Designing Databases for Historical Research
A1. Database design concepts
After working through this Handbook it is hoped that you will have a good understanding of the complex relationship between historical sources, information and data, and will be aware of the translation processes that are required when moving from one to the others. The informational contents of historical sources need to be converted – often in multiple ways – before they can be used as data, and a number of methodological decisions will need to be taken as this is done. Unlike the more mechanical aspects of using databases in historical research, such as building tables, linking records or running aggregate analyses, this translation process is not only difficult to learn other than through experience, it is also likely to be a substantially different process for every historian doing it. Each historian has different materials, different projects and different research aims, and so the databases they build will (or should) address these in the way that best fits their specific purpose. This ‘modelling’ of historical data is a difficult process, but happily the difficulties that arise are those that by and large are faced by historians in their everyday, non-database, work, meaning that you will be well equipped to deal with them. The modelling of data is also possibly the most interesting and enjoyable aspect of using databases in historical research, although perhaps this is only a relative response to the long hours of data entry that follow the initial design of the database!
This Handbook does not require the use of database software, although it will show the occasional screenshot of a database for the purposes of illustration. Instead it will spend most of its sections on discussing sources and research questions, and how these need to be recast when interacting with a database with its strictures and rules. The exercises that are offered will not have right or wrong ‘answers’, just as there is no right or wrong way to design a database of historical sources. Or rather, it may be more accurate to suggest that while there is no right way to design a database, there are a number of (if not wrong exactly) unhelpful ways to design a database, and this Handbook will focus as much on the latter as on the former.