For one, most GM-centric games do not necessarily facilitate ‘high context’ player characters–oftentimes they hem in those tendencies by encouraging players to ‘go along with the plot’ (i.e. GM context). For two, I have run and played in plenty of GM-centric games where there are high-context non-player character elements–from the maguffin everyone is trying to get their hands on to the the nemesis that challenges the party again and again. The more tightly focused the setting, the more the case this ends up being, as the players encounter again and again the same cast.

Also, context seems like a ‘distributed’ feature of the game. It isn’t in a single player or character but in the sum total of interactions around that element. A good GM can weave the same elements into a game over and over, from a number of angles, enhancing its interconnectedness, its context with each go round. She can drop an element with clear connections premade, so it comes onto the scene with a lot of weight. Contrariwise, A bad GM can leave the players largely out of this process so that their characters remain largely ‘unconnected’ and so are denied a lot of context-power (which is starting to sound like a better way to describe a certain kind of railroading).

In this way, the GM has some pretty powerful tools for inserting context that are denied to the players, simply because of when these components are generated. Since the GM can introduce these components at a later time in the game, after previous context has been established, the GM is in a better position to use this context-inducing method than the players who generated their components prior to the context-establishment, but had just as great an influence in establishing the current context.

Also, because the GM can introduce more than one component at any given time, a method of reinforcement could be use to lend context to another component. A player with equal authorship rights may have the ability to say, “Hey look, that house is haunted.” but a GM (with the authorship rights that most games reserve for the GM) can instantly spring up a whole town, filled with low-context characters that are built around enforcing this idea of the haunted house that the GM reveals. Even if the player has said “that house is haunted” for several weaks now, it may still not have the context that a GM could put it through with only an hour and all the authorship tools at his disposal.

But I’m still not sure I buy the whole authorship = context thing. I grant that an increased opportunity for authorship (such as players playing a single character over time) grants increased opportunity to increase context (by capitalizing on pivotal situations or gradual build-up over time). But I still don’t buy that there’s a one to one relationship.

You’ve also discussed that there are various levels of the qualities of context, most notably those of showed and told contexts, but you haven’t yet explored their relationship to each other (other than to say one is superior) or their relationship to this master agenda you describe above about finding suitable techniques of interaction between components of different context levels.