On 14 May 2024, I published my Index to Roscher’s Lexicon of Greek and Roman Mythology on Zenodo . While it is already useful in and of itself, my main goal with this is a matching between Roscher’s entries and databases about Greek, Roman and other mythologies. To this end, a matching with Wikidata seems like the best step as this project already incorporates links to Theoi.com, Myths on Maps, ToposText, MANTO, Mythoskop and other mythological databases.

Searching for Roscher’s 15,780 headwords and finding the closest match in each case (if one already exists) is a tremendous task for any single person, which is why I have decided to take two different and somewhat complementary approaches. One is to create a dataset from Roscher’s Lexicon from scratch, and I have chosen FactGrid for this (as outlined in my blogpost from December 2023). The other involves Wikipedia, particularly the German language edition as it uses Roscher’s Lexicon as a reference in more than 2,200 articles. The German Wikipedia’s Liste der Gestalten der griechischen Mythologie is even more copious: After its initial creation in 2004, it was continuously edited until in the summer of 2019, a Wikipedia editor by the name of ELexikon took it upon himself to expand the list using Roscher’s Lexicon as a reference. ELexikon added links to Roscher’s Lexicon as well as Pauly-Wissowa articles and in some cases, even references to primary sources to the the list and greatly expanded it, listing not only individuals which already had a Wikipedia article but also those who haven’t yet. The result is, basically, a reduced index for Roscher’s Lexicon, augmented with Pauly-Wissowa references, which may serve as a To Do List for Wikipedia’s volunteer editors. Lots remains to be done there, so if you can spare the time, you are welcome to write a few articles from there.

Apart from the Liste, Roscher’s Lexicon is used as a reference in more than 2,200 Wikipedia articles. These cover divine epithets, deities, mythical persons, monsters, concepts (like metamorphosis or emperor’s cult), groups and pairs of mythical characters and so on. At my request, a helpful Wikipedia editor by the name of Wurgl created a list of all occurrences where Wikipedia employed Roscher’s Lexicon as a reference, complete with corresponding Wikidata IDs.

I then edited this list to check if all matches were truly referring to the same concepts. Sometimes, a number of different Roscher articles were given as references in a Wikipedia article. In these cases I chose to find the closest match and ignore or excise the others. Out of the initial 2,388 Wikipedia articles, 2,342 remained that were close matches to the Roscher articles they were quoting. I added the Wikidata IDs from these 2,342 articles to my Index File (Tab A, column P) where up to that point only 28 IDs were recorded. In the same process I discovered that I had overlooked three articles (Thaleia 3–5) which I instantly added to my Index File.

For a workflow which will take new Wikipedia articles into account, I am conferring with ELexikon. Until then, having 2,380 out of 15,780 articles matched to Wikidata is a good start. More importantly, it yields a few interesting lessons in terms of data modelling: Not every Roscher article has a straightforward match on Wikipedia / Wikidata, because

1. not every Roscher article deals with a clear-cut subject (e.g. “Sternbilder, Sternglaube und Sternsymbolik” – roughly matching asterism and astrology);
2. a few Roscher articles only deal with certain aspects of a subject (like “Herakles in der Kunst“, “Thespiaden in der Kunst“, “Mars in der Kunst“), or restrict it to a specific cultural context (like “Sterne (bei den Babyloniern)“).
3. some Roscher articles split up mythical persons which already in antiquity were supposed to be variants of one and the same character (cf. Hyrtakos, who is treated as a single character across 12 language editions of Wikipedia, like in Pauly-Wissowa, but split in two by Roscher)
4. groups and pairs of mythical persons are not dealt with uniformly in Roscher’s Lexicon or Wikipedia. While some groups have an article for themselves, others (like Kratos and Bia, or Philemon and Baucis) are treated under one of the two names with the other one just serving as a cross-reference. The most straightforward way to deal with this is to find or create Wikidata items for the parts of the whole and link those as the closest matches.

All of these aspects command careful consideration. For somebody with limited time but lots of love for Greek mythology, this is indeed a Pandora’s box. And knowing myself, after peering in it I cannot and will not put the lid back on.