That kind of gets to the crux of my last post. If that's not what you mean, what do you mean? I think there's a very basic non-understanding between the parties - or at least on my part - as to what the other side means by "plot" and how it relates to how you run an adventure.
Once again I can only speak for myself, but when I say "plot", I mean the series of obstacles, events and situations that I expect the PCs to encounter. There is usually some sort of sequence, in as much as there are introductory elements and finale elements. Depending on the game (and how much time I have), the plot may be more or less tightly plotted, but there is probably an idea of a "default" plot that the PCs can safely follow and expect to have interesting challenges with a dramatically satisfying conclusion. There is almost always some deviation from my initial plan, but it mostly (80-90% of the time) ends up with a final confrontation that is something like what I initially conceived.
My games also have what I think of as less plotted (or more "sandbox") periods in which I run scenes (you are at place X and you have interaction Y) to help PC development, advance a story or introduce an NPC, and the PCs have a broader opportunity to pursue their own agendas. (Generally, these agendas are following up on plot threads that I had dropped but one or more players finds important.) During these less plotted periods, I have a plan as to how I'm going to run the session (mostly to make sure that each player there has something to do), but I don't have any underlying agenda or preference about how things would turn out.
Stepping back a bit, I guess I think of plot as something created principally by the GM in which the GM is making a significant effort (either behind the scenes or with the cooperation of the players) to ensure that the issue at stake reaches a dramatically satisfying resolution. (The resolution can be success or defeat for the PCs -- either one can be satisfying.) In contrast, I think of sandboxing as a more organic process that allows for more player whim in determining what happens. In my opinion, a good GM blends both design philosophies - changing plots to accomodate player motivations and adding plot into sandbox situations to make sure that these organic stories have a satisfying conclusion.
Either one can be disfunctional. In general, a plot is disfunctional when there is no buy-in for the players. Either they don't care about what's at stake or they thing the "assumed plan" that would bring them to the obstacles is stupid. In contrast, a sandbox is disfunctional when the players spend a lot of time working on something that they care about only to discover that it didn't really matter or never could have succeeded.
My synthesis of these two approaches is to say that a good game needs two things:
(1) Goals with buy-in from both the players and their characters
(2) Interesting and satisfying things that happen when the players pursue those goals
And I think you can come about it either way. You can use a sandbox in which players develop their goals on their own (and the buy-in comes naturally). In a sandbox, you have a challenge of creating situations that will result in interesting in satisfying things. Alternatively, you can write a plot that intrinsicly involves interesting and satisfying things that occur and then try to sell the PCs on the motivation.
Both approaches have their strengths and weaknesses, and either approach will generate an awful game if you don't address both points. A plotted game is terrible if the GM screws up and writes a plot that PCs don't want to do, and a sanbox game is terrible if the GM abdicates his responsibility to make sure that something interesting happens (or, at least, is very likely to happen).
And, of course, both approaches are delightful (if a matter of taste) when performed well.
-KS
