Non-Euclidean Justifications for Character Classes

Comments on yesterday's post brought up the old question of what makes a new character class necessary. Why have a Sage class when knowledge is something that a character of any class can have? Why a Thief class when a character of any class can steal, hide, etc.? (If memory serves, the example provided in the AD&D 2nd edition DMG was the Assassin: why have an Assassin class when a member of any class can kill for money?)

On the one hand, there is a strange conservatism, almost a reactionary flavour, to this argument: a default against having new classes, as though new classes is per se to be frowned upon. Admittedly, this is a position which has its attractions for me. But it isn't likely to convince doubters. What arguments might?

At first glance the strongest argument for the existence of classes, which is basically a variant of niche-protection: a class should mean a type of PC who has exclusive purview over one or more particular activities. Wizards for magic, Clerics for healing and turning undead (perhaps working better when exclusively able to do the latter), Fighters for fighting.

Except, er, members of any class can fight.

And, except Elves, Dwarves and Halflings, who could be described as being special cases were it not for the fact that this would make for just as many special cases as "standard" Basic D&D classes. (In fact, if you include the Fighter, it would make Wizards and Clerics the special cases in being classes-based-on-exclusive-abilities. Not to mention the fact that Elves also cast spells. But I've made my point, I think).

What other justifications might exist for creating a character class, then? One other argument standing a fighting chance might be that of fidelity to source material. There are wizards in fantasy literature, and there are warrior/fighter types - hence Wizards and Fighters. And there are Rangers (Aragorn), and there are also Elves and Dwarves, and Halflings. But Clerics? I'm also not convinced that there are many Druids appearing in fantasy fiction, at least those who are described as such, nor Paladins.

A better argument could simply be functionalism, which is somewhat related to the argument from niche-protection. Here, the idea would be that a class can exist where there is an actual need for it. You need a class to heal and turn undead. Hence Clerics. You need a class to be able to fight. Hence Fighters. You need a class to be able to cast spells. Hence Wizards. But don't Dwarves and Elves complicate both of the latter, creating functional redundancy with the existence of two classes doing more-or-less the same things?

The truth is, you are a fool if you think D&D is based on anything like rational or logical principles. D&D is resolutely non-Euclidean. Euclidean geometry is based on the notion that you can identify a small number of irrefutable apex principles from which all rules logically follow. This is not just an idea from the realm of mathematics: for somebody like Descartes or Rawls, it's also how you should organise anything, more or less - whether a role-playing game or an entire society or system of justice.

D&D is not like that. It's rules are a tangled mess with no coherent underlying principles: it is, in effect more of a result of Darwinian selection combined with inherent inertia on the part of players and rule-makers. Why does an iteration of D&D have the classes it does? Partly because it has always had them (Wizards, Clerics, Fighters). Partly because people clearly just want them (Rangers, Paladins, Druids). Partly because they are what works (you need there to be Fighters in a combat-heavy game; for some players, it's just easier to have a Thief class who disarms traps rather than having them puzzle them out for themselves). Partly because of fidelity to a shared notion of what fantasy fiction is (Elves, Dwarves). And partly because the creators of a new edition for obvious reasons are compelled not to stray too far from what has gone before.

It seems to me to be no accident that the more designers tend to try to start from apex principles and work from there, the less effective and interesting D&D becomes - 4th edition being the paradigm example. A slightly different way of putting all this is what Jeff Rients once said in what is probably still the single most intelligent blog post to ever emerge from what we now call the OSR - "every RPG theory fails the moment D&D comes into consideration".