Message 00544 [Homepage] [Navigation]
Thread: oxenT00544 Message: 1/1 L0 [In index]
[First in Thread] [Last in Thread] [Date Next] [Date Prev]
[Next in Thread] [Prev in Thread] [Next Thread] [Prev Thread]

[ox-en] Ilkka Tuomi * Internet, Innovation, and Open Source: Actors in the Network


I just (skip) read

which has been on my reading list for quite a while. It's an
interesting text presenting some figures about the Linux kernel
development and the social processes visible(?) in it.

It looks at Free Software through the glasses of two social theories
and I find it results in a quite interesting view. Particularly I find
the concepts of punctualization and that of sedimentation interesting
because I guess it has a connection to dominion ("Herrschaft" - don't
know whether this is actually the correct translation). (We had rather
heated debates on dominion, Free Software and so on in the German
list.) However, I don't see very clear here yet.

Because I find it really interesting, I include the last section from
the article.

						Mit Freien Grüßen


--- 8< --- 8< --- 8< --- 8< --- 8< --- 8< --- 8< --- 8< --- 8< --- 8< ---

Two complementary approaches to innovation and socio-technical change
have been used above to interpret the evolution of Linux operating
system kernel. First we argued that knowledge is located and develops
in communities that are organized around practices. Knowledge is
tightly linked to technologies used in these practices, and to the
system of meanings which the community uses to communicate and make
sense of the world. This "community-centric" view has earlier been
used to discuss creation of meaning and knowledge by Bakhtin and by
Fleck, and more recently to describe social learning, by Schön and by
Engeström, and more specifically, socialization to existing traditions
and practices by Lave and Wenger and others. Actor-network theory, in
turn, has been used to describe the evolution of socio-technical
systems, often focusing on the struggles and strategies of
appropriating and creating power in the network.

The present paper puts these approaches in a context of an ecology of
communities and a modular technical architecture. Specifically, we
have tried to show how the changes and dynamics of technological
architectures reflect tensions that are created in developing the
system in question. We have put actor-networks inside communities of
practice, and briefly described how communities become actors in a
network of communities.

We have therefore also changed the way actor-networks and communities
of practice have often been interpreted. The theory of actor-networks
has the problem that it potentially makes humans and non-humans too
symmetrical. It is as if machines, tools, and technologies would have
their own motives and will in the same way as humans. This assumption,
of course, would require careful discussion on the nature of motives,
and here activity theory can bring useful insights (Miettinen, 1999).
Such discussion, however, leads to a view that sees motives as
grounded to social practice, division of labor, and tool-mediated
activity. The locus of activity can then be found in a community that
organizes itself around the specific practice in question. By arguing
that communities are special and fundamental types of actors in an
actor-network we can describe what makes the evolution of
actor-networks possible and how such evolution occurs.

On the other hand, by utilizing the concepts of translation,
punctualization, and resource as described in the actor-network
theory, we can better understand the evolution of practices and
communities. This is important for understanding technological change,
as new technologies are always appropriated by integrating them into
social practice. Indeed, one can argue that innovation occurs only
when social practice changes. Often such change results from
appropriation of a new tool, which reorganizes the practices of a
community. The key to innovation, therefore, is in those social
communication and learning processes that underlie change in social
practices. The accumulated learnings, however, also become partly
embedded and sedimented into the architecture of the technological
artifact that is developed in the process.

Social practices, however, are interlinked in the ecology of
communities. Resources and tools that are used in a given practice are
produced in other practices. It is not always possible to change
social practice without breaking those translations processes that
make a community a resource to other actors. Change is difficult
especially when the same translation process is used by several
actors. As the evolution of Linux shows, one way to solve this problem
is to sediment resources, institutionalize practices, and stop

The history of Linux, however, also shows that effective translation
mechanisms can lead to rapid growth. The problem of managing
interfaces between modules has led to relatively standardized ways of
building and using interfaces. This, in turn, means that modules can
easily be added to the system. Furthermore, these standardized
translation mechanisms mean that modules can be relatively easily used
by different actors even when the modules change.

Linux is therefore in many ways open to combinatorial innovation.
Standardized interfaces and translation processes generate smooth
module boundaries and facilitate rapid recombination. The source code
itself can be sometimes reused, but more importantly, the learning
that is represented in the source code can be reapplied in different
contexts without major problems. As a result, the various communities
that develop the different parts of the Linux kernel become very
mobile. In this way, the solution to the problem of translation leads
to an ecology of communities that can readily reconfigure its

In the Linux development, the Schumpeterian creative destruction
destroys pieces of code, but competence and experience are reorganized
with little waste. In this sense, one could argue that the Linux
development model and the Silicon Valley innovation model (Kenney,
2000) have similar characteristics. As motives and values emerge and
become articulated within specific social contexts, one could expect
that the cultures of Silicon Valley and Linux development communities
are relatively easy to integrate without major conflicts. The main
difference, of course, is that Silicon Valley has a venture capital
driven entrepreneurial culture, whereas the economic sphere has been
relatively invisible in the Linux development. This tension is
actively being managed by the Open Source Initiative (c.f. Raymond,
1999). Indeed, the Open Source Initiative can be seen as one more
organizational form or a community that springs up in the evolution of
Linux to repair social damage that is created when these two
relatively similar cultures collide and create conflicts in the
developer community.

As the analysis of the evolution of Linux shows, rapid growth requires
that the core is institutionalized and that some of the translation
processes are taken for granted. In this model, innovation happens in
periphery. It is interesting that such peripheries are conventionally
described as frontiers. We could ask, however, whether - and in what
sense - progress results from moving the boundaries of periphery, or
whether this simply is one strategy to reduce change in the core.


Thread: oxenT00544 Message: 1/1 L0 [In index]
Message 00544 [Homepage] [Navigation]