Applied Semiotics/Sémiotique appliquée 1:3 (1997), 203-230
The State of Current Theatre Research
Patrice Pavis
Université de Paris VIII
1. Review of Research
1.1 Before Semiology: Dramaturgical Analysis
Performance analysis obviously does not simply date from the age of
structuralism and semiology. Any spectator who comments on a performance
analyses it ipso facto, since he selects, names, prioritises and examines
one particular element as opposed to another and establishes links
between these elements. When spectators comment on the performance,
they do not have to verbalise the unsayable; they try rather to find a few
landmarks. The description most often takes the form of the narration of
a story (plot or fabula) or at least an account of the most remarkable stage
events which facilitates an understanding of the materials used, a natural
segmentation of the performance and a highlighting of the most powerful
or chosen moments in the mise en scne.
The tradition of dramaturgical analysis goes back to Diderot (De la
poésie dramatique, 1758) and Lessing (La dramatique de Hambourg, 1767) in
whose work we find remarkable descriptions of acting and stage effects.
Brecht therefore only renews a tradition already established in Germany,
that of the Dramaturg (the director's literary and theatrical advisor, now
known as the dramaturge). He offers dramaturgical analyses on the theatre
and particular productions of his time which reveal much about the
general conception of mise en scne. In France, this same approach is
found in critical theorists such as Roland Barthes or Bernard Dort. Their
analyses are always based on ideological and aesthetic mechanisms found
Patrice PAVIS
204 AS / SA
in the production. Until the 1960s, this mode of description dominated
all others due to its breadth of vision, its precision and the compromise it
managed to find between meticulous observation and interpretation.
Dramaturgical analysis, aware that it is not exhaustive although not necessarily
aware that it also contributes to performance, offers an initial synthetic
approach to performance; it underlines the main structures of a
performance while avoiding a fragmented perception of it.
The following chapter shall show that there are numerous other tools
at our disposal with which we might examine a performance; and it is
useful to link all these different methods and increase our sources of information.
It is advisable to start by putting forward the possibility of
performance semiology, which is both a well established `science' and a
field of research in the process of being `restructured.'
1.2 Analysis and Semiology
The stock phrase `performance analysis' is perhaps not the most suitable
term. Analysing means deconstructing, fragmenting or breaking up
the performance's continuum into small pieces or tiny units, producing
more of a `butchered' effect or a mise en pices rather than an overall understanding
of the mise en scne through the mise en scne. However, the
spectator needs to see and thus describe the totality or at least the ensemble
of systems which are already structured and organised into what we
now call the mise en scne. In this respect, it does not make sense to
speak of an analysis of the mise en scne, since mise en scne is, by definition,
a synthetic system of options and organising principles and is not,
unlike the final performance, the empirical and concrete object of future
analysis. Mise en scne is an abstract and theoretical concept, a more or
less homogenous network of choices and constraints which is sometimes
called metatext (Pavis, 1985), performance text or testo spettacolare (de
Marinis, 1982). The metatext is an unwritten text which encompasses all
the various choices the director has consciously or unconsciously made
during the rehearsal process, options which are visible in the final product
(or which can sometimes be found in the production book which is
not, however, identical to the metatext). The performance text is the mise
en scne considered not as an empirical object, but as an abstract system,
an organised ensemble of signs. These types of texts ­ in the semiological
The State of Current Theatre Research
Vol. 1 (N 3) 205o
and etymological sense of fabric/texture and network ­ provide a key to
possible ways of reading performance, but they must not be confused
with the empirical object: the performance in all its materiality and concrete
situation of enunciation. Performance analysis ­ whether it be
dramaturgical analysis or a mere description of fragments or details ­ necessarily
leads to an understanding of the mise en scne which itself
groups and systematically organises the different materials of the empirical
object, the performance. In this respect, it seems easier ­ and perhaps
more obvious ­ to suggest some general hypotheses on the way mise en
scne works rather than provide exhaustive so-called objective descriptions
of the heterogenous aspects of the performance. But these hypotheses
obviously are only worth as much as the person who proposes them,
i.e the spectator/analyst in search of a global understanding of the mise
en scne. They do not have the apparent objectivity of empirical observation,
nor the absolute universality of abstract theory; they oscillate between
close yet fragmentary description and general, uncertain theory,
between formless signifiers and polysemic signifieds.
Moreover, one should determine for whom the analysis is meant, its
aim and in what `spirit' it is undertaken. It seems to have two main functions:
reporting and reconstruction.
1.3 Two Types of Analysis
Reporting-analysis could have live radio sports broadcasting as its
model; such an analysis would comment on the performance in progress,
as in a soccer game, indicating what is happening on stage between the
`players', highlighting their strategies, recording the result and `goals' of
both teams. This would mean approaching the performance `from
within,' in the heat of the action, reproducing the detail and impact of the
events, experiencing there and then everything that moves the spectator
throughout the performance, determining the punctum (Barthes, 1980) and
how the spectator is emotionally and cognitively implicated in the dynamics
of the acting, the waves of meaning and sensations which have been
generated by the multiple, simultaneous signs.
Ideally, reporting-analysis should be carried out during the performance;
the spectator reacts immediately and becomes conscious of his
Patrice PAVIS
206 AS / SA
reactions just after he has expressed them; he notates the `emotional punctuation,'
both the punctuation of the mise en scne and his own reception
of it. Although this practice is uncommon in a performance, apart from
recording on a medical level the spectator's physiological or summary
psychological reactions, such reporting nevertheless does penetrate the
live experience, and observes in situ the spectator's reactions to the stage
events. Most traces of these reactions are lost however, as in Western culture,
the theatre spectator (at least the adult spectator) is not exactly permitted
to express his impressions, reactions and thoughts during a performance;
he is expected to wait until the end of the performance to express
them. Therefore, an important part of these immediate impressions is lost
for ever, or at least buried beneath a memory and rationalisation a posteriori
of past emotions. One of the tasks of performance analysis is to note
when and how such emotions arise and how they influence meaning and
the senses. Dramatic criticism, insofar as it is immediate and spontaneous,
sometimes retains a valuable trace of these early impressions when it
describes the performance as a metaphor of first impressions.
Reconstruction-analysis, on the other hand, is a speciality of the West
which is inclined to conserve and store documents and to maintain historical
monuments. It is in some respect similar to historical reconstructions
of past productions. It is always done post festum; it collects clues,
relies or documents from the performance as well as artists' statements of
their intentions which were written down during the performance's preparation
and all mechanical recordings from all angles and in all possible
forms. Such a studium (Barthes, 1980: 50) is endless, but the difficulty ­ as
we shall see in chapter 2 of the book whose first chapter is constituted by
this article ­ is to use all these documents in such a way so as to restore
some of the audience's aesthetic experience. A performance, whether it
dates from yesterday or the time of the Greeks, is lost forever, and we can
no longer have an aesthetic experience of it nor have access to its living
materiality. Henceforth, we must settle for a mediated and abstract relationship
with the aesthetic object and aesthetic experience; it is a relationship
which no longer allows for an examination of objective, aesthetic
data, but which, at best, permits an understanding of the artist's intentions
and their impact on the audience. Whether we are dealing with a
production which has actually been seen by the analyst or a reconstruction
of a past performance, we only really ever reconstruct a few of its
main principles and not the authentic event. Once these principles are
established, the performance text becomes an object of knowledge, a theoretical
object substituted for that empirical object which was once the
The State of Current Theatre Research
Vol. 1 (N 3) 207o
performance itself. However useful and important the statement of such
principles, intentions or effects may be, they still cannot be considered as
performance analysis; they serve rather as a theoretical framework which
the analyst will use to relate in detail certain aspects of the performance.
Reconstruction-analysis is particularly concerned with the study of
the performance's context and the nature and extent of this context or
contexts. This context might be the place where a performance is held,
the audience on a given night including that audience's expectations and
socio- cultural composition, but it might also be the place and concrete
circumstances of the performance itself. Obviously it is not easy to restore
these contexts and the modes of behaviour which created them.
Richard Schechner's notion of "restoration of behavior" allows us to
imagine and restore `the actors' behaviour or that of all artists involved in
various performances. But these contexts and modes of behaviour are
extremely variable, potentially infinite and absolutely immeasurable:
"Even if human memory can be improved upon by the use of film or
exact notation, a performance always happens within several contexts,
and these are not easily controllable. The social circumstances change"
(Schechner, 1985: 43) These contexts, considered either as restored behaviour,
socio-cultural codes or social context (Pavis, 1980: 250-94), place
us in a semiological or, more precisely, a socio-semiotic perspective. This
is obviously not the only existing methodology for performance analysis,
but it offers a whole ensemble of techniques which we ought to consider.
Rather than give a regimented exposé of all these theories (already studied,
in any case, by Kowzan, Übersfeld, de Marinis, and Elam), an overall
review shall be given here if only to measure what we have acquired and
learned from semiology, even if it is all rather obvious now. The early
stages of semiology shall be evaluated, including its debits and debates, in
the hope of moving beyond some dogmatic prejudices, prejudices which
we shall mention here so as not to succumb to them a second time.
1.4. Semiology: Rise and Crisis
Since the sixties and seventies, the structural analysis of the narrative
has been applied to various literary and artistic fields ­ fairy tales, comic
strips, film, the plastic arts, etc. Theatre, as text and performance, has not
escaped such systematic analysis and the very first theories sought to test
Patrice PAVIS
208 AS / SA
the hypothesis of a specific theatre sign (Kowzan, 1992). The semiology
of literature and theatre takes the stage as a means of moving beyond the
impressionism and relativism of so-called traditional criticism which is
more interested in the text rather than performance. In reaction to the
somewhat vague discourse of dramatic criticism, it sometimes believed
that it had found a universal model in cybernetics and the theory of information:
but such a model often remains slave to a linear model of communication.
It leads to an extremely naive conception of theatre as a piece
of information which is coded by the director, to be then decoded by the
spectator, as if it were a mere question of transmitting a message with as
little loss as possible! Semiology is not a `code-sharing' between author-actor-director
and the spectator who is then expected to mechanically
decode those signs directed his way. This caricature promptly disappeared
in later analyses inspired by semiology, despite a persistent and
fairly malicious tendency to attribute it to semiology by constructing a
semiology of a particularly narrow and limited sort.
Theatre semiology established itself as the dominant academic discourse
of the seventies because theatre (since Artaud) felt the need to be
treated as a discipline in and for itself, as an autonomous language and
not as a branch of literature. Its principal concern has thus been to start
with the stage, with the large moments or stage units, and to examine the
text as it is enunciated on stage. As a result, the semiology of the text was
neglected or even disqualified, the text and the stage were radically separated
as were dramaturgical analysis and `theatrical language.' But now the
dramatic text is making a marked comeback: theatre is no longer simply
considered as a performance space, but once again, albeit in a different
way, as textual practice. Once again, we talk of theatricality, but theatricality
as found in the text (Bernard, 1976/1995) and in language (Finter,
1985). We must examine this return to the staged text (see chapter 5, part
2), to stage pragmatics where the text is spoken and unspoken.
The rise of semiology in the human sciences coincides with the sort
of criticism, most often Marxist, of so-called bourgeois ideology. Barthes
(1964) best exemplifies this alliance between linguistics and sociology:
Tired with the slippery, oratorical nature of ideological denunciations, I was
dazzled when reading Saussure (it was in 1956), that an elegant method (as is
said of the solution to a mathematical problem) could be developed for the
analysis of social symbols, class distinctions and the cunning mechanism of
ideology.
The State of Current Theatre Research
Vol. 1 (N 3) 209o
What has changed since the dazzle and optimism of the sixties is the
belief that semiology necessarily has a socio-critical perspective, the confidence
in a radical criticism of ideology. Since the fall of empires and `socialist'
ideologies, semiology has often been accused of having compromised
too much with ideologists and `masters of meaning.' Such criticism is all
the more violent now that we are no longer interested in denouncing false
consciousness, and generally condemn any established system, discipline,
or theatre which claims to represent reality. For many critics, nothing remains
of these ruins of ideology and militant critical thinking but a
sceptical relativism and a vague meditation on the `end of history' and the
futility of theory. Such a vision is rather limited and pointless when the
theoretical reflection strives systematically to describe a production in all
its components: such a task is obviously very demanding and requires
considerable patience.
Such a task is all the more demanding given that post-modern criticism
accuses semiology of camouflaging notions of intention and authority
behind the concept of signification, to which post-modernism opposes
the ideas of opening and non-representation. Thus, the very model
of the sign is called into question.
This misunderstanding is not a recent one, as semiology has always
been accused of mechanically applying the linguistic model to fields other
than literature, notably social and artistic practices. However, since
Saussure and Barthes, it ought to be clear that even if these other practices
are based on the signifier/signified opposition, they cannot be reduced
to a grid where non-linguistic signifiers are automatically translated
into linguistic signifiers. Nothing forces, nor allows for that matter, the
spectator to translate the experience of contemplating a light, a system of
gestures or music into words that can then be integrated into the global
meaning of the stage. The objection made by the Québecois theorist,
Rodrigue Villeneuve (1989: 25), seems to be inappropriate: "what seems
to be difficult to accept," he writes, "is this general attitude, more or less
clearly manifest, that confines everything that is not linguistic translation
of the stage object to an indefinable zone." Semiology, even Saussurian
semiology, is not obsessed with describing everything, which would, in
any case, be impossible, still less with translating everything into words.
There are, in fact, several questions mixed into this critique of the
sign. Certainly, the Saussurian model of the sign is binary (signifier /signified)
and not ternary like that of Peirce (1978) (representation, object,
interpretant). In this way, we can consider theatre performance as an en-
Patrice PAVIS
1. To use the title of Mark Johnson's Book, The Body in the Mind (1987), Univ. of Chicago Press.
210 AS / SA
semble of signifiers which only have meaning as a series of differences.
As regards the possible signifiers which arise from this series of differences,
we can link them with those signifieds in other semiological systems
borrowed from the referent (or real world); thus this world can be
included in our consciousness as a series of semiological systems already
preformed and pre-constructed by culture and language. Thanks to
Saussure, therefore, we can grasp meaning as the construction of a signification
and not as the naive communication of a preexisting significa-
tion.
However, once we actually perceive the materiality of the Performance
(the signifiers, in Saussurian terms), nothing prevents us from positioning
ourselves in the pre-linguistic, "just before Language," of grasping
"the body in the mind;" this is, after all, what we actually do when we1
watch a dance, a gesture or any signifier not yet contaminated or transcribed
by language. Another question is how to link these pre-linguistic
impressions to the other elements of the performance, notably the linguistic
and narrative elements. We will later examine (in chapter 1, Part 2)
if it is possible to do this without a binary semiology, if the spectator's
gaze or desire is not always directed, channelled, vectorised by signs.
There are many other misunderstandings about this initial semiology
which need to be clarified. hose with which we have just dealt here are
easily dissipated; however, this first phase of semiological research has
many other limitations which we must try to overcome by imagining all
possible solutions.
1.5. Limitations of Classical Semiology
Minimal units
Theatre performance cannot be segmented, like natural languages,
into a limited series of units or phonemes, where rules of combination
could demonstrate all the possibilities. It is impossible then to transpose
The State of Current Theatre Research
Vol. 1 (N 3) 211o
the linguistic model of phonemes and morphemes onto the plane of what
is now metaphorically known (since Artaud) as `theatre language.' It is no
use trying to isolate, in the continuum of the performance, minimal elements
distinguished as the smallest units of time and space. Such careful
examination is only interesting if it does not leave out important clues to
help us understand the performance; it does not explain how the signs
work and the minimal unit is not or is no longer the philosopher's stone
which will segment the performance as if by magic.
Categories of signs
Analysis should not concern itself with establishing a repertory or a
system of signs which would provide a framework for any performance
and which could be found in every production. Such a system does not
exist and enumerating signs or types of signs proves nothing, whether it
be a semiotic typology of signs (Pavis, 1975) or a classification of performance
types: thus, there is no point in drawing up a `dead list' of the categories
of signs used in any (Western, text-based) performance, categories
which the analyst would have to pinpoint, noting when and how often
the signs appear. The categories of Kowzan (1970), Elam (1980) or
Fischer-Lichte (1983) only deal with the elements which belong to the
`average' Western theatre performance, notably that of illusion and bourgeois
realism. Unfortunately, these subdivisions limit performance rather
than shed any light on the subject. ]bey force us to think in old-fashioned,
ready-made categories which any avant-garde or indeed any mise en scne
systematically calls into question. For instance, the way in which these
categories radically separate the human system of the (animate) subject
and system of the (inanimate) object is no longer relevant in current theatre
practice: the human body is sometimes treated as inert material (Butoh
dance) and an object often replaces and signifies a human presence (an
item of clothing, for example, or a prop associated with a person). This is
the reason why we will not use these old categories in this current work,
even if they are part of our Western cultural heritage and even if we have
difficulty in doing without them when discussing theatre.
These categories inherited from the classical, indeed antiquated European
theatre tradition, whose aesthetics and division of labour no longer
have much in common with current practice, shall be replaced with trans-
Patrice PAVIS
212 AS / SA
verse systems or categories, such as the system of chronotopes (see Chapter
3, Part 2), the system of vectorisation, or other transverse tools which
enable us to move beyond a compartmentalised vision of performance.
1.6. New Departures
A desemiotics?
More recent attempts have tried to set themselves apart from that
initial sort of semiotics which was too taxonomic and fragmentary by
seeking alternatives in the hermeneutic and pragmatic German tradition
or in phenomenology (Ingarden, Derrida, Carlson, States, Gardner).
Here, it has been most often a question of going beyond Saussurian
binarism and "closure of representation," (Derrida, 1967: 341-68) and of
suggesting a "generalized desemiotics" and a "theatre of energies"
(Lyotard, 1973) instead of a theatre of signs. Lyotard (1973: 104), the
most articulate representative of this tendency, has, himself written a radical
criticism of the sign from Brecht to Artaud; but his criticism of theatre
as "taking place" (lieu-tenance) and representation, and his proposal of an
energetic theatre unfortunately remain rather undeveloped and questionable:
"It [energetic theatre] does not have to suggest that this means that;
it does not have to say it either, as Brecht wished. It has to produce the
greatest intensity (by excess or default) of what is there, without intention.
Here is my question: is this possible, and if so, how?
Lyotard's desemiotics, it seems, is even less possible and feasible than
semiotics, but it is commendable in that it questions the notion of the
sign, at least the fixed sign, as linked to language and taking the place of
the whole materiality of the performance. This is what phenomenology
also sets out to do, criticizing the segmentation of everything into signs
and thus the semiological function of representation and performance. In
phenomenology, the perception of the performance event is global, so all
semiological segmentation is absurd. The problem with semiotics is that
in addressing theatre as a system of codes, it necessarily dissects the perceptual
impression theatre makes on the spectator. And as Merleau-Ponty
has said, "it is impossible (...) to decompose a perception, to make it into
The State of Current Theatre Research
2. Quoted by States (1987), from Merleau-Ponty in The Primacy of Perception and Other Essays..., ed.
James M. Edie, Northern University Press, p.15.
Vol. 1 (N 3) 213o
a collection of sensations, because in it the whole is prior to the parts."2
This is the source of anti-analysis reactions and the beginnings of a
globalising phase where the aim is to provide syntheses rather than reading
grids.
Global Understanding and Vectorisation
These various critiques of classical semiological analysis led, in the
1980s, to a globalizing phase: performance considered as a series of syntheses
or frameworks. The mise en scne, in the structuralist sense of the
term, became the key notion in a new theory able to synthesize stage options,
dramaturgical choices and the performance's main structures. Instead
of fragmenting perception, differentiating sensations, multiplying
signification, and thus, arbitrarily segmenting the signifier in order to
translate it into possible signifieds, the signifiers are imagined as those
which anticipate possible signifieds; the notion of individualized signs is
replaced with series of signs grouped according to a process of
vectorisation. Vectorisation is at once a methodological, mnemo-technical
and dramaturgical method which links networks of signs. It consists
of associating and connecting different signs in a network within which
each sign has meaning only in relation to other signs. Let us suppose that,
like in Chekhov's Seagull, a gun appears on a wall: the spectators relate it
to other indices and the moment it disappears and they hear gunshots,
they have no doubt that the depressed and suicidal hero has just put an
end to his life. Such networks are like nets which hold the production
together and prevent it from being permanently fragmented.
This globalisation, however, is not risk free, since analysis looks for a
kind of secret `key' to the mise en scne, when this term is taken in its
centred and concentrated form, thereby excluding theatre practices based
on "decentering," arbitrariness and chance. To avoid an all too coherent
closure of mise en scne and its analysis, it should be made clear that mise
en scne ­ both production and reception ­ never comes ready made, but
that there is ­ and this is our hypothesis ­ a vector-isation: certain signs
Patrice PAVIS
214 AS / SA
or moments in the performance are linked dynamically together and there
is a network of meaning which links these moments and makes their interaction
relevant- we can only describe the guiding lines in the
dramaturgy and the main stage options without excluding those pragmatic
decisions which digress from the main idea. Since it is impossible to
be exhaustive, it is of interest to replace the signs and vectors in a guiding
schema which constantly evolves: thus the spectator does not run the risk
of being submerged by insignificant details. It is better to reconstruct the
network, understanding its orientation and guiding lines than be left with
a disorganised mass of surplus material or useless recorded documents.
Thus the description of a performance always oscillates between a
totalizing demand for synthesis and empirical individualization, between
order and chaos, between abstraction and materiality.
The Experience of Materiality
The spectators concretely experience the materiality when they perceive
the various materials and forms in the performance, provided that
they remain on the side of the signifier, i.e provided that they resist the
temptation to immediately translate everything into signifieds. Whether it
is a question of the presence and corporeality of the actor, the texture of
his voice, or some kind of music, colour or rhythm, the spectators are at
first submerged in an aesthetic experience and the material event; they do
not have to reduce this experience to words, they savour rather the
"erotic in the theatre process" (Lehmann, 1989: 48) without trying to reduce
the performance to a series of signs, as, according to Bert States
(1987: 7) semiotics sometimes does: "What is disturbing, if anything,
about semiotics is not its narrowness, but its almost imperialistic confidence
in its product: that is, its implicit belief that you have exhausted a
thing's interest when you have explained how it works as a sign." Such an
objection is worth considering: performance should be treated both as
materiality and as potential meaning, and should never be reduced to an
abstract and fixed sign. When the spectators observe a gesture, a particular
space or listens to music, they appreciate the materiality of the performance
for as long as possible; they are at first touched, surprised and silenced
by these things in front of them which offer themselves to the
spectators in their being-there before becoming completely integrated
The State of Current Theatre Research
Vol. 1 (N 3) 215o
into the rest of the performance and evaporating into an immaterial signified.
But sooner or later, the spectators' desires are bound to become
vectorised, the arrow is bound to reach its target, transforming the object
of desire into a signified. Reading the signs in a performance thus means,
paradoxically, resisting their sublimation: the question is for how long?
Desublimation
Getting to grips ­ body to body ­ with the performance's materiality
is thus to be taken literally: the analyst returns to the `body' of the performance,
moving beyond the kind of sublimation which comes about
whenever we use signs; he is fully absorbed by the aesthetic experience
and by the material aspects presented on stage. He desperately tries to
overcome the "blindness of many semiologists when it comes to the material
force of aesthetic signifiers" (Villeneuve, 1989: 25); he is aware of
the "phenomenon of non-intentionality, of libidinal investment of events,
of the sensual materiality of signifiers, which make it impossible not to
consider the corporeality of things, structures and living beings through
which signifiers are produced in the theatre" (Lehmann, 1989: 48). To
experience aesthetically a circus, a performance or any production using
several kinds of materials, we must be open to the impressions such materiality
might create and resist giving it meaning. This comes naturally to
children as well as those who watch a performance from a cultural tradition
different from their own.
The current trend in performance analysis is thus a return to the material
and concrete reality of the stage, a desublimated return to the body
of the performance. This breaks with the abstract idea of the mise en
scne as sublimation of the body, as idealised abstract schema. We shall
see in the chapters dealing with the different components of performance
how this materiality can be revealed in detail and how we can trace the
vectorisation which organises this materiality in what can be called
`space-time-action.' Suffice it to point out here that it is possible to follow
the traces of the trajectory and energy in any movement or utterance; it is
possible to closely determine the breath, rhythm and the `voice' and `path'
of the performed text. This "logic of sensation" (De-leuze), this movement
which moves the text and nerves the spectator, this displacing of
affects and attention can only be grasped and felt if we refrain from
Patrice PAVIS
216 AS / SA
resublimating them in a univocal written trace, reduced to a signified or a
secret code.
However, this insistence on the material side of signs is always carried
out in the structured framework of the organised and channelled event
which is the performance, i.e according to a certain vectorisation.
Semiotisation and desemiotisation are thus both antithetical and complementary
operations of a work of art and an aesthetic experience of it. We
must bear this in mind when we evaluate methods of performance analysis.
Hence, the double criteria when describing performance: we must, on
the one hand, return to the `body' of the performance, but on the other
hand, distance oneself from it and draw the outline and itinerary of the
performance from the point of view of the desiring subject. Such is the
current state of analysis: the progress it has made is remarkable, but new
developments are still necessary. Semiology must take advantage of this
sceptical age into which it is drawn by its post-modern demon. Semiologists
are still faced with many methodological problems and questions
which have not yet been answered; they will be systematically outlined
here.
2. Unanswered Questions
However annoying these limits and unanswered questions may be,
they reflect the difficulty of adapting analysis for performances which are
constantly changing, defying all interpretation and demanding new strate-
gies.
2.1. Experience or Reconstruction?
Is the average spectator's experience of seeing a performance only
once sufficient for analysis? In principle, yes, and this unique experience
should be the golden rule when examining a performance, itself unique
and organised in terms of the ephemeral and the singular. However, the
The State of Current Theatre Research
Vol. 1 (N 3) 217o
temptation to cheat is great ­ to use, for example, the experience of a
spectator who has seen the performance more than once (which radically
changes the normal position of the receptor) and to artificially reconstruct
the performance using substituted, indeed left over material from
the theatre act: photos, audio-visual recordings, statements preceding the
performance (notes, preliminary plans and proposals, artists' statements
of their intentions or interviews following the premiere). It is essential to
distinguish carefully between the artists' intentions or statements on the
one hand, and on the other, the artistic result, the final product viewed by
the audience which should be our sole focus. We must make a clear distinction
between (1) statements of intention (documents outlining the
project, a commentary, an interview etc.), (2) the paratext (or the ensemble
of texts which are written around the actual dramatic text, notably
stage directions), (3) mechanical recordings of the performance (soundtrack,
images, videos or films), (4) the technical notation carried out after
the performance, (5) the semiological analysis of signs and the network of
vectors, an hermeneutic interpretation of the work, and the critical discourse
which follows the performance.
The kind of analysis desired here does not of course exclude the historical
reconstruction of past performances and all the disciplines which
that involves, but it is based above all on the unique and individual experience
of the spectator watching the stage event, an experience which theory
endeavours to generalise in a method of analysis.
2.2. Segmentation
Segmentation remains the main issue for performance analysis. If it is
agreed that nothing would be gained by producing an `atomization' of the
performance into minimal units, one does not yet know what the dimension
of the macro-units of the performance should be. Unfortunately,
performance is still often segmented according to the text (according to
its standard division into lines, scenes and acts); it is rarely based on observable
units in the performance. A text-based segmentation, however,
does not necessarily correspond to the dynamics of the performance. The
latter has its own rhetorical framework, its breaking points and pauses
which provide the only adequate reference points for any segmentation
of the performance.
Patrice PAVIS
218 AS / SA
In reaction to this text-based segmentation (i.e. a segmentation suggested
by the text), theory has naturally looked for units based on the
stage actions of the performance. But here again, researchers do not always
resist the philological (or text-centric) temptation to reduce the acting
to units which are marked in the text where it is possible (or, in their
view where it is essential) to note an actor's move. Such segmentation,
which absolutely insists on movements and dramaturgical units coinciding
at very precise moments anchored in the text, arbitrarily privileges
one signifying system (visible, marked moves) and imposes a text-based
segmentation on the rest of the performance. Rather than establishing a
purely textual segmentation, one should choose a mode of segmentation
based on the global rhythm of the performance, the rhythm of physical
actions and the musical composition of the mise en scne: i.e. according
to the temporal sequence of the rhythmic frameworks. In brief, one
should take into account those sequences when text and stage move `out
of sync,' and above all respect the possible vectorisations in the nose en
seine as a whole.
2.3. Textual Concretisation
In the same logocentric way, there is (or was) a tendency to see the
production as the stage concretisation of the preexisting dramatic text, to
deduct it from the reading of the dramatic text, which would find its
concretisation in the performance. As regards classical works which are
produced over and over again, of course one can understand this need to
refer back to the text in order to compare the series of possible stage
concretisations (Pavis, 1985). But that again is a philological, "bookish"
attitude ­ to use Lehmann's (1989: 43) merciless term ­ where the stage is
examined from a textual point of view: in actual fact, the two fields cannot
really be compared. Performance analysis takes the completed empirical
object as its starting point and does not attempt to go back to what
might have generated it. We must consider the stage as an autonomous
field which, contrary to Ingarden's view, does not have to concretise, materialise
or negate a preexisting dramatic text, and is thus "an artistic practice
which cannot be foreseen and predetermined from the perspective of
the text" (Lehmann, 1989: 44). We shall return to this relationship be-
The State of Current Theatre Research
Vol. 1 (N 3) 219o
tween the text and performance if only to overcome this false opposition
(part 2, chapter 5).
2.4. The Status of the Text
The status of the text in the production, which might be called the
staged text, should be questioned here. The words spoken by the actor
(or any other kind of stage utterance) must be analysed in the way they
are concretely stated on stage, coloured by the voice of the actor and the
interpretation of the scene, and not in the way we would analyse them if
we had read the written text. Text and performance are no longer thought
of as having a cause and effect relationship, but as two relatively independent
ensembles which do not always necessarily work together for the
sake of illustration. redundance or commentary.
2.5. The Narratological Model
Performance analysis can also have recourse to narratology, which
serves to identify the various components of the performance and make
the dynamics of the stage events explicit. But here again, as with the segmentation
of the performance, the narratological model should not be
based on the text alone, but also on the stage events; it should be neither
too universal nor too closely modelled on a parochial case. If narratology
is particularly well-developed for the, analysis of narrative and film, it is
not the case with theatre; perhaps because theatre, particularly Western
theatre is too often considered in an all too unilateral way in terms of mimesis
as opposed to diegesis. Instead of asking ourselves what is represented
mimetically, we shall examine what is narrated, how, by whom and
from which perspective (Barko, Burgess 1988). Theatre is not a world full
of mimetic signs, but a narrative which uses signs. Dynamic research on
Patrice PAVIS
3. See also Pepito Mateo, "Pepito Matéo, Conteur?," Dire, no 15.
4. See Tzvetan Todorov, Théorie des formalistes russes., Paris, Seuil, 1965.
220 AS / SA
the storyteller in the theatre (Haddad, 1982) clearly shows that the actor3
can also narrate and that narratology would be very useful for
dramaturgy.
A first step would involve arranging the different rhythms in the stage
system, locating the rhythmic frameworks and seeing the resultant
rhythm. For those performances which tell a story in a figurative way,
where the spectator follows the `sensory motor logic' of the action and
the plot, we can take inspiration from the Stanislavskian notions of physical
action, through-line of action or superobjective. A theory of vectors
which groups and energizes entire moments in the performance shall be
suggested later. These vectorised figures are found within the restricting
yet clarifying framework of the orientated action, the plot (fable or
fabula) and the way it is chronologically presented in the subject (two
Russian formalist notions which should be recovered from the prop
store!).4
2.6. The Question of Subjectivity
Semiology was set up as a means of avoiding an impressionistic discourse
on performance. As a simple notation of signs, it eliminated the
spectator's subjective gaze which is never neutral, considering the object
of analysis by means of a conceptual and methodological apparatus.
However, this fragile gaze, whether masculine or feminine, should never
be totally eliminated, but examined in its relationship with the stage and
the actor, above all, in order to grasp intuitively "what is indefinable in
the acting, the obscure emergence of emotion" (Behnamou, 1988: 10).
But how can we determine and note down the occurrence of such emotion?
At best we can imagine ­ bearing in mind the cinema's perspective
on theatrical reality ­ that the analysts gaze is comparable, albeit metaphorically,
to that of film apparatus: point of view, distance, scale of
shots, framings, montage, connections made through free associations,
The State of Current Theatre Research
Vol. 1 (N 3) 221o
etc. In this way, theatre analyses gain from the language of cinema, which
itself has been influenced by a particular logic of the human gaze.
But the subjective gaze of the (film or dramaturgical) lens is not so
much a fugitive impression which is mechanically grasped, but rather a
way for the spectators to experience the movements of the perceived object
aesthetically, i.e. follow the movements of the actor-dancer and the
overall dynamics of the performance corporeally. Like Barba (1992: 101),
we are thinking of "those few spectators capable of following or accompanying
the actor in the dance of `thought-in-action.'" In our view, these
spectators should not be so rare, they should even be the general rule:
theatre lovers who are capable of feeling and understanding the sensations
and movements of their own body, of perceiving the
`thought-in-action', the body of the performers and the performance as
an auto-biography in the strict sense of the term, i.e as a writing of the
actor's body as much as the spectators, a writing which inscribes itself in
the scene (to be) described.
2.7. The Non-Representable
The stage event, for that matter, is not always easy to describe because
the signs of the acting in current practice are often tiny, almost imperceptible
and always ambiguous, if not unreadable: intonations, gazes,
restrained rather than manifest gestures are so many fleeting moments
where meaning is suggested, but difficult to read and scarcely
externalised. How do we notice signs which are scarcely materialised, if
not by intuition and by a `body to body', sensory-driven experience of the
performance? The rather unscientific and unsemiological term, energy,
can be useful in an attempt to determine what this non-representable
phenomenon is: by his presence, movement, and phrasing, the actor or
dancer releases an energy which directly reaches the spectator. Such a
quality makes all the difference and contributes to the whole aesthetic
experience as well as the development of meaning. In reaction to a hegemonic,
visual culture of the obvious, we are trying here to identify the
non-representable which is essentially, but not exclusively, the invisible,
as rhythm, that which is heard, or kinesthetically perceived, i.e beyond
Patrice PAVIS
222 AS / SA
over obvious visual signs or units that are largely visible. Here we must
read the body, as does Foster (1988: 58), as that of a dancer, for example:
Literacy in the dance begins with seeing, hearing and feeling how the body
moves. The of dances must learn to see and feel rhythm in movement to comprehend
the three-dimensionality of the body, to sense its anatomical capabilities
and its relation to gravity, to identify the gestures and shapes made by the
body, and even to re-identify them when they are performed by different dancers.
The reader must also notice changes in the tensile qualities of movement the
dynamics and effort with which it is performed ­ and be able to trace the
path of dancers from one part of the performance area to another.
All of these tentative questions, as yet unanswered, clearly demonstrate
how theory and performance analysis, in trying to unbolt every
door, have moved away from a semiotics of communication and codes,
and how little the semantic model of the sign and levels of meaning is
adapted to contemporary performance. On the other hand, we already
foresee a more flexible model for the working of signs and their vectors
where the signifier is not sublimated into an immaterial signified, and
where the guiding lines of the vectors are clearly indicated (the following
chapters will come back to this): This vectorisation of desire ­ as much
the desire of the body of the performance as that of the spectator ­ becomes
a possible model for analysis, as soon as we are able to see
vectorisation as that which organises the performance, opening it out to
contradictory perspectives without confining it within a definitive struc-
ture.
In order for this model to develop further, the conditions for analysis
and the most appropriate theoretical fields for it to be put to use most
effectively need to be outlined.
3. Updating Theory
The fundamental question in any analysis is to know why and for
whom we are doing the analysis and which method is the most suitable.
Whereas the review in the printed or audiovisual media is addressed to a
large audience which wishes above all to be informed and advised, the
semiological review, which demands a longer time for reflection and a
The State of Current Theatre Research
Vol. 1 (N 3) 223o
more sophisticated conceptual apparatus, is practically always addressed
to other theatre theorists and intellectuals, to other colleagues entrenched
in the same somewhat frosty, studious or fetishistic relationship with the
stage. Theatre artists are rarely users of analysis, either through fear of
being `revealed,' through a vague fear of theory or some primal anti-intellectualism,
through disinterest or lack of time or curiosity. The question
is not how to interest them in our theories, but ­ all modesty aside how
will our theories influence their practice in the same way that their
practice has given rise to our theories. To this tragic misunderstanding is
added the fact that research is nearly always carried out in isolation by a
group of specialists working within the same critical tradition, often unaware
of other traditions. It is well known that between French semiology,
Dutch empiricism, Swedish audience reception studies, English
pragmatism, German hermeneutics or Italian historiography, there is
practically no exchange. Alas, poor Erasmus.
Despite all these obstacles and the rather discouraging list of unanswered
questions above, it seems that analysis, `twenty years on,' could
begin over again on more solid ground if we were to take advantage of
some well established, sophisticated disciplines like sociology and anthropology
which have almost always been developed far from the concrete
case of texts and performances. We will limit ourselves as a reminder to
listing five fields of research which are best discussed using concrete analyses
as found in various other chapters in my new book.
3.1. Productive-Receptive Theory
For an analysis which is interested in both the final product of the
mise en scne and its origins, we have to invent a theory which considers
production as well as reception; a theory which is neither partial nor unilateral
like the research on the creative process in literature and theatre or
the aesthetics of reception (where everything is based on different readings
of a work by different audiences). We must invent a model which
combines aesthetics of production and reception, a model which studies
their dialectical interaction, which looks at both the anticipated reception
of the production and the activity of the spectator in the act of reception
(Pavis, 1985: 281-297).
Patrice PAVIS
224 AS / SA
There is indeed a real danger, as Thies Lehmann points out, of simply
transferring the production problems to the realm of reception, naively
expecting the spectators to resolve them all with a wave of a magic wand,
as if they were endowed with all theoretical power. Productive-receptive
theory attempts to distribute evenly the process of shaping forms and
signs among productive and receptive instances; it assumes that one cannot
ignore the other, that in fact they work artfully in tandem, generating
strategies and routes of greater or lesser negotiability. This productive-receptive
concept results in an interactive strategy where we produce
as a creator and receive as a spectator. Such a strategy prevents us from
returning to the debate about the intentionality of the creator-producer
and the subjectivity of the spectator-receiver. We must look for their mutual
seduction (rather than reduction); this seduction is familiar to those
cultures involved in intercultural exchange; they surrender to it without
hesitation and not without pleasure.
In the same way, it hardly seems worthwhile to reintroduce the subjectivity/objectivity
polarity in order to associate subjectivity with the artist
and spectator, and objectivity with the work of art. Evidently, it is the
subject who analyses and evaluates, but to say that analysis is subjective is
not only banal, it also presupposes the existence of an objectivity on
which everyone might finally agree, and which would be the common,
lasting reference, the object finally trapped in the flight of desire.
3.2. Socio-Semiotics
Another field which needs to be developed is the kind of semiology
which is interested in ideological questions and observes how signs are
anchored and constituted in a whole socio-economico-cultural context
Empirical studies on audiences have (or should have) understood that an
examination of the cognitive, emotional and semiological mechanisms
used by the spectator to create meaning cannot be neglected (Schoenmakers,
ed., 1986 and Sauter, 1988). Is the semio-cognitive approach
compatible with the sociological and ideological approach? A method
such as soco-semiotics explicitly asks that crucial question. Sociosemiotics
differs from reception theory stemming from German Rezeptionsästhetik
and American reader-response criticism, both of which unfortunately
neglect the ideological plurality of the reader or spectator be-
The State of Current Theatre Research
5. Phenomenology applied to theatre can be found in Bert States, Great Reckonings in Little Rooms
(1987), Marvin Carlson, Theories of the Theatre (1987) and Stanton Gardner, Bodied Space (1994).
Vol. 1 (N 3) 225o
cause they presuppose an `ideal,' isolated, individual reader rather than an
ideological, cultural intersection of tensions and contradictions corresponding
to conflicting tendencies and groups (chapter 2, part 3).
3.3. Between Socio-semiotics and Cultural Anthropology
In the last few years, socio-semiotics has moved towards cultural anthropology
which encompasses the cultural and relational dimension of
performance. The development of an intercultural theatre (Brook,
Grotowski, Barba) over the past few years has accelerated the challenge
of purely linguistic and semiotic instruments for analysis. Intercultural
semiotics encourages us to relativize our choices, priorities and habits
when analysing a performance. It warns us, for instance, against our obsession
of describing a visible and readable space, of looking for and
quantitatively processing information and redundancies, of valuing all
that deviates from the norm and shows originality. Such semiotics may
heal us of our inability to understand the phenomena of hearing, voice,
time and rhythm, our inability to follow several parallel actions and to
assess the energy of an actor. Without demagogically renouncing our
Western cultural habits, we should acknowledge how our ethnocentric or
Eurocentric gaze influences and often distorts our perception, and how
much we will gain by changing our perspective and tools for analysis
(Pavis, 1990).
3.4. Phenomenology
The basis of phenomenological thought is that any experience of5
perception has a form or gestalt which contains organized, defined
Patrice PAVIS
226 AS / SA
wholes standing out against a background. The spectator's perception
tends to look for the most balanced, simple and regular form to distinguish
different ensembles with clearly defined contours: the ensembles
are in a relation of hierarchy, but they are, nonetheless, globally perceived
by the human eye and understanding, as Tindenmans (1983: 53) suggests:
Perception is fundamentally a constructive rather than a receptive or simply
analytical act. [...] A really satisfying theory of mental processes can only exist,
however, if we find an equally important place for theories of motivation, of
personality and of social interaction. Is it a risk to say that when people watch a
performance, they also look continuously for points of recognition, for causal
connections between events.
Phenomenology provides an image of the stage processes which is at
the same time a theory of action and a theory of the perceiver's appropriation
of the performance (he seizes everything). "Theatre does not `reach'
someone, someone has the theatre `reach' him" (Tindenmans, 1983: 55).
Whether we are thinking conceptually, looking at a painting or watching a
performance, eye and mind are active and not merely recording. `To think
means to try, to operate, to "reform, regulated only by an experimental
control in which only the most highly `wrought' phenomena intervene,
phenomena which our mechanisms produce rather than record"
(Merleau-Ponty, 1964: 10). Likewise, the spectators produce their perceptions
and the connections between them, instead of merely notating
them. `This phenomenological perspective is a valuable invitation to
move interactively through the highways and byways of performance and
meaning.
3.5. Theories of Vectors
To a relative degree, moreover, one must `follow the arrows' on such
walks, for our movement is channelled by the `arrow of desire' (an arrow
which seeks, but does not find), but also by an arrow which outlines a
way through the performance according to vectors which, as we shall see
in the chapter on the actor, organize and dynamise the whole perfor-
The State of Current Theatre Research
Vol. 1 (N 3) 227o
mance). This idea of an open yet coherent network will enable us to take
into account the necessary updating of theory while addressing different
methods of analysis and the various components of performance. This
open and coherent network of vectors will allow us to accept the necessary
renewal of theories, while retaining the general framework of
vectorisation. Semiology remains a discipline ­ in the sense of an ethical
and methodological guiding rule ­ which we naturally use to observe
theatre performances. It must always be enriched ­ without losing any of
its rigour if possible ­ by studying the mechanisms of (sociological) necessity
and (psychoanalytical) desire from the perspective of an anthropology
of the actor and of the spectator.
Bibliography
Barko, Ivan and Burgess, Bruce, La dynamique des points de vue dans le texte du
théâtre, Paris, Lettres modernes, 1988.
Barthes, Roland, Essais critiques, Paris, Seuil, 1964.
Benhamou, Anne-Françoise, Méandres d'un enseignement atypique,'
Théâtre/Public, no. 82-83, 1988.
Carlson, Marvin, Theories of the Theatre, Ithaca, Cornell UP, 1984.
De Marinis, Marco, Semiotica del teatro. L'analisi testuale dello spettacolo, Milano,
Bompiani, 1982.
Derrida, Jacques, `Le théâtre de la cruauté et la clôture de la représentation,' L'Ecriture et
la différence, Paris, Seuil, 1967.
Dort, Bernard, Lecture de Brecht, Paris, Seuil, 1960.
Elam, Keir, The Semiotics of Theatre and Drama, London, Methuen, 1980.
Patrice PAVIS
228 AS / SA
Finter, Helga, "Théâtre expdrimental et sémiologie du théâtre: la théâtralisation de la
voix,' in Féral, J., et al., eds., Théâtralité, Ecriture et mise en scne, Hurtubise, 1985.
Fischer-Lichte, Erika, `Die Zeichensprache des Theaters,' in Möhrmann, R., ed.,
Theaterwissenschaft heute, Eine Einfuhrung, Berlin, D. Reimer Verlag, 1990.
Foster, Susan, Reading Dancing, Berkeley, University of California Press, 1983.
Gardner, Stanton, Bodied Spaces, Ithaca, Cornell UP, 1994.
Ingarden, Roman, Das literarische Kunstwerk, Tübingen, Niemeyer, 1931.
Kowzan, Tadeusz, `Le signe au théâtre,' Diogne, vol. LXI, 1968.
-- , `L'art du spectacle dans un systme général des arts,' Etudes philosophiques, January
1970.
-- , Sémiologie du théâtre, Paris, Nathan, 1992.
Lehmann, Hans-Thies, `Die Inszenierung: Probleme ihrer Analyse,' Zeitschrift für
Semiotik, 11:1, 1989.
Lyotard, Jean-François, `La dent, la paume,' Des dispositifs pulsionnels, Paris,
UGE, 1973.
Paul, Arno, `Theater als Kommunikations-prozeß,' Theaterwissenschaft im
deutschsprachigen Raum, Darmstadt, Wissenschaftliche Buchgemeinschaft, 1981.
Pavis, Patrice, Problmes de sémiologie théâtrale, Presses de l'Université du Québec (and Doctoral
Thesis, Lyon), 1975.
Pavis, Patrice, Marivaux  l'épreuve de la scne, Paris, Publications de la Sorbonne, 1986.
Pavis, Patrice, Voix et images de la scne. Pour une sémiologie de la réception. Lille, Presses
Universitaires, 1985.
Pavis, Patrice, Ed. Le Théâtre au croisement des cultures, Paris, Corti, 1990.
Pradier, Jean-Marie, `Ethnoscénologie: la profondeur des émergences,'
Internationale de l'imaginaire, no. 5, 1996.
Sauter, Wilmar, New Directions in Audience Research, Utrecht, Instituut voor
theaterwetenschap, 1988.
Schoenmakers, Henri, ed., Performance Theory, Utrecht, Instituut voor
Theaterwetenschap, 1986.
States, Bert, Great Reckonings in Little Rooms, Berkeley, Univ. of California Press, 1987.
Tindemans, Carlos, `L'Analyse de la représentation théâtrale. Quelques réflexions
méthodologiques,' Théâtre de toujours, d'Aristote  Kalisky. Hommages  Paul
Delsemme, Bruxelles, ed. de l'Université, 1983.
Übersfeld, Anne, Lire le théâtre, Paris, Editions Sociales, 1977.
The State of Current Theatre Research
Vol. 1 (N 3) 229o
Übersfeld, Anne, L'Ecole du spectateur, Paris, Editions Sociales, 1981.
Villeneuve, Rodrigue, `Les îles incertaines. L'objet de la sémiotique théâtrale,' Protée, 17:
1, 1989.