0.32 Scholarly information management: a proposal  (Page 2/11)

The model I propose is called Scholarsource and it is divided into three parts, which correspond to three subquestions: how is it possible? who can do it? and how can the information be organized and managed? The three corresponding parts are:

1. The Conditions of Possibility of Scholarship
2. Scholarly Communities on the Web
3. Scholarly Information Management

The three parts of the model are each imbued with a different status. The first point—the conditions of possibility of scholarship—is a necessary requirement for any environment that aims to support humanities scholarship. The second and third points, on the contrary, indicate only a possible way to realize scholarship in the digital era and can be replaced by different strategies. In some ways the first point is more philosophical, while the second is a sociological one and the third is more technical; but, as usual, these disciplinary distinctions are not very precise.

Conditions of possibility

Borrowing the phrase from Immanuel Kant, but using it in a non-Kantian sense, “conditions of possibility of scholarship” is used here to mean the principles that undergird scholarship. It is those general rules without which either our infrastructure will not function or it will produce something different from scholarship. By way of explanation, let us have a look at a successful example of transposition of a traditional activity into a digital environment: eBay. eBay is a digital infrastructure for selling and buying. It was made possible because its inventors could identify and reproduce in a digital environment all of the key requirements for a successful business relationship, that is trust—trust in payment and in merchandise delivery. Once successful in ensuring trust, additional features could be added, such as price comparisons, email reminders and advanced searching. Without the trust rating system, though, all of the additional features would have been useless because people would probably not have used eBay at all. Research infrastructures in the humanities have in many cases been driven more by capacity than by exigency, with each advance in technology inspiring a new set of aspirations and plans and producing new sophisticated features. Before adding additional features, though, we have to ensure that the key requirements for scholarship are fulfilled. In other words, we have to start by identifying the conditions necessary for conducting scholarship, and only then will we have the basis upon which to develop and evaluate digital infrastructures for the humanities. Three of these requirements—Quoting, Consensus and Dissemination/Preservation—will be discussed here.

Quoting is the first requirement for the activity we call scholarship. Scholarship is a conversation based on hypothesis, arguments and facts. We should remember that facts, in the humanities, are often contained in documents, in texts. Emma Bovary drinks arsenic and dies. That is a fact. What exactly this fact means might be a matter of interpretation and dispute, but that she “drinks arsenic” is indeed a fact. But to be sure of this, you must be able to consult and quote the first edition of Flaubert’s Madame Bovary . Quoting requires stability of bibliographic references and, most of all, stability of texts. Printed texts can normally ensure both of these. But what about on the Web? On the Web, this type of quoting can be quite difficult. Web pages change every day, appearing, disappearing, reappearing under other names and addresses. Nevertheless, it is certainly not impossible to create special systems, like little islands in the Web, to ensure the stability of electronic documents and web addresses. Technical solutions exist. From a technological point of view, the URL/DNS technologies are perfectly sufficient to ensure the stability of web addresses There are plenty of initiatives to ensure the stability of Web addresses, such as DOIs (Digital Object Identifiers) and many more. This is a hot topic in the librarian community. My personal opinion is that trying to tackle this problem by inventing a new naming system is fundamentally useless because the URL (Uniform Resource Locator) or more precisely, the URI (Uniform Resource Identifier) can already identify documents in a stable and univocal manner. What’s more, DOIs are managed by a commercial organization that has the same if not more chances of disappearing as each single repository that manages its own URIs; anyway, it doesn’t give any more guarantee of stability than IANA (Internet Assigned Numbers Authority) and it is a contractor ICANN (Internet Corporation for Assigned Names and Numbers) who manages IP addresses and domain names. In the end, DOIs are “identifiers of identifiers” that just shift the problem to a new layer, and their supporters seem to ignore the story of the anthropologist and the Indian: “What does the world rest on?” the anthropologist asks the Indian. “The Great World Tortoise.” “And what does the Great World Tortoise stand on?” “Another tortoise…” and a simple checksum system is able to verify that documents were not changed over the time. But the existence of technical solutions alone is not sufficient. To create an island in the Web where documents and their addresses are stable requires both the decision to install these technologies and a strong commitment not to alter them over the time. We already have the technical solutions, but we need coherent scientific policy decisions. Digital libraries should simply have the same policy as that of prestigious traditional libraries, which are not used to lose or alter the content of their books, or to change their signatures (at least not without writing a table of concordance with the old one).