If the player tells me what his PC does, I don't have the option to decide if it's part of the shared fiction. It IS part of the shared fiction at the moment the player told me what his PC did.
<snip>
D&D allows the following.
1) DM contribution to the shared fiction without any players.
2) DM and player collaborative contribution to the shared fiction.(You don't need the DM to agree to this.)
I don't really get what is going on in (2). If the GM doesn't have to agree, where is the collaboration?
But anyway, there have been posters on this board - eg @Saelorn - who have expressly denied what you assert in the first part of this quote. They have said that an action declaration is simply a statement of what the player
wants to be part of the shared fiction, but that it doesn't become part of the shared fiction until the GM approves it.
I also think it is, perhaps, continuing to conflate "setting" and "fiction." The former is a subset of the latter, and everything the players have their characters do changes the fiction.
The second sentence is true. But what I think tends to matter is
what is the nature and degree of such change? If it is confined to adding into the fiction that the PC performed a certain bodily movement, or that such a bodily movement had some local result,
but then the GM "manipulates" other offscreen elements to confine or minimise or negate any broader ramifications of these things the player hasn't changed the fiction very much.
There are various ways that my italicised thing in the previous paragraph can happen. One I associate with some modules/APs: the GM is instructed to do it in order to keep things "on track" - eg if the BBEG is killed early than a henchman takes over and keeps the plot going; if the PCs fail to save the world then some helpful NPC turns up and set things right; etc.
Another one I associate with sandboxing/"living world", in which the GM draws upon his/her sole-authored fiction that has not yet been revealed to the players (I also often call this
secret backstory) to determine that the consequence of what the player successfully had his/her PC do is different from, or less significant than, the player intended. Eg the player does all the right things to learn the fate of his/her PC's long lost brother, and the GM reveals that it was a completely prosaic death in a minor plague five years earlier, with nothing more of consequence that follows.
I think that to the extent the GM is playing to discover the players' conceptions of their characters, the players have say over that; seems to mean the players are deciding what is emerging.
<snip>
The players can change my world--and have done so. Granted, in play that's been entirely through character action/s, but there've been changes based around backstories as well.
The next questions I would want to ask is
who decided what the changes were? and
what sorts of topics did the changes pertain to?
In most games that aren't using the AP-type techniques I mentioned above, the PCs can (say) set fire to a forest or kill a bandit leader. But who decides what happens - eg does killing the bandit leader cause the bandits to be cowed and flee? Or to fight back in revenge? Or to seek service with the victorious PC? In some RPGing, these sorts of things might be the stakes of actions declared by players for their PCs and so the GM would not be the one to unilaterally decided what the change is. In other RPGing the player can have his/her PC kill the bandit leader, but deciding what results from that is entirely up to the GM. In both cases it is true to say that
the player changed the world but in the second case the causal influence was indirect - a prompt to the GM to do some authoring of the change in question.
As to topics: can the players make changes such as
establishing that a 1,000 year old dungeon exists? Or can they only make changes that involve the future of the imagined world?
Note that both the above ways in which RPGs can differ are completely consistent with
in play that's entirely through character actions. Burning Wheel depends entirely on character actions, and allows for players to decide not just that change will occur but what it will be (via the role of intent as well as task in resolving character actions) and permits players to make changes on topics like the presence of ancient dungeons (because of the way it resolves actions like
I search in these hills for an ancient dungeon).
But to get the scope of player-change-of-setting found in BW, the role of the players has to go beyond
presenting a conception of their characters. We have to look at how action declaration is understood and resolved.