Monthly Archives: February 2015

How to identify (and to cure) an error in the database

Following our advisory board workshop and the valuable feedback we received from the participants, we have recently been able to finalise the database, thanks to the successful work by our project’s Digital Humanities Developer, Matthew Groves.

As one can expect, there were numerous considerations that surfaced only once the database was at a stage where I have been able to create the first prosopographical entries. Thus, the HTML-based designer system allows different methods of entering data, such as a multiple-choice format, a multiple-entry one, a free text option as well as the possibility to choose from, and add to, several lists. This naturally leads to further questions, such as of how to define the differences between unknown and n/a (say, if it’s a special category that is applicable to few cases only, and there really is no information extant in the sources to answer that question at all). How to deal with a data entry system that requires one to enter a numerical time-span, when all we can really say is that at some unknown moment in time someone writes a letter addressed to someone else. All this is completely independent from the difficulties in the interpretation of the sources. The combination of these considerations naturally required different review sessions.

In this context, a hilarious obstacle occurred that for several days delayed completion of the database. While entering more and more data for a given exile case, at one point a message popped up and kept on preventing the data from being updated to the system: ‘An error has occurred during the creation of item “Eusebius of Vercelli”’. After extensive investigation, we found that the message came up every single time when I tried to update the field ‘Religious affiliation’, with its associated content (‘Nicaean’), and that this had caused the database to reject the new data as erroneous. It can’t be that the database is heretical, can it?

At any rate, we are scheduled to explore the possibilities of visualising the early database content soon and are therefore convinced that at this stage there is want for further clarification and assurance that we are still on the right way.