Subject Classifications

An important consideration for our Books and Borrowing database is how useful it will be for researchers. Not just us, but anyone who would like to use our data in the future. One question our team has been working on is how to classify the thousands of entries our database will contain.

Our original vision was to have a system that would honour early modern systems of knowledge and create an infrastructure that an enlightened library keeper travelling forward in time would recognise. Our virtual shelves would be a super library, bringing all of our partner libraries into one place. And, importantly, it would follow the principles of Naudé, use the intellectual mapping of knowledge detailed in the Encyclopédie, and consider the bibliographical conclusions of Dibden in the early nineteenth century. We discounted modern systems of classification as too complex and not in synch with how our readers would have perceived their books.

Classifying Knowledge in the Encyclopédie

To this end, we created a comprehensive spreadsheet. The top end categories would reflect the traditions of the eighteenth- and early nineteenth-century library while more detailed sub-categories would provide nuance. We modelled our ideas on the occupations system that we have adapted from the work of our colleagues at the Atlantic Worlds project. With this system, we can select multiple tags for our borrowers and it seemed a good place to start with our subject classifications.

The classification spreadsheet looked beautiful (if you like that sort of thing).

Then we tried it in practice.

In a ‘Classifications Challenge’, we took examples from libraries that we are working on and circulated them to the team. Where would each of us put a certain title? Would we agree on how to classify texts? Surely the large amount of sub-categories would offer clarity?

The Zoom discussion was lively. It showed that the complexity of the proposed structure would not work in real life and that our differing perspectives meant that we placed the some of the same books in different categories. Also, we soon realised that we certainly would not have time to read every book under discussion in order to make a solid judgment.

We turned to Professor Andrew Pettegree of the Universal Short Title Catalogue (USTC) for advice. Prof Pettegree generously explained how the USTC’s subject classifications came about and it was clear that for projects of our scale, less is more.

Our dream of a modern universal library was over. We re-visited and streamlined our system. Our system now comprises twenty-one main subject classifications and, as a compromise, we will use up to three categories per title. This will be most useful when trying to assign collected works to categories. Our ‘what goes into’ each category is still comprehensive, but we won’t be using the sub-categories as individual tags.

We won’t know until much later in the project if we have been successful at identifying the necessary categories for our borrowing records. We have decided to leave the work of assigning categories until closer to the end of the project when we have a better idea of the sorts of books our borrowers favoured and what they might have thought of them.

List of subject for Books and Borrowing in wordcloud form


T. F. Dibdin, The Library Companion; or, the young man’s guide, and the old man’s comfort in the choice of a library (London, 1824)

Jonathan I. Israel, ‘Libraries and Enlightenment’, in Radical Enlightenment: Philosophy and the Making of Modernity, 1650-1750 (OUP, 2002; Oxford Scholarship Online, Oct. 2011, doi:10.1093/acprof:oso/9780198206088.003.0006)

Gabriel Naudé, Instructions concerning erecting of a library, tr. John Evelyn (London: Printed for G. Bedle, and T. Collins, 1661)