You could call it Deus Ex Machina, the hand of god, or if you are really against the concept you might even call it "railroading". I've wondered many times over the course of the past year whether it even constitutes a mechanism.
If you're not familiar with the concept of GM fiat it works pretty simply like this.
The GM simply decides whether your idea will work, the motivation between this choice is usually based on where the action will take the game or story. It can be done well, but it often gets a bad rap because the concept is usually associated with GMs who do it poorly.
In traditional rolplaying games, a group of players gathers to play through a story. There is a subtle difference; they don't gather to communally tell a story, they gather to put a group of characters through a series of set pieces pre-defined by the GM. When a GM uses GM fiat as one of the mechanisms for their game, they simply allow characters to take the actions that will logically lead the story from one set piece to the next, while they make any other actions difficult for the characters to engage.
Difficulty can be defined a few ways; psychologically (it is implied that really bad things will happen if the characters choose to follow the specified course of action), mechanically (any action that takes the characters beyond the predefined scenario faces a target number or difficulty much higher than it really should be), interpretively (any successful actions that push beyond the scope of the pre-defined story seem to have less effect than they should) or even blatantly ("No, it doesn't work!"). It's this last case that really makes a GM stand out as a user of fiat. The first few cases, if used carefully and subtlely, might fly completely below a playing group's collective radar.
Like tables in an earlier mechanism, I've had my attitude to GM fiat change over the years. I once thought that te idea was simply the hallmark of a bad GM...that a good GM could craft a decent story from the actions of their players rather than forcing their players down specific storylines. I looked at the hundreds of modules on a gaming store shelf, never thinking that I'd play straight through them because this type of pre-defined linear narrative belonged in novels, not in games.
I've played in live games where the GM fiat became noticeable, and then blatantly obvious to me...only to end up frustrating me out of the game. Yet other players have really loved these games. I thought it was perhaps naivety on their part, but came to realise that it all links into the concept of illusionism within a game, and determining how much of my creative voice I was willing to give over to a core visionary.
If the core visionary is telling a story that makes sense in my eyes, or is weaving a tale that I'm finding interesting, I'm happy to sit back and take the ride. Engaging enough to keep my interest levels active, but not trying to rock the boat too much.
Looking back on GM fiat in that light, it is actually a useful tool for driving a story forward, especially where more modern games might get mired in a bog as multiple players pull against one another with the actions of their characters. So GM fiat can include the introduction of any element that drives a plot toward a predefined goal. This interpretation means that I'm as guilty of GM fiat as the next GM, I just hide my efforts in the colour of the setting, and the flavour of the actions, trying to subtly use psychological, mechanical and interpretive means to push a story to an outcome I'd like to see.
I have been developing an idea of game analysis through vectors over the past couple of days, defining game concepts through vectors passing from, to and through specific scene nodes. The place of GM fiat pushes a vector in a specific direction, or places boundaries on there a story vector might head...more of this should hopefully start making sense next year, as the gaming vector model is explored, detailed and discussed through this blog.