Abstract for "In Memory" (continued)

In the last post I gave a simple abstract and talked about two important problems to be dealt with regarding it; how narrative can be represented, and how the world should be manipulated. These problems should be seen naturally as something interconnected. We could view them as concentric circles, with the concept in the center, representation of narrative in the middle, and means of interaction as the periphery, but this runs into many problems, most notably that it marginalizes the importance of interactivity and the user alike.
More importantly, I would say that these are ill-defined. How the narrative is represented is very closely tied to the processes by which it is re-defined by the user. In turn, to create the right kind of representation will take some idea of what the player might be doing when exploring the world. Therefore, I imagine that it will require carefully looking at both at the same time, oftentimes going back and forth in altering the two problems. With that in mind, I will now address the issues:

Representing the Storyworld & Narrative

This is a heavy task in many more ways than one. Interactive storytelling has been written about a great many times, and very few of these books offer more than wish lists and general gobbledygook. Even these few have had a lot of trouble getting very much off the ground at all, meaning that any kind of a fully-equipped storytelling engine is not an option. However, there most likely is no need for an engine of this complexity. Instead of explaining this, I'd like to for now leave it at the fact that we are working only with images and how narrative is constructed with them. The concept may grow, but it's safe to start here and see what kind of system emerges, therefore I will start with the most basic solutions and go outward:

The most simple solution that comes to mind is one based purely on connections between images. If we imagine images as neurons or modules of neurons, depending on how precise or approximate we want to be, then all information and function is based on the connections between parts (this is admittedly presumptuous, so any neurologists out there, feel free to bash me). Similarly, what if we defined a storyline as an ordering of some selection of images. For those familiar with graph theory, the graph S would be the storyworld, the vertex-set I would be the total collection of images stored in the storyworld, and the edge-set N would represent the narrative created by the player. This is obviously an inadequate model, but it's a good place to start from, and thus I'll further elaborate before continuing.
How would the player interact in order to create this set? Already, there's a question of how the user's interactions are related to the data structure chosen. What images would be appropriate for a player to explore through, which ones could he collect to make a narrative out of? Perhaps there is a space he walks through with memories picked up as items. But then again, how does he put them in order? Is it simply in the order he picks them up, or does he determine the order even after he's picked them up?
Furthermore, how would this series of images be interpreted? Without any context, it's nothing more than a slide-show. How would context be taken into account? Using a ton of if-then statements regarding specific contexts so that no matter what order its in it would make sense would be a nightmare. Even a storyline of 10 images would mean a total of 10! (10 factorial) different total permutations to take into account. For the less mathematically inclined reader, that's 3628800 different ways 10 images could be ordered into a story. There may be some way to simplify the patterns, but nothing obvious. The idea that connections between images is the key is arguable, but for the time being, this over-simplified version of it should be snuffed.

Seeing how many questions about the user's interaction the previous proposal for a model made, it would probably be a good idea to take a detour and think a little bit about what the point of the user's interactions are. I won't name anything too specific, because in the end, what the player "does" in a layman's sense is an interface that should reflect the inner processes of the game. However, on a deeper level, it would be wise to look at how the player functionally interacts with the system.
The first question that comes to mind is "what is the player trying to accomplish?" Clearly, the player is not trying to just construct a narrative out of thin air. Well, if the player is jettisoned into a world with no precedent, and that world is the world of the protagonist's mind, then the player's goal is to re-establish a sense of self. To get that out of the way is both a load off of my shoulders and a doorway to a myriad of different possibilities for gameplay.
The next logical step would be to explore the possibilities for how the player could interact with this world to create a "self". From there, we can agree upon what units this can be measured in, and in turn once again look at possibilities for modeling the system. The problem is surprising me with how interconnected everything is, and I'm finding my brain somewhat tangled. Excuse me for a somewhat abrupt and vague ending, I'll return to writing after a little bit of time sleeping on it.