further limitations. It is in fact obvious how even works that are perfectly duplicatable can vary in an unrepeatable manner, because of variations in the practical context, a change in the receiver, or the like. Even if the text-in-itself recurs in an identical way, what will always inevitably vary is the text-in-situation (thus constituting an unrepeatable event), that is, the "interplay of communication acts" (Schmidt 1973) triggered by the work with the help of a complex set of contextual factors. Consider, for example, the copies of a film projected in different theaters and for different audiences, or readers of various copies of a literary text. Conversely, one could also take, for example, a film seen again by an individual spectator after a lapse of time, or a novel read several times by the same person in different circumstances. More generally,
it is impossible that two events are identical in every conceivable aspect, given that they must be distinguished if only for their location in time and space. To claim, therefore that one is witnessing a repetition it is necessary to presume the irrelevance or accidental quality of the spatiotemporal setting and the possibility that the same thing might occur in different places and moments in time. (Amsteramski, 1981: 76)
We, must conclude from this that even in the case of works that are technically reproducible (and hence reproduced), contextual reproducibility is al-: ways impossible. All of this is of course valid a fortiori for the theater, where the unrepeatability of the contextual factors (changes of audience, physical space, and the like) is added to and increases the unrepeatability of the co-textual element which has been amply discussed up to this point.24
2.2.2. Theater and Everyday Ldte
To return to the two principles articulated above and to the pragmatic definition of theatrical performance that can be derived fxm them, I would like to point out how this definition allows the field of inquiry a broad, even provocative, scope (including, for example, the spectrum of nonrepresen-tational theatrical phenomena already mentioned, in addition to events not usually considered "theatrical," such as a folk celebration, a game of soccer, or a military parade).25 Yet the same definition also provides a reliable criterion for making a clear distinction between theatrical performance and other important performance genres which are associated with it according to Kowzan's definition, and which we could describe as "nontheatrical." Cinema and television are the most obvious of these nontheatrical genres, the production of which, like theatrical production, is necessarily implemented through space and time. Because of its technical makeup, cinema never provides simultaneity between textual production and its communi-
■ ■  .. ....
■<v;
■f
■ ■ : ,■ :.:       ; ■     . ... ■
AM,
''yymmii
ynvm.y;^
54
The Semiotics of Performance
cation. Although television can provide this simultaneity (through liv broadcasting), it does not permit the real physical encounter of sender an addressee, also as a result of the specific technical characteristics of the me dium. Moreover, in contrast with the unrepeatability of theatrical perfoj mance (its merely "partial replicatability"), works created through the me dium of film and television are reproducible in full, as we have just seen This pragmatic definition also clearly separates theatrical performanc from other forms of artistic communication (such as literature, painting and music), bringing it closer in essence to the interactions of daily life/ This similarity is quite meaningful, and in my opinion far from dangerous if correctly understood. It has already been confirmed in the field of micrc sociology in Erving Goffman's famous hypothesis of "daily life as perfo] mance."
[Goffman] explicitly draws the analogy between normal interactions experienced by subjects and events that happen on the stage; he speaks of perspectives of theatrical representation, of principles of a dramatic type; "the common social relationship is in itself organized like a staged scene, with an exchange of theatrically exaggerated actions, counteractions and exiting 'lines.'" (Wolf, 1979: 88-89, citing Goffman 1959)
In one of his latest books, Goffman has taken up and developed this aspet from other perspectives:
Often what the speaker tries to do is to present dramas to an audience rather than to convey information to the addressee. It seems that we devote less time to conveying information than to performing. This meatricality is not based on the simple display of feelings. . . . The parallel between theatrical performance and conversation is much deeper than that. The point is that ordinarily when an individual says something, he does not state it as a bare assertion of fact. He re-relates it, he re-executes a string of already determined events, for the involvement of his listeners (1974: 508; cited in Wolf 90)
If correctly understood in its metaphorical sense and in its practical ut age,27 Goffman's "dramatic metaphor" proves to be extremely valuable t scholars interested in the type of macro-interaction that constitutes thea rical communication. Goffman does not simply tell us that daily life als involves fiction, lies, pretense, mise-en-scene, or that "representation ; not confined to the realm of fiction but constitutes an important, essentia device in our daily life." He also informs us, above all, that truth and lie; honesty and fiction, pretense and real acts, communication and seductior and display and concealment constitute the basic materials that theatria interactions are made of, just like everyday life. If life is also theater (o: more precisely, if it can be read and analyzed through the theatrical met* phor), then theater itself also involves real life, actions, transformation! and behavior.
A character staged in a theater is not real in some ways, nor does it have the same kind of real consequences as does the thoroughly contrived character performed by a confidence man; but the successful staging of either of these
569870
The Performance Text' 55
types of false figures involves use of real techniques—the same techniques by which everyday persons sustain thir real social situations. Those who conduct face-to-face interaction on a theater's stage must meet the key requirement of real situations; they must expressively sustain a definition of the situation.28 (Goff-man 1959: 254-55, emphasis added)
I fully realize that criticism that can be leveled against my definition of theatrical performance and also against its objective limits. In particular, I could be reproached for limiting myself to a definition of theatrical performance exclusively on the pragmatic-contextual level, and for severely neglecting the co-textual aspects, on the level of both expression and content. But the reasons that militate in favor of my choice are many, (a) A definition of theatrical performance based on content is clearly impossible. As Ertel has recently pointed out (1979:166), there are no meanings specific to the theater, or which theater is delegated to transmit. But the risk of a content-oriented approach to the theater is, above all, bound to bring the dramatic text to the fore, linking it to the very definition of theater itself. This seems to happen, for example (as we have seen in chapter 1), when it is claimed that the class of theatrical performances is composed exclusively of productions of dramatic texts.29 (b) Neglecting the semantic plane for the pragmatic one, the statement for the utterance, the "what" for the "how," my definition intends to restate and corifirm a fact that is emerging with increasing clarity from theoretical reflections and experimental practices in contemporary theater. This means that theater is more about production than product, process than outcome. It is also obvious by now that theater can never be reduced to its performances, since there are certain kinds of "performance-less" theater (if we consider theatrical animation, the para-theatrical activities of Grotowski in the 1970s, and the like). When faced with these kinds of theater, traditional critical approaches are completely inadequate and .inappropriate since they tend to judge from the outside (and from above) the creative products of a reality (groups, actors) about which nothing is known, and whose working strategies, customs, and cultural traditions are completely ignored (or apparently ignored, given the well-known habit of "critical distance"), (c) The final advantage that can be attributed to my definition is, as we have already seen, that it enlarges the limits of the genre to the greatest extent possible, avoiding the danger of excessive specificity, and thus including (perhaps even merely in a passing way) various phenomena traditionally excluded from consideration but which our current sensibility both as audience members and theater scholars would consider as "theatrical," i.e., as linked in some way or from some perspective to the theatrical universe. We can certainly include East Asian rituals of trance or possession, or religious celebrations in the Western tradition, spectator sports, and military displays, but we could also consider the widespread trend toward performance that has been occurring for the past several years in the artistic experiments of the avant-garde. In connection with this trend, the visual arts, music, and literature itself (in the case of "poetry readings") have tended in recent times to come to-
MS
56
The Semiotics of Performance
gether at least partially under the label of theater—a label that spans different genres—working on the elements of theatricality inherent in the principles defined above, physical display in public, the copresence of sender and addressee, simultaneity of production and communication, and the ephemeral and unrepeatable nature of the work of art understood as event and action.
2.3. COMPLETENESS AND COHERENCE OF THE PERFORMANCE TEXT
Since I have now established the meaning of /theatrical performance/ as it is used throughout this volume, I will return to the hypotheses of textual analyses that I began to formulate in 2.1. First of all, we must determine under what conditions a theatrical performance (i.e., a theatrical event satisfying the two requirements articulated in 2.2.) constitutes a text. In the last section I investigated the properties that constitute the theatricality of performance. I must now examine the properties that constitute its textual-ity, i.e., the properties that transform a theatrical performance—as understood within the paradigms of textual semiotics—into a performance text.
To begin, I must refer back to the previously mentioned generalization of the notion of /text/ in semiotics. My approach to performance is located within a field of textual semiotics of which Textlinguistik constitutes only one sector, even if this is currently the most developed sector in the entire field. I would argue that the extension of the categories, theoretical hypotheses, and models of textual linguistics (or discourse analysis) beyond the area in which they were initially elaborated, and above all their application to texts that are mainly, if not entirely, nonverbal, cannot be regarded as easily implemented or to be taken for granted, but must be evaluated one case at a time, unless we intend to proceed in a purely metaphorical way.
Paul Bouissac is undoubtedly one of the scholars who has tried most diligently to utilize the tools and concepts of textual semiotics in the study of nonverbal or mixed commuriication. Focusing particularly on circus acts in Circus and Culture (1976), he attempts to define a "semiotic text" in a sufficiently general way, i.e., in a way that would transcend the particular characteristics of a given expressive medium. Thus, having made the initial claim that "a text can be defined, independently of its physical and symbolic components, by certain formal properties," Bouissac proposes the following definition of /text/:
any permanent set of ordered elements (sentences, objects, or actions, or any combination of these) whose copresence (or collocation) is considered by an encoder and/or a decoder as being related in some capacity to one another through the mediation of a logico-semantic system. (126)
This definition is a more abstract reformulation of some "empirical conditions for the identification of a text" expressed by Bouissac on an earlier occasion.
The Performance Text 57
1. Clear-cut boundaries that isolate the message as such, i.e., nonambiguous and formal marks that delimit a finite set of meaningful interwoven elements.
2. A direct or an indirect endurance over time which makes it possible to "read" a message again and again as an "immutable synchrony" or an "ach-ronic present"; the possibility of repetition can be assured because the material of the message is invariable or because it is possible to memorize it exactly. 3. A deep system of relations ensuring the surface coherence of the message; in other words, a structure that accounts for its understandability, i.e., the possibility of building up from its elements a network of relations that includes all the actual terms as well as their relations. (90-91)
Leaving aside the second condition (direct or rndirect duration in tirne) which does not appear to be a preliminary property of all {semiotic) texts but merely a particular property of some,301 will pause to consider the other two conditions, (a) completeness of the text, and (b) coherence of the "elements" which constitute it. According to claims widely accepted both in linguistic theory and textual semiotics, completeness and coherence represent the principal and constitutive characteristics of textuality. These are the qualities that make a text a text (Conte, ed. 1977: 42). Examining completeness and coherence in relation to theatrical performance, in order to see if and how performance possesses these properties, is tantamount to asking if and how a single performance can be enucleated from the continuum of theatrical productions that are part of the "general text" of a given synchrony.31
2.3.1. The Delimitation of the Performance Text
Generally, we can consider completeness as a property which is not hrima-nent to units of communication, but which is assigned by the intention (or the attributed intention) of the senders and receivers. This issue concerns the subjective and therefore pragmatic character (context-dependent) of every operation of textual delimitation. Certainly, there are "signs of initial and final demarcation," but these—as Segre notes in connection with verbal texts (1978:132)—are in reality "evaluated by the observer only with the help of clues of a contextual nature." In effect, to cite Segre again, "the limits of the text are determined by the scholar of the text. ... It is not therefore possible to define general rules of textual demarcation" (132-33).32
To return to the present issue, it seems reasonable to suppose that an entire theatrical performance can be considered a complete unit of communication by both poles of the interaction (or at least by the sender). Nevertheless, apart from the possibility of taking into consideration performance texts that are smaller or larger than a single performance (which I will discuss later), it should be pointed out'that the actual delimitation of a single theatrical occurrence (and hence of a performance text corresponding to a single, entire performance) is not as easily or intuitively achieved as might appear possible at first glance, as is also true for verbal texts, whether written or oral.33 In fact, there are numerous examples from ancient times onward of so-called installment texts, or texts that are spread discontinuously
MR
58
The Semiotics of Performance
in time and/or space beyond the limits that Western culture has, from a certain point in time onward, customarily considered "normal" for a single theatrical occurrence. Examples of these serialized performances can be found in the puppet theaters of Sicily and Naples, as well as in the tradition of the Neapolitan sceneggiate. Among performances larger than a single performance event, we could include Renaissance festivals, at one time the structuring framework for games and theatrical entertainment, which were spread out in a discontinuous way in time and space and yet also constituted a "global" performance, the Great Theater of the World; or the medieval mystery plays and Sacre Rappresentazioni, which, as we know, were often performed over a period of days, at different intervals, and sometimes in several places, simultaneously or sequentially,34 But these references could also include, with better justification, non-Western theatrical traditions, and East Asian traditions in particular. A particularly noteworthy example is the performance format of one day in Noh theater (see Sief-fert i960).
In these cases we are presented with "performances" that fully deserve to be considered unified and complete, if only for the fact that they are so considered by their senders and/or receivers. Yet they do not fulfill the requirements of spatiotemporal demarcation (also linked to the participants' behavior during the communicative act) which Western culture has imposed as "normal," and as the only manner of execution imaginable. I am alluding of course to the traditional indicators or markers for the startoi a performance. These are (i) the audience arrives at the performance site, [(ii) the house lights are dimmed], (iii) the curtain is raised and/or the actors emerge, and so on.35 Traditional indicators or markers for the end of the performance are: (i) the actors leave the stage, (ii) [the curtain falls and the house lights are turned on], [(iii) the actors reappear to thank the audience], (iv) the audience leaves the performance site.36
To take a contemporary example, let us consider the works of the American director Robert Wilson, which up to this point have all lasted a great deal longer than is customary. The Life and Times of Joseph Stalin (1973) went on for twelve hours, Overture (1972) for twenty-four hours, and Mountain and Gardenia Terrace, which was presented at the Persepoli Festival in 1972 on the slopes of a mountain, lasted seven days and nights. Because of their extreme length (and sometimes also because they cover so much physical space), and because of the unusual slowness and repetitiveness that characterize them, these performances are inevitably experienced in-a-very discontinuous way. Indeed they demand this kind of reception, and are not damaged by it. The spectators converse with each other/ get distracted, eat, even fall asleep. Most important, they come and go constantly, and are absent from the performance for varying periods of time. After observing the audience's behavior during a performance of Wilson's Stalin in a New York theater (which lasted from seven o'clock in the evening to seven the following morning), Richard Schechner (1975) mentioned the phenomenon of "selective inattention," pointing out that an unusual pattern of movement
The Performance Text
59
was created between the auditorium and the adjacent space (set up as a .waiting room for the audience), with audience members passing continuously from one space to the other:
What happened during Stalin was unusual in orthodox American theater, but common in many parts of the world. People chose the parts of Wilson's work that they wanted to listen to carefully and the parts they could miss. Going into the LePerq space to rest, exchange opinions, have a drink, or get ready to go back into the auditorium or elsewhere added a dimension to the experience. The social goal of the circuit between the auditorium and the adjoining space was just as important in Stalin as the aesthetic viewpoint. (Scheduler 108)
Clearly, here as in many other similar cases, the traditional criterion based on the audience's entrance and departure proves inadequate, and we.must look for other markers denning the beginning and the end.37 Usually, however; an audience member who leaves early because of boredom, .disappointment, or some other reason does not come back to the auditorium. For this individual the performance is finished. Similarly, if an irate audience should abandon the performance space during a hypothetical
■■■ "avant-garde'' performance, this would also mark a conclusion. From that moment onward, theatrical communication as such comes to an end, even if the performance continues.
Toconclude, all that is left is to confirm what I have already said: it is not
■'[possible to articulate general rules for the external demarcation of the performance 'text (as is true for all other kinds of texts). The (delimitation) definition of a performance text is obviously founded on specific markers manifest by the text, but, in the long run, when the text is approached for the purpose of analysis, the discretionary powers of the analyst also come into play.38 Thus the performance text is a performance unit which the analyst's intention
: ■(or. the intention of the ordinary audience member) designates as semiotically com-
■-. plete.
2.3.2. Levels of Coherence in the Performance Text
What I have said regarding completeness is also true of the second requirement constitutive of textuality: coherence. This is the attribute that enables us to distinguish a text, understood as the unitary assembly of components on various levels, from a casual, disorderly conglomeration of elements. The debate on the conditions and levels of coherence of a text is still very heated and features a broad spectrum of different conceptions, depending on the various theoretical models called into play. It is nevertheless possible to note the emergence of various common or widespread tendencies regarding some of the fundamental issues. For example, coherence is viewed less and less often as a quality immanent within texts, and is increasingly understood as an element which the receiver assigns to texts, and which therefore derives from operations situated on the pragmatic level and linked to the context of production and reception. More precisely,
villll
60
The Semiotics of Performance
textual semiotics is shedding light on the fact that the coherence of a text depends on and derives from the determination of its topic (or theme), or on the practical operation through which the receiver (reader or listener in the case of verbal texts) establishes the "subject" (or focus) of the text in question:
Once it has been determined how the selection of the topic is a practical operation which, though inscribed in the text, must be actualized and rendered operative by the reader or listener, we can also better understand the problem of the text's coherence. In fact, the decision of whether a discourse is coherent or not is closely related to the viewpoint from which we read the discourse in question. A text's coherence is not an absolute given, but it is always a coherence with respect to something, a coherence with respect to the matter under discussion; in other words, coherence with respect to the topic. (Manetti and Violi 1979: 39-40)
Eco also affirms that the act of identifying the topic, and hence assigning coherence, is an inferential process, or more precisely, in Peirce's term, a task of abduction:39
To single out the topic means to posit a hypothesis regarding the symmetry of textual behavior. This type of symmetry is also what imposes the limits and conditions of textual coherence.40 (1979: 90)
Naturally, since it is possible to establish hierarchies of topics in a text, one can also distinguish different levels of coherence in a text: levels situated between the plane of linear or surface coherence (also called textual cohesion), resulting from the correct actualization of co-references, pronominalizations, and the like, and the level of overall or complex coherence, linked to the identification of the macro-topic or the macrostructure (van Dijk 1977a).41 In this practical approach to the coherence of texts, genre or textual type clearly plays a determining role. As Garavelli Mortara has recently observed, taking his cue from Segre: "It is only within each type [of text] that rules of cohesion can be reasonably formulated" (1979:16).
Now, before considering the coherence of the particular type of text constituted by performance—keeping the strictly pragmatic framework that discourse analysis tends to impose in this case, as we have already seen—I would like to point out that the question of the types and levels of coherence of the performance text must be discussed in rather different and specific ways, because of the diversity and specificity of many of its other textual characteristics (which I will examine in detail in the course of this book). In particular, I fully agree with Guarino's statement criticizing the fact that the issue of textual coherence has been treated up to now almost exclusively in thematic or semantic terms:
In order to deal efficiently and in a theoretically correct manner with the problem of textual coherence, which reinsritutes the terms of the codes and gives structure to their hierarchies, the priorities which have up to now been accorded to the semantic aspect must be revised and critiqued. Both Greimas's concept of the isotope, founded as it is on a principle of the redundance of
The Performance Text
61
semantic categories, as well as the models formulated by theorists of textual semiotics are incapable of taking into account the level of combination of different codes. This means the cohesion—for the sake of expediency I will call this type of coherence "superficial" when compared to the "deep" coherence of the content—that is implemented between the different components of theatrical expression. (Guarino 1978: 8)
As we will see in detail in 2.5., the performance text is characterized by a double heterogeneity, in its expressive media as well as its codes. Therefore, in the case of the performance text it is of foremost importance to single out the network of relations, interactions, and hierarchies which organize the various codes in order to recognize/assign a generic semantic coherence. I will call this network the textual structure of performance, and will discuss it in detail in chapter 3. The textual structure of performance, insofar as it is a more or less hierarchized network of codes, is what grants the performance text its coherence on the level of code relations, on which I believe the concrete cohesion between the different expressive elements of performance depends. Though coherence between the codes is not the only level of coherence within theatrical performance, it is nevertheless the most important, transforming a performance into a performance text, and allowing it to be modeled according to the paradigms of textual semiotics. This level can thus be called textual coherence.
As for its pragmatic character, this emerges from the nature of the textual structure that gives rise to it. As we will see in the next chapter, the textual structure of performance is not a concrete entity existing prior to critical analysis, but a theoretical "construction" of the latter. It follows that even in the case of performance, textual coherence is not a property immanent to the text. It is analytical-interpretive analysis42 that gives (can give) textual coherence to theatrical performances (constituting them as performance texts), even those presenting premeditated and explicit signs of "incoherence" during their enactment, such as "happenings," street theater, or more-recent performance-art events. Thus, on the level of code relations, coherence is found in performances where the receiver is able to "construct" a textual structure of performance. This is not impossible even in the extreme examples just cited (and in other similar ones) if we do not understand the textual structure of performance to mean simply a strictly unifying integration of codes, but see it instead as an overall pattern capable of covering a wide spectrum of relationships and co-textual interactions, ranging from simple juxtapositions to hierarchized interrelationships (Guarino 176).43
As already mentioned, the level instituted by textual structure is the principal but not the only level of coherence that can be attributed to the performance text. For the performance text and other linguistic texts, one can also hypothesize a (deep) semantic coherence and a (superficial) syntactic cohesion. Semantic coherence involves the eventual thematic unity of the text, its total meaning (which can be expressed in macro-propositions), and syntactic cohesion is assured by cohesive agents similar to those employed
62
The Semiotics of Performance
for the same purpose in verbal texts: anaphoric and cataphoric mechanisms, co-references, repetitions, paraphrases, and so forth.44 Contrary to what Guarino seems to propose, I would argue that the level of coherence between codes must not be confused with syntactic .cohesion, which connects semiotic units, not codes, and is therefore more superficial than the other level. We are nevertheless dealing with strictly interdependent planes; syntactic coherence cannot be obtained without coherence between the codes and vice versa. This is not the case, however, when dealing with deep semantic coherence. One can imagine cases of semantically "deformed" performances (variety shows, curtain raisers, etc.), and vice versa ("happenings" deliberately featuring discontinuity on the syntactic level and between the codes, and nevertheless endowed with semantic coherence if only on the level of what Segre has called "motivational paraphrases") (1978) (see note 40). These examples confirm what I have already stated regarding the necessity of carrying out an analysis of a text's rules of coherence within a genre or textual type.
2.4. THE PERFORMANCE TEXT: BETWEEN PRESENCE AND ABSENCE
In the preceding section I examined the performance text's relationship to the two characteristics that constitute textuality (completeness and coherence), i.e., traits which make a text a text, and which must therefore be possessed by all texts. Now I will describe the textual characteristics proper to the performance text itself. Having investigated if and under what conditions a theatrical performance can be considered a text, I must now ask: "what kind of text is a ■performance?" and, "what are its specific textual properties?"45 I will first offer some observations on the degree of "consistency" and persistence of the performance text, as well as on its suitability for analysis, since there has been some confusion regarding these issues. To this end, I will return to the previously postponed question of the (direct or indirect) duration of texts in time, and hence of their availability for repeated fruition at will.
I feel at least some degree of perplexity in the face of the assertion by Gulli Pugliatti that "nothing in research or cultural work, in books, exists 'as such'" (1978: 621). She continues her argument with the claim that the position of semiotic scholars vis-a-vis their "object" of study is not sustan-tially different from that of other scholars in humanities or the physical sciences. Because it is undoubtedly true that every material (empirical) object becomes "present" as a theoretical object only when assumed within the paradigms of a given discipline, and only to the extent to which the theoretical and technical tools of the discipline are capable of constituting it (and of "constructing" it) as such. Before this occurs, the object is only formless raw material, "absent" from the scientific perspective.46 It is also true, however, that this indisputable circumstance cannot erase the profound difference that exists in relation to the respective objects of mquiry,
The Performance Text
63
between the position of a semiotician studying theater and that of a semi-otician studying cinema, literature, or painting (just to limit ourselves to the field of artistic endeavors). A simple shift of levels does not abolish that difference. In fact, for the semiotician specializing in theater, the object is doubly absent. In the first instance, it is missing as a scientific or epistemo-logical object (since the discipline has not yet been able to constitute it as such: a situation capable of resolution, or already in the process of being resolved, and one that is nevertheless not exclusive to the semiotics of theater); in the second instance, it is absent as a material, prescientific object. This second absence47 is certainly particular to theatrical performance, at least in the field of the so-called artistic languages, the production of which generally involves a very different kind of presence and persistence. A literary text, a painting, a film—insofar as they are empirical objects—all exhibit in different ways and to different degrees a capacity for repeated fruition and analysis (a capacity that is greater in the case of the novel, less in the case of a film text)."*8 This capacity is entirely lacking in theatrical performance. As we have seen, at the end of the temporally limited process of production/reception (a process that is not controlled by the audience),49 a performance "dissolves," leaving behind only its "traces" which are more or less consistent according to each instance: script, director's notes, filmed, photographic, or televisual documentation, descriptions by audience members, critical reviews, and the like. Apart from its unrepeatahil-iiy, in the sense of its being "unreprodueible" (see above 2.2.1.)/ performance is also unrepeatable in the sense that it is not lasting or persistent, does not lend itself to repeated fruition, and cannot be repeated at will. As Pagnini points out, "in theater the act of utterance dominates absolutely over the statement" (19B0: 95).
It is reasonable to suppose that this second type of "absence" (meaning the ephemeral character of theater's material presence) is accountable at least in part for the poor state of theater semiotics at the present moment. The unrepeatability (in the double sense just explained) of theatrical performance (of a given performance event), and hence its absence as empirical object at the moment of analytic endeavor, are also the main causes of its absence as scientific object, that is, in the present case, as performance text. This explains the fact that the field of theater semiotics lags behind other branches of aesthetic semiotics, such as the semiotics of music, cinema, and the figurative arts.
2.4.1. "Present" Performances and "Absent" Performances
I nevertheless disagree with Ruffini when he generalizes in a rather metaphysical way about the element of absence characteristic of theatrical performance without distmguishmg different cases and situations. I will now briefly attempt to clarify some of the necessary custmctions..This involves verifying whether the element of material absence, though common to all theatrical events, might nevertheless allow such internal variations as to
64
The Semiotics of Performance
permit different modalities for the recuperation of a given performance, in fact, for its "reconstitution" as a discursive occurrence. Indeed, the documentation relative to a performance event (the general class of things that I invoked earlier as "traces" of performance) can vary substantially from case to case, both in quantity and quality,but generally dirrurushing as one goes back in time. Predictably, however, the procedures for textual reconstitution change in a very relevant way when, for example, we move from a Renaissance performance (the staging of Bibbiena's La Calandria in Urbino in 1513, let us say) to one of Luca Ronconi's contemporary offerings, which can be viewed directly and/or mdirectly through video recording. I thus propose that we distinguish between (a) (partially) present performances and (b) absent performances.
(a) (Partially) present performances. I have decided to use this term for those performances which the scholar/audience member can view directly and/or through video broadcast or recording.
(b) Absent performances. This term will designate performances where no access to direct and/or indirect viewing (through audiovisual recording) is possible for the scholar/audience member.50
The distinction prompts some immediate observations: (1) This corresponds only in part to the distinction between contemporary performances and performances from the past (which is one of the reasons I prefer it). If it is indeed true that all performances that are "partially present" are, for obvious reasons, xonlemporary, it is nevertheless equally true that not all contemporary performances are_of necessity "partially present," according to my definition. For the analyst or the common theatergoer, only those that can be seen directly or indirectly with audiovisual aid are in fact "partially present."
--------(2) My classification is of a binary type, and hence has the advantage of
leaving no margin of uncertainty. According to this scheme, a performance is either present or not present. In reality, however, the documentary status of performances is much more nuanced and offers a range of intermediate stages, including, at one extreme, the absolute lack of direct documentation,51 and the possibility for direct and/or mdirect audiovisual viewing at the other. From this perspective, it can also happen that a performance from the past that is more amply documented than a contemporary performance is more "present" than the latter, yet only in the graduated sense of the term. This is certainly the case with certain comedies and celebrations from the Renaissance for which we can often find excellent literary and iconographic documents: plans, budgets, reports, chronicles; drawings, etchings, paintings, and the like.52 Consider, for example, the research on the Roman festivals of the sixteenth and seventeenth centuries carried out by Cruciani and others. Cruciani's study provides a "reconstruction" of the festivals of 1513 in particular (1968, 1972). Similarly, within the context of her larger work of the Roman carnival of the seventeenth century, Boiteux has provided a study of the very famous and well-documented Giostra del saraceno performed on February 25,1634 (Boiteux 1977). Carandini and Fa-
The Performance Text
65
giolo delTArco (1977-1978) in particular have undertaken a systematic analysis of L'Effimero barocco that offers a valuable inventory of the main festive events of the seventeenth century. The study provides a carefully designed chart, taking into account the date, occasion, and location of these events; the relationship between commissioner, executor, and audience; a description of the development and physical trappings of the celebrations, as well as notes on allegorical meanings and cultural relationships; literary and visual sources; and a bibliography.53 But we could also consider many productions mounted by the "fathers" of the modern stage, from Craig to Mejerchol'd to Brecht. These frequently offer a considerable amount of material: director's notebooks, interviews, photographs, . drawings and sketches, articles from newspapers and magazines. The first example that comes to mind is the famous Hamlet staged by Stanislavski at the Moscow Theater of the Arts. This took three years of work to mount (1909-1911). Apart from the comments by Stanislavski himself in his book My Life in Art (1924), many sources of written information are available on the play, most of which remain unpublished. These include a copy of Hamlet with director's notes written by Craig (1909); the correspondence betwen Craig and Stanislavski; Craig's Day-Boohs I and II; notebooks containing the English transcription of conversations between Craig and Stanislavski; notebooks with the Russian version of the same conversations; a series of seventy pastel and charcoal drawings prepared by Craig for the Moscow show; photographs of characters and scenes in the performance; models of the sets, reproductions of those used on the Moscow stage, which were created later by Craig's son, Edward Anthony; and seven sketches by Stanislavski regarding changes he made in Craig's scenes.54
(3) One might object that the mere viewing of a performance (without the subsequent possibility of audiovisual recording) does not constitute a sufficient condition for claiming that a performance is "partially present." Now, it is beyond doubt that a filmed performance with accompanying sound track (or a video recording executed with more than one camera and standardized sound) remains up to this point the most complete and faithful way of documenting a performance.55 I would however argue that, even in the absence of an audiovisual recording, a theatrical event which can be viewed directly (in the theater) and which is also closely synchronous to our time offers access to many other forms of direct documentation, internal and external (photographs, director's notes, interviews with the director and with other persons involved in the production, the press book, descriptions by the audience), which, given their quality and sheer volume, are often capable of compensating more or less successfully for the absence of the basic document. Having said this, I must repeat that the optimal conditions for an analytical approach to a performance (beyond direct viewing and reviewing, and using other documentary materials) should involve the possibility of working with audiovisual recordings of good quality.
As already mentioned, I would argue that the profound qualitative and
66
The Semiotics of Performance
quantitative differences in documentation existing between performances that I term "partially present" and those that I term "absent" are such to require the analyst to adopt very different techniques in order to proceed with the textual recuperation, reconstitution, or analysis of the performance. One may use the term description/transcription56 only in the case of performances that are "partially present," that is, performances whose texts are in some way available for description/transcription, since they are viewable or re-viewable directly, and/or through audiovisual recording. Therefore, only in the case of "partially present" theatrical events is it possible to attempt the procedure that leads from description/transcription to an analysis of the textual structure of performance without losing the two-dimensional aspect (in both spatial and temporal terms) that is characteristic of performance.
In the case of an "absent" performance, instead of describing/ transcribing a discursive entity which does not exist as such, it is the analyst's task to reconstitute it. Regarding the procedures for "recuperating" a performance, the opposition partial presence/absence (meaning, in general, contemporary performances/past performances) changes into the opposition description (or transcription)lreconstitution. Kowzan correctly argues:
In this work it is necessary to distinguish between two very different situations which correspond to two contrary moments in time. Must we reconstitute a performance in the past which we were unable to see? Or must one freeze a contemporary performance? (1978: 52, emphasis added)
"To reconstruct" (or reconstitute) an "absent" performance, and in particular a performance from the past, means to attempt a contextual retrieval of that text, through recourse to indirect documents (direct sources being scarce, and generally "illegible" on their own). This would involve consulting other texts from the same cultural synchrony, which a careful historical investigation would indicate as relevant to the performance under analysis. These documents involve both theatrical materials (records of scene paintings, architectural design, dramatic texts, information on choreography and costumes) and nontheatrical materials (documents of a pictorial, literary, rhetorical, or philosophical character, or materials on town planning, etc.). These texts constitute in general what I refer to in 2.7. as the cultural context of performance, and, taken together with other elements, they also constitute the general (cultural) text.57
2.4.2. Rutfini: Contextual Restoration
A very well-argued proposal on the issue of the contextual retrieval of an "absent" performance can be found in Ruffrni's writings, especially his 1976 volume. ^
Starting with the basic concept of the "absence of the signified" {in his view, the meaning of an expression—but perhaps it would be better to say its referential or extensional "truth"—cannot be attained in or with lan-
The Performance Text
67
guage, but can be only indefinitely deferred, postponed through the unlimited play of meanings),58 Ruffini (1976a) arrives at a definition of the content of a given text (called i-Text [i-T] as a "set of expressions") as the set of descriptive expressions of the classes obtained by dividing up the i-T on the basis of a relationship of equivalence ("the meaning of E [i.e., an expression of the i-T] is nothing other than its equivalency class" [8]).59 It follows that the adequacy of a certain descriptive expression E to describe a certain equivalency class is not founded on the superiority of its linguistic level with respect to a generic expression of that class Ei (in fact, for Ruffini, these belong to the same metalinguistic level, "in which language reveals its capacity to transcribe itself and therefore to describe itself" [12]) but only on its actual or virtual belonging to that class itself.
With regard to the analysis of an i-T, a formulation of this kind leads to some quite important and particularly fruitful consequences for the particular example of the i-T provided by the performance text.
(i) The description60 of an i-T is resolved in its "rewriting"—constituted, in fact, by the set of descriptive expressions of the equivalency classes: obviously this and the division of equivalency that underlies it "involves . . . a great deal of partiality which, at the same time, provides the only guarantee of the legitimacy and vahdity of the analysis" (24).
(ii) Therq is no longer any real separation or hierarchy dividing artistic work and criticism {or science) since all texts can be located at the level of the semiotic analyses that can be conducted on them: "Only by confusing a relationship of precedence with a relationship of cause and effect can one consider the artistic work as necessary cause for its rewriting." At this juncture we find an issue that is extremely important for an "absent" text like the performance text: "The rewriting of an i-text largely leaves out of consideration its integral (entire? complete?) persistence, that is, its existence at the moment in which analysis intervenes" (I.e.).
(iii) Accepting the impossibility of attaining "objective" contents, from now on in the analysis of texts no more true or false descriptions will be offered, but only descriptions, or rather, rewritings, which are more or less effective. Their effectiveness will depend on the degree of relationship which they manage to establish between the i-T in question and the sum of all the other synchronous i-texts which make up the general (cultural) text.61 More precisely this effectiveness will be assessed "by the number of 'disparate' phenomena which transcription is capable of 'realizing'; that is . . . by the number of expressions from outside texts which can be acquired in the equivalency classes" (26).
(iv) The analysis of a text is "nothing other than a contextualization of this text to a greater or lesser extent." It is clear, in fact, that a greater contextualization of the i-T in question, that is, of its being placed in relationship to a greater number of external i-texts, will increase the number of expressions of its equivalency classes. Generally speaking, this will necessarily involve a diminution of the component parts of class definitions, "that is, in semic terms, a recognition of a smaller number of present
68
The Semiotics of Performance
(pertinent) semes in expressions of the class. According to Ruffni, it is therefore possible to argue that contextualization is associated essentially with the creation of a relative theoretical model" (20).62
I shall return later to the problem barely mentioned by Ruffini regarding the relationship between the aesthetic text (a performance text or some other kind) and the general text of culture. For the moment, I must insist on the characteristics of the operation of contextualization and its consequences in order to reconstitute performances from the past. According to Ruffini:
[In contextualization] the basic text (i-T) becomes the site of a continuous centrifugal movement toward the outside, toward other i-texts, and, vice versa, it is also the site of a continuous, centripetal return, from the cultural text [the general text] toward the text under analysis. (19)
According to Ruffini, if an analysis of the text in terms of contextual rewriting is always the only theoretically correct and productive path, in the case of such an extreme example of textuality as the performance text this also seems the only possible one. Thus, we see the emergence of a strong hypothesis for the reconstruction of past performances conceived in terms of contextual retrieval and hence necessitating a multidisciplinary approach:
The i-text, as we have seen, is a subset of the cultural text. But how does such a subset present itself for the scholar who does not wish to equate theater with dramatic writing? As a space in which the boundaries are unknown, but which is still an almost empty space. The scholar is aware of having access to only a small quantity of the expressions that constituted the i-text. Often these are scattered, fragmentary, heterogeneous elements. Here, the intrinsically temporary and relative character of every division constitutes a theoretical foundation for the analysis. In each case we can form equivalency classes. But the possibility of implementing them, of making them useful, is linked entirely to the contextualization of the i-text. Theatrical research thus finds that its peculiar handicap (the scarcity and fragmentary quality of the elements which constitute its texts) is reinstated by that very element that, paradoxically, is often considered another and not less negligible handicap: the composite nature of the performance event. Indeed, while the linking of a literary i-text to an architectural, pictorial, or other form of i-text might seem "critically foolhardy," in the case of a theatrical i-text, this is natural and necessary. For example, the "poverty" of a theatrical i-text from the Renaissance period can (and must) be compensated for by acqtiiring i-texts from architecture, painting, town planning, rhetoric, treatises on manners, and so on. (24-25)
Ruffini's 1978 volume (which I have already referred to many times) again presents his theoretical and methodological proposal for a reconstruction of performances from the past ("absent" performances, in my terminology) through contextual analysis in the same terms as the study discussed above. However, more importantly, Ruffini had previously offered a very interesting, practical argument in another article written in 1976, sig-
The Performance Text
69
rriflcantly entitled "Contextual Analysis of La Calandria among Performances Given in Urbino during 1513." This is dedicated to the "reconstruction" of the theatrical site of the famous Urbino production of Bibbiena's play (Ruffini 1976, 1976b).
Because of lack of space, I will not discuss in detail Ruffini's method, hypotheses, and conclusions in his very close analysis where philological rigor is combined with great intellectual insight. I will limit myself to only one example, mentioned by Ruffini (1978a) in his introductory essay: the octagonal construction placed at the center of the stage in the Urbino production of Lo Calandia, according to the testimony of Baldassare Cas-tiglione, who provides a very precise description of this construction with careful attention even to ornamental details. If the analysis of this textual unit had been limited exclusively to the performance text of origin, its co-text in effect, we would have tiad to^arrive-at a definition that "would have recorded all the constitutive parts .of the description of the object, without having to include different expressions" (1978a: 28). In no way, in fact, could we have grasped the pertinence of the buttding's "octagonal-ity," much less its "omamentality" or colors. This becomes possible by opening up our analysis to the context, and by the subsequent acquisition by i-text of other expressions from the same cultural synchrony. Linking the performance text of the 1513 staging of La Calandria to other texts from the same period and from other centuries (passages by Vifruvius regarding the foundation of the city, la Sforzinda del Filarete with its eight towers, illustrated documents on the "towers of the winds," graphic interpretations of Vetruvius's city by Fra Giocondo, Barbaro, etc.) allows us to recognize a relationship of equivalence between the Urbino temple and other contextual expressions. It permits us to "model" the octagonal edifice and, on the basis of the very precise yet useful initial description, to draw a small number of pertinent constitutive features: "sacred object" + "octagonal" + "central." Only at this point does the textual expression constituted by  the  octagonal  construction  described by Castiglione become really meaningful for us. According to Ruffini's concluding hypothesis, we recognize the "symbol," the "figurant" {to repeat a term used by Schefer [1969]) of the model of a city, of the "ideal city," of the Utopian city envisioned in Renaissance culture and in the art of the fifteenth and sixteenth centuries, a city which found its "natural" site of projection in theater.
I would argue that Ruffini's contextual analysis constitutes the most rigorous and convincing study to emerge up to this point in the subfield of theatrical historiography dedicated to the reconstruction of performances from the past. I have only a few marginal (though still significant) reservations to make regarding his work, without intending to challenge its overall persuasiveness. More precisely, I would like to make two general observations of a theoretical character, and one specific argument (that is specific to the problem of theater) of a methodological character.
(1) I am in agreement with Ruffini's opposition to the notion of the "sig-
., ....... r^yyyyy.
' '■■yyyLyyy^yyy.y.
: ■■■■yyyyymy ?
—ayy^yy-M
'-yy-b,:-.
yy-
•V.
wmuB
yyy -y
y/y:'^m^:y^
yyymm
y -;^Äfft .^:-:.u^:^ym ^y
yy.,yy
. ■r:My0:':-yy]^ry
y^y&c^M y--y<--. ■yyr:0m: ;yy,y.---^
-■■\ y. y%€:^:y^:yy.
mym
yyywym
^yä^yy^y.y-fy ■    -yr:r-:-:yy-'yy^
.---Mr,-*-}
mmmmyy
:-yy yyyyyy+myy^
70
The Semiotics of Performance
rtified" (meaning), which he pursues in the wake of the French theorists of "absence" (but which is grounded above all in the semiotic-linguistic work of philosophers and logicians such as Peirce, Wittgenstein, Carnap, and Quine), as long as he is engaged in rejecting all aspects of a metaphysics of content and truth, stating with an appropriately radical approach the inevitable partial and relative character of all semantic descriptions and all textual analyses. This also means unmasking the periodically recurring temptations of gnoseological essentialism and realism (even in the field of semiotics), according to which semiotic analysis consists in the recognition of objective meanings (of "truths") that are immanent to expression (the text) and preexist analysis. I part company with Ruffini when, in the midst of his earnest debate against crudely referentialist and realist conceptions of signification, he seems to broaden the target of his own criticism to include all theories of meaning, as if it were only possible to conceive of the signified (and denotation as well) in referential and extensional terms (for Frege, only as Bedeutung and not also as Sinn). In my opinion, it is quite possible to avoid a rigidly referentialist and extensional conception of content without having to repudiate the very notion of meaning. This is what Eco succeeds in doing in his theory of the codes {1975), by.defining meaning as a cultural unit and projecting it—using Peirce's category of the "mterpretant"—into a dynamics of "unlimited semiosis." The process demonstrates
how signification (as well as communication), by means of continual shiftmgs which refer a sign back to another sign or string of signs, circumscribes cultural units in an asymptotic fashion, without ever allowing one to touch them directly, though making them accessible through other units. Thus a cultural unit never obliges one to replace it by means of something which is not a semi-otic entity, and never asks to be explained by some Platonic, psychic, or object entity. Semiosis explains things by itself. (Eco-1975: 104)
What emerges from this is that meaning is not only nonreferential but also relational or oppositional, which is to say, non-substantivized. A content (interpretant) is defined as such with respect to other expressions that convey it, but is—or can be—an expression for other contents or interpret-ants in other sign functions.63 Now, this meaning/cultural unit can be considered an "extralinguistic reality" (Ruffini occasionally seems to consider it as such) only to become confused with the referent/4 that is, with objects, events, or states of the world which are thought to correspond to the content of a sign function, but which the intensional semantics of a theory of codes must absolutely exclude from its ken as something that contaminates its theoretical purity (Eco 1975: 88-91).65 To conclude this point, I would argue that the notion of meaning as synonymy, present in Carnap and Quine, is, in spite of Ruffini's claim, altogether compatible with the notion of meaning as a "cultural unit inserted into a semantic system characterized by contradictoriness and recursion as the result of uruirnited
The Performance Text
71
semiosis, and in the process of continuous restructuration" (Ruffini 1976a: 8-9).
(2) While elurimating all hierarchies between creative work and criticism, that is, between language and metalanguage, and quite appropriately viewing the readings of a text as its partial, relative "rewritings," Ruffini nevertheless goes as far as to postulate an unlimited multiplicity of these rewritings (e.g., 1976a: 24, where he speaks of "mfinite possible rewritings"), thus aligning himself with the well-known views of Bar-thes (1970 and 1973a). I cannot agree with this, however, and I will address the issue in greater depth in 3.4. For the moment I will limit myself to asserting that textual analysis is always the result of a dialectic between fidelity and freedom, between the constraints of the text and the choices of the analyst. In the next chapter I will address the issue of limited multiplicity and pertinent readings in order not to devalue either of the two poles of textual analysis. By pertinent readings I mean interpretations that are capable of assigning a textual structure to the theatrical event in question. To clarify, I would say that the possible readings of any performance text are at least indefinite if not infinite. This is not the case for the textual structures that can be attributed to them, which must instead be conceived in necessarily limited numbers (and, if possible, in very small numbers), in the same way as'the relative (pertinent) interpretations that "construct" them.66
(3) The third reservation is, as I have said, of a specific type, and brings us back to the main focus of the present chapter. It seems indeed that the validity and, above all, the necessity of the method of contextual reconstruction is limited to "absent" performances, especially performances from the past. In these cases, the usual, sometimes almost total, absence of direct documentation (with the obvious exception of the dramatic text) means, on the one hand, that indirect texts (both theatrical and extra-theatrical) must be consulted and, on the other hand, that the scholar is obliged to work on individual portions of the text, on partial texts—like Ruffini himself in the case of La Cahndria's stage sets—abandoning the goal of a unitary and supposedly comprehensive analysis of the theatrical event.
In the case of "partially present" performances, I would argue by contrast that, if one proceeds first of all from the description/transcription of the play and then, on the basis of this, to the discovery of its textual structure, the analyst retains the possibility of safeguarding textual unity and completeness, at least theoretically.67 As Bettetini correctly observes:
The quest for the textual structure of a theatrical performance should already to a great degree be "produced" and assisted by the process of transcription, which would successfully achieve a graphic representation of the performance through accurate symbolization techniques. (1977: 21)
I would now like to devote some brief attention to the theoretical and tech-
72 The Semiotics of Performance
nical problems posed by what Bettetini calls the "graphic transcription" of performance.
2.4.3. Description/Transcription
As already mentioned, the goal of description/transcription is to provide an adequate substitute during the analytic process for the absence of the performance itself by providing the critic with a metalinguistic reconstruction that could serve as the basis for an appropriate textual analysis. Up to now I have spoken of /description-transcription/ only in a general way. With the help of Bouissac's study (1973: 13-15), we can distinguish between (a) descriptions used for the "translation" of verbal language of a more or less technical kind; and (b) transcriptions used by "original systems of analysis and symbolic notation."68 Obviously, these chstinctions concern only the paralinguistic and nonverbal texts within performance, and not the verbal text. In fact, if we begin with the supersegmental elements (intonations, accents, timbre, pauses),69 the written record of the verbal text should not present too many problems, as I have already shown in 1.5.
According to Bouissac (1973), all known attempts to provide a graphic reconstruction of performance are classified as "descriptions" (I should note that these attempts have not been numerous), since they are either predominantly or exclusively verbal. Take, for example, Stefan Brecht's minute reconstructions,70 or the even more ambitious analytical project by scholars at the Centre National de Recherche Srientifique in France, whose work has been published in a series entitled Les votes de la creations theatrale and already comprises eight volumes. Above all, these projects explicitly express the overall goal of providing a multilevel performance score in the most comprehensive and "objective" way. Despite this goal, or perhaps because of it, the results seem far from successful.
Let us take, for example, the description offered by Serge Ouaknine of Grotowski's production of the Constant Prince (Calderon-Slowacki) in the first volume of the CNRS project. At least in the more analytically described sequences (the first twelve of the play), Ouaknine elaborates a kind of script or scenario, legible both vertically (the temporal axis) and horizontally (the spatial axis).71 The central part of this description is taken up by the verbal text (Grotowski's adaptation of the Slovak version of Calderon's play), marked with appropriate indicators referring to the two lateral columns to the left and to the right of the text. The right-hand column offers a verbal description of the sound track, excluding words (music, intonations, noises), and the left-hand column provides a description of the nonverbal texts (mime, gestures, movements).72 To fill the remaining spaces in each page of the volume, Ouaknine adds various visual devices with goal of achieving an adequate transcription of the performance. These devices are of three different kinds: (1) Sketches of the layout of the stage, indicating the position of the actors, their principal move-
The Performance Text
73
merits, and verbal stage directions giving the basic actions within a particular sequence; (2) drawings of the spatial layout of the stage as seen from the front, with the actors sketched in charcoal; and (3) photographs (see figure 1).73
Sequence 10; Intermezzo (where all the acton ate left to (heir own devices)
£4 Eu^t^W,
r
Tabvdakt In temp 1« the jealoue episode with a long sigh, and pretendsliumli iIiiTiji luliii ■iiuiflln ulimi u ■liaainillnfiiiiii a section of her blidt dress). Mourn grata Pan* ly the collar and pushes bar forward eontenptuoualy. She walk* In a mechanical maimer. The Kmc take* the red cloth from Dtw Faatuw and leturna ittoFzn^whoputs ItonatthnighttweRamaiihlla. Don rsnanand T*«ouo*ht gHde ty and men edt followed by Moutar. Theampm-theatcrlaTiow empty. stretches out on the platform near the
rr&nbHng Don VwtHwc, and reveals hi* plan in ■ condHstcey but ambiguous tone.
Tw Kmc - Ahtb:
239JI Prince, behind these rocks,
and over Iha^ bdimd file mouttatni
Ifaavenude preparations foe i hunt.
FpfnrF the tun wwt—ww—
lla crown amid pearUand cor all,
huntsmen wSl rush forth from thatapot
and will hunt down the tiger.
It will be an amusementfor you, my Prince.
[ Handwritten nntatfon in diagram read*: "He Prince dcesnot move"]
Figure 1. Jerzy Grotowski's The Constant Prince, reconstructed by Serge Ouaknine.
Ouaknine's effort is interesting (like others mentioned in the eight-volume Voies),74 though pursued in a manner that was not totally systematic. The expedient of adding nonverbal texts to the transcription of the verbal text only occasionally succeeds in preserving the two-dimensionality of the performance, which is to say, in synclironizing the partial texts with each other. This is not to mention the fact that it represents a decidedly ad hoc expedient, unusable in the case of nonverbal performances. Furthermore, the verbal description leaves too much room for ambiguity and inaccuracy (even though it pretends to be a comprehensively effective tool). As for the nonverbal devices, these are still too empirical
74
The Semiotics of Performance
and imprecise to provide a foundation for a satisfying system of multiple notation.
I must repeat, however, that actual transcriptions of (entire) performances are unheard of, and it is easy to imagine why, given the enormous theoretical and technical difficulties presented by such an endeavor. Nev ertheless, at least since the second half of the nineteenth century—from tht beginnings of modern theatrical directing—there have been some notabk developments in the methodology of recording the movements of actors or stage, and particularly in recording the relationship between the actors themselves as well as between the actors and objects on stage. (It is no se cret that the Utopian fantasy of comprehensive notation was shared by verj different personalities among the leading theoreticians of theater during the first half of the twentieth century, from Appia to Moholy-Nagy anc Artaud.)
In almost all cases, however, we are dealing with developments o an empirical or utilitarian nature, conceived and carried out with the goa of helping the director's memory, without any pretense of srientific accu racy or comprehensiveness. These notations are often mixed with othe: material which the director and his collaborators put together whil< preparing to stage a play (script, director's notes and plans, rehearsal jour nal, etc,)/ and which sometimes become a "model" of that production Such is the case with some of Mejerchol'd's productions, Craig's notebook! and model stages, Reinhardt's Regiebiicher, and especially Brecht's Modell biicher.75
The "system of notation"76 devised by N. Ivanov in the U.S.S.R. durinf the 1920s seems less empirical and more rigorous than others (even if it wa conceived primarily as a tool to assist the director's task). Ivanov's systen combines the linear transcription of the actors' movements in musical no ration with diagrams which trace these movements on a map of the perfor mance space (figure 2). With this double system Ivanov (1926: 90) claims tt "map out every move the actor makes on the stage, with an approximatioi at the same time of the presence or absence of objects on the stage." Hav ing thus transcribed the actors' movements as well as the direction of thes movements, Ivanov suggests that the duration and rhythm of movements cai be indicated by the musical symbols of the quarter note, eighth note, am thirty-second note (95), which Jacques Dalcroze, the inventor of rhythmics had previously identified, respectively, as "the rhythm of a person's strid while walking normally. . .rapidly. . . and at a running pace."77 Notwith standing the ingenuity of this notational method, its usefulness and accu racy will obviously decrease in direct proportion to an increase in the num ber of actors and objects on the stage. Moreover, this method transcribe only some of the fundamental parameters of movement (length, directiori duration), omitting other no less important ones, such as the "style" c gestures, position of the limbs, facial expression, etc. To summarize, we ar dealing with a partial system, capable of recording certain portions of a pei formance with only limited accuracy. Nevertheless, within these limits
The Performance Text
Argente: No, this is too much, I'm going before the judges.1
Satpin: For the love of God, sir!1
Argente No, no, I shall add nothing.1
Scapin: Just a tiny mule, sir.3
Argente: I shall not give as much as an ass. (Turns)
Scapin: Consider...4
Argente: No, I prefer going to the law.'
Scapin: Oh, sir, consider your words, and what they mean for you. * Think of all the ins and outs of the law. Just think how many appeals you will have to go through.7 How tiring all these preventive efforts are.1 Think of the ferocious beasts through whose daws you will have to pass:' sergeants, attorneys, counsels, registrars, substitutes, reporters, judges, and their clerks.M There's not one of these people who will not oppose the most straightforward case in the world for the merest trifle.11 A police commissioner will serve you with forged deeds, upon which you shall be condemned without knowing itB Your attorney will come to terms with fhe other side, and sell you for ready cash. ° Your counsel, won over in the same manner," will be missing when your cause has to be pleaded,11 or adduce arguments* that will thwart your case,17 or have no relevance to it.B
©©©©© ©   © © ©©$>
	At	_	
		-   / I	
			
^HV=- -1! ,_ T	~ -1-- ■»<?l«rJ		—•
16	4" -		
75
Figure 2. N. Ivanov: Transcription of the simultaneous movements of two characters in MoHere's The Swindles of Scapin, XL, 8.
IBB
76
The Semiotics of Performance
Ivanov's method offers a certain degree of usefulness, provided it is not thought of as a complete system of notation, which is not what it claims to be.
What I have already said regarding the absence of adequate multiple notational systems for performance can be modified at least in part if we take into account the transcription of individual partial performance texts: above all, the paralinguistic and kinesic texts. The notation of oral and gestural expression which originated in the Renaissance, if not earlier, belongs to a very old tradition. Bouissac (1973) has partially reconstructed the tradition of describmg/transcribing bodily movement, and, more recently, Kowzan (1979} has'dedicatedsome attention to both traditions. I would recommend these two studies for a rather detailed examination of the main attempts to provide a notational system for recitation and gestural expression, although this issue does not directly concern the problem under discussion (which is the transcription of performance).78 I should add, however, that one of the notational systems has enjoyed a degree of popularity and notoriety that far exceeds the scope in which and for which ii was conceived. I am referring to Labanotation, a method of dance notation developed by Rudolf von Laban, a choreographer and painter of Austriar origin, which was later perfected by A. Knust and P. Pollenz. Goodmar has praised the rigor and effectiveness of this system (see Laban 1926 1956).79
This brief account of the history of theatrical transcription and of its present state does not inspire much optimism. Kowzan, at the conclusior of his already cited "very incomplete panoramic view of the code system: of intonation and bodily movement," observes "a general failure in thi case of most of these attempts" (10). Faced with this picture, the very pos sibility of completely transcribing an entire performance seems questionable I do not intend to provide a detailed commentary on the relative technica questions nor to propose my own system of notation. Before ending nr discussion of this issue, however, I would like to make a few comments 01 method that may have some relevance in the light of what has been elat orated up to this point, (i) Inihe first place, it appears that we must aban don all naive and dangerous attempts at exhaustiveness, all expectations c total "physical" restoration like those that characterized the earliest filmj transcriptions80 and which, as I have-already observed, are still typical c much research in the field of theater studies (from the already cited CNR project in France to frequent contributions to the Drama Review in th United States). It is not possible, but above all it is neither necessary nc useful, to record everything, to aspire to a form of all-inclusive notation.1 According to Bettetini:
An inquiry conducted in this way, which is totally attentive to all the semiotic transformations within the performance, proves completely impossible and, above all, is of little value for the purpose of defining a concept which is in itself as ambiguous and as relative as the concept of "textual structure." (1977: 15-16)
021
The Performance Text
77
Instead we must proceed according to weU-defined and carefully delimited areas of pertinency, following precise analytical and interpretive hypotheses (on the textual structure of performance, on the hierarchy of codes, etc.). On this basis, we must attempt to "enucleate the important moments in the signifying chain, to discover the main stages of the process of semiosis" (Pavis 1978: 51). Bettetini (18) correctly invokes predi-rected transcriptions that would be capable of filtering the signifying material of performance.82 Moreover, as we will see in the following chapters, even the most strictly analytical phase of the study of performance (the phase relating to the actualization of codes, subcodes, and textual structures) will be inevitably incomplete and biased. As Metz claims for cinema (1973), we must work on partial, even doubly partial, models; which is to say, models that are not valid for all performances or for everything in performance. I would add: models that are not valid at all times, i.e., provisional models.
(ii) My second comment concerns the working procedure outlined here.
I
				1
viewing the performance (live aceess and/or audiovisual replay)		graphic description/		textual analysis
		transcription		
				
We are dealing with an absolutely nonlinear process. Through the integrated use of methodologies of an inductive and deductive nature, this process allows us to formulate initial hypotheses on the textual structure of a performance immediately after the performance has been viewed. On this basis, the work of description/transcription is set in motion, and the scholar is also provided with the possibility of returning repeatedly to the same text (whether through live viewing, audiovisual playback, or both) with the goal of verifying and checking hypotheses that emerge along the way in the various phases of the study.
2.5. THE DOUBLE HETEROGENEITY OF THE PERFORMANCE TEXT
The crisis triggered in semiotic research during the 1970s around the notion of "sign" and "message" and the emphasis, in turn, on the notions of "sign-function" and "text" is linked in a parallel way to a growing awareness that there are no truly linearox homogeneous languages. This means that languages are not regulated by a single formal system, but display a fairly high level of internal heterogeneity, necessarily resulting from the copres-ence of several codes, many of which are dissimilar to each other and un-specific. Obviously, this is true even in the case of verbal language which
78
The Semiotics of Performance
"always represents the confluence of heterogeneous semiotic references, for example, of a linguistic, metrical, or intonational model, and so forth" (Garroru 1972: 356). Verbal language can thus be considered linear and homogeneous only if "we study it hypothetically in its formal, homogeneous aspects." Homogeneity (like linearity) is not a property of so-called privileged languages, but is only a convenient theoretical abstraction (Garroni, l.c; Metz 1969: 174).83
The possession of multiple codes (with resulting heterogeneity) is thus a constitutive property of all texts—which are considered texts precisely because of this multiplicity of codes—and not only of the few texts produced in the field of aesthetic semiosis that are considered particularly "complex." Nevertheless, certain kinds of texts, including the performance text, are characterized by two different types of heterogeneity which should not be confused.
(1) a heterogeneity of codes (or multiplicity of codes in the proper sense) produced by the presence of several codes which, I must repeat, is common to all languages and all texts to varying degrees;
(2) an expressive heterogeneity (or multiplicity of concrete forms of expression) proper only to texts that are both composite in their manifestations and "heterogeneous from the standpoint of their codes" (Metz 1970:136) insofar as they simultaneously employ several concrete forms of expression, which may or may not belong to different sensory spheres.
For examples of doubly heterogeneous but monosensory texts, we can consider a radio performance, consisting of words, sounds, and noises, or a silent film, which uses different expressive materials of a visual type (two, according to Metz [1971]): images and titles. Staying within the scope of artistic practices, we find examples of texts that are, by contrast, doubly heterogeneous—and also mvolving more than one avenue of sensory transmission—in sound cinema and television broadcasting, in addition to, above all, theatrical performance. The performance text is thus multi-coded, which is to say heterogeneous, in its codes, multimedia (or "composite"),84 and also multidimensional, since it involves more than one channel of perception: in all cases, at least two (the acoustic and the visual),85 sometimes three (the acoustic, the visual, and the tactile, as in performances that involve the active participation of the audience), more rarely four or five, as in the "happenings" of the 1960s, or more recently, in performance art, where smell and taste are sometimes activated.86
2.6. PERFORMANCE TEXTS SHORTER OR LONGER THAN ONE PERFORMANCE
*  2.6.1. Partial Texts and Segments of the Performance Text
Since the performance text is not only multi-coded but also multidi-
The Performance Text
79
mensional and created from diverse expressive media, it manifests a number of textual levels that are almost always materially divergent: the available verbal text, intonations and accents, mime, gestures, costumes, music, stage sets, and so on. Each of the "layers" that can be distinguished by taking a cross-section of the performance text may be considered a full-fledged text in itself, since it reveals an ultimate multiplicity, even simultaneity, in its codes. We thus discover that the performance text is actually a text constituted by texts, i.e., it is a macro-text. As I already mentioned, to distinguish these smaller texts from the performance text proper, I propose to use the term /partial performance texts/ or simply partial texts to describe the single but superimposed levels into which a performance can be divided along the syntagmatic axis.87 A partial text is the part of the performance text distinguished by a single concrete form of expression, and hence composed of what Barthes calls "typical signs" (1964: 44). In my opinion, the separability of the partial texts is always assured, in principle, by the diversity of expressive media. The sharp distinction that I draw between the texts and the codes displayed there (see 3.1.) in fact allows us to assume that the kinds of transformations undergone by the codes when integrated into the textual structure of performance does not preclude the separation of the partial texts. I would say that this separation should still be facilitated, at least in theory, by the multidimensionality of the performance text.88
In addition to its capacity for being cross-sectioned along the horizontal axis, the performance text can be segmented vertically, according to its "density." The terms partial segments and performance segments describe the "parts" obtained by splitting the performance vertically, along the axis of its temporal development, i.e., at the level of a single partial text (single-text segmentation) or at the level of all the partial texts in operation during the interval of time under consideration (multi-text segmentation). The oral verbal text or the mimetic text of a "scene" in performance X are thus examples of partial segments. As a (non-ordered) whole of superimposed and variously interrelated partial segments, the same "scene" constitutes, on the other hand, a performance segment. I should note that the partial texts in a theatrical performance will participate in creating the performance segment in either a continuous or discontinuous way, depending on whether or not these partial texts are continuously present in the given interval of time. It is also important to point out that the vertical segmentation of a performance text can be implemented by (or at least begin from) the subdivisions planned by the sender and explicitly designated by appropriate markers (the unities of traditional drama, sequence, scene, and act, for example),89 and can also create new divisions and articulations, for the sake of analytic convenience, or according to the level singled out for analysis. The two options are not alternatives but can be implemented and integrated as appropriate. As I have stated about the external demarcation of the performance, and about its delimitation, it should be clear that even the
80
The Semiotics of Performance
internal demarcation of the performance text largely depends on the discretional powers of the receiver and it is not possible to formulate general rules about this.90
The process of single-text or multiple-text segmentation of the entire performance text requires the smgling out of its irunimal pertinent units. Beginning with the divisions inscribed within the performance, segmentation primarily serves to make available to the analyst concrete "portions" of the performance text, in which one can readily observe the relation of signifying systems, and upon which in turn the operations necessary for dis-tmguishing the code unities can be carried out.91
2.6.2. Groups and Classes of Performance Texts
In addition to the case of performance texts that are smaller than a single performance, we must consider the contrasting instance of performance texts larger than a single performance. These must not be confused with performances produced "in installments," or with the "framing-device" performances mentioned in 2.3.1. Following the example of Metz (1971) on cinema and Corti (1976) on literature, we can see that in some cases it may be appropriate to consider a group or class of performances that are different from each other but are in some way linked as a single performance text (a macro-performance text), and to treat them accordingly. The foUowing are examples: (i) the work an individual director, or part of that work (the productions of Antoine at the Theatre Libre, the early work of Mejerchol'd, or his biomechanical phase; Strehler's Brechtian productions); (ii) the work of an individual actor, or part of that work; (iii) a theatrical genre, such as melodrama, ballet, mime, musical comedy, animation; (iv) a stylistic "school" or an avant-garde movement: symbolist performance, dada theater, expressionist drama, "third theater," the post-avant-garde, body art; (v) the production of a given historical period or ciutural-geographical area: Renaissance theater in Italy, Golden Age theater in Spain, or the theater of the Weimar Republic.92
Obviously, this is still a very sketchy proposal, the adequacy and usefulness of which can be assessed only after making adjustments and concrete experiments. It is dear, however, that in an approach of this type, genre (in the broadest sense) is a key concept. This has been the focus of research in mass cornmunications for some time now (as well as in literature and, more recentiy, in music and pamting), and will be discussed in greater depth below in chapters 4 and 7.
2.7. CO-TEXTUAL AND •   CONTEXTUAL ASPECTS
The characteristics of the performance text examined in the preceding paragraphs are part of what might be called its internal co-textual aspects,93 to use the terminology of Textlinguistik. Let us now consider its contextual aspects, which, as already mentioned, will be divided into the "performance con-
The Performance Text
81
text" and the "cultural context" (and in making this distinction I am departing somewhat from Petrofi's interpretive model).94
(a) The performance context is constituted by the conditions of production and reception of the performance text, and by their constitutive elements.95 Within the scope of the performance context, we should also include the process of production in its various phases and aspects (adaptation of the text, training, rehearsals, scene planning, etc.). Taking this into account is of considerable importance to the scholar's ultimate achievement of the goal of a more complete understanding of the performance itself.96
(b) The cultural context is made up of the sum of synchronous cultural texts, which are both theatrical (performance texts or partial texts relating to mime, choreography, scenography, dramaturgy, etc.) and extra-theatrical (literary, pictorial, architectural, rhetorical, philosophical, or related to city plarming), and which can be brought to bear on the performance text in question or on one of its partial texts. These texts in. furn are part of the general (cultural) text, meaning the complex system of culture, or the entire set of all synchronous texts (Kristeva 1969: 97ff.). /
Given the ephemeral nature of the performance tepd and its lack of temporal durability, both the exploration of connections'between a given performance text and other synchronous texts both theatrical and nontheatri-cal (a process which goes under the rubric of intertextuality) and the task of contextualizing it within the general text are sometimes indispensable steps in the restoring the performance to "presence" and thus making it available for analysis (as already seen in 2.4.2.). These steps are always preliminary and essential for the proper understanding of a performance text (even in the case of the ordinary audience member).97
I will return in chapter 5 both to the contextual analysis of the performance text and to the phenomenon of theatrical intertextuality. Here, however, I wish to add just a few general remarks regarding the notion of intertextuality, while referring to two writers who, above all others, have placed this concept at the heart of their work: Julia Kristeva and Eliseo Veron. For Kristeva (1969, 1970), intertextuality is the implicit or explicit presence within a text of other texts. It is the largely deliberate positioning of a creative work at the center of a rich network of echoes and references to other works. Thanks to intertextuality and the "differences" that institute or mark it, the text ceases to present itself as a "closed" entity and reveals itself instead as an unending process of production:98
The text is thus a productive process, and this means: (a) that its relationship to the language in which it is situated is redistributive (destructive/constructive), and because of this, it is analyzable through logical rather than purely linguistic categories; (b) that it is a permutation of texts, an intertextuality: in the space of a text numerous statements taken from other texts are interwoven and neutralized. (1969: 97)
In a "typology of texts" founded on the recognition of the "specificity of diverse textual organizations," located within (and in relation to) the gen-
ii
82
The Semiotics of Performance
eral text of which they are a part, the intertextual function necessarily play a major part, which Kristeva (97-98) calls /ideologeme/. In Le texte du roma (1970) she locates this as the third "generator" (of quotations and plagie risms) within the process of generating a text, along with the "generator c narrative complexes" and the "generator of actants."99
According to Ver6n (1973: 93-94), we must consider at least three diffei ent dimensions of intertextuaKty: (i) the intertextual operations "within certain discursive universe": for example within cinema, or literature; (ii the intertextual operations "between universes of different discourses": re ciprocal influences between cinema and television or between daily am weekly newspapers (the majority of the discourses of the latter are incorri prehensible if one does not take into account their metalinguistic relation t< the texts in daily newspapers); (iii) intertextual functions—"in the proces of producing a certain discourse from other relatively autonomous dis courses, which, while fonctioning as moments or stages of production, di not appear on the surface of the 'produced' or 'finished' discourse." Exam pies are the relationship between the sketch and the painting, between thi screenplay and the film, or between an architectural drawing and its fina realization. Veron has in fact used the term "deep intertextuality" in thi regard, "since we are dealing with texts which, though used as part of thi process of the producing other texts, never arrive at the level of the soda consumption of discourses (or, if they do, this is rare, and occurs througl very limited channels).
Qearly, we are faced with two conceptions which mutually validah and corroborate each other. But the most mteresting fact is that botl Kristeva and Verdn conceive of intertextuality not only as the validation o a fact, of an objective datum, but also and above all as a methodologica tool, as a way of positioning oneself in relation to the text: in sum, as ai interpretive "construction." Naturally, in this strong sense, intertextuaKtj is presented as a central moment in the contextual analysis of the creative work within the general text.