the vast bulk of
@pemerton's posts (or an oft recurring underlying motif) has mainly been challenging related ideas pertaining to GM vs. player authorship in sandbox play in regards to creating a consistent or living world.
<snip>
It is not that GM-authored is bad, but, rather, that player-authored is equally valid. He is primarily "hostile" about anything that questions the validity of his own player-driven play, particularly coming from more traditional GM-fiat perspectives
My main interest is in reflecting, in plain terms, on who is authoring the shared fiction. So I always take it as a premise that the fiction has to come from somewhere.
There are two particular things related to that. One is unpacking the metaphor of "exploration". In the literal sense, one explores something by brining it into interaction with ones sensory and cognitive capacities (eg peeking around a corner; looking in a cupboard; cresting a hill). And the knowledge gained is the result of a causal process whereby the "external world" impinges upon the explorers brain/mind.
In RPGing the only process that resembles this, on the player side, is
prompting the GM to say stuff. Relative to any given player, the GM is part of "the external world" and the things s/he says impinge upon the player's brain/mind and generate new knowledge/ideas. But the player is
not in any causal relationship with an imaginary world. Not even by proxy - the GM is not a model, nor running a model. The GM is making authorship decisions.
The other thing is action declaration. My view is that action declaration and resolution is often under-examined, or is explained adopting premises that need not be true, or under investigation turn out to be quite odd (like the idea that because a
player is responsible for some fictional element, that must mean the
character controlled by that player is causally creating that element, at that moment, in the gameworld). My rote example for this: as moments of authorship by way of gameplay,
I search for a secret door - hey, look, I found one! and
I attack the Orc with my sword - hey, look, I killed it! are no different. Both are action declarations that result in the fiction taking on a new "shape" or new content. But it is very hard to get clear discussion about, this, because the first gets framed as "changing the setting" (it is now established that a secret door is part of it) where as the latter is not (
even though it is now established that a dead Orc is part of it). And this tends to go back to notions of "exploration" of an "objective world" - as if authorship by way of action declaration is (ipso facto) correlative to causation in the actual world by performing actual, causally efficacious, actions.
When we try and unpack what is actually different about the secret door and the Orc actions, it turns out to be something to do with
who gets to author what topic of fiction, and this is related to
GM authority by way of prep/notes.
To relate this to
@hawkeyefan's comment about the limitations imposed in his 5e D&D game by describing the rival NPC, I think everyone agrees that a game in which the GM had already pre-authored that the Orc will, or will not, die might count as a railroad. (I don't think it has to, but it clearly might and I think typically would). So how is the game in which the GM has already pre-authored that no secret door will be found different? I'm not saying there is no answer to that question, but I don't see how any good answer can be given that doesn't engage with the question of who gets to author what. You can't answer it just by pointing out that
characters can't spontaneously create secret doors, because outside of passwall-type spells everyone agrees with that and there games don't feature that sort of thing.