a version of this Foam Publication appears in “A Guide to Good Practice in Collaborative Working Methods and New Media Tools Creation” as “Formalising Operational Adaptive Methodologies or Growing Stories within Stories” http://ahds.ac.uk/creating/guides/new-media-tools/kuzmanovic.htm

In the following pages I will describe the situations that have given rise to the formation of FoAM, a cultural laboratory based in Brussels and Amsterdam, as well as look at the issues around public perception and reflection on multidisciplinary projects. Since the challenges and successes of collaborative projects can be best illustrated through case studies, due to a lack of conclusive “how to” manuals, I will discuss TGarden, a research and production project for responsive environments in mixed reality.

Interdisciplinary projects are a highly praised praxis - at least in theory. In practice, the ecstasy of collaboration frequently becomes a delirium of communication between strategists sending smoke signals to the other camps from their disciplinary seats, but not actually moving their positions any closer to each other. When a small operator finds itself in between the camps, it can't help being mesmerised by the new perspectives opening in this space. In the smoke signals there is no shortage of engaging philosophy espoused both by artists and scientists. Unfortunately, these theories are only rarely translated into practice. The first difficulties encountered are often the bureaucratic challenges, such as dealing with as many administrative procedures as there are disciplines involved. Many grass-roots initiatives avoid the economic and political debates around multidisciplinary endeavours by working on small, garage/kitchen type projects, which usually have a limited amount of structural and strategic impact within larger cultural and scientific networks. As soon as these initiatives grow slightly bigger (in terms of people involved, technology used or the ambitions pursued) they usually slip between the cracks of rules and rigid boundaries between disciplinary realities. Alternatively, these initiatives can become institutions that with time become too inert to cope with the fast changes in the scientific, technological and cultural fields. The second option tends to be preferred in the political and economic circles, as the initiative appears to become more reliable with its size and clear hierarchic structure (especially administrative and financial).

“In panic, people try to replace the lost order of the organic process by artificial forms of order based on control.” - Christopher Alexander

As a sole operator, an artist - albeit independent and free to pursue his/her experimental activities ñ does not have much chance of becoming a respected player in the structures that decide on the strategies for social, political and economic development in the fields of multidisciplinary cultural production. The known methods for infiltrating these worlds by individuals are artist residencies or creative appointments in research institutes and businesses. As enticing this perspective might appear, the artist often ends up either as a marginalized member of the community (as a rarity exhibited in the institution's Wunderkamer), struggling for the acceptance of her methodologies, or is exploited for his/her implementation skills, without a significant impact on the strategic level of the organisation (artists are often hired to make the scientific, technological or business proposals look “nice”). It should be mentioned that artists are not widely known for their willingness to adapt to an imposed pre-existing order, and this can often become an important factor in the interaction difficulties.

In summary of this incomplete story, being an independent artist will give you the freedom, but usually not the tools to develop your works; working inside businesses or research centres can isolate you from the issues and methodologies that are crucial for ongoing artistic development; but becoming an institution to facilitate your interdisciplinary work can eventually lead to stagnation of the fresh ideas needed for a productive creative process.

In order to maximise the advantages of working in multidisciplinary contexts, together with scientific and business communities, while remaining autonomous and flexible, we have found a need for an entity to function as a mediator between the different disciplines and their imposed ruling principles (or “religions”). This entity should provide a legal and financial framework for partnerships with public and private research institutions, educational structures and businesses, while remaining an enticing experimental territory, allowing its dwellers to grow their own worlds.

“There is no solid status quo, only a series of relative realities, personal to each of the characters, any or all of which are frail, and subject to eruptions from other states and conditions.” - Clive Barker

The challenge involved in multidisciplinary collaboration, and more specifically the crossover between art and science, is that its substance, its subject matter is not continuous; there are no predefined starting points or outcomes. This in itself does not pose a great difficulty to its practitioners, but it becomes a barrier for evaluators of multidisciplinary projects, who usually come from either traditional artistic or traditional scientific contexts.

The main issue of concern is that there are no consistent classification or evaluation mechanisms for the results of these projects. For example: a responsive environment that could comfortably span approximately 20 existing colloquia will be represented as an installation in a visual arts context; as a performance in a sound or theatre world; as a demo session on visual tracking by computer vision experts; or as a statistical usability lab for social scientists. However, although a responsive environment does encompass all these (and frequently more) parts, its strength lies not in their sum, but in the way they relate to each other and the whole. This whole is gradually formed through the process of the interaction between research methods, creative inputs, tools, hardware, software, responsive materials, ephemeral states of the media worlds, analysis of real world and simulated data, historical contexts, and last but not least the users, players or inhabitants of the responsive environment. The strength of successful interdisciplinary collaborative processes is that the rigid borders between these parts erode into malleable edges, where at the points of intersection the substances of the parts form a new alloy.

The political and economic structures assigned the task of promoting innovation and creation today seem to be more concerned with preserving the state of the separate substances, rather than encouraging the interaction between them. It is also true that the organisations operating on the cusp between scientific research and artistic practice have not yet developed a valid cross-disciplinary critical framework that would make their efforts credible to both the cultural and the scientific worlds. The language is used to reflect on these works is often so overflowing with neologisms, jargon or “catch phrases as theoretical objects”, which make the language even less understandable to professionals from traditional disciplines. The terminology used to describe and promote the advances in multidisciplinary cultural investigations should not force the subject matter into premature classifications or definitions, but should emerge out of the collaborative process: the professional jargons of both the arts and the sciences should allow room for a clear language grounded in practice, that can be qualitatively judged by all knowledge domains, or at least be understandable with a minimum of translation.

“In theory there is no difference between theory and practice. In practice there is.” - Yogi Berra

http://fo.am/tgarden/

TGarden was set up in 2000 by SPONGE, based in San Francisco, and FoAM, based in Brussels, with the intention of developing a long-term project with multiple trajectories, partners and outcomes. The concept underlying TGarden is that of a “topological garden”: a malleable mixed reality environment in which the media (sound, visuals, fabrics) are shaped by the movement and social interaction of its visitors.

TGarden began as an ambitious enterprise spanning several continents, institutions and independent collaborators. Its production cycles in 2000 and 2001 have both resulted in environments-in-progress that were inspiring for their visitors, but disappointing for most of the developers, whose expectations were reduced with each cycle of implementation. Through a survey of visitors' and collaborators' reports at the end of these production cycles, a range of opinions have been collated, and some conclusions drawn about the positive and negative practices in distributed collaborations, where the collaboration involves small organisations, larger institutions, international funding bodies, universities and private businesses. The following paragraphs will give a summary of this survey, looking at the interdisciplinary collaboration process in the past two production cycles focusing on the three principal impediments to the successful realisation of the project: communication, implementation and funding.

Frequently, agreement on the methods and a common (operational) language can take up to thirty percent of the development time in interdisciplinary collaborations. The success of the projects tends to be more contingent on the motivation of the participants to truly listen to each other's points of view, rather than on top-down management of the project. All participants must have at least a basic understanding of all disciplines (and the more important intricacies), as well as a good understanding of how the different outcomes and processes should influence each other at any moment during the development cycle.

In TGarden the communication between the partner organisations became a bottle-neck for development. Although the rhetoric of all partners encouraged horizontal coordination structures, open development, and communication in all forms, only a few have followed this rhetoric through into practice. Much of the crucial information, such as design decisions or production delays (or misunderstandings) important for other developers in co-dependent fields, was kept within the organizations, only to be explosively released when it was too late for other partners to collectively find other solutions to the problems at hand.

The lack of funding made remote, distributed collaboration a necessity for TGarden. The necessary co-present work was scheduled for development workshops at key points of the process and long testing periods were planned before each of the public showings. Most “verbal” communication was to be done by email or through a discussion and content development board. A CVS repository (version control system) for documents, media and software was established to enable concurrent development, while enabling the process to be documented and evaluated at the end of the production cycle. However, this type of collaboration, although inexpensive, demands a high degree of flexibility and commitment, which most TGarden partners proved unable to provide. Many agreements were reached on the phone or in meetings of small groups of people that would often not be passed on to the rest of the collaborators, and therefore were not acted upon.

Communication between collaborators from different disciplines, which should have led towards a more integrated approach to the problems, often became a competition of personal priorities and disciplinary ontologies, with the consequence that co-dependent components have never been properly integrated, and as a result of the collaboration, the responsive environment remained fragmented (the sound, visuals and costumes all lead their own lives, without having too much experiential overlap). The differences in approaches to problem solving between disciplines, as well as the professional and cultural skills of collaborators, have led to miscommunications and misunderstandings between the different partners. This subsequently resulted in additional unnecessary losses of time, money and resources in already tight schedules and budgets.

Several instances of TGarden have been presented during the project's development, including: SIGGraph in New Orleans, Medi@terra in Athens, Ars Electronica in Linz and Las Palmas (in collaboration with V2) in Rotterdam.

The presentation area consisted of three spaces: the waiting room, the dressing rooms and the play space. The visitors would gather in the waiting room, where they chose a play costume from an array of garments designed to influence the movement of the wearer. Then, two to five players were led to their dressing rooms, where they were dressed in the garments and equipped with movement sensors (accelerometers) and wireless transmitters (Body Synth or Ipaqs in different instances). After the dressing process was complete, the players were guided into the play space, where the visuals (projected on the floor) and the sound (played through a surround system) were manipulated by the players' body movements, gestures and position in space. The media environment was designed to evolve through the interaction of the players, but this part of the system was never completely implemented due to the lack of testing and calibration time in fully built play spaces.

Responsive environments in which the media are synthesized on the fly (or in real time) demand the simultaneous development of tools for the construction of the media, as well as the media itself. This development process requires recursive experiment cycles in spaces with the same set-up as the presentation spaces. The tests must be conducted by both tool (system) and media developers, who are often the same people, preferably in the presence of the whole developers' team. This has proved more complex to achieve in practice than we anticipated. The presentation venues do not plan for extensive testing periods, usually due to either inability to grasp the significance of such experimentation or due to budget issues (the presentation equipment, such as projectors and sound systems, is often too costly to rent for extended periods of time) or sometimes both.

The solution investigated by the venues in previous TGarden production cycles was to conduct the testing as experimental labs open to the public. This would function well in the final stages of development, when all the systems would function properly, but still needed to be “choreographed” to form a coherent whole. Due to the venue restrictions and to the delays in production of separate components, the TGarden team has not managed to successfully work undisturbed in a fully built and equipped space, to have a fully functioning system necessary for the improvement of the tools and the experiential quality of the environment. The tight deadlines of an artistic production (demanding a presentable public spectacle every few months) pushed the TGarden team into a frenzy of impromptu development, which compromised many of the ideas presented during the conceptual phase. The delays in the development of integral parts of the system, and later unsatisfactory experimentation and presentation conditions increased the tension of the team in the already difficult situation. These problems were mainly due to communication, coordination and funding problems.

One of the major problems of the TGarden production was that the level of funding and institutional support was too low for the bold ambitions that the core team had during the conceptual phase of the project. The budget cuts (which included the bankruptcy of Starlab NV, FoAM's previous host institute) resulted in the loss of a few crucial collaborators and infrastructure, forcing the improvisation on several acute parts of the system (often with only a basic understanding of the complete system on the part of the developers doing the improvisation).

As the project was designed as both a short-term artistic production, and a longer-term technological and sociological research project, the funding needed to be acquired from both cultural and technological fields. The artistic endeavours were funded appropriately, although with long delays, but the technology development needed larger funds and longer development and testing periods to cover the full scope outlined in the TGarden technology specification. The scientific grants generally provide more room for iterative development: from a research question to a first prototype, to usability testing, to evaluation, back to the research question, to the second prototype and so on. Although cultural funds did not consider such technological research an eligible cost, and suggested that this part of the project should be funded by a scientific framework, the technology funds didn't find TGarden's artistic undertakings to have a strong enough basis for economic and industrial development.

Despite the difficulties in the two past production cycles of TGarden, the artistic and scientific goals are still an inspiration, driving several partners to continue pursuing them. The TGarden 2001 cycle concluded with a usability research and documentation, with a survey of both visitors' and developers' responses. The visitors provided a colourful pool of information on their experience of the environment, while the developers focussed more on collaboration methods and the usability of the tools, the components, the media and the overall behaviour of the system The usability research provided a range of helpful insights, which need to be further analysed and put into practice in subsequent versions of the project.

The visitors' responses have been valuable in the evaluation of the project, but as expected, a wide range of reactions was noted, from “the room is a stupid machine with a complicated interface” to “the media and the system are too complex” to “this is like a really good bar” to “I want to have this in my house and play for hours”. Trying to form a conclusion based on all such generalisations would make the project's success nearly impossible. For later versions of the project a new documentation system is being designed in order to better understand the relationships between (1) the phenomenological experience of the players, (2) players' actual (physical) movements and actions, (3) the received and analysed sensor data and (4) media generation. This will allow the researchers to compare the visitors' responses to the recordings of the above components of the experience simultaneously. In this way, we hope to provide usable study material in the fields of HCI (Human Computer Interaction) and HCHI (Human Computer Human Interaction) for professionals who develop, research or evaluate these types of environments, as well as working towards a qualitative new audience development for multidisciplinary creative practices.

For the next cycles of TGarden development, prototypes of the working system must be made ready months before the public showings, by developing the hardware and software tools in a progressive, iterative fashion. The system should be tested and tuned in incremental steps, so that the component development and the connections between the components increase in complexity, but remain compatible and consistent.

The design goals are still relatively ambitious for a collaboration coordinated by a small organisation, but once agreed upon, all collaborators are asked to remain as close to the specified design as possible, especially when the co-dependent components are developed on remote sites. The quality of design and implementation is tempered by the interaction between both stages. The coordination of this interaction takes less effort if the agreements are followed through - a good design changes little during the implementation process. This requirement is quite a challenge in an artistic collaboration, as artists are used to changing their designs on the fly, without consulting the rest of the group (a trait often shared with programmers). In TGarden, the discrepancy between design and implementation led to a fragmentation of systems and media, and resulted in an often incoherent experience for the visitors.

Collaborative design does take more time and energy from the people involved. When done well, the designers are forced to communicate their ideas clearly from the beginning of the process, which later results in truly usable designs that can be easily integrated with the other components of the environment.

In txOom, one of TGarden's derivative projects, all design developments are continually updated on a CSCD (Computer Supported Collaborative Design) website in the form of text, collaborative sketches, images, movies and sounds. The software is shared through a public CVS (Concurrent Versions System) repository. The material development (mostly in textile and electronics hardware) tends to need more co-present work, as it is difficult to share the copies of material samples and prototypes. Therefore, more design workshops are scheduled for the collaborators from the fields dependent on material construction.

In order to improve the communication process, the projects that derived from TGarden place more importance on both streamlined co-present development workshops and distributed collaboration. The goals of the collaboration are set clearly in the beginning of the process, with as many possible bottle-necks anticipated, co-dependent development coordinated and methodologies agreed upon. Getting all the “heads and hearts” working in the same direction during the initial workshops is a bare necessity in distributed collaborations, together with careful nurturing of commitments and motivations of all involved, without discriminations based on disciplines or levels of education (misplaced elitism and traditional hierarchies tend to destroy these projects' core principles). In order to avoid any discrimination of disciplines or partners, the project's coordinators need appropriate tools, allowing them to grasp the interdependencies of the components and to anticipate possible problems due to delays in one (or more) development branches. These tools should be available online, accessible and understandable to all partners.

In order to avoid funding problems of the scale experienced in 2001, the initiators of TGarden, sponge and FoAM , have decided to split the project in several smaller derivatives for the next stages of the project's development. Each of these derivatives will focus on a different research and development agenda, but still share the acquired knowledge and the available results with the TGarden team. Shared CVS repository and content development sites remain operational and are being updated by the groups involved in further development. Information resources are open to all partners in the TGarden consortium, as well as interested third parties. It should be noted that this consortium is not a legal entity, but an informal assembly of partner institutions and individuals involved in the development of TGarden and its derivatives.

The research and production agendas have been clearly defined and separated, and appropriate funding sources have been outlined for both lines of development.

A few examples:

• How can we develop materials that respond to human gestures and social interaction; materials that change their physical and chemical characteristics, causing transformation of their shape, texture, colour?
• How does a computer system perceive human behaviour?
• How can a system improve its gesture tracking and analysis by learning and forgetting about human interaction within a responsive environment?
• How can we improve the coupling of sensory experiences through integrated sonification and visualisation of behaviour?
• What methods of ethnographic and phenomenological research can we apply to learn more about the usability of responsive environments, and the creation of meaning in interactive artworks?

• How can aesthetic and functional requirements be met in the integration of materials, garments and objects with hardware, such as sensing, controlling and actuating devices?
• Can we use biomimetics in the design of media and materials as a connecting principle between the real and the virtual?
• Can we wear an environment, sculpt media, mould time?
• How can we make an intermedia system that does not differentiate between sound and image?
• What kind of interface design is necessary for the system to become a useful tool for artists from a variety of disciplines?
• How do we design human-computer interaction to be understandable, but not simplistic?
• What kind of commercial products can be developed from the TGarden research results?

The partners involved in TGarden are exploring these and other questions in several derivative projects, such as txOom by FoAM and TCostume by Georgia Institute of Technology. Some of the results of these projects will be integrated in future productions by other partners, or may follow new research tangents not derived from the original project. In this way, the TGarden team will benefit from the open nature of the project, as the resources for research and development will be distributed across a growing number of partners.

“It is probably true quite generally that in the history of human thinking the most fruitful developments frequently take place at those points where two different lines of thought meet. These lines may have their roots in quite different parts of human culture, in different times or different cultural environments or different religious traditions: hence if they actually meet, that is, if they are at least so much related to each other that a real interaction can take place, then one may hope that new and interesting developments may follow.” - Werner Heisenberg

© maja kuzmanovic 2002