(tlg 0006, 0011, 0019, 0085) Omitted paratexts (lists of the dramatis personae and argumenta)

[juliajbeine]

Dear @lcerrato and @AlisonBabeu ,

going through the Greek drama files, I noticed that you provide the text starting from the first scene (e.g., the first episode). The preceding title page, argumentum, list of the dramatis personae, or title header are usually not included in your files.

For example, see this PDL file and this source edition.

I was wondering if there was a reason for omitting these text parts from the source edition. For instance, for GreekDraCor, we could generate the castList element from the list of the dramatis personae, etc.

My colleagues I would be grateful if you could give us some feedback on this matter.

[lcerrato]

Hi @juliajbeine
This predates my work, but I believe the editors followed the TLG convention and it was not the usual practice to include any later editorial interventions such as the arguments or cast lists. This holds true for Latin as well.

[juliajbeine]

Hi @lcerrato
Thank you for your quick reply. I was wondering if you plan to add these paratexts in the future or if you stick to the described practice. In GreekDraCor, we plan to revise and update the files based on your repo, thus, that information would be relevant to us.

[lcerrato]

@juliajbeine
I would be unlikely to add these in the near term. If another editor wished to add them, I'd be happy to review and merge.

EpiDoc is not very friendly to drama so I don't know what would best in this case.

This tends to affect things like word counts so there would have to be a practice that distinguished these from the main text.

[juliajbeine]

@lcerrato, I discussed this with my colleague @cmil, and we will add and encode the paratexts in front elements, which would allow you to distinguish the paratexts from the main text for the word count. We will create a pull request for you to review once we are done.

[lcerrato]

@juliajbeine I don't think front matter can be used in EpiDoc/CTS. Other such things are generally put into notes, I think. I would suggest trying one text, but front/back matters is a widespread issue with all Perseus/OGL texts.

[juliajbeine]

@lcerrato, okay, let’s try one text first, then.

[juliajbeine]

@lcerrato, as a test, I added and encoded the paratexts of Euripides’ “Heraclidae” (see #1736). The front element seems to be generally allowed in EpiDoc (see https://epidoc.stoa.org/gl/latest/ref-front.html). In the front element, I encoded the argumentum in a div type="textpart" subtype="argument" element. Moreover, I encoded the list of the dramatis personae in a castList element, but the castList element seems not to be allowed in EpiDoc. Is there another suitable tag for the list of the dramatis personae?
I am looking forward to your reply.

PS: oXygen seems to have changed the breaks in the XML file, which might be confusing in the diff view on GitHub.

[juliajbeine]

@lcerrato, I opened a new pull request for a version without the breaks previously added by oXygen: #1737.

[michael-roe]

I noticed this issue as well.

This is basically an editorial decision on the part of the Perseus project as to what to encode, and I can sympathize with their reasons for doing it.

For an experiment with large language models I was thinking of doing, it would have been useful to have the front matter as well. (E.g. are large language models better at answering questions about a text if they have access to a plot summary and list of characters as well as the body text?) This one use case doesn’t seem significant enough to justify including the front matter, but maybe there’s a lot of people out there who want the front matter for other reasons.