In my paper, instead of talking about the virtual university, I want to
explore some relations between three distinct domains: 1) the philosophical
theory of communicative action, 2) education, 3) computer-mediated
communication. I hope to show that these theoretical relations make a
strong point for the use of free/open-source software in computer-supported
learning.
2 Habermas and claims of validity
In his "Theory of communicative action" [2] the German philosopher Jürgen
Habermas investigated into the grounds of human discourse. He built upon
speech act theory, extending its sociological implications. Habermas
asks what claims a responsible speaker of a human language implicitly or
explicitly makes, when he adresses another person. Essentially Habermas
distinguishes three kinds of validity claims that are at least implicitly
present in all human communicative action. Each of them finds its origin in
one of the essential dimensions in which man orients himself through
language: the objectif world of perceptual and cognitif facts, the
subjectif world of emotions and intentions, and the normatif world of
social conventions. Each of them can become foregrounded in different
classes of utterances, but even if they stay in the background they are
implicitly present. Each of them can become the object of contestation by
the addressee of the utterance.
With respect to the objective world our communicative action raises the
claim of being true, that does not mean that we garanty the truth of our
propositions absolutely, and it does not exclude the possibility of
fiction or irony. It means that when we responsibly utter a proposition, we
have to be able to explicit on what conditions our belief in its truth is
grounded, or to describe the frame of reference inside of which the
proposition can be held for "true". (I hope you won't question me on all
the logical problems that this implies).
With respect to the subjective world our words usually claim to be
sincere. The sincerity of an utterance cannot be questioned immediately and
absolutely, since the inner world of the speaker is only accessible to
him. But it can be doubted on grounds of anterior previous of the speaker,
and it often has to be realized in subsequent behavior. (A promise has to be
kept)
With respect to the normative world, our utterances pretend to be
just. Living in society means being grounded in a complex set of
conventions, and implicitly when we speak, we rely on a subset of them. Our
utterance can be questioned, either in its being faithful to a given norm,
or by contesting the justice of the norms it is relying on.
What is important for Habermas, is that these claims of validity can be
critiziced in a rational manner, and his theory tries to analyze explicitly
the proceedings of this critique in order to establish a framework for
democratic and pluralistic discourse. Even if we have to take for granted
these claims in the course of everyday life, only if they can be justified
in principle, put into question and renegociated, our discourse can pretend
to give equal rights to all participants. In other words, the definition of
the frame, in which we communicate, i.e. the world we refer to, the
subjectivity we express, and the norms we accept, has to be, in principle,
the result of a rational and social discussion.
There is a fourth class of validity claims, Habermas only mentions
in passing because it seems so fundamental that it does not need
elaboration. With each communicative action a speaker claims to be
understood. My hypthesis is that in mediated communicative action this
fourth claim which seemed trivial to Habermas becomes problematic, but that
we can use his theory to evaluate claims of understandability of mediated
communication. In technologically mediated communication understandability
has to be guaranteed by the apparatus of the media. Normally we take it for
granted that these apparatus work and that we can rely on them, but we
would also insist that the institution that is supplying or controlling
the mediating technology is accountable for how it works, and what it does
with information generated by our communicative action.
3 data formats as objects of a Habermasian discourse
I think that extending this Habermasian argument to computer-mediated
communication we are lead to assert first of all the necessity of open
standards for data formats and communication protocols, and secondly the
advantages of free or open source software.
CMC depends on a complex interaction of different systems, subsystems,
programms and processes. This interaction works because there are data
structures which guarantee that information is transmitted and interpreted
correctly.
The user normally is protected from this complexity, but in
order to ground his claim that what he communicates is understandable by
his addressee, he must be entitled to demand a justification for the
workings of all the components of the system, and he should be able, in
principle, to give an account to his addressee in what conditions his
message can be decoded and understood. That does not mean that each user
has to understand all the complexity of computer communication systems, but
that this complexity can be publicly examined by experts who can explain to
the larger public the implications of these systems.
The argument in question is not the same in respect to open standards and
open software: Using open standards in computer communication means that
the user when sending a message does not have to force his addressee to use
the same computer software as he does, he can instead refer to the open
specifications of the standard, and to the multiplicty of programms that
are able to understand that standard. Using open software goes one step
further: It gives the user or the expert he is asking for advice the
possibility to understand different aspects of his communicative action,
take responsibilty on them, and to contribute to the discussion about their
technological and social quality and to efforts to improve them.
4 software, knowledge, education
I'd like to show in the following how this argument can be applied to
computer-supported learning:
When we construct, communicate, question and validate knowledge, we do not
only consider its content, but also its form. This is true not only for
technologically mediated knowledge, but also for human speach (think of the
importance of rhetoric) and writing (think of the elaborate forms in which
knowledge is usually presented as articles and books). Learning to work
with knowledge doesn't mean only to acquire content, but to master form.
Computer-supported learning uses a great number of new forms for
knowledge production, presentation and communication. Think of hypertext,
think of data bases, think of mailing lists. Each of them implies the
definition of data structures, communication protocols and user
interfaces. A user would not want to know the details of their technical
implementations, but I want to argue that in order to work with them
constructively and responsibly a certain consciousness about these forms is
required, and that these consciousness can best be cultivated in open
computer systems. I'll speak about the communicability, the construction,
and the manipulation of knowlege in turn.
communicability of knowledge: When we use computers for learning, we might
be a solitary student, who keeps his drafts, his essays, his annotations
for himself. But today it is widely acknowledged that learning is mostly a
collective and social practice. We enhance our learning, when we take part
in collaborative projects, where we share our thoughts, get feedback, and
might even write collectively. Computer-supported learning as CMC in
general is confronted with the problem of compatibility and
interoperability. The Internet has been built on open standards that permit
a large number of individuals to be connected without depending on one
vendor's system or software. But what is true for some fundamental
protocols of the Internet, is unfortunately not yet established for
specialized applications as document preparation, or educational
computing. Open standards for these applications exist, and I think that
the efforts in academic production of documents and software should be
concentrated on these open standards.
construction of knowledge: recent theories in collaborative learning
suggest that students learn best when they have the opportunity to
integrate knowledge they are confronted with in there own frames of
reference. Computer-supported learning can enhance a student's playful
interaction with knowledge, if it permits him to combine texts in new ways, to
test hypothesis with educational software, to confront different
perspectives on a given subject. The forms of knowledge can be hidden from
the learner in black boxes, and probably in certain stages it is preferable
if he is not directly confronted with them. But in order to evaluate their
validity he should be able, in principle, to understand how they work,
i.e. to construct perspectives on them that are not given in the user
interface of the software he uses. Norm Friesen [1] (referring to recent
literature in educational technology) expresses this idea in the
term "grey boxes", i.e. "components that would reveal more of their
"implementations" to their users." A user working with a system whose
informational objects and data formats are open -
open for investigation, experimentation and improvement by the user - can
define his own way of structuring and presenting knowledge.
manipulation of knowledge: computerized knowledge has maybe one decisive
advantage over traditional media, it can be manipulated in multiple
ways. So when we evaluate validity claims of computerized knowledge on its
media level, we not only can ask "In what contexts is it communicable and
understandable?" or "How has it been constructed?" but also "What can be done
with it?" How can it be printed, interacted with, questioned, remodeled,
combined, searched through? Again I suggest that open formats are most
capable of giving learners the freedom to manipulate knowledge at their
will, to print it according to their preferences, to choose the interface
they prefer, to highlight its constructedness, to mould it in other
formats, to create links to other knowledge, to extract relevant
information from it.
5 GNU/Linux and the virtues of ASCII text
I want to conclude this paper by showing how my Habermasian argument for
public software in education can be sustained by the GNU/Linux operating
system.
On the one hand it seems to me that the actual open-source developper and
user communities are examples for productive, virtual learning communities,
where knowledge is exchanged on collaborative grounds, in open formats, and
constructed in a way to be questionable, combinable and manipulable by
every competent individual. The tools used by successful open-source
projects (like CVS for version-control and collaborative text managment,
diff and patch for the economical distribution of updates, texinfo for the
production of documentation that can be read both as an online hypertext
and as finely type-set manuals) are examples for state-of-the-art systems
for collaboration and document preparation (I admit that much can be done
to improve the user interfaces to these programs). The information gathered by
users in Usenet fora and documentation repositories as the "Linux
Documentation Project" give another striking example of the potentials of
knowledge construction communities that are socially and technically
open. I do not pretend that every open-source project works on this
grounds, and that this ideal communicative practice is not very often more
an ideal horizon than an actual practice. But I think that this horizon is
effective by animating the brains, the hearts and the fingers of many
members of the open-source community.
On the other hand - and I will elaborate more thoroughly on that - working
on a GNU/Linux system seems to imply some principles that can be seen as
a realization of my theoretical argument. The main principles are the
respect of open standards for data formats and the prefered use of ASCII
files for data storage instead of binary formats. Both these principles
have consequences on the three aspects of digital knowledge mentioned
abvoe: communicability, construction and manipulation.
Learning to work on a GNU/Linux system can imply to learn about the
communicability of data. If data formats are defined by open standards, the
learner gains access to and can share knowledge with others working in
different digital learning environments. If data is stored in ASCII files,
it can often even be communicated to people without a specialized interface
for a quick examination with every text editor.
Data stored in ASCII files and in open formats is not only more
universally communicable, it can also empower the learner to develop his
own way of constructing it. Firstly because he often can choose between
different interfaces to these formats, secondly because some open systems
allow the user to develop customized formats without sacrificing
communicability. This interaction between universal and open formats and
individual definition of text and data structures is best exemplified in
document preparation systems as Latex or SGML, where the user can decide to
stick to well designed and useable document types or to develop his own and
still be able to communicate them accompanied by his document type
definitions. Evidently this argument applies also to the use of these
systems on non-open platforms, but it seems to me that on a GNU/Linux
system the computer literacy necessary to use them proficiently can best be
developped.
The user's opportunities to manipulate digitalized knowledge is
intrinsically linked with his freedom to communicate it and his choice of
tools in constructing it. Open formats for ASCII files can be accessed and
manipulated by a large range of tools, developped by users and available in
many different online forums.
Working with open standards on an open computer system gives the user the
possibility to examine and also to formulate Habermasian validity claims
for communicative actions - and I think constructing and sharing
knowledge have to be seen as communicative action. It is often objected to
this line of argument that this competence is not accessible for the
average end user. But I think that this objection can be refuted.
Freedoms always exist in principle, and actual hinderances to make use of
them do not reduce their importance. The user of free software can decide
to engage in this criticial examination by acquiring the necessary
technical expertise. But he can also decide to delegate responsibility for
specific questions to trusted experts, and in every community of practice,
every member implicitly gives trust to many others. In the case of free
software this trust can be a free decision instead of the lock-in that
results from closed and proprietary data formats.
Thus, an individual learner can reap the benefits of open-source
computer-supported learning systems, if he can take part in a community of
learners where the examination of validity claims and the definition of
communicable and customized data formats is a collaborative
undertaking. Digital knowledge becomes the object of, simultaneously, social
debate and individual engagement.
Jürgen Habermas.
Theorie des kommunikativen Handelns, volume 1.
Handlungsrationalität und gesellschaftliche Rationalisierung.
Suhrkamp, 1981.
Trans. by Thomas McCarthy: Theory of Communicative Action. Vol.1.
Reason and the Rationalization of Society. Boston: Beacon Press.