I’m a bit late to the party on this, but Emily Short is attempting an interesting experiment in interactive fiction construction. Alabaster is a ‘fractured fairy tale’, retelling the story Snow White with a halloween flavour. The interesting thing about the project, however, is its mode of composition.
To quote from the introduction to the game:
Alabaster allows you to play along, but also to create your own conversation when you find things that the game does not cover. Whenever you ask a conversation question that cannot be answered from the data in the game, you will be invited to write a new piece of conversation of your own.
If you accept, you’ll be asked to provide dialogue for the player and for the interlocutor, and to answer some questions about how you expect the conversation to flow.
The new text is stored in a file and users are invited to send these files back to Emily for inclusion in the game. Several iterative releases have been made, incorporating more and more user material (to the point where the game is getting somewhat slow in responding to input).
Does it work? Well, as a player the current release feels somewhat awkward and undirected — but that is to be expected of any work in development. I don’t think Emily expects it necessarily to be regarded as a complete work at this stage. Her plan is to freeze it in a more few days and add some finishing touches thereafter. So I will reserve judgement until then.
As a contributor, I find myself resistant to making changes to the game. I didn’t know the creative direction of the project and how my additions would fit with the rest of the game. If this were a pure CYOA, then my additions could simply exist as an isolated branch and there would be no need for concern, but it this is not the case here. Adding a new ‘quip’ does not immediately close off other responses in the text. If I made contributions without taking the rest of the text into account, it would become incoherent and schitzophrenic. Inevitably I need to refer to the source code to know how to best make my additions, which is something I think Emily is trying to avoid.
I have tried a similar but slightly less amibitious experiment on LambdaMOO. The rooms that I have built contain a small amount of ‘spy’ code which records the failed commands of visitors. Commands that are not implemented still give the normal “I don’t understand that.” response, but they are logged for my later perusal. From time to time I check these logs to see if there are any interesting additions to make. I have found this to be an excellent way to gather synonyms for common actions and to inspire new ways to interact with my work. Of course, I have the advantage that the work is online and so I don’t require my users to send me files, but I’ve often wondered whether something similar couldn’t be done in Inform.
The difference, I think, is that I am retaining sole authorship of the work and my world is more of an “IF Art” piece than an interactive narrative. I think it would be very difficult for multiple authors to write a coherent story in this fashion. The danger would always be for it to split into multiple independent single-author threads.
I am reminded at this point of the Lexicon game. I have seen instances of this game which managed to create a unified text, but even so the ‘narrative’ of these texts are mostly implied and their ‘encyclopedic’ nature explicit admits multiple points of view. The rules of citation require the writers to constantly refer to and extend each others’ work, which helps to act as a unifying force.
I have a personal motto: “The one on the ground should not contradict the one who is flying.”, so I applaud Emily’s experiment and shall reserve judgement until it comes to ground. Good flying.