Need of @towhom attribute

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Need of @towhom attribute

José Calvo

Dear TEI community,

I have been thinking a while to write this email and while attending the DH17 Conference in Montreal I decided to do it finally.

Similar to the attribute @who, I think there should be another one for the target of the communication: @towhom, or @targetperson or @corresptarget or something similar.

This is not only a theoretical need. I have been encoding in the last years the Bible in Spanish, marking (together with basic structural information like books, chapters, pericopes and verses) references in names to people, places and groups, and the direct speech. At the beginning I only marked who was communicating and how (oral, written, prayer, dream…). But the text was pointing me out very clearly that there was something missing in my encoding: to whom. In the majority of the cases is not only true that is clear to a human reader whom is being talked to; there are also very clear references normally in the previous verses. One example (Mat 3.13-14, here in English) of John the Baptist (#per18) talking to Jesus (#per1):

<ab xml:id="b.MAT.3.13" type="verse" n="13">Entonces <rs key="#per1">Jesús</rs> vino de <rs key="#pla7">Galilea</rs> al <rs key="#pla9">Jordán</rs>, a <rs key="#per18">Juan</rs>, para ser bautizado por él.</ab>

<ab xml:id="b.MAT.3.14" type="verse" n="14">Pero <rs key="#per18">Juan</rs> procuraba Impedírselo diciendo: <q type="oral" who="#per18" corresp="#per1">Yo necesito ser bautizado por ti, ¿y Tú vienes a Mí?</q></ab>

Until now, I have used the attribute @corresp, which is not an awful decision but probably not the best that it could exist. Of course this attribute shouldn't be mandatory, there are many times in which the people talk without a clear target or we don't know who it could be.

If the Bible is not your kind of text, I have other examples from The picture of Dorian Gray, encoded in a similar way in this GitHub Repository:

<ab type="direct-speech"><said who="#BasilHallward" corresp="#HenryWottom">Yes, that is his name. I didn’t intend to tell it to you.</said></ab>
<ab type="direct-speech"><said who="#HenryWottom" corresp="#BasilHallward">But why not?</said></ab>

Now, at the DH17 Conference there was a great number of papers and panels about networks applied to different areas, one of them literary texts. I can mention projects using networks for representing interactions between characters in USA, Canada, Germany, UK, France, Spain, Switzerland, Ireland, Russia… And these are just on the top of my head. Networks are on fire. When these projects discuss what format should they use for the annotation of this information, the arguments for TEI would be stronger if there was nice attributes for that and they didn't have to customize it. Therefor we would have more data in TEI.

I wanted to share it with the list and hear your opinion about it.

Best regards,

José Calvo


Digital humanities and Spanish Philology
University of Würzburg, Department for Literary Computing

www.MoreThanBooks.eu