Henchmen also get paid wages. Page 51, Henchmen Monthly Fee. Even a 0-level henchman gets 12 gp/month (heavy infantry wages), which is pretty cozy, especially since a knight’s retinue (non-yeomen anyway) probably get upkeep as well. So the knight is paying that.
And I disagree with Jard, above: I think you can totally model knights as vassals, and the numbers work out pretty perfectly; one of the first things I did with ACKS was run the numbers for fiefs 1-4 square miles in size. Mostly, it’s not worth doing that modelling, and using henchmen and garrisons is easier and faster and thus better, but vassal knights (as opposed to household knights, who are paid henchmen or garrison members) would definitely be domain-holding vassals.
Since I love those numbers, here they are!
-
A half-fee (1 square mile) can be fortified with a townhouse, and has a population of 41-125 people in civilized areas: your basic small manor village. This is basically the smallest administrative unit in most feudal-type societies: a single village clustered around a larger house. They probably won’t have knights, but bailiffs or even just a reeve, and just a few yeomen as military. With average income, they could produce a 2nd-level ruler at most (so, bailiffs are 1st or 2nd level fighters).
-
A knight’s fee (2 square miles) can be fortified with a townhouse or wood building and maybe a palisade, has a population of up to 250 people, and in civilized areas has a ruler of no more than 4th level (so knights are 3rd or 4th level fighters).
-
A large knight’s fee (3 square miles) requires a wood building and palisade, or a stone building - a typical fortified manor, basically. Population up to 375 people, maximum 4th level ruler still.
-
A double knight’s fee (4 square miles) would require a wood or stone building with a palisade (a bigger, better manor), has a population of up to 500 - a really big village, probably a bunch of surrounding farms - and has a ruler of up to 5th level. So, wealthy and famous knights are 5th-level fighters.
-
A small-average barony (32 square miles, 1 6-mile hex) requires a small, medium, or large tower to fortify, and has a ruler of up to 8th level (on average income in a maximum-sized civilized barony).
That all gets us some possibly useful numbers on non-adventuring nobility: knights are mostly 3rd or 4th level, maybe 5th, and barons are generally no more than 8th level.
Now, none of that is necessary, or necessarily even useful, to model, but it all works out well at that level. And if the spirit of N. Robin Crosby is with you, you will model all of that, for every realm, throughout an entire campaign setting, and make tables of it! … I love HârnMaster for its obsessive detail.
Heck, honestly, if I ran multiple campaigns focusing on the same single small kingdom (like my own original campaign setting is shaping up to be), I really might model individual knights’ manors for at least some duchies, maybe even the whole kingdom. Spreadsheets would make it all pretty manageable, after all… from base income and domain type you can derive average per-square-mile incomes… but for most campaigns, it’s not worth it.
Vassals & Henchmen
A vassal does not need to be a henchman. Not at all. In many cases, it’s not possible.
Let’s say you’re the heir to the kingdom your adventurer father carved out of the wilderness, but happen to only have Charisma 10, and your negligent father has only gotten you up to 4th level (not giving you a good enough domain, not taking you adventuring enough, etc.) and you don’t have Leadership (so, max 4 henchmen). Your father had Charisma 18 and Leadership, and his 8 henchmen were his dukes and direct vassals; they each had various amounts of vassals. Your father dies and you become the king, but you can only have 4 henchmen. What happens?
Well, you become king and those dukes are still your vassals… but none of them automatically become your henchmen at all. The ones who are lower level than you (maybe some of the duchies were inherited already) probably do, although recruitment rolls (probably with bonuses; maybe equal to the Loyalty modifier your father had with them, since you’re the legitimate heir) may be needed. The rest might, but at the very least, there’d be penalties to the rolls, and to their loyalty, I think.
Does that mean the non-henchman dukes become independent? No. Most likely they’ll swear fealty, and then act according to their personality, goals, etc. If they’re loyal to the kingdom itself, and to your father’s memory, they’ll probably try to help you succeed. If they’re not, they may betray you. In really bad cases (or with really bad rolls), they might rebel. You could even adapt the syndicate “succession” version of reaction rolls here, with some vassals going with the majority, some rebelling, some becoming loyal…
My interpretation of the differences between henchman vassals and non-henchmen vassals is that only henchman vassals give you a free duty each month; and that henchmen use Loyalty/Morale rolls (e.g. mostly roll for calamities, etc.), while non-henchmen make reaction rolls when you want them to do much of anything: in short, henchman vassals are easy to manage, others aren’t.
Most realms would consist of henchman and non-henchman vassals, which creates instability and leads to infighting; which is historically accurate of the periods the game is mostly based on. It’s especially appropriate to have a crisis during succession, where the vassals splinter off; after William I conquered England, there was no English reign free of internecine warfare or succession strife for something like 400 years!
Historically, kings often held well in excess of 6 direct vassals - far, far in excess, actually! In England, most barons were direct vassals of the King (AFAIK, to a degree the term baron meant at times a “count” who was vassal to the King rather than a Duke?). That would mean a lot of loyalty rolls at lot of the time, which would lead to a lot of barons not doing as they’re told (historically accurate!), and (once a plurality of barons are disloyal or rebellious) revolts, like John Lackland faced.
ACKS rules don’t limit you or act as a straightjacket; they’re tools that facilitate interesting play. I hope that I’ve shown above that the rules for henchmen and vassals can be used to facilitate historically plausible and, most importantly, interesting scenarios. Judges who prefer randomness (like me) can use these rules to organically create succession crisises, barons’/dukes’ revolts, and the like.
Of course, the real take-away from all that rambling is whatever you think will improve your campaign!