Erik over at Tenkar's Tavern has a blog up about alignments, and whether folks prefer the AD&D 9-alignment system or the OD&D/BX/BECMI 3-alignment system (which ignores the Holmes 5-alignment system, but we'll let that pass for now).
This got me thinking about how people actually use alignment in their games. I mean, yes, we have alignment restrictions on certain classes--Paladins have to be Lawful Good, Thieves cannot be Lawful, Druids must be Neutral, etc. But what exactly does this mean?
I have a really good friend who abhors alignment. He hates the entire concept of it. He uses it in games that I run, because I use it, but he very often questions things like the disallowance of a good assassin. "How," he asks, "can a skills set be good or evil?" My take on it insofar as the assassin goes, and I've laid it out, but he still disagrees, is that when one's skills set is murder, it's impossible to call oneself good, unless you have foresworn the use of said skills set...in which case you're no longer an assassin.
In some cases, however, I agree with him--there's nothing inherent, for example, to the Ranger class that requires a good alignment, and so there should be no reason to disallow evil rangers. The only reason the alignment restriction is there is that Aragorn is the source material for the Ranger. I could go on, but the point, I expect, is made. For some classes alignment restrictions are somewhat justified. For others they seem arbitrary.
But the real question I'd like to examine here is, how do you think alignment should be used in game? Is it a restrictive set of behavioral conditions within which characters must operate? Is it a broad set of guidelines for governing your character's actions? Or, is it best applied as a descriptor of the character you have in mind?
In short: does alignment define your character, or describe him?
There's room for many shades of gray in the interpretation here, but in my games I prefer not to force hardcore restrictions on my players based upon one or two words on their character sheet. What is more important to me is that players are consistent in the way they play their character--that is to say, your fighter shouldn't rail against wholesale slaughter one session, then the next kill a bunch of orc children while justifying it as, "well, they're orcs," unless there's some pre-established reason for him to have such an abiding hatred of orcs as for it to be reasonable for this to be a sudden exception to the, "wholesale slaughter is wrong" rant.
So for me, alignment is an important tool, but is a descriptor. It tells me, as DM, what to expect out of your character until I get to know the character. At some point, however, that designation has to fade into the background in favor of the fact that yes, you've played your character in a rational, logical and consistent way which has shown growth and evolution, but not outright contradiction.
The rules, in many cases, seem to point towards the opposite intent, and I can understand why this rankles some folks. A lot of character requirements revolve around alignment--should a Paladin, for example, behave in a way that is not lawful and good, he loses all the benefits of Paladinhood and becomes a fighter. Okay, I'll grant that this is an extreme example, since the entire point of Paladinhood is to be a paragon of goodness and law: Paladins are the Superman of D&D. But others see similar restrictions: Rangers lose some of their abilities if they become non-good. Druids cease to be druids if they ever fail to behave as a True Neutral. It goes on and on, and in some cases one can see the rationale behind the behavioral restrictions, but still--the main thrust seems to be towards treating alignment as a definer rather than a descriptor.
AD&D is an interesting step in the development of role playing, as it really represents a sort of transitional period between the pure gamesmanship of OD&D and the more open and epic character-based play of second edition (which then struck an odd sort of balance with mechanically-defined character-based play in third ed). In OD&D, for example, the onus was more upon the players than the characters. It was the players who used their brains to solve puzzles and traps, regardless of the numbers on their character sheets, which in OD&D meant very little BESIDES being in-game descriptors. In AD&D, with ability scores taking on a more important role, the character you created took on a more personalized and important position in play. It was in AD&D that you likely would first hear a DM say, "I don't know that your character's bright enough to think of that."
Things like alignment functioned as rigid descriptors of this character idea. Gygax seems to have felt that DMs needed these sorts of player restrictions to help them maintain control of their game, lest the players run amok. Could this be an early example of the "bad player vs. bad DM" issue that seems to constantly find its way to the forefront of game design these days? The "how many rules do we need to make sure DMs don't screw players and vice-versa," debate? In any case, I feel that keeping too tight of a reign on alignment as a descriptor results in an exceptionally limited number of character concepts. At very least there needs to be a broad range of behavior within each of the alignments, which is probably why in later publications TSR came out with a number of different systems for tracking alignment--my personal favorite had overlapping spheres with Neutral in the middle, Law and Chaos on the sides, and Good and Evil at the top and bottom. But in the end I never used any of these tracking systems, because as I said, what was more important was that you were consistent in the way you played your character.
But then, I always favored Good vs. Evil over Law vs. Chaos, and I never used alignment language, either.
What about you? What are your thoughts on how (or if) alignment should be used in D&D?