0.32 Scholarly information management: a proposal  (Page 6/11)

1. Research Libraries . The most successful strategy is to dedicate some libraries to a single research topic. While a general library allows users to consult numerous collections dealing with a wide variety of subjects, the purpose of a research library is to focus on a single subject and to provide scholars with access to all the primary documents and reference works they need to conduct research therein.
2. Physical Arrangement . Open-shelf libraries often arrange their books in a way that puts the primary sources next to the relevant secondary sources. In the Dewey classification, for example, the critical essays on an author usually follow his collected works.
3. Cataloguing . Independently from the physical arrangement of the books, catalogues of primary and secondary sources or subject catalogues help scholars retrieve a relevant publication and relocate the information according to their research needs.

Digital libraries can do even better. Not only can they unite the collections of different libraries, but they can also easily reconfigure their holdings according to any scheme, taking into account the different status of a given source within different research contexts. If a scholar enters our network of semantic digital research libraries through the door of, for example, Plato scholars (using the Plato Source ontology), the information and resources would appear to him in a certain configuration. For example, Plato’s Dialogs would appear as primary sources and articles by Nietzsche on Plato would be listed under the secondary sources and be accompanied by other critical essays on Plato. But if the scholar enters through the Nietzsche door (using the Nietzsche Source ontology), the same material would be presented in a different way—with Nietzsche’s articles on Plato appearing as primary sources (within the class “published works”) and related to all critical essays and other secondary sources on Nietzsche (not on Plato), while Plato’s Dialogs would be included in the class Nietzsche’s “personal library.” So in this way we could transpose the structure of traditional scholarship onto the Web, preserving the different epistemic values and relationships which scholars attribute to their sources, and improving the way in which the documents can be dynamically rearranged according to these relationships. Furthermore, all digital objects would appear as generic resources having the same epistemic status and the user could search them using a minimum set of shared, standard metadata, such as title, author, date of publication, etc. In this way our infrastructure can be very specialized and targeted to the needs of specialist scholarly communities, and at the same time be fully interoperable with general digital libraries and aggregators. So the general library will serve all kinds of readers and ensure interoperability while the specialized research libraries (concerning Plato, Nietzsche, Wittgenstein, etc.) can permit scholars to find their way in an electronic environment structured according to the standard classification used in their communities. The same thing would, of course, presumably be possible concerning theoretical ontologies. A first example of the Scholarship Ontology has been produced and formalized within the HyperNietzsche and the Discovery project. Within the Discovery project were also written and tested some philosophical ontologies devoted to Nietzsche, Wittgenstein, ancient and early modern philosophy.