Authority, conflict, and D&D

A few months ago, an old gaming buddy of mine named Marc, having listened to me at length about RPG theory, expressed an interest in playing D&D with me. I agreed, thinking it an opportunity to show, rather than tell, all the stuff I'd been talking about. Given that we're on opposite sides of the U.S., we resorted to playing over the phone.
A few ground rules: we used the Kicker technique, it would just be the two of us playing, we would use 3rd edition rules (sticking as close to play-as-written as possible), and there would be no GM, only two PCs.
That last bit was probably what got us in trouble, and what prompted me to think a lot about conflict, authority in the gaming group, and so on.
In the Mask playtesting campaign I've been running this year, I noticed that reliable sources of antagonism have given the game the fuel it needs to stay high-energy, to stay focused. Without dedicated antagonist characters, it's drifted into listless, low-key play that largely hinges on Exploration for its own sake.
This is essentially what happened in our D&D game, as you might expect. To start, we only had one Kicker, and it corresponded to his character, not mine. As a result, it was up to him, by and large, to drive play through decision-making. At least, it was supposed to be: I've managed to get pretty heavily into player-centered RPing, the sort of play where, as i stated above, it's up to the players to act as protagonists, making decisions that drive the plot along. Antagonism provides something for the PCs to push against, a sort of context for their actions. Without any kind of push-back against what they want to do, there's no tension; Lumpley has already explained the non-relationship between tension and winning the game here, so I will refer you to that link instead of elaborating here. But, on that note, antagonism isn't always personified. A target number to perform a task can be the "push-back" you need to give a little more shape to the story, but real-live enemies work better, and are much more interesting.
Anyway, we were going somewhere interesting, and I have to admit that Marc and I came up with some cool stuff (he's a nobleman's son who's given up his birthright for a career in music; I'm an elf-turned-mercenary who's hiding out from agents of my former king). The problem(s) were: inconsistent antagonism and very one-sided creativity. The two relate, of course; Marc, having had a bog-standard, GM-centered RPing career, kept looking to me to decide everything, but because I had a PC of my own, I had a conflict of interests to deal with. I could come up with where to go and what to do, and suggest new areas to Explore, but without anyone else dedicated to stopping me doing as I please, I hesitated constantly, afraid that I was "doing it wrong".
A big, big chunk of all this: creative authority went pretty much undiscussed, despite my efforts to the contrary. This happens a bit in my sessions of Mask, but because there's a fairly traditional, if player-centered, set of roles in place (the regular players mainly portray a single character each; the gm-player portrays nearly everybody else), I can always say, "Yeah, well, it IS up to you guys right now to decide what happens. It's right in the rules!"
Traditional roleplayers will want to push back against any attempts to empower them, of course; there's a vague, underlying notion that there is a "real" game world out there, but only the GM can see it, so he must relate it to the players in a manner of his choosing. Working with this false assumption, you have a range of benevolent-to-maniacal GMing styles, largely differing on the degree to which they a) share information with the players and b) manipulate player ignorance.
These may seem like the same thing, but they aren't. I daresay you could make a two-axis graph that has "level of information shared" and "level of manipulation" as the axes. Or, well, I suppose ignorance-manipulation is an inherently dysfunctional, Illusionist technique, as it's entirely predicated on the implicit assumption behind "sharing" information, i.e. that the GM is actually scrying the game world and revealing it, like the Oracle at Delphi, for the players. This isn't what's happening, and every opportunity you can take to show that the GM is making it up moment-to-moment (no matter how much prep is involved), the better it is for players, and the more likely they'll chime in and be creative along with you.

Okay, big money: this is all leading to a re-evaluation of the Wisdom stat in Mask: it occurs to me that there was something I was going for with it, a sort of "say, do I have any good dirt on this guy?" kind of resource that acted, a la Dune, like the slow knife to Wit's swift knife. To unpack that metaphor, think about it like this: in most games with physical stats, you have something like dexterity and something like strength. One is your finesse, and the other is raw power. That was my original, albeit vague, divisor for each "stat pair" (two social, two physical). In actual play, my "dexterity" stat, Self-Discipline, acts as a sort of social/physical hybrid for everything related to self-control (not responding to taunts; keeping your balance), Strength is purely physical, and Wit is purely social.
Given that, I'm free to experiment with Wisdom. Because Wisdom had no clear role, however, that left us with only three stats to go with the nine skills, and meant that sometimes stat-skill combos got shoehorned into uses that felt, at best, forced. Okay, okay: Wisdom's new function, which I've managed to implement successfully, is to create pure information about the Shared Imagined Space. While "I use my Strength+Banditry to torture him into talking" is certainly creating information about the SIS, it's doing it within the context of the stakes set for the dice roll.
To use the above example, a successful roll there would mean "I torture him, and he gives us the information we need". A failure would mean something akin to "He resists, or gives us bad information, or something else happens". Like usual. But if you used Wisdom instead (along with a skill, presumably), then success would mean "I introduce something relevant to the situation that is useful or interesting."
That was vague. What I mean is this: rather than asking the GM, a la Mother May I, if this or that fact is correct about the situation, you simply say, "This is true. Now, let's roll!" It's just like any other roll, in that respect: declare, roll, continue. But, instead of the stakes referring to actions and events, they refer to information about the events at hand. Because a standard stakes-setting roll works with what has already been established, but it's still so darn flexible, Wisdom actually doesn't/shouldn't come up all that often.
You can see I'm having a hard time defining this. Okay: if you, the GM, find yourself in a situation in which a player wants to find out information, ask/determine if they're looking for color or control. If it's color, you might as well agree to it. If the player is looking for control of the situation, and it hasn't been established that what they're asking is obvious, then the player should roll to make it true.
Hopefully that's clear, to you and to me.