Warning, long rambly post up ahead: I Have been thinking a fair bit on where the research directions could go, and here is a run through of a couple of ideas which are in my head at the moment.
There are a couple of things to consider in terms of research or technical goals to accomplish.
Research goals are intended to answer a broad research question, and understand a larger picture by looking at various examples, and maybe developing a theory or process or method for doing something. The method in this case would be one for adaptation of certain types of narrative forms.
Adaptation is an extremely broad subject, and so it would be important to narrow the field down significantly. From the perspective of advancing the medium of games, we can look at narratives that are not traditionally adapted into games and enumerate those: dramas, comedy, social narratives, romances, etc. Even still, this is very broad, and we could do with something that is finer. Maybe not by limiting ourselves to a genre, but by limiting ourselves to a type of story or type of framing of a problem.
So, looking into a method for adapting certain kinds of social dramas is very potent. However, the examples would necessarily be very limited. They might be various, but they would need to be small in scale. Questions arise over how the method or approach would be technically shaped. I could see it as a software library, but a library would need to be focused in its scope.
A larger perspective look at the project would leverage knowledge from the Statics project, and exploring how people use models, and build and run simulations to understand things (physical things, social things, any abstract concepts). That would be a supercedent of the adaptation problem.
Still, it would also be possible to look deeper and more specifically, and try to accomplish some technical goals, say, “Build a Pride and Prejudice game,” but this is dangerous. While it would be important, innovative, and necessarily new, it’s dangerous in the sense that developing a method for this specific setting would not necessarily be generalizable and would introduce very challenging specific technical details that might not fit nicely with other components.
An example of a technical solution is Facade, which is very successful at being something new, but has posed many problems in attempts to extend the character behavior language, ABL, to other projects and domains.
Clearly, there is a degree of freedom here, there’s no real unique solution. We can do something very theoretical with limited technical accomplishments, or something very technical with limited theoretical accomplishments. It seems like a compromise is necessary here, but the question becomes what is optimal, and where does it fit in this larger picture?