Kristóf Nyíri sent us two papers as a starting point for a discussion. I want to make two comments referring to those papers. The first comment deals with the relationship between the technology of writing and knowledge representation in general, and the second one with hypertext (as an example of a technology of writing) in particular.
1. The Target Paper. Here I will comment on points (3) and (4).
I think it is quite obvious that the technology of writing influences human forms of knowledge representation. On the other hand this relationship seems to be rather complex. Many features of knowledge representation which are supposed to be typical for the electronic age already existed before the introduction of computer technology. Kristóf himself often quotes Musil or Wittgenstein as early examples for networked, distributed thinking and writing. Apparently, the influence of technology on the process of writing is not direct and one-dimensional. There might be something like a philosophy of networking influencing both the technology of writing and the forms of knowledge representation. Rand Spiro, a well-known scientist in the hypertext community, argues that hypertext technology is a method for coping with the complexity of modern life. In the context of his theory, complexity, and not technology as such, is the reason for changes in knowledge representation. Complexity is, then, regarded here as a sort of philosophical idea which brings about technological and intellectual changes at the same time.
2. In his paper ``Electronic Networking and the Unity of Knowledge'' Kristóf writes: ``The now fashionable idea of treating all retrievable information as mere raw material out of which users might freely establish their own preferred hypertext structures is an acceptance of the fragmented state of knowledge, not a solution to the problems it creates.''
I agree with Kristóf insofar as it is one feature of Hypertext that it enables readers to find their own paths through the jungle of knowledge. In some situations, this can be very confusing. Nevertheless, Hypertext is more than that. Hypertext also enables its authors to represent the structure of a text more explicitly. Overview cards in a graphical form can fulfill this function. If authors use typed links they have to consider very carefully the exact nature of the links they want to create. If a hypertext document is well structured, such a form of representation can give the reader an overview of the material presented that is much better than that provided by traditional books. Additionally, such an approach forces authors to consider the overall organization of their text to a greater extent than is the case with traditional linear text, where relationships between different parts of the text are very often represented merely implicitly.
Hypertext is still a very ``young'' medium, therefore it remains to be seen if future hypertext documents will be a fragmented and disordered mess or a well-structured but flexible object. It seems quite plausible that the two kinds of development will each affect human reasoning in a different manner. I think this shows another major problem when talking about the influence of electronic networking on knowledge representation - there are many different ways of using a computer.