[Note: part of this assignment was to devise a fictional organizational context for the project. ]
As members of the Society for Dramatic appreciation, it has become more and more obvious that undergraduate students at this University do not have the appropriate tools to interact with the dramatic works they are reading, studying, and watching in their classes. Knowing this, we have begun work on a a text encoding project that will facilitate the interaction between the student, instructor, and the dramatic text.
To begin our project we narrowed our first effort to marking up dramatic works by female dramatists writing before 1913. Currently, the little know works by female authors is a "hot topic" in academia. We hope that by choosing these items we can appeal to a select audience and then be able to open up the project to more canonical works.
To promote better understanding of drama among college undergraduates by encouraging multidimensional interaction with dramatic text.
Choosing how a dtd should look can often be a challenging, long process. One of our first choice when in the modeling process was whether to have a strict or more flexible dtd. Dramatic works have many standard and familiar elements that occur in almost every play. Knowing about a play's complex, hierarchical structure would initially make a person think that the dtd needs to be more strict than a dtd for another literary form would be. After some consultation, the production team for the dtd for drama decided to make a more flexible dtd. The reason for this decision was to support works of creative expression and not hinder the creative work that is playing with structure in non-traditional way.
Many modeling decisions go into creating a dtd that will apply to dramatic works. As a creative form, drama is structurally very complicated and interesting--drama is a work meant to be acted on stage, but the instructions for acting is written on paper and later published. Converting these works to an electronic format again changes the medium of this versatile work. Modeling a dramatic work can be an intricate process of the variable nature of the structure and content of a dramatic work. Challenges in modeling From the beginning, our group decided that we would rather make a looser dtd that allowed the creativity of the playwright to stand out instead of making a strict dtd that would stifle a writer's way of expression. Our first hard decision to make was how we would break up the foundational levels of our dtd. The root element, "play", has four children: "head", "front", "text", and "back".
In any project the inclusion of metainformation can help with information retrieval and general documentation/project record keeping. The "head" element is a container for metainformation tags. In order to make our model more understandable and transferable to other projects and organizations, we decided to base our information on the standards set forward by DublinCore. The names of the tags included in the head will be based on the names of the DublinCore elements. Currently, this portion of the dtd is being put aside and will be modeled more completely later. After the inclusion of the metainformation, the actual markup of the text will occur.
The structure of a play or dramatic work can be broken up into the front matter, the body of the text--that includes the actual play, and the back matter. Structurally these different sections seem discreet and by dividing them up it may be easier to The "front" tag is used to represent a work's front matter. The front matter includes distinct general sections that help to enhance and clarify the work within the body of the text, such as cast list or prologue. The "text" tag is a container for the structure of plays, most specifically acts and scenes. The tag "back" encloses different sections of the back matter which may include epilogue.
While modeling these first two levels of the drama hierarchy it soon became apparent that each of these three distinct structures of drama would include some specific sections of text that only applied to each specific structure, but there were also some other sections that were rarely used or could be located in different structures (confusing sentence). One of the largest decisions we made was to include a "section" tag that could be included in each of the three levels--"front", "text", and "back". "Section" is also a child of the "text" tag on the same level as "act" and "scene". When the "section" tag is chosen the encoder will then fill in what type of section actually is, for example, postscript or chorus. While this chose may allow for more encoding errors and freedom, this decision conforms to our original choice to have a looser dtd. This flexibility, especially within "act" or "scene", allows for encoding situations like in Shakespeare's Hamlet or A Midsummer's Night Dream where there is a play found within a play.
One place the "section" tag should not be used is for a cast list. The castlist functions in a different way than other generic sections. Cast lists are exclusively located in the front matter and can be broken down into smaller divisions of information. For these reasons, the element "castList" was added to ensure that the unique structural and informational unit of the castlist will be preserved, while other sections that are not so structurally limited can be used with the generic "section" tag.
After these basic and foundational structural sections were outlined more debatable and specific choices were modeled. One example is the "line" tag. Choosing what to do with the actual dialogue found within an dramatic piece was a hard decision. First of all, what should it be called? Are these paragraphs or lines? Is the dialogue written in prose or verse? How does this distinction change the way a play would be marked-up? All these questioned were asked and our solution was to look at the dramatic form from the perspective to an insider. When an actor learns his or her part in a play, that person learns her or his lines. In that case, each individual section of spoken text should be called a line. Preserving the language of the dramatic community was a main concern when establishing our tagging guidelines and the use of the term "line" indicates that concern. Within a line there are many things that should be identified. Some options that we discussed were marking up a line and having the speaker of that line as an attribute or using the "line" tag as a container. For presentational issues, we chose to use the "line" tag as a container because it may be easier to pull out the speaker's name and place it in the left margin in a way that imitates printed dramatic works. After choosing to wrap the name of speaker of a line in a tag named "speaker", another discussion, probably our biggest argument, was about if spoken part of the line should be wrapped in a tag as well. One argument was that it was unnecessary to wrap the spoken part of the text within the line because it served no function. Another argument was that wrapping the spoken text was not only necessary, but tidy and general good mark-up. In the end it was decided that the spoken text should be wrapped because of the hierarchy. If the spoken text were not marked, then the "speaker" tag would be on an entirely different hierarchical level than the rest of the material in the line. This structure would show the encoder that the spoken part of the text was of less importance and would do injustice to the genre. In the end, it was decided that the tag "line" has the children "speaker" and "spoken".
Part of the objectives of this project was to include sections that would help facilitate professor to student communication. The way we incorporated it was to include "note", "comment", and "explication" tags. Each of these tags, while seemingly the same, serve distinct functions. The "note" tag is used to express footnotes that may have been present in another version of the text being encoded. The tag "comment" is a professor's thoughts and interpretation about the text. The "explication" tag is also one made by the professor, but this one is for basic explanation about what is going on and not suppose to be a professor's opinions. By including these three, our group was able to fulfill our goals for the project.
Another modeling issue that was brought up was about the "title" tag. At first, we thought "title" would only be used at the beginning of different structural sections, but after close examination of different dramatic works, it became obvious that they could be scattered at different parts throughout the text. This made us change some of our original modeling.
Where to put mixed content is something else that we had to consider. While there was an argument to be made that at the act, scene, and section level there should only be element content, this seemed too restrictive. Especially at the section level, which can be used in the front and back matter to note sections of the text such as a dedication, first performance, and other material that is distinct from the body of the play, forcing all content into elements seemed insufficient. We added a paragraph element in part to cope with this, but even still decided to allow mixed content at the section level. We decided that especially during encoding, it would be better to allow mixed content and possibly tag it later, than to have too strict a model and force tag abuse or the removal of text.