In addition to supplying a collection of pharmacopeias, many of which are not already available in libraries, the HP team is working to convert records into an actionable dataset. This dataset will consist, on the one hand, of record attributes and metadata (date, location, type), and, on the other, fully parsed text in which each word is associated with a higher order lexical entity and, ultimately, a supralinguistic semantic concept. By way of example, the inventory of Batrona Maurella (Marseille, 1428), includes an entry for "azafetida."
This is one of the local spellings of "asafoetida." Through a series of database linkages, the instance recorded in Batrona Maurella's shop inventory can be linked to the lexical headword "asafoetida." Since other languages have different ways of referring to the conceptual entity we call "asafoetida," all the relevant headwords can be ultimately be connected under the supralinguistic concept for the drug. Inter alia, these data relations will make it possible for scholars to search for a concept and identify all instances in the corpus, regardless of language or spelling.
We hope that the dataset will enable these forms of analysis, along with others that may emerge during discussion:
- Relative Frequency. The frequency of appearance of simples and compounds will vary over time and space, as products enter the medical marketplace or fade in importance.
- Geographic Distribution. The distribution of products, at least for functional pharmacopeais, is made possible by georeferencing the location of record. This technique makes it possible for researchers to trace phenomena such as the rising popularity of New World medicaments in Europe or the transregional movement of indigenous American medicines.
- Co-occurrence. This approach considers durable patterns that may emerge in assemblages as one considers which medicaments that systematically co-occur in lists of products.