Contexte et enjeux du projet. Background, state of the art, issues and hypothesis.

De Wiki @ Brest

Décrire le contexte économique, social, réglementaire… dans lequel se situe le projet en présentant une analyse des enjeux sociaux, économiques, environnementaux, industriels… Donner si possible des arguments chiffrés, par exemple, pertinence et portée du projet par rapport à la demande économique (analyse du marché, analyse des tendances), analyse de la concurrence, indicateurs de réduction de coûts, perspectives de marchés (champs d’application, …). Indicateurs des gains environnementaux, cycle de vie… Décrire le contexte et les enjeux scientifiques dans lequel se situe le projet en présentant un état de l’art national et international en incluant les références nécessaires.

Global context.

Traditional intellectual protection framework allows producers to finance the initial phase of the innovation process, anticipating returns in the diffusion phase, in which they expect a legal monopoly to produce and commercialize it. Open Source movement shows that, for some knowledge intensive/cumulative innovation, this can be counter-productive, as the diffusion phase allow the producer to finance its initial investment without having a monopoly, but thanks to the feed backs and the joint needs generated by new users and uses.

By connecting a large amount of individuals around a common project and without any closure, open software is a way to take advantage of a fantastic potential of distributed skills. And by return developers not only expect the benefits they will draw from the availability of the public good that constitutes the open source program, but also will benefit from individual learning effects (Foray and Zimmermann, 2001). Due to the open context of working, learning through contributing to a free software project can be considered as more efficient than in the boundaries of a close enterprise. This can be explained by the multiple interactions a developer benefits with a wide variety of programmers using a wide scope of methods and programming styles. This “learning by interacting” represents an other complementary dimension of individual incentives. Such variety is also a source of quality for programming goals.

Then it is clear that the status of intellectual property, as it is conceived in open source software, allows the development and working of a remarkably efficient production process in which all members benefit from the whole community efforts, while continuing to manage their proper activities on a decentralised way. Beyond the sole case of software development it is to be foreseen that such principles of working of “distant community” [1] could have many applications in other knowledge intensive fields of activities. Diverse cases have emerged recently that could foreshadow a wider range of applications.

A first illustration is given by the “Wikipedia” project of encyclopaedia. The project is born in 2001 in the US Internet company Bomis. It counts now for more than one million of entries, in a very large scope of languages : mainly English but also French (62 000 entries), German (100 000), Japanese, … It is one of the most visited sites in the world. It is supported by gifts and voluntary contributions and a strict postcontrol permits to avoid content drift. It is the responsibility of a core group of “administrators” endowed with the trust from the whole set of the contributors. Telabotanica3 project on botany is another example of collective production and sharing of knowledge. There are also some propositions for transposing the open source software model into biotechnologies [2]. However, these initiatives have variable impacts on the way such good or knowledge are produced. In the chip production field, which use VHDL language to design these chips, there are some libraries of simulated componants [3], sometimes free. Initiatives to co-produce Free circuits are old [4], but they had little impact on the industry, at least for the moment.

An other significant illustration of this need to get out of a strict intellectual property rights protection is given by the recent iCommons initiative. It is an international movement born in the United-States, that is at the origin of the Creative Commons tool “to give authors free tools to enable them to mark their content with the freedom they intend their work to carry, while reserving the rights the author believes must be reserved.(…) It has attracted musicians, academics, authors, film-makers and researchers internationally who want a simpler way to exercise their rights without rejecting the protection of copyright altogether”(L.Lessig, 2004)4. Creative Commons Licences began to be translated and adapted to a wide range of national juridical contexts : Netherlands, Taiwan, Australia, Sweden, France … and have still been adopted by more than two millions of creative works in the world.

Today, the biggest challenge for intellectual property rights mutualisation lies undoubtedly in the field of knowledge intensive activities with a high level of complexity of the knowledge base to be mastered and combinatorial aims of informations and skills. A significant example is the one of life sciences and industry. Joly and Hervieu (2003) plead for a high degree of knowledge resources mutualisation in the field of genomics in Europe, not for intellectual property renouncing, but by organising a collective system of management of intellectual property. This will permit to reinforce the competitive position of European firm facing US multinationals and, by pooling basic technologies, to avoid innovation clamping.

What is to be investigated.

To do so, we must better understand the the functioning of these communities of practice mediated by the Internet. We define community of practice as "Des groupes d’individus partageant le même intérêt, le même ensemble de pbs, ou une passion autour d’un thème spécifique et qui approfondissent leur connaissance et leur expertise en interagissant régulièrement »[5] (Wenger et al, 2002, p. 4, the translation is ours and we stressed the "et")

The example of Open source studies proves that this can be made, but appeal for complementary analysis at individual level (the role and the incentives of the participants), at organizational level (how these people coordinate to co-produce and diffuse the pieces knowledge) and at institutional level (how the existing institutions, being classical organizations such as the firms, the public research institutions, or social framework, such as the Intellectual Property Right, interact with these communities).

The roles of the participants.

The open source development model generally doesn’t involve a fixed and pre-established division of labour among the developers. Demands for contributions to components production or improvement are usually sent out towards a broad and open population of developers in order to improve the chance to get a good and fitted solution and to benefit for that from a large variety of skills and approaches. In addition, improvements or modifications can be proposed spontaneously and developers often already have on their own shelves the answer to the raised question (Von Hippel, 2001). Then it is the responsibility of a core group of developers to establish the relevant orientations of the development, select the contributions, articulate them in the global architecture of the program and to supervise a large part of the developments involved in the program structure[6]. This is the condition of coherence and quality of the collective effort, it is an essential task of coordination of the community production. But an open source community doesn’t restrict to the sole developers but also includes a category of “frontiers users” (Kogut and Metiu, 2001) that are not able to contribute to the software development but have an essential role in terms of source of innovations (Von Hippel, 1988) and above all testing and debugging base. “Hence, it is not modularity that gives open source a distinctive source of advantage, because it too relies on hierarchical development. Rather the source of its advantage lies in concurrence of development and debugging. In spite of its unglamorous nature, maintenance alone represents anywhere 50-80% of the software budget. The largest part of the developer community are not involved with code writing, but with code-debugging” (Kogut and Metiu, 2001).

Is this division of labour, with different roles assigned or endorsed by the participants, a constant and a condition of success for these communities?

How do these communities work?

Demazière, Horn and Jullien (2004) use the expression of “distant community” referring to the fact that the individuals involved are not only dispersed in geographical or organisational terms but also to their heterogeneity regarding their individual profiles. First developers can occupy various professional positions : students, universities or research centres employees, workforce of private enterprises linked to open source software, personnel of other enterprises… For those different categories there is a different balance between open source development activity and remunerated work, inducing a plurality of the juridical and temporal conditions of the contributions. This statutory variety has to be added to the formerly mentioned variety of skills, programming styles and methods, but also to the cultural context and ethical dimensions of motivations to contribute to the open source development. However, in spite of this diversity, all the contributors “share the feeling to belong to a specific community motivated by a strong common identity”, that counterbalance the weakness of their direct interactions and their dispersion (Ibidem). The cement of this cohesiveness is clearly the status of intellectual property that is a guarantee against deviations or misappropriation of the collective effort.

But for doing so, such a core group needs to be devoid of private economic concern into the project. That’s the reason why individuals of the core group are chosen from the sole consideration of their technical competences and are co-opted by the developers group alike in the scientific community. In this way the developers’ community preserve all the characteristics and merits of an epistemic community: cognitive goal, commonly accepted structural authority, synergy of individual variety, individuals’ knowledge accumulation based on their own experiences, recruitment with regard to agent’s contribution to the common goal (Amin et Cohendet, 2004). From the other side the way of working of open source production, that is allowed by the principles of open intellectual property management, endows the community with the same coordination efficiency, in terms of collective production consistency, than a formal organisation but while preserving its flexibility and the wide extent of its resources. It forms a very motivating mode of decentralised coordination very similar to firms cooperation networks but more flexible and never entangled by intellectual property conflicts.

What are the internal rules for selecting a contribution, and more broadly organizing the community? What are, if any, the invariants of these rules or in the organisational structure for a mediated community of practice to work? This is the second major concern of the project, and in Human and Social Sciences: to understand new forms of interactions mediated by web technologies. Online epistemic communities are representative of these distant and asynchronous situations of interactions.

For example, in open source software (OSS) communities, design is almost exclusively done at distant locations and in an asynchronous mode. It is usually assumed that OSS developers rarely meet face-to-face. Design is essentially mediated by tools such as discussion lists, forums, and platforms for cooperative work. The traces of interactions are archived on the projects platforms. Online epistemic communities have been studied for ten years in various disciplinary fields: economy, sociology, linguistics, psychology and computer science. Although this research focus is at the convergence of various disciplinary fields, there have been, so far, very few pluridisciplinary attempts to study these new forms of online organization. There is a need for jointly analysing it/them according to three dimensions relevant to understand such interactions, a social dimension, a cognitive dimension and an interactive dimension.

In which context these communities can succed?

Levine (2004) emphasizes the spreading of collectives that “operate in concert to accomplish innovation goals that may have great economic significance, and accomplishment of those goals is often their primary raison d’être.

The significance is clear: real products are created, services are offered, and economic rents are appropriated or lost”. Such collective, that Von Hippel and Von Krogh (2002) call the “private collective” innovation model, can be mainly observed today as carrying their interactions on an Internet framework, but as put in evidence by Allen (1983) this type of “collective invention” doesn’t require modern communication tools and can be observed in a more traditional context. In the British iron and steel industry (Cleveland district) in the second half of 19th century, companies have revealed freely among competitors technical information related to blast furnace design in order to allow an experimental incremental advance to improve the efficiency of industrial smokestacks. Osterloh and Rota (2004) also point out two other examples of “private collective” innovation models. With the Homebrew Computer Club formed at Stanford University in 1975 members exchanged ideas and projects in order to explore the potential applications of the already emerging microprocessor technology. Steve Wozniak, Steve Jobs and the other founders of Apple met in the club. An other example they give is the one of the early developments of flat panel display technology in the late 60s. Following Spencer (2003), a majority of the firms actives in this field have published at least one scientific paper then sharing a part of their RD results. These firms are precisely those who got the highest innovative performances (in terms of the value of patent portfolio).

What does the Internet change for the emergence and the economical impact of such industries? In Jullien et Zimmermann (2006), Jullien (2008), we have been able to explain the involvement of firms into Open source communities analysing the market of such firms, the way knowledge is built and used to propose products and services in the computer industry. Finally, it is today argued that, in industries with strong network effects (Farrell, 1992), and characterized by sequential and complementary innovation (Bessen & Maskin 2000), a low level of intellectual protection favours the cumulativeness of innovations and parallel investigation based of competing approaches, and thus improve the technical progress.

Can we measure a correlation between the intensity of the network effects and of the complementary/cumulativeness of innovation in industries and the emergence of online communities of practice? How these experiences use the actual intellectual property Right framework and what are the evolution of this framework needed?

In this project, we will try to answer to these questions, at these three levels (individual, community, and institutions), knowing that they need to be studied in parallel to make the knowledge of online communities of practice progress.
  1. D. Demazière, F. Horn and N. Jullien, 2006, How free software developers work. The mobilization of “distant communities”, working paper, M@rsouin. A former version (in French) has been published in Cahier lillois d'économie et de sociologie, 2005, n°46, pp 171-194.
  2. See Janet Hope, 2008, BioBazaar: Biotechnology and the Open Source Revolution (Harvard University Press, janvier 2008, ISBN-13: 978-0674026353). the BiOS (Biotechs Open Source) initiative,
  4. A list of them:
  5. « group of people sharing the same interest, the same problems, or a passion for a specific theme and who deepen their knowledge and expertise interacting regularly
  6. Following Mocus and al. (2000) about the Apache project “despite broad overall participation in the project, almost all new functionality is implemented and maintained by the core group” (p.268).