> On Oct 22, 2018, at 10:05 AM, Joosen Vanessa <[log in to unmask]> wrote:
> Dear TEI-L,
> For a research project on the construction of age in British and Dutch children’s literature my colleagues and I are looking for the best way to annotate fictional texts with information about speakers and subjects. We are all new to coding with TEI and have the feeling we are making things more difficult than they can be.
> These are the issues that we are currently struggling with:
> 1) we want to tag information about characters who use direct speech: we want to add their age and the certainty with which the age can be assigned. We can identify the speaker with the who-attribute, but we still need to find an easy way to add the speaker’s age, which may vary within one book and thus cannot be included in the external description of the speaker.
I’d use particDesc to describe the character. To solve the problem of character who age during a narrative, I can see multiple possibilities:
1 Give an age range for the character, in order to avoid the tricky question of granularity of age. (In the dialog of a single scene, each time a character speaks that character will be a few seconds older than before — I would not like to have to indicate a character’s age with that level of precision.)
2 In cases where the project feels able to specify distinct age estimates for a character at different points in the narrative (e.g. because the narrative skips forward several years in time), I would be inclined to use multile particDesc elements for the character at different ages, with @sameAs or some other linking attribute to signal that they are distinct descriptions of the same character. I would then use the @decls attribute on the largest possible containing elements in order to associate the applicable particDesc elements with portions of the narrative.
In some cases, of course, it may not be possible to segment the narrative in this way. In one of Ray Bradbury’s stories (the title of which I cannot recall), a time-traveling protagonist encounters himself as a boy and tries to give the boy advice. At every point in the relevant scene, the character is present both as a boy and as a man, so both particDesc elements will be relevant.
> 2) if passages give information about characters, we want to identify those characters, their age and the certainty with which the age is assigned. Is there an element to annotate the subject of a passage to which we can add a character’s name and age?
I’d use a @ana for this. You might create an ‘interp’ element (or just a note, if you wish) to capture the notion of a “passage giving information about characters”, give it an ID (say ‘pgiac’), and then use @ana to point to it, as well as to the relevant particDesc elements: the paragraph
<p @ana=‘#pgiac #Malcolm’>Malcolm was the landlord’s son, an only child. …</p>
is then identified explicitly as a passage giving information about the character Malcolm.
If one preferred, one could make distinct interp elements for passages giving information about each character: <p @ana=“pgia-Malcolm’>Malcolm …</p> — on the whole, don’t think that that approach has any great advantages over the alternative.
> 3) if the passage, either direct or indirect, gives a statement about a specific age or age group, we want to annotate this information as well (similar to the previous question but here the subject is a life stage instead of a character who inadvertently has an age). Any suggestions here?
> Are there tags and respective attributes available that we can use to annotate all of the above information? We have looked for them, but may have missed them. Or should we look into creating our own tags and attributes in a customized schema/ODD? If so, is there a way we can get help with this? We would like to have this done as quickly and neatly as possible and we feel we lack the experience to do so. Any advice or help you can give us would be much appreciated.
It is certainly possible to create your own element and attribute types in a TEI customization. They can make certain kinds of annotation more convenient or more direct, and they can make it simpler to check your data for consistency and (some forms of) correctness. It’s not obvious to me that custom element or attribute types will provide much advantage to you, based on my possibly faulty understanding of your description.
It may matter what you are hoping to do, or make it possible for other people to do, with the markup. I guess the initial requirement is to be able to find passages which provide information about characters, and in particular passages which give ages, so that you can more easily look for patterns in the way people of a given age are described. The use of particDesc and @decls and @ana should make that relatively straightforward for anyone with basic skills in XPath.
I hope this helps.
C. M. Sperberg-McQueen
Black Mesa Technologies LLC
[log in to unmask]