Introduction

New information and communication technologies make possible the simultaneous creation of artworks upon a virtual setting.Such a possibility raises questions regarding collective interactive creation in the pictorial domain.Can a collective creation process produce works of art? How can one adjust the process in order to improve the artistic character of the created works?

A common feature of these new environments of artistic creation is that the “artist is no longer viewed as the creator of objects, but as the creator of situations where the public’s creativity can be unveiled”. The question that arises today regarding these creative contexts is: who is the creator? Indeed, a number of players can lay claim to this role: the one who conceives the work, the one who produces it, and even the “specta(c)tor” (spectator/actor), to use a term coined by Réjean Dumouchel. The present paper examines such questions by drawing on the CATI (Création Artistique sur Toile Interactive—Artistic Creation on Interactive Canvas) creative environment, a tool inspired by the Surrealists’Exquisite Corpse technique.

Techno-logical evolution

As a prelude to these questions, let’s take a look at the evolution of pictorial creation techniques.Pictorial creation upon a virtual support introduces a significant change in artistic practices.Yet, we can sketch a fictitious “techno-logical” course of development regarding the relationship between pictorial creation and its supports that makes it possible to consider electronic creation as one mode among others for inscribing images. The various stages in this evolution are the result of what can be seen as a progressive loosening of constraints: constraints concerning the support, the creation process, the shapes, the layout, etc. This fiction is thought of in a cumulative sense: new technologies build upon possibilities introduced by those preceding it.

One starts with the physical world, the real. To this is added the image of the real — the imagined, figurative representation that one finds, for example, in cave paintings. Next, the fixed support (cave walls) is extended to movable objects found in the environment which have an ascribed function, such as an amphora. In the following stage, the support is no longer limited to practically useful objects, but becomes a support specific to artistic expression: tablets, canvases. Technology introduces the digital, computer-generated image, and the means of processing information (via computers), which in turn make possible the creation and manipulation of images. With the arrival of advanced sensors, such as spatial location sensors, the modalities of interaction have also moved forward. The artwork is the product of the environment and the spectactor’s interactionswith it. Lastly, the introduction of high-speed communication between networked computers opens up the possibility for the simultaneous production of works among several participants, i.e. collective creation.

The CATI collective drawing environment

In order to illustrate this point, we’ll examine the collective pictorial creation environment of CATI.[2]

The creation process

CATI is a creation environment designed for use primarily by non-artists (and non-computer experts). The creation process is modeled around the surrealists’ Exquisite Corpse experiment from the turn of the century, but extends it along two spatial dimensions and one temporal dimension.

The Exquisite Corpse was a party game and just one of the techniques perfected by the Surrealists for exploring the “mystique of accident.” Georges Hugnet, in his Petite anthologie poétique du surréalisme (1934), describes it as follows: “There are five people around a table. Each person writes on a sheet of paper, without letting the others see, a noun to be used as the subject of a sentence.  The sheet is then folded in order to cover the word, and transmitted to the person on the left, at the same time that one receives a folded sheet from the neighbor on the right. Below the noun that remains hidden, one then writes an adjective. The same process is followed for the verb, then for another noun serving as the direct object, etc.” The example, which became a classic and provided the name for the creation process, is taken from the very first sentence created : The exquisite corpse will drink the young wine. The same process was applied to the creation of drawings, in which each participant’s task was to draw, for example, a part of the body.

Several people (usually three) can participate in the CATI process, which takes place within a grid of boxes (usually nine boxes arranged in a 3x3 matrix).

Figure 1:  Canvas-Grid

In the first drawing stage, each participant can draw only within his/her own pre-assigned boxes. He/she sees only the edges (margins) of other participants’ drawings whose boxes are contiguous with his/her own. Each participant is assigned a creation window containing a drawing area and a set of tools, such as: a pencil, paint brush, eraser, a drawing tool for basic geometric shapes, text, image, color palette, and line styles. The creation process involves several successive stages, each one resembling the previous one. At the start of each stage, participants are able to see the entire drawing currently being produced.However, during the drawing stage, one cannot see the others’ drawings, except along the margins whose size and dimension can be adjusted. A participant can only draw within boxes assigned to him/her. The creation process usually stops after three complete stages.

The coordinator of the creation process has access to a control window in which he/she can set parameters (e.g. usable tool palettes), assign boxes to participants, and see the overall image currently being created from the individual drawings.

The example of the coordinator screen shows the coordinator’s control window, which includes the following elements:

• on the upper left-hand side is the overall image currently being created from the individual drawings
• to the right is a set of definable parameters and the start/stop buttons
• 2/3 of the way down is the assignment grid for participants (in this case, 3 participants and 9 boxes)
• at the bottom one finds the list of tools and palettes available to participants

The example of the creator screen shows the window that participants use:

• occupying most of the screen is the drawing area (on a light background), including margins
• to the left of the drawing area are buttons for activating the available tools
• to the right are the settings for selecting colors, line styles, etc.

CATI fits within the context of cooperative creation. One of the first digital productions was Olivier Auber’s poietic generator for pictorial creation among several participants. In this case, each person can express himself/herself within one small part of the overall work. A difference with CATI is the fact that within Auber’s environment, the entire work is always visible to participants, whereas in CATI participants see only part of the overall work, just as in the game of the Exquisite Corpse.

The principle of the Exquisite Corpse has been used on several occasions to create texts or images: one such as example is found at artslab.ucsd. An important and original aspect of CATI is that it puts in place a number of heuristics for focusing in on “artistic quality”: heuristics that take into account the fact that participants cannot see the overall work.

Other creation environments, such as Work.de,whose main advantage derives from a creation tool capable of integrating high-quality images, differ from CATI in that their creation process is neither simultaneous nor in real time.

The artistic quality of a work

While using CATI we noticed that some drawings could be qualified as artistic (in a very subjective sense, of course) while others made up a category that one might call “cacographic” (in the sense of pictorial writing). It is not my aim to enter into a discussion over the artistic nature of a work. Rather, I want to note what seemed to us to constitute a necessary (but not sufficient) condition: namely, the ability to trigger an emotion. We thus sought ways of promoting the creation of drawings that, on the face of it, did not show signs of belonging to the cacographic category. Different heuristics resulted from the search for such methods.

The first heuristic (H1) concerns the background which, if not the same for all the boxes, can lead to a reading structured in terms of boxes rather than in terms of the overall drawing itself.

A second heuristic (H2) has to do with image content. In order to promote greater homogeneity of content (which is optional), it seems useful to choose a common theme (as was the case to some extent with the Exquisite Corpse). We therefore explored a variety of themes: some more concrete (water, a party), some more abstract (geometric shapes); and yet others more conceptual (fantasy, euphoria).

A third heuristic (H3) concerns the drawing’s homogeneity as understood in terms of correspondence between the box-contents: continuity of lines, shapes, color selection, etc. We considered two ways for improving such homogeneity:

• by using margins:the margins allow the illustrator to control the drawing’s continuity; their relative size can be adjusted (from 0% to 100%)
• by using box assignments: two neighboring boxes that are not homogeneous can be assigned to the same person in the hope that he/she realizes what the problem is and makes the necessary adaptations.In CATI, assignment is entrusted to a human coordinator; it would be interesting to automate all, or part, of this task.

In order to illustrate our approach, three examples of drawings created with CATI are shown below.he drawings are the work of Irène Charon, Olivier Hardy and Alain Grumbach. Example 1 was produced using the heuristics H1, H2 and H3: the theme was: "euphoria".

A fourth heuristic (H4), which is somewhat more subtle, concerns what we called “style”.CATI makes it possible to influence the creation process with the aim of imitating an artist’s style.This is done by selecting certain tool palettes (lines, basic shapes, etc.) and colors similar to those used by an artist. For a Mondrian style, only squares and horizontal and vertical rectangles are made available; the color palette is made up of three primary colors (which can be chosen), as well as black and white. It is thus possible to create a drawing that resembles a painting by Mondrian, but which consists, of course, of its own specific significant elements. Example 2 was created using heuristics H1, H3 and H4, and applying the Mondrian style.

A fifth heuristic (H5) has a graphic form: initiating a drawing. In a preliminary step it is possible to begin a drawing in one box (usually the center box which is adjacent to the other eight) and then let the illustrators elaborate upon it within the remaining eight boxes. At the end of the process, some adjustment can be made to the starting box. Initiating the drawing is the coordinator’s task. Such a method was used in conjunction with heuristics H1, H3 and H5, with the coordinator first drawing inside the central box. Figure 2 below shows the initial drawing. I invite the reader to elaborate upon the drawing within the other eight boxes, using his or her own imagination, before looking at the actual result shown in Figure 3

Figure 2: Initiating a drawing

H1 through H5 constitute the set of heuristics used by CATI. They are in keeping with a traditional view regarding “works of art”. Yet, one could have chosen other heuristics, and thus explored the possibility of producing artworks belonging to a potentially new genre. One such possibility, which seems quite interesting, is found in Maurice Benayoun’s installation The Tunnel under the Atlantic. Benayoun combined a set of dynamic images that came into play during the process. The tunnel under the Atlantic is a tunnel of images commemorating the ties between Canada and France. At each end of the tunnel a spectactor digs with the help of a joystick. In this way he/she is able to move about within a virtual gallery whose walls are covered with images. The system chooses these images based on the spectactor’s presumed tastes: tastes that are assessed according to the amount of time spent in front of each image. The heuristic used here leads to automatic interventions during the process. In relation to CATI, such an intervention could be applied to the assignment of boxes: a (brief) analysis of the box-contents (tools and palettes used) could make it possible to influence the assignment of boxes, and hence the creation process itself.

Within the context of this creation process, another question arises: Can one produce a new type of artwork? The heuristics presented so far tend to reproduce features that traditionally characterize works of art, which constitutes first step towards collective creation. Nevertheless, the question is worth raising, though the answer remains open.

The “creators”

Having examined the creation process, I’d now like to turn attention towards the creator of the artwork by asking the following question:

Who is the creator of such a collective interactive work?

This question provides an opportunity for clarifying who the contributors are in a collective creation process.>

The poetics of the open work establishes a new type of relationship between the artist and his public.

Several actors can lay claim to the role of creator. The first creator is the designer of the tool, who is closely followed by the developer (the programmer).

An installation by Catherine Ikam and Louis Fleri [Ikam 2000] consists of a space in which a spectactor (or “participant”, “interpreter”) can move about. The image of a young woman’s face is projected upon one of the walls; the facial expressions change in relation to the spectactor’s position and movements. In this installation, the role of the designer is important since it is he/she who defined the behavior of the created images.

In CATI, the designer has a limited role in the work: he/she creates a tool without having to worry about the artistic nature of the “creatable” objects. Instead, this worry is the responsibility of those who assume the role of spectactor, i.e. the illustrators. Moreover, as we already saw in the discussion on heuristics, the creation process presupposes the intervention of a “coordinator” whose role is to manage the use of heuristics: to choose a theme and background, to assign boxes, control the time, etc. The coordinator can either be human or automated, as in the case of Benayoun’s The Tunnel under the Atlantic.

Collective interactive creation thus involves the cooperation of several participants: the designer (or author who had the idea); the developer (who actually produces the computer tool); the coordinator (who manages the creation process); the observers; and the actors (illustrators). To the terminology corresponding to this set of functions we can add the “creator”, the subject of this section’s opening question, who brings together under one role the different functions (with the exception of that belonging to the pure, non-acting, observer).

Conclusion

The process of artistic creation has evolved from its original state as much in terms of its tools, forms and styles, as in terms of content and participants. This evolution is now passing through an important phase as a result of new possibilities opened up by modern technologies.

Artistic creation, in this context, is oriented towards modes that draw on the contribution of (a) spectactor(s). The designer envisions and creates an environment that is shaped by the spectactor. If an artwork created in such a manner is not to become mere cacography, the designer must take care to respect certain constraints concerning the potential creation process used by the spectactor. These constraints are necessary for the artistic creation process to function, but they are not, of course, sufficient. The spectactor’s intuition takes care of the rest—with greater or lesser success.