This is actually kind of fascinating to me. I've certainly encountered players in real life who have a predilection towards being led around by the nose, but the belief being professed by several different people in this thread that all action must originate with the GM is truly bizarre to me. Surely no one actually plays the game this way? I'm having difficulty even imagining what playing at a game table like that must look like.
2-If the DM has described a pointless courtyard, he is leading his PCs by by the nose. He led them into a pointless courtyard. Because I guarantee you, I didn't ask to go there! If the DM had listened to me, I'd be at court, or hunting a beast or whatever feels right.
3-I never professed all actions originate from the DM. The scene originates from the DM. The Players act in it. If the DM sets boring scene, you got a problem.
I keep saying I would want out of such a useless scene as fast as humanly possible and you keep saying I like being led by the nose. Let's say I cared about your approval, how could I earn it? Have my PC sit down under the tree, eat an apple and start singing? How is having my PC immediately lockpicking the door and getting away from that courtyard where nothing is happening the equivalent of 'wanting to be led by the nose'?
Mal, your reply here makes sense to me.
The way I approach the game (almost always as a GM) is this: the players don't only want their PCs to act - they want their PCs to act in a way that is meaningul both from the ingame perspective, and from their own out-of-game/metagame/gameplaying perspective.
Therefore, as a GM, when I design a part of the gameworld - a place, a creature, an NPC, a nefarious plot, whatever it might be - and whether I do that in advance or (as often happens given my frequent lack of prep time) on the fly, I try to make sure both (i) that it gives the PCs an opportunity to act - a pit to jump, a lock to pick, a statue to investigate, or whatever it might be - and (ii) that in undertaking these sorts of actions the PCs will have made and acted on a decision that is meaningful in both the sense noted above.
That doesn't mean that I've predetermined what the players' or PCs' responses will be.
For example, if I have a player whose PC is a priest of the Raven Queen, then when it comes time to stick a random statue in a crypt and describe it to the players I am more likely to choose a statue of Orcus than a statue of Melora because not only does a statue of Orcus provide the same range of actions as a statue of Melora - break it, look under it, cover it in a shroud, walk on past it indifferently, etc - but it provides an additinal range of options, such as channeling divine power from the Raven Queen to cleanse, and it also means that whichever action the player has his/her PC take, the consequences are likely to be meaningful for that player - because by choosing to play a priest of the Raven Queen, the player has signalled that elements of the game dealing with death and undeath are meaningful to him/her.
When I think of the GM's role in originating scenes in D&D, this is the sort of thing I think of. (Other games have more formal rules for linking the setting of the scene to various elements of the PCs, either by directly empowering the players in scene setting or by obliging the GM to include certain elements that are drawn from the PCs. D&D doesn't.)
I know from experience that it is possible to run a sandboxy game in which the constraints on scene-setting are derived more heavily from considerations of gameworld consistency, hence making metagame considerations - such as what the players might find interesting - less important. Also in my experience, this sort of game obliges the players to work harder to "find the adventure" - if they want to find statues of Orcus to respond to, they will have to hunt out the Orcus cult rather than rely upon the GM making sure an Orcus statue turns up in the very next crypt visited.
I used to prefer this sort of play because I though it better served a goal of verisimilitude. I also played a game that tended to encourage this sort of play through its action resolution system (Rolemaster). Now I tend to dislike it for the sorts of reasons that Mal gives in the quote above. I like to frame scenes a bit more aggressively in order to cut straight to the chase (or at least to my best guess of where the chase might be found - every GM can have a bad day where you think a scene will excite the players but in fact it doesn't).
I think this issue of sandbox approach to scene framing vs more metagame-y aprroaches to scene framing is pretty orthogonal to the issue of player power or leading players by the nose. In a good sandbox, the GM has a large degree of power, because they tend (in the typical sandbox game) to describe the world and adjudicate attempts by the PCs to move through it. The players obviously have some power - they can initiate actions for their PCs. In a sandbox game in which powerful scrying and movement magic is available, the players gain more power as their PCs have a much wide field of action within the gameworld.
In a good game with hard scene framing the GM has a large degree of power, because they get to originate the scenes, but the players also have some power - they have a major influence over how any given scene resolves, and they also have at least indirect influence over scene framing in the first place, because a good GM will be building scenes that pick up on elements of the PCs that the players are keen to explore further in the game.
