My own view is that the most important anchor between fiction and mechanics in 4e is keywords. So the divine keyword means something in the fiction - namely, that this particular character has been bestowed with power by the gods - and I think that creates scope within the fiction for NPCs to distinguish between a rogue or a knight who is lofty but not divinely empowered, and a "paladin" who is a divinely empowered warrior. The distinction between STR cleric and STR paladin, though, is not going to be one that can easily be drawn within the fiction. (On this account, blackguards are a fictional as well as metagame category - they are the armed and armoured STR guys who use power that is both divine and shadow!)
I don't know if I consider keywords to be
the most important anchor between fiction and mechanics in 4e... but I can agree that some can/do differentiate classes (especially as concerns power source)from each other.
On the other hand I think this view ignores the sum of the whole by only looking at the specific parts. A strength cleric isn't different from a strength paladin because of a keyword... it's different because the class has different abilities, skills, and so on. A paladin is not taught ritual casting, Healer's Lore or Healing Word... while a cleric was never trained to use plate armor, call down a divine challenge or have to be the same alignment as his deity.
Of course, NPCs/monsters generally do not have power source keywords for their powers. Which makes identifying NPC "paladins" a bit trickier - some interpolation is required. Similar considerations arise in relation to paragon paths, which are often described in terms suggesting that there are orders of NPCs who follow the path, even though - mechanically - there is no obvious way to build such NPCs.
There is an NPC Paladin classs in DMG 1 we established this earlier in the thread...so as long as we don't narrow our criteria to "only keywords" it's quite easy to identify a NPC paladin... it's a character who the DM has built using the NPC Paladin class... so there is a mechanical way to build a Paladin NPC. Also, contrary to your assertion, DMG 1, while warning against it for every NPC, very much allows NPC's to be stated out as full characters... if they are considered by the DM to be important. So both of your points here seem a little off.
I think the game expects the participants to paper over the cracks a bit here. Is there anything important at stake here? To put it another way, is there any GM in the world who is going to blanche at a PC who is not built using the paladin class, but who fights with armour and weapons, who uses divine powers to do so (or at least to help with doing so), and who, in the fiction, calls him-/herself a paladin? Is there any real life GM who is going to insist that the player of that PC is doing it wrong, because you can only truly call yourself a paladin in the fiction if you are built using the paladin class?
First, I am not arguing for or against what any particular DM chooses or doesn't choose to do in their campaign (though I see nothing inherently wrong with playing up the fact that classes are discrete things, such as orders with specific training, skills and abilities in the world... for a good example of how this can be done check out the Earthdawn game.). What I was talking about were the default assumptions of the game... and IMO, the default assumptions seem to be that classes are actual archetypes in the gameworld akin to orders or disciplines, while builds are more specific concepts as opposed to packages of generic abilities.
Is a cavalier's embodiment of a virtue part of the fiction or not? Perhaps, a bit like a warlock's pact.
It's very much part of the fiction... as evidenced in the blackguard description and the mechanics as the cavaliers powers are based around whichever virtue he picks. Again any DM or player can reskin whatever they want but I am concerned with the default assumptions of the game... not whether they can or cannot be changed.
But this doesn't undermine my claim that, in the fiction, there is no inherent difference between a STR cleric and a STR paladin from the PHB. Neither has a class feature that singles them out as distinctively related to the metaphysics of the gameworld in the way that a cavlier's virtue or a warlock's pact does.
Well I can think of two fictional/mechanical differrences between the cleric and paladin...
1. A paladin must select a deity to serve... a cleric may select a god, pantheon or even philosophy. This right here creates a difference in both the gameworld fiction and mechanics of the two classes.
2. Paladins must have the same alignment as their deity when created... cleric's do not necessarily have to abide by this restriction if following an unaligned deity... or following a philosophy(which has no alignment).
I'm inclined to think that the same reasoning applies to the WIS cleric and Avenger. Invokers, on the other hand, do have an inherent difference (eg the impliments they use, and the reasons for that).
Again, an Avenger must serve a single deity... not a pantheon or a philosophy according to the class write up... also an unaligned avenger can serve any deity... an unaligned cleric cannot.
Among the Arcane PCs there are clear differences: spellbooks for wizards, bloodlines for sorcerers, pacts for warlocks, and the absence of any of these things for bards.
I see, so are you making the argument that ony some classes exist within the gameworld? That the designers/developers are inconsistant with their determination of this? Or what exactly?
But the martial PCs in this respect are like the divine ones - there is no inherent fictional difference between a STR ranger, a warlord and a fighter, for example, or between a DEX ranger and a rogue. And given that it was fighters and rangers that were the main focus of the discussion upthread, I don't see that this discussion of paladins and blackguards undermines the points made in relation to them.
Actually this discussion (at least the one I was involved in) was about whether classes were overarching archetypes or just packages of generic abilities, and whether combat role should be tied to them... I think I've proven that classes aren't just generic packages when you really examine them, and you apparently agree at least as far as some of the classes go. Now my argument didn't specifically focus on Rangers and Fighters, and I think martial is harder to draw the distinction because it is concerned with the mundane and mostly defined by combat styles, weapon types and skills as opposed to service to otherworldly forces or ideals (divine) or magical power and the techniques of wielding it(arcane).
The first fictional/mechanical difference in martial characters would be their respective variance in skills. For some reason in order to be a highly skilled warrior I have to be a Rogue or Ranger...Fighters, and to a lesser extent Warlords just aren't trained in a diverse number of skills... mechanically this shouldn't be the case if they are just generic packages of abilities... however it very much speaks to the archetype of the combat focused(whether in melee or tactics) warrior... vs. the crafty hunter or clever rogue.
The second fictional/mechanical difference is in fighting style/weapon use. You see this enforced by the builds available under each class as well as the weapons that can be used with the powers of the particular classes.
On a final note here's a line from the Rules Compendium, under Class and Race on page 76... Emphasis mine. It seems that either the developer and designers agree with the ascertion that classes are a specific thing in the gameworld or are just being sloppy with their language and the expression of ideas because the paragraph below equates clas with a particular vocation.
The first decision to make in character creation is picking the characters class and race. Many different types of heroes inhabit the world: sneaky rogues, clever wizards, burly fighters and more. Race defines a character's basic appearance and natural talents, and class is the character's vocation.