Le récit interactif, tables rondes, 6 décembre 2000 — ENSAD-ARi, labEi, CIREN
  Le récit interactif à l'École nationale supérieure des arts décoratifs, Paris

Le récit interactif | Langage et écritures: Jacques MORIZOT, Jean-Pierre BALPE, Anne CAUQUELIN, Georges LEGRADY, François Soulages, Liliane TERRIER || Images et dispositifs: Jean-Louis BOISSIER, Grahame WEINBREN, Raymond BELLOUR, Anne-Marie DUGUET, Timothy MURRAY

Le récit interactif :
Images et dispositifs

English Text
Jean-Louis BOISSIER — La perspective interactive (Visibilité-Lisibilité- Jouabilité)

I first discuss how a perspective of visibility has been—always and singularly with the rise of numerical technologies—associated with a perspective of readability. Readability of space, symbolically rebuilt, but also readability of text, of the textual landscape set in a network. Then I will explain why, what I propose to call the "interactive perspective", depends on playability.

(1) The site does not exist in 2010

(2) Roland Recht,
La Lettre de Humboldt, Bourgois, 1989, p. 148

(3) Paul Virilio, L’Espace critique, Bourgois, 1984,
p. 28; L’Inertie polaire, Bourgois, 1990, p. 21
1 — A perspective of Visibility

In a three-dimensional virtual space, when a camera, virtual itself, is being moved, an appropriate perspective builds an image that corresponds to a point of view, to an angle and a framework. This classical perspective has no originality other than being automatic and interactive. Yet it brings to the fore a fundamental duality inherent in the perspectivist framework, that is, either working in a synthesis mode or a capture mode. Those two words are chosen intentionally from the vocabulary of computing even though this duality belongs to history of perspective. Moreover, I make the assumption that the association between synthesis and capture is a constituent of the system of perspective. I shall retain this duality when I later suggest other kinds of perspective.

If the representation of a space by perspective first seems to be the application of a method to an external reality, as a way of rendering it rationally, it was originally conceived in a constructive and fictional project. Thus La tavoletta of Brunelleschi, which combined construction and confrontation to reality, had in fact the functions of verification and demonstration. The perspectivist calculation that applies to some three-dimensional virtual entities plays an essential part in the building and uses of this virtual world because it allows its analogical representation. Only this perspectivist calculation reveals it to the sight, but it intervenes afterwards and, in short, optionally. As a reference in the artistic field, the works of Jeffrey Shaw could be cited: The Legible City (1989), The Virtual Museum (1991) or The Golden Calf (1994).

The classical perspective is predominate in three-dimensional computer-generated images since it is constitutive of their visibility. It should be underlined that the perspective that the computer calculates intervenes in a constructivist mode, consistent with the environment in which it is used, even though this environment is itself the result of an analogical or numerical input, for example, like a satellite landscape or an anatomical view. But precisely, apart from the assumed light and projection that will literally make it an image, this type of virtual object always has a modeling phase or at least of synthetic construction.

I mention perspective in computer generated images to establish a link with, and, at the same time, make a distinction from optical perspective that operates in photography, cinema and video. With the camera, the perspective is actually due to the capture itself. Nevertheless the two modes, photographic capture and algorithmic construction, meet in several numerical processes, like in QuickTime panoramas or the recent 360-degree video system of Be Here (1) . In this case, calculation operates to produce an anamorphosis and to display a view in which the perspective seems orthodox and consistent with the mobile frame exploring the global image. Miroslav Rogala, Zoe Beloff or Jean-Marie Dallet have all worked at this junction of photography and recalculated view.

This direction is certainly very stimulating, resonating with several attempts to contest the central perspective, the computer calculation being able to offer some totally heterogeneous views. We might cite reversed perspectives or refer to the subjective sphere of a child done by Tamás Waliczky in The Garden (1992) and in The Way (1994); or the optically impossible mathematical views displayed in Michel Bret's films. To illustrate this question of the image requiring some advanced calculations we might also cite some four-dimensional objects made visible, optically understandable in the two-dimensional projection of a screen. All that is needed, in a manner of speaking, is to film themselves being projected on a three-dimensional surface.

Again it should be underscored that the numerical virtual space, like the transformations and behaviour occurring in it, exists outside any particular system of perspective. Yet, the potential for its appearance requires some perspective, itself potential, which will be the first version of what I call "interactive perspective". Perspective because a perceptible image has to be produced. Interactive because it depends on the nature of the observer.

In Susan Amkraut and Michael Girard's installation called
Menagerie, using a system of stereoscopic vision in virtual reality called BOOM, some schematic figures of animals run towards the observer, avoiding or escaping from him. This is not only a three-dimensional stage to be explored but also a site for simulating behaviour. In that respect, a metaphor of circus and animals is relevant. In order for the animals to exercise the tropism that leads them towards you, to locate yourself—a moving and reacting observer— to let the spectacle take place, an area has to be delimited: a simple circle representing the ring. From this delimited area duration emerges, temporality, allowing it to be called an "interactive spectacle". How is such an event represented? One might say that perspective integrates not only the will to see but also the will to interact.

When Menagerie was shown within the context of the Revue virtuelle at the Pompidou Centre in January 1994, I thought that this installation would be both a technological and esthetic prototype of a generation of artistic works to come. But today, video games, and industrial and commercial products, provide a significant cultural, and why not artistic, posterity. In them are found the same characteristics: site, creatures with programmed behaviour— including, for example, their simulated emotions—but also able to learn by themselves new behaviour, some avatars ready to welcome the players.

The Japanese artist Seiko Mikami's installation Molecular Informatics (1996) uses a device for detecting the eye movements of two visitors. They are facing each other without directly seeing one another because they are wearing video-stereoscopic viewing helmets. Yet they have a sort of visual conversation. Each of them can actually see in front of him an unfolding broken line punctuated with little spheres, which evokes the symbolic representation of a molecule. This representation follows the successive regards and changes of direction of one’s own eyes. One can then try to meet in the virtual space, the molecular structure generated by the other's eye.

Masaki Fujihata works with settings having a similar perspective, in the literal etymological meaning of "looking through", displaying in real time some real lines of the virtual eye's vanishing lines. The virtual spaces of his installations Global Interior Project (1996) or Nuzzle Afar (1999) are accessible through a local network or on the Internet.

Thus a dimension referring to relational behaviour is added to a perspective referring to the optical. In this second mode of interactive perspective, interaction plays the same role as geometry in optical perspective. To put it another way: if perspective makes it possible to capture or build a visual representation, then interactive perspective is able to capture and induce interactions, to describe interactions.

According to Roland Recht's analysis devoted to the beginnings of photography, La lettre de Humbolt (Humbolt's Letter) (2) , photography has its foundations in the individual, diversified, free way of seeing, liberated from the illusionist and centered vision of the end of 18th century, and from Romanticism, focused upon landscapes. Adopting the system of composition, the image is accepted as part of a wider, complex and moving universe. But now the veduta, the window opens flat, on the surface of computer screens. The "depth of time" and the "trans-appearance" announced by Paul Virilio (3) are now operating in the Web's depths.

Every perspective does not require a horizon to be drawn. For example, the Chinese traditional pictorial space, which is incorrectly said to be without perspective, or a parallel projection to given direction, described by Poncelet in 1822, considering it as a central projection whose center is thrown back to infinity. With such vanishing lines, we would be able to stand here and elsewhere at the same time.

(4) André Corboz,
"La Suisse comme hyperville", in
Le Visiteur, N°6, automne 2000,
Société Française des Architectes, Paris,
pp. 112-129

(5) cf. Hervé Morin, "Des logiciels savent lire et classer les images ", Le Monde,
19 juin 1999,
et "Le X traqué sur Internet par la reconnaissance
d’images ",
Le Monde,
5 novembre 2000

(6) Jean-Jacques Rousseau, Dialogues. Rousseau juge de Jean-Jaques, Histoire du précédent écrit, Copie du billet circulaire dont il est parlé dans l’écrit précédent, O.C., t. 1, La Pléiade, pp. 990-992.

2. A Perspective of Readability

At this point, we shift to a perspective of readability. There is perspective in flatness. The screen knows how to show depth. Optically, a landscape should catch light and should organize according to the visible —and readable— perspective. Yet, a labyrinth is also entitled to have a perspective free of any comprehensive view. The interactive perspective foresees not only a landscape made out of code and language but also made out of signs and images. From now on, this perspective does not only concern the image, but a complex set of relations and virtualities. Incidentally, a theory of urban space proposes the metaphor of hypertext to designate its actual rhizomic spreading as a hyperville (4) .

In the hypertextual space, in a landscape made out of texts, the vanishing line of a language stimulation will certainly encounter some points of intersection. The "text" is not necessarily made out of words. It could concern recognition of shapes like that which automatically selects any painting from an image bank wh juin ich, for example, looks like a Gauguin, or which locates signs of a pornographic site (5) .

We again confirm that the system of designations, which most interactive programs on screens are based on, builds depth. If it avoids temptation of an illusory total immersion, it might be able to escape both stereotypes of stereoscopy and the doxa of a "depth of thought". The emancipation of the way viewing by “perspectors”, claimed by the philosophy of the Enlightenment, would correspond today to a desire for surfing on the screens' surface, to use vehicles whose search engines invent landscapes, made-to-measure.

In support of these comments one could cite the work of Knobowtic Research. This group of artists based in Cologne design some critical and utopian quasi-landscapes, made out of data recombinations and of connections to the informational and natural world. Another example could be provided by Philip Pocock and Christoph Keller, Florian Würst, Florian Wenz, Udo Noll. Their work available on the Internet,
Arctic Circle, Tropic of Cancer, Equator, consists of a set of real itineraries followed by some "explorers" traveling together or alone, gathering information from day to day for the web site and supplemented by data exchange over the web.

In my experiment called Le billet circulaire
(1997), inspired by a note from Jean-Jacques Rousseau (6), I imagined an "aerial perspective", the blurred cloud of the countless notes sent by the "intelligent agent" R.— a new version of the "perspector"—to all the correspondents who were "unknown whose appearance would please me most", found in the Internet space by the look of their name (having a central double consonant in it)..

(7) Gilles Deleuze et Claire Parnet, Dialogues, p. 69]

(8) Gilles Deleuze, L’Image Temps, Minuit, p. 92

(9) Gilles Deleuze, L’Image Mouvement, Minuit, p. 22

(10) Jean-Louis Boissier, Moments de Jean-Jacques Rousseau, Gallimard, 2000]

(11) Gilles Deleuze, L’Image Mouvement, Minuit, 1983, p. 12

3 — The Relation Image

To support this idea, the research to which I refer is being carried out in the Laboratoire de Esthetique interactivité of the University of Paris 8, and in my own experiments. Those experiments have the peculiarity of considering the interactive dimension of recorded images, in continuity with cinema and video.

I am trying to describe and test an image-relation whose function should be presentation of an interaction, comparable in a way to what Deleuze says about cinema: cinema is not just an image to which we have added movement, but immediately a movement-image (image-mouvement), a time-image (image-temps), in a relation of inverse subordination, a direct presentation of time.

A relation is narrative. The first historical use of this word is for a story, a report. A report like when something is being reported, when one makes a report of an inquiry, before becoming simply a relation, a logical relationship. A two-fold aspect of this notion of relationship, capture and synthesis, has always interested me. The history of this word provides a surprise: a relation (in French) is first what "tales" relate and then what "connects".

The relation characterizes objects, multiple thoughts regarded as a unique material or intellectual act. Deleuze recalls this in all his books, from Empirisme et subjectivité to Dialogues, or in L'Image-Mouvement, which is closer to what is being discussed here: "The relation is not a property of objects, it is always outside of its terms." More precisely "'Pierre is smaller than Paul', 'the glass is on the table': the relation is not interior to either of those words, which would then become the subject, or to them both. Moreover, a relation could change without having its words changed. (7)"

We say right away that this affirmation suits us fine, because if the relation has a shape, a shape with the power to transform the set it arranges, one understands that this shape can be modeled, it can be enter into an artistic process, and so does the relation-image, whose theoretical use I am trying to verify here.

Image-relation belongs to the couples built upon the word image) by Deleuze to try to understand cinema: movement-image (image-mouvement), time-image (image-temps), and of course perception-image (image-perception), action-image (image-action), affection-image (image-affection), drive-image (image-pulsion), but also memory-image (image-souvenir), dream-image (image-rêve) and world-image (image-monde) (8). Relation-images, like change-images (images-changement) or duration-images (images-durée), for Deleuze, are themselves constitutive of time-image (image-temps), beyond movement itself (9). I certainly am not trying to add a chapter to Deleuzian studies; but simply though a linguistic declension on the Deleuzian pattern, to link my proposition concerning interactive images, that do not have their own name, not yet their own name, to the cinema itself—cinema, which has its own distinguished name.

Thus we use the hypothesis which considers the cinema as a particular case of a more general system, of which the interactive-image will now be the prototype, for which computing provides variability.

In cinema on a computer, the logical relation to the real— what gives cinema its power, in its production and, symmetrically, in its reception—is not canceled or reduced, but is transformed, enhanced. Furthermore, this variability opens up several levels of interactivity. Because the degree of openness of an interactive object, has no link with any kind of indetermination but on the contrary, an improvement of its management of internal relations which provides greater autonomy, that of being able to respond to the impact of external interactions and changes of its environment.

Software is what provides logic. In the cinema, the main logical line is the one that regulates the recording of appearances in the flow of time. The camera and the projector, symmetrical machines, moreover, were originally amalgamated, following the same rule: linking static images in a determinate order and at a determinate speed. Interactivity could then be a way of augmenting, varying this logic, to the point of making cinema and video “software” ("logiciels") whose particularity would be in the formula: "24 (or 25) images per second, set in their chronological order". Interactive cinema frees potential variability of cinematographic parameters, even making this variability become its main lever. The interactive-video material on which we conduct experiments is mounted as close as possible of the photograms, cinema's discreet component elements, or of the chrono-photography, which might constitute a new genealogical branch.

A first factor of variability—succession and cadence— is therefore determined. With computers, the order of images, the cadence of their display can be varied and modulated, not only with what can still be called capture of reality, but also with the re-capture given by an interactive reading. In the animated image, interactivity deals with a collection of images, just like a bank of images, following rules of order and exploration. In that way interactivity is seen as a constituent part of the image, making itself the image.

In our CD-ROM, Moments de Jean-Jacques Rousseau (10) , sequences strictly speaking have no real beginning or end. Yet they do have some entries — we systematically kept two of them— at both ends of the panoramic or zoom movements on which they are built — but then continuing in endless loops, swinging or circular, which reverse, branch off, leading to other loops, depending on upon linked action of the direction and reading,

A second factor of variability is related to the camera-projector relationship. In the cinema's early beginnings, at a period, which I consider naïve, the spectator's position and relation to the screen were decisive for the manner of conception, beforehand, during the shooting. The projector's fixity and the continuous scrolling of the film had to correspond to a fixity and continuity of the camera. In that respect, it should be recalled that the first forms of tracking were shot by cameras on a train and this seemed to require a cinema hall shaped like a train wagon. The same kind of naïvety is often seen today in some forms of simulated spectacle or virtual reality.

However, when the camera frees itself from the projector, then a process of differentiation begins, establishing foundations of cinema as writing and art. Deleuze underlines it at the beginning of L'Image-Mouvement:

"On the one hand, shooting was stationary, the shot was therefore spatially and formally stationary; on the other hand, the cinema camera was amalgamated with the projector, endowed with an abstract regular speed. The cinema's evolution, the conquest of its own essence or novelty, happened with the editing process, the mobile camera and the liberation of shooting from projection. Then the shot won't belongs to a spatial category any more but to a temporal one; and cuttings will then become mobile and no longer immobile (11)".

At the same time as the main story, there could be a story in a minor mode of the mobile projection. Three examples can be fond in contemporary art: Gary Hills’s installation Beacon (1990), the show-installation conceived by the group Dumb Type (1994), or Jeffrey Shaw's virtual reality pieces Eve and Place, a User's Manual (1995).

If the author is next to the camera to set its movements, the spectator, beside the projector, could feel untitled to move the projector. I believe that work on the image should not be confused with a supposed or tangible immersion in the image. The most legitimate location that we could be offer the spectator, without allowing him get inside the representation, would be our location, which appears to be in the interactive representation of the settings itself.

We have found out that once a sequence's photograms are separated and arranged, it is possible to build up an animation which projects the movement of the camera on the computer's screen surface. The frame of the image and of the computer are separated. What was filmed won't change location in comparison with the screen but the image's frame recorded during a camera’s movement will be reproduced as a horizontal translation on the screen. The effect would then be similar to a cinema or video projector's beam sweeping across the screen according to movements captured by the camera during the shooting.

For example, all the sequences in the CD-ROM Moments de Jean-Jacques Rousseau were made using a digital video camera installed on an extremely precise electrical rotating head, made especially by us to create such interactive panoramas. The beginning and ending positions having been carefully marked out, the sequence of the images lends themselves to being displayed on the screen, with looping or internal branches. Above all, because of this equivalent of a moving view, and only by controlling the image's movement in this way, will the reader be able to explore the contents of each sequence.

Every deviation from the classical cinematographic model will find their descriptive, narrative and visual pertinence, only by the access given to the reader. The variability of interactive video not only permits numerous possibilities for branching, suspense and release, but also the possibility of a general dissociation between the recorded time and the restored (reception or reading) time.

In the interactive panoramic example, the camera's movement is time spent which coincides with the filmed subject’s own time. If necessary, it is possible to reserve, suspend or reverse it. But specifically the two strata of time cannot be dissociated. In suspended loop, on the contrary, the filmed time tends to get detached from the observer's time.

In the interactive image, or more precisely for me, the reading time can be separated from the shooting time during the interactive assembling of photograms. In the Rousseau experiment, for example, both times coincide, with a certain plasticity, only when the observer designates one or the other side of the screen and then commands the transformation and movement of the image. This brief and unstable situation, where the observer's temporality catches up with the filmed subject own temporality gives, by way of contrast, the feeling of a real manual capture of the circulation of time. Such a projection of time in time could be the basis for some ideas of depth of time and relational perspective, of which interactivity would be the principle.

(12) cf. Artifices 2, catalogue, Saint-Denis, 1992

(13) Roland Barthes, "Écrire la lecture" 1970, O.C., t. II, Seuil, pp. 961-963.

(14) Libération,
15 décembre 1987

4 — Perspective of Playability

To the optical perspective, which naturally is the first mechanism, I propose to demonstrate how an interactive perspective is added. In brief: how is (optical) perspective put in (interactive) perspective?

If this relation is considered as a shape, it is easy to understand that the image-relation should be produced by a new kind of perspective. An optically referenced perspective is joined to a dimension referenced to relational behaviour. In this interactive perspective, interactivity plays the same part as geometry in optical perspective. Again, if perspective makes it possible to build a visual representation, then the interactive perspective makes it possible to capture or model relations.

This interactive perspective projects relations in a relational space, setting them slantwise, so they can be identified, and one might say, playable. Finally, the game is the essential word to define the performance which realizes the image-relation. Our reader is at the same time a spectator, but also someone else. What is expected is that he plays the game, but a game freed from plain amusement and entertainment, which would have the value of an exercise or an interpretation, but not using it as a tool, staying on the side of art.

Matt Mullican's virtual reality performance installation Five Into One (1991), takes the Game Boy as a reference (like Campbell’s Soup) (12). Having a purely topographic appearance, his "map as a city" presents an uncommon topology, since it changes each time the spectator steps across an invisible border. As with any map, we read as much as we see in it. But above all, it is the most appreciated when playing it live.

This is in fact what Roland Barthes says in favor of the "text that we write (in us, for us) when we read":

"Opening the text, unfolding the system of its reading, is not only asking and showing that it can be freely interpreted; above all and more radically, it leads to recognizing the fact that there is no objective or subjective truth in reading, but only a ludic truth; yet the game should not be understood as an entertainment, but as a work [...] (13)"

Because, after all is said and done, it's in the performance of relaying that the image-relation acts to build the landscape and story of interactive objects in the specificity of interactive objects. If the landscape-story comes from both visibility and readability, there will be an interactive perspective only with a certain amount of playability. In scholarly language one might call it “performativity” (performativité). But let's finally adopt playability, a word chosen by computer game enthusiasts. Yet, it should be admitted that this word has found its cultural and, perhaps, theoretical, pertinence. It is also true that it is better to designate something by the possibilities that it opens, rather than naming it too quickly. "The football field, this place where players are playing", Marguerite Duras said to Michel Platini (14), thus designating a site by the possibilities in it. If one wants to catch the pertinence of the interactive perspective, it should be thought as something that simultaneously links production and reception in the three modes of visibility, readability and playability. Playability (after and with visibility and readability) makes a site. A site "bi-faced, actual and virtual", to use Deleuze's expression. From the virtual (by interactivity) to playability, the possibilities (the interactions) of what can be played are added, without amalgamating with it. In the virtual space of playability in which interactive perspective appears, another perspective is being realized, in the sense of what is being planned and eventually happens.