0.32 Scholarly information management: a proposal  (Page 10/11)

To understand why the HyperNietzsche interface was not satisfying, let us try to consider the principles on which it was built: we will see that the difficulty here was probably not the design of the sidebar, but the organization of the content—that is, the general structure of knowledge that this design was expected to express. In the print culture, scholarly knowledge came under the form of well-defined genres shaped by the physical structure of the book: treatises, critical editions, journals, collected papers, catalogues, etc. The problem is that these genres stored in the same container heterogeneous kinds of information. For example, a critical edition contains in a single book several types of scholarly contributions: manuscript transcriptions, texts editions, philological commentaries, critical commentaries, cross references, bibliographical references, introductive or critical essays, and so on. From a logical point of view—and even more from an information technology perspective—this way of collecting and mixing different types of scholarly contributions is not satisfying because then it is difficult to query, assembly and redeploy them according to different purposes. In theory, digital technologies will allow users to collect and compare different editions or translations of the same texts, or to read all the philological commentaries concerning a certain text but excluding the philosophical ones, or to create a diagram showing all the cross references concerning a certain text, etc.; but for this to be accomplished the different kinds of scholarly contributions and their parts need to have been clearly distinguished previously. Otherwise, as happens in digitization projects like Google Books and many others, digital technologies cannot deploy all their possibilities and the user is only allowed to search words, getting endless lists of occurrences without being able to retrieve and compare the information he needs. In order to allow advanced scholarly information retrieval, in HyperNietzsche I established a scholarly ontology containing a catalogue of all the different types of primary and secondary sources used by Nietzsche specialists (see above, “Ontologies”), and built the database which powered the HyperNietzsche website on this ontology. In this way it was possible to perform all the kinds of queries needed for the dynamic contextualization, e.g., to retrieve all philosophical commentaries concerning a certain page of Nietzsche, or all reviews concerning a certain article, etc. As we explained, these queries would have not been possible without such a disassembling of the machine of scholarship into its constitutive parts that were hidden in the form of a book. But what we didn’t understand at that time was that this way of structuring information, which was completely appropriate to construct the database, could not be suitable for interface design. We reassembled the machine of scholarship in a fully hypertextextual way, transposing the logical structure of dynamic contextualization directly in the interface, abstracting from old forms of knowledge organization like editions or journals. This was not a good idea. The use, the manipulation, the construction of knowledge objects don’t depend on logic, but on history. Scholars cannot work well if their materials are organized in conceptual structures which are too innovative, too different from the long-term scholarships practices of work with objects showing a certain layout and presenting a certain affordability. Even when new media permit a different and more logical organization of content, at the beginning new media mimic the old ones: it is well known that the first printed books imitated manuscripts books, and the first CD-ROMs tried to reproduce the look and feel of printed books.