For me a comparison would be this: the PCs meet a dragon and decide to try and defeat it in combat. Do they succeed or not - ie is it true or false in the fiction that the PCs beat the dragon? One way would be for everyone to sit around, think hard about what the PCs can do and what the dragon can do, communicate frankly about this, and reach agreement. Another way would be to play the game. I prefer playing the game to work out what happens to the dragon; and I prefer playing the game to work out whether or not a particular player's evaluative conception of his/her PC, as expressed through the play of that PC, is viable/worthwhile/good/bad/admirable/despicable. In my experience the results can be interesting.
I find this a poor comparison. We have plenty of rules for combat with the dragon, enabling an objective determination of whether the Dragon or the PC's won. We don't argue over whether the Dragon should or should not be able to fly. We accept that Dragons can fly. We don't assert that, through sheer power of belief, a given PC should stand unharmed in the inferno of the Dragon's breath weapon - we look to the rules.
Many of the alignment arguments come back to debating the rules. "Well, is 'respect for all life' really Good, or is it OK to adopt a system of capital punishment for capital crimes? My Paladin thinks that it is not only Good to execute perpetrators of certain heinous crimes, but Evil not to do so." Whether I think Dragonfire should cause damage or not, the rules say it does. Whether I think capital punishment is morally right or morally wrong, the rules say "Respect for Life is a Good ideal", so capital punishment is a compromise of that ideal. I think the biggest issue with alignment arises when we assert that no compromise of ideals can exist. The Paladin must respect life and protect the innocent, so when he must either defend the innocent with violence, and possibly cause the death of another person, that should not mean "Oh, you compromise the ideals of Good either way, you have fallen", but that there is room for compromise of one ideal to the betterment of another. I doubt any gaming group would assert that the Paladin should not take up arms. But in our real world, many pacifists would outright reject this form of resistance.
But that the universe responds in that way itself has implications for the truth or falsehood of certain moral and cosmological conceptions. For instance, if the universe is neutral as between being good or being evil - as is the case in Planescape, where both are valid alignments able to shape the planes by way of belief - then the typical mindset of a paladin (according to which the universe is on the side of good, and providence will ensure that honour and duty align to reinforce rather than oppose one another) has already been refuted, and the paladin is consequently self-deluded.
Who says the mindset of the Paladin is that the universe is on the side of good? The Paladin devotes his life to exemplifying the virtues of Good and opposing evil at every turn. He trains diligently in martial skills so he can oppose evil. That strikes me as someone who believes Good is not foreordained to victory, but that the Virtuous must take up arms and fight to their last breath to see that virtue succeeds, and evil does not win.
As I said upthread, this can work in some games - eg in a Conan game a paladin would be self-deluded - but doesn't work for the sort of fantasy game I default to, which is romantic fantasy a la Tolkien and King Arthur, not REH-esque.
I think there are two questions conflated here. One is whether alignment can be a useful tool, perhaps as a role playing guide. The second is whether it should have a tangible impact on the mechanics by which the characters play, highlighted by the loss of Paladin powers for failure to adhere to certain alignment tenets.
In those romantic fantasies, adherence to the tenets of Good and Law carry rewards, and a fall from grace carries a cost to the character, so I think this argument suggests the enforcement, not the abandonment, of alignments in the romantic fantasy, and suggests they should impact the mechanics. In a Conan game, we may well still have virtuous Knights and vile, evil Sorcerers, but the cause of Good does not reward its followers with power (yet Evil still does, as sorcerers make pacts with vile entities from other worlds - Good is its own reward, but Evil bribes its followers).
This paragraph begins by talking about the player - and whether it is hard or easy to roleplay a particular character - and then ends up talking about the PC - who is maintaining certain principles. Talking in that way already involves many assumptions about playstyle that are not true for the way I play or GM the game. For instance, I don't expect my players to find it hard to roleplay their PCs being challenged. I hope that they will find it easy and enjoyable, and challenging only in the sense of setting an intellectual and aesthetic goal to aspire to - but in that respect, playing your PC being rewarded can be equally challenging, although often less dramatically engaging.
Is the problem, then, with the alignment system or the player's choice of character? If I do not want my character to be beholden to any third party for his powers and abilities, then I should not select a class which derives its powers from adherence to the tenets of Law and Good, and is rewarded by divine favour (ie the Paladin), but rather I should choose a class whose abilities are inherent, and do not trust to any outside force. The Paladin's abilities are linked to his adherence to an alignment as clearly as a spellcaster's spells are linked to his spellbook and use of components.
Most suggestions I see for replacing alignment bind the Paladin instead to some form of moral code. Let's return to your "chastity" character. If that character were a Paladin in a game where one of the tenets of the Paladin's Code is remaining chaste to remain pure, then I see nothing wrong with your character re-evaluating his priorities and perhaps deciding that he does not consider his purity dependent on chastity. But his character mechanics still say "he must adhere to his code to maintain his powers". So the character either needs to find another code which will grant the same, or similar, powers without a requirement of chastity (and presumably requiring some other code be maintained), perhaps also changing his powers/mechanics in some way, or forego his powers because he is no longer willing to meet the requirements of having those powers continue to be granted by the external force which grants those powers.
I contrast this issue, by the way, with your "accidental death" Paladin. In that case, you are unhappy that his powers are not immediately struck from him due to a failing the character perceives as justifying such punishment, but the deity does not. In the case of the vow of chastity, you are unhappy when failure to abide by the restrictions imposed does carry consequences. Clearly, you do not find this to be inconsistent. I do. To the player who is unhappy with mechanics of a character depending on maintaining some behavioural requirement (be it alignment, a code of honor, a vow of chastity, the regular sacrifice of an innocent, loss of control under a full moon or any other restriction or requirement),
don't select a class, archetype or ability that imposes such requirements. This doesn't seem so tough to me. Play an Honourable, Pious or Chaste fighter whose decision to abandon honour, piety or chastity has no mechanical repercussions.
They don't. That's part of the point of playing a paladin - finding out what your own conception of "honour" might require, and finding out what happens when that comes into collision with other participants' conceptions.
It seems like your issue is less one of "Alignment", as we can replace the word "honour" above with "Law" or "Good" and have the same expectations. To me, your issue is more around who decides what happens, yourself or the GM, and whether the concept is embedded within the rules or not. The rules are quite clear that when the Paladin crosses the line, he loses his powers. That, then, is "what happens" when his conception of the ideals by which he lives "comes into collision" with the conceptions laid down by the rules, as interpreted by the GM. If we remove "alignment" and require the Paladin live by his honour, with no other change to the game system, then the GM decides whether his has upheld the requirements of his honour, and you do not. In another group, that decision might be made by the table as a whole (and in all games I've played, the GM listens to the opinions of those at the table, especially when the call is not clear and obvious). From your comments here, and in prior posts, it seems you want the question of whether the Paladin's actions are, or are not, honourable to be somehow decided by a die roll, so if your "Honour" skill his high enough, somehow striking an enemy from behind in a dark alley is sufficiently "honourable" to satisfy the code.
Actually, as I posted upthread, alignment rules seem to have been invented precisely for players whose PCs were pawns on a chess board. They introduced an extra constraint on the play of those pawns, in return for better access to certain benefits (like hirelings, healing and resurrection).
I think the alignment rules may well have been motivated, at least in part, by a desire to include some rules parameters encouraging the players to treat their characters as more than a pawn on a chessboard, or at least that this may have motivated some of the evolution of the alignment rules. This is no different than psychological complications in later game systems. However, I believe the origins of alignment are found in the influence Michael Moorcock's writings had on the early roots of the game.
Conversely, once you are playing living, breathing characters why do you need alignment? You just play your character. If one of your character's "weaknesses" is that s/he won't fight dirty then just play him/her that way - though the whole idea that not fighting dirty is a weakness rather than a strength strikes me as wildly misguided unless the focus of the game is on nothing but extracting benefits from others via the most efficient application of physical force.
It is a weakness in a case where fighting dirty would be to the player's, or character's, advantage. If we are letting matters be decided by the dice, and our PC can gain an advantage by throwing sand in his opponent's face, or by sniping with a crossbow rather than facing him in open combat, then an advantage is foregone by not "fighting dirty", whatever we perceive that to be. Hopefully, the game also incorporates advantages to the character's honour perhaps trust, respect and goodwill which assists him outside combat), but then the lack of similar trust, respect and goodwill extended to the sniper might well be seen as a
disadvantage of his choices.
Lastly, If there is any point to the alignment system (specifically the "your powers are contingent upon it"), it is that the GM should frame conflict that juxtaposes its competing ideas and priorities. As such, I don't look at the classic D&D alignment system as having any advantage here. Inherent to it is, when presented with a hard choice, if you misprioritize your hierarchy of ideals (specifically in the GM's eyes) and commit an action that champions one over another, you may lose access to your archetype (class) defining features (your powers).
This, to me, comes down to interpretation in play. When presented with a hard choice, my view is that the character does not have clear guidance on how to proceed. The fact is that had choices have no easy, right answers. Given that, I consider it inappropriate to impose negative consequences to any choice made. The Paladin should not fall because he was forced to make a hard choice, and rank his priorities. He should fall when he fails to make an obvious choice, even where that choice may be disadvantageous. But the campaign ground rules should be discussed at the table, no different from a ruling on mechanics. If Orcs in our game are irredeemably evil and fair game for slaughter by the forces of Good, this should be made known to everyone, and they should get reminded when the player forgets what their character would clearly know. If, on the other hand, they are sentient beings capable of moral choice, then the players should know/be reminded of that as well, and treat Orc prisoners no different than human prisoners. To me, the problem is much less "alignment as a concept" than it is an adversarial, "gotcha!" style of play where the GM considers it his job not to set interesting challenges for the Paladin (since he is the example we keep coming back to) but to trick him, or place him in untenable situations, driving him to fail.
