If this is the theme of the storyworld, then I have to define what it means. This definition requires more than just a bunch of hand-waving; I must define it in operational terms, in terms of the actions available to the characters. How does a king behave differently than a warrior?
Wait a minute! If I succeed in coming up with behavioral differences between warrior and king, then I have reduced the storyworld to a puzzle. Pick the king-option instead of the warrior-option. Oops.
OK, so what I really need are verbs that can work as either warrior-type verbs or king-type verbs, depending on the circumstances. I think I can come up with one clear example: use of physical violence. Suppose that there is some sort of nasty confrontation between TestCharacter and Aggressor. TestCharacter has an option to punch Aggressor in the face. This is a warrior-action if the punch's primary purpose is for TestCharacter to assert himself. It is a king-action if the punch's primary purpose is to assert higher principles.
That much makes sense. But now I have to examine this matter of "primary purpose". It requires something about the intentions of the subject. I could separate a punch-verb into two parts: punch-angry and punch-discipline, but that takes us right back to the earlier problem of the puzzle. No, the same verb MUST have an intrinsic ambiguity. How then can I resolve the ambiguity computationally?
First, I must dispense with any reference to the intentions of TestCharacter. After all, I don't know those intentions; moreover, we are concerned with the image as king or warrior, not the reality. Therefore, we must infer the intention from the circumstances. In other words, we must make fine moral evaluations of the punch-act. How justifiably aggrieved is Aggressor? Would a reasonable observer conclude that Aggressor's actions are wholly or partly justified?
The real focus, then, is on the mental model of the both characters that would be built by a reasonable and objective observer. That mental model MUST consider the initial relationships of the two characters, and then consider the entire thread leading up to the evaluated action. This is intrinsically anticipatory behavior -- this could get complicated.
What if the observer imagines himself in the position of each of the antagonists and asks himself, "what would I do in that situation?" If the antagonist's action matches the imagined action, then the observer concludes that the antagonist has behaved properly. If the antagonist's action does not match the imagined action, then the observer concludes that the antagonist is behaving improperly in some fashion. But how do we assign a value to this impropriety? Can it be either positive or negative? Does matching constitute kingliness, and deviation constitute warriorness? And what about the values of this "objective and reasonable observer" -- should they be subjective to each actual witness, or objective by some universal standard? If a timid person observes an antagonist acting boldly, will the timid person conclude that the action was kinglily bold or warriorly rash?
My inclination is to go with the objective template of kingly behavior, because the alternative is so dauntingly complex. The protagonist would have to "play to the audience" of whomever is present, adjusting his actions to meet the expectations of the audience. This would be entirely too difficult for any protagonist. In other words, we create an imaginary character (perhaps we could use Fate for this role) whose personality characteristics are those of the Perfect King. Aha! We also set aside another character, whose personality characteristics are those of the Perfect Warrior. We then run the imaginary reaction of the Perfect King as well as the anticipated reaction of the Perfect Warrior. We compare these actions with the actions taken by the antagonists, and the degree of matching constitutes the "Kingliness" value and the "Warriorness" value.