Ovinomancer
No flips for you!
One thing that should be reiterated here in this excellent essay is that there IS NO CAUSAL PROCESS WITHIN THE FICTION. The fact is no such person as Sherlock Holmes, no person with characteristics similar to him, can exist in the real world. This isn't even limited by just ordinary physical constraints (IE nobody can focus their attention well enough or remember things so reliably as to perform the feats attributed to him). It extends to LOGICAL POSSIBILITY as well, fiction need not even abide by the basic tenants of logic. Things can both exist and not exist, be in two places at once, have mutually exclusive characteristics, etc. within fiction. Not only that, but this HAPPENS ALL THE TIME. Mostly we don't notice. We suspend disbelief and we simply accept the fiction's conceits as given.
There's nothing remarkable about this when we're talking about a fixed passive form of story where the reader simply participates by reading and imagining what is told by the author. However, when we get into RPG THEORY then its VERY VERY IMPORTANT to understand this! What it means is that the ONLY THING THAT MATTERS is who, by rule/convention/whatever, is able to assert elements of the fiction. There is no 'fictional causation', it doesn't exist, it is, at best, a convention to pretend that it exists, and that only certain participants are bound by it. It is this convention, the practice of RPG game play, which is the subject of RPG game theory, which is what we are discussing here.
Every time people talk about what is 'in the fiction' except as it pertains to how they will relate it to play procedures, is just not significant. What is significant is 'what are those procedures and how do they work?' In particular how does pre-authoring content work, why is it done, and what effect does it have on play processes? (since that was the question of the OP).
Most of the recent thread has passed me by, but I wanted to touch on this argument as it seems foundational to many.
I fail to understand how saying fiction has no causal process is remotely relevant when the argument then becomes 'except as stipulated in the rules of the RPG.' The first argument may be true (there's still an open philosophical debate as to whether fictional things are real), but it fails to remain relevant when it's then accepted that RPGs treat some fictions as causal by convention.
If the argument is that the exact fiction authored is meaningless and it's the authoring the matters seems very, very premature when you then limit the nature of what is authored by already authored fiction. If "fictional positioning" has any meaning, it's causal to what can be authored to the fiction at that point. If "genre tropes" has any meaning, it's causal to what can be authored to the fiction at that point. If you use fiction, even by convention, to constrain how new fiction can be authored (and even who can author the fiction, characters not involved in a scene have little input in most RPGs) then you're accepting that fiction has causal power.
I can see a rebuttal that takes the form: ah, but it isn't the fiction that does this, it's the agreement on rules that does this. And that gets around a good bit of it, but it doesn't account for the fact that we still check the fiction to see what's allowed. You can't say, for instance, "new fiction must not contradict the existing fictional positioning" and stop there -- you still need the existing fictional positioning, which is still fiction. The rules define which bits of fiction gain causal power over new fiction, but there's still fiction involved that is doing work. I don't see a coherent position that excludes fiction as having any causal power that then utilizes existing fiction to limit future fiction.
What am I missing, here? Because, at the moment, this seems like one of those things that sounds really smart and relevatory, but it actually isn't. If you want to say that fiction has no causal power then you cannot reference fiction as a limit of authorship, as that's using fiction to apply causal power. And you can't cut the fiction out of that by saying it's convention or rules, as the actual limits on authorship depend on the nature of the fiction referenced.