"How is there a village witch?" Or, States Should Be Hoarding Magical Power: A Problem for World-Building

This is a long post. I'm going to lay out the structure in advance to help preserve the thread of the argument. The parts are:

1. I list a bunch of things that we know about people and technology
2. I note some common elements of how pre-industrial states work
3. I establish various features that are typical in magic systems (especially those of fantasy TTRPGs)
4. I claim certain desiderata for fantasy settings (especially those of fantasy TTRPGs)
5. I point out how those things we know about people/pre-modern states and those commonalities of magic systems together keep us from achieving the desiderata :(
6. I mention a few unfortunate corollaries to the problem

Thank you to the folks on Mastodon (https://dice.camp/@Periaptgames) who debated some of my initial thoughts on this topic and helped to firm up my ideas.

In the next post, I'll go on to present and analyse various possible solutions to this problem.

PART ONE: THINGS ABOUT PEOPLE AND TECHNOLOGY

People, as a whole, are optimisers and innovators and they recognise when things are useful and they do what they can to improve their lot in life. They are not (individually or collectively) perfect at any of these things, but most people are pretty good at them. If that weren't the case - if a setting's people didn't create new things, refine things, solve problems, use technologies, and try to make their own lives better - they wouldn't feel like people.

People, with rare exceptions, are great adopters of technology. When a farmer is presented with a new animal-drawn plough, he's not going to upgrade from pulling one himself if he can. If the three-field system is going to improve his harvest, he's going to switch to that. When metallurgy makes steel armour possible, soldiers are going to want that (if they can afford it). They're going to favour javelins over rocks and powder weapons over javelins and laser pistols over powder weapons.

For a useful technology to spread, all it needs is for people to do what they always do: travel and trade a bit, and adopt that technology once it is available nearby (by which I mean local people know it exists, know how to get it, and know how to use it). A more useful technology is likely to spread faster and see wider uptake. In the long term, it's practically tautological that useful technologies will see widespread use.

People's mindsets are affected by culture, and cultures are affected by their component people's mindsets. Ideas and perspectives change. Modern civilisations recognise many things as being utterly unacceptable that were everyday in many places and times: slavery, rule by might, execution of political opponents, torture for information or punishment, torture of animals for sport, etc.

When something is useful, some people will use it - even if the thing is regarded as terrifying, or gross, or unethical, or dangerous. Sometimes if something is discovered to be useful, the culture will change so that the negative valence is reduced or disappears.

There are enough people who enjoy the benefits of power that, in every place and time, SOMEONE will find a way to be in charge. Ruthless and strong-willed people and people with a leg up (in power terms) will tend to be the ones who end up in charge.

PART TWO: THINGS ABOUT PRE-INDUSTRIAL STATES

I think it is realistic to view states (maybe excepting some states very recently in history) as primarily a product of, and a tool to perpetuate, individuals or groups attempting to monopolise organised violence. All that is needed for a state to form is that someone wins the competition of applying violence to get their way. In early peoples, better nutrition and a touch of ruthlessness might be enough to tip the balance: the big mean guy gets the best food and develops the best muscles and, as the guy who can beat the snot out of everyone else, sets the rules. He leverages this to make use of other people's labour to get weapons and armour and housing and the best food for his children, and all of a sudden you have the start of the institution of military aristocracy which essentially all early states comprise.

Side note: I think it's actually safe to generalise 'the state is an attempt to monopolise organised violence' to 'the state is an attempt to control as much as possible of the tools of power: organised violence, formal alliances, wealth and taxes, land ownership, high-calorie and nutritionally dense foods, education, administrative functionality, military technology and the people who make it, infrastructure, personal security, etc'.

I am not going to provide a lot of arguments in support of this view; I think it is more or less self-evident. As a source I will pass on a recommendation originally from https://acoup.blog/ for Patricia Crone's book, 'Pre-Industrial Societies: Anatomy of the Pre-Modern World'.

So, individuals are motivated to use organised violence as a tool to improve their own life, and the lives of the people close to them (spouse, children, other family, friends, cronies, clan, caste, backers, etc). In pre-modern history, the mechanics of that basically go like this:

1. Organised violence gets you control of fertile land and other people's labour
2. Control of fertile land and other people's labour gets you a food/wealth surplus
3. Food/wealth surplus, possibly with an intermediate step of 'trade', gets you supporters and specialists who improve your life by providing things like arts and entertainment, medicine, physical security, high-quality goods, robust shelter, etc

I'm not saying that the state is perfect at using organised violence and other tools of power to amass resources and improve the lives of its components (people in the military aristocracy), but it is almost certainly better at that than anybody else in the vicinity.

When a thing is both scarce and desirable, the people who succeed in using organised violence and other levers of power (i.e., the state) will usually get that thing. In practise, the outcome may not be "the state has all of the thing" but it's probably at least "the state has disproportionately much of the thing and tries to prevent its enemies getting the thing".

When the scarce desirable thing is a useful military technology, all of this is doubly important, because states that don't secure enough of the scarce desirable military technology will be destroyed or outcompeted by states that do.

I should clarify that I'm using "the state gets the thing" as a shorthand. It can mean that the state "possesses the thing", but also "oversees the thing, can demand the thing from its populace at will, indirectly controls who gets the thing and who doesn't, limits access to the thing, prevents its enemies getting the thing, etc."

Finally, let's consider what happens when "the thing" is a person, or a person's service. There are lots of different methods the state can use to "get the thing" then. One way to categorise these methods is:
(a) Positive incentives
(b) Soft or hard coercion
(c) Cultural programming
Or as I'll be flippantly calling them, Benefits, Blackmail, and Brainwashing (BBBs).

Let's take a look at three examples of the state acquiring military technologies using BBBs.

1. Men-at-arms (or knights, i.e., men-at-arms with a title) were a useful military technology. They were raised to fight from childhood. This often took the form of a child being sent off to be a page at some other knight's property. By raising a man-at-arms to serve their liege you are culturally programming them; you are providing positive incentives (it is prestigious and beats the base case of being a starving peasant); there is some coercion involved (the father, also a man-at-arms, is obliged to fight for his liege - note that his child is effectively a hostage of one of his liege's allies).

2. Longbowman were a useful military technology, famously so in medieval England. Cultural programming in the form of legislation and social expectation required young men to do archery practise and banned certain other more 'frivolous' pastimes. Armies were (before the rise of professional armies) distributed, hierarchical, and mustered when needed: a village or estate or shire would muster a certain number of archers to accompany their liege when he went at his liege's command. There was little element of choice (so, coercion). Soldiers were, in theory, paid, and the major 'perk' of the job was looting (a positive incentive).

3. Arrows were a useful military technology. The Parliament Rolls of Medieval England reveal that makers of pattens (wooden overshoes) were forbidden from using aspen, which was regarded as the best wood for arrows, on pain of a (colossal) 100 shilling penalty. This was an Act of the Parliament of Henry V in 1416, in response to a petition by arrow-makers cautioning that source wood was running low; the same Act required arrow-makers to sell arrows at lower prices. As the primary buyer of arrows, the English Crown had a huge interest in manufacture of cheap, plentiful, good quality arrows. Other legislation penalised low-quality steel in arrowheads, dictated what were acceptable feathers to use in fletching, and so on. Artisans were financially both incentivised and coerced to provide the Crown with arrows.

An aside

In early polities, there was usually very little difference between the state and the individual ruler, who usually had final say in matters of legislation, judiciary, religion, taxation, and warfare. That's why I've been saying things like "the ruler is the state" or "the state is made up of a few ruthless powerful people". This means that the ruler's personal wealth was, essentially, indistinguishable from the state's wealth.

Contrary to popular belief, historical rulers weren't always the most wealthy of the people in their land. Often this was because cultures treated exorbitant spending by rulers as a sign of nobility ("largesse"); sometimes it was because of financial mismanagement. There are lots of examples of rulers having to borrow money and lots of examples of merchants "as rich as princes". Wealth was a signifier of military power but it wasn't necessary for it because of, amongst other things, BBBs: if the state itself was skint it could often still rely on members of its carefully cultivated ruling class and their cronies to provide soldiers and equipment. Why? Because those supporters stood to benefit from doing so, could be coerced if needed, and the culture they grew up in said that they should as part of fealty and vassalage.

PART THREE: THINGS ABOUT MAGIC SYSTEMS

Here are some things I think are very common in magic systems, and nearly ubiquitous in TTRPG magic systems:

1. At least some people can use magic.

• Using magic may entail casting spells, applying magical powers, producing magical items, calling down miracles, creating magical materials, subverting existing magic, waving wands, performing lengthy rituals, negotiating for the intervention of magical entities, or whatever.

2. Magic is at least somewhat consistent or predictable.

• A magic-user has at least some expectation about what is likely to happen when they use magic, and are correct more often than they're not.
• (As opposed to there being a simple dichotomy of "use magic and something completely out of your control happens / don't use magic".)

 3. Either magic is capable of unique feats, or it can improve on mundane feats, or both.

• By "unique feats" I mean bring about an outcome that isn't possible to bring about with mundane work using the non-magical technologies available.
• By "improve on mundane feats" I mean reproduce those feats in a way that is strictly better along at least one axis (time, effort, cost, expedience, secrecy, scale, durability, whatever) even if it is worse along other axes. If you can snap your fingers and conjure a stone wall but doing so exhausts you worse than if you'd built the wall by hand, that still counts - that's a useful tool in the toolbox because it has emergency and military applications.

I think that together, these features are enough for us to call magic "a useful technology". Notably, if either (1) or (2) is false, it's not a technology, because people can't reproducibly use it for specific effects. If (3) is false, it's not useful: some irrational/ignorant/desperate/curious people might get involved with magic, but it has pretty much no worthwhile impact on the world.

The conclusion: in most magic systems, and in almost all TTRPG magic systems, magic is a useful technology.

Let's add that it is also very common for magic to be a useful military technology, which is to say that in almost every setting where magic is useful, it has useful direct or indirect military applications. I think it is very easy to understate this.

For example, in just core fifth edition D&D (which I really doubt is an outlier in this regard), the following are cantrips (i.e. trivial spells infinitely repeatable by even the lowest tier of mage):

• Any number of ranged weapons that require no ammunition, some of them invisible or inflicting lingering harm
• Instant lights for signalling or night operations
• Perfect mending of damaged weapons, armour pieces, or fortifications
• Instant fire
• Absolutely secure messages capable of crossing 120 feet and minor obstacles
• Brief improved resilience, aim, or ability for a soldier
• Instant, guaranteed effective, battlefield trauma medicine
• Tripling the volume of orders delivered by voice

Availability of even a small number of these technologies would have significantly changed military tactics (and in some cases, operations) for most of pre-industrial history. Once you start getting into level one spells, which the worst tier of mage can cast a few times per day, you have the potential to radically change pretty much every aspect of warfare.

PART FOUR: DESIDERATA FOR FANTASY SETTINGS

We generally want a magic system and a fantasy setting (especially for the purposes of TTRPGs) to have the following features. This is of course a subjective claim, but a conservative one; I don't think any of them will be controversial.

1. Magic is neither incredibly scarce (three spellcasters in a kingdom of a million) or incredibly common (9 out of 10 people learn a little magic; half go on to specialise as full time mages).

2. There are magic-users in roles that aren't focused on military power, intensive economic support, state administration, spycraft, and political influence. That is, there can be hedge wizards, village witches, small-time mages, adventuring sorcerers, reclusive warlocks, runaway apprentices, obsessive thaumaturgy wonks, small town militia magicians, travelling entertainers with a knack for illusion, magical confectioners, dabbling alchemists, toothache-cure-makers, and so on. To put it another way, we want it to be possible that most magic-users - say, 80%, or even 90% or 95% - are private individuals, rather than being co-opted into the apparatus of the state.

3. It's possible for a state to exist that's not run exclusively or near-exclusively by magic-users at the top (although some might be). That is, the state being a magocracy is possible but not necessary.

4. The world has verisimilitude. It feels real; things in it make sense; its rules are self-consistent and in principle discoverable; things work like in our real world unless there's a specific reason they don't; people with power mostly aren't holding the Idiot Ball; consequences logically follow from antecedents.

Some specific settings won't care about (1) or (2) or (3). I think they're rare, especially in fantasy TTRPGs. Some game tables won't care about (4). That's sufficiently antithetical to my TTRPG preferences that I'm not going to consider it further (sorry).

PART FIVE: THE PROBLEM

Now let's synthesise what we've laid out. You've probably already seen where this is going and are wishing I'd done it in half the word count.

Importing the following premises from earlier...

1. People optimise, innovate, spread and adopt technology, and want to use it to make their lives better.
2. States form naturally, and where they don't, other states take over.
3. Pre-industrial states are individuals or groups attempting to control as much as possible of the levers of power: principally organised violence, but also auxiliary things like "useful military technologies and the people who create/invent/maintain wield those technologies."
4. States are not perfect at getting and keeping control of power, but are generally decent at it. They're made up of (often ruthless) people, who when presented with a big juicy prize will make sweeping changes to get it, including changes to society/culture/the state apparatus. States who aren't good at controlling the levers of power are replaced over time - one way or another - by states that are.
5. Controlling the levers of power may mean possessing them, overseeing them, the capacity to demand them from parts of the populace at will, limiting access to them, denying them to enemies, etc.
6. The state uses what I flippantly called BBBs - Benefits, Blackmail, and Brainwashing - to control the human component of the levers of power.

Then in most interesting fantasy TTRPG settings...

 

7. Some people can use magic in a way that is predictable or consistent, and there are notable benefits to using magic; therefore magic is a useful technology. In most setting's magic is a VERY useful technology with direct applications as a lever of power that may be wielded for administration, military action, subjugation, communication, economic improvement, information gathering, life extension, trust building and verification, cost avoidance, etc.

 
8. The setting is verisimilitudinous, magic is a modestly scarce resource, the state doesn't control all (or even most) magic, and magic-users themselves don't (necessarily) run everything.

My argument is that (1-7) appear to preclude (8).

(1-7) imply that the only plausible way for magic to be distributed amongst people is for magic-users to be substantially under the thumb of the state, or to themselves comprise the state.

Magic is, basically, too desirable. The state has too much interest in magic as a useful technology. Mundane or mixed magical/mundane state powers will use the power they have amassed to get control of magic and magic-users in various small ways, thereby getting more power, which they will use to grab more magic, and repeat, continuing to consolidate power in that way. Magic has so much upside that there's basically no case where you, the overlord, say: "Oh that guy who can conjure wine, send secret messages across the country, improve crops, change the weather, heal grievous wounds, and set things on fire with his mind? I already have some guys who can do that so I'm happy to just let him do his own thing."

There's a natural selection problem implied by premises (3,4,7). If powerful/ruthless people fail to seize such an important technology as magic, then other powerful/ruthless people who succeed will eat their lunch (in the case of individuals, usurp them personally; in the case of kingdoms, invade and loot them; in the case of economies, outcompete them; in the case of cultures, destroy them).

That's if magic-users can be bullied by state actors. Alternatively, if magic is SO powerful that mundane states can't get leverage over magic-users and thereby acquire them as useful technology, then the only plausible system of government is magocracy: the most powerful/ruthless local mages will seize control, because they are people and want power and resources too, becoming rulers/patrons with less powerful/ruthless mages as vassals/clients, and so on down a hierarchy until you reach mundane subordinates. If magical power overwhelms mundane power, all it takes is one magic-user to get the ball rolling. People tend to form traditions and institutions, which means that the only thing particularly likely to overturn a magocracy once established is a paradigm-shifting outside force: mages stop being born, invaders from an even more powerful magocracy, etc.

I think the most plausible world where (1-7) are true is one where (8) is not: some enormous percentage of magic-users are courtiers, high-status servants, mercenaries, bureaucrats or otherwise employed by or involved in the state. Even in a world with unusually disorganised/incompetent/poor/incoherent state apparatuses, you're going to end up with either "most wizards are officials/vassals" or "wizards are mostly in charge".

• For every mage dabbling in orchard magic or enchanting baskets, there's ten mage-knights specialised in military magic.
• For every town warlock delighting children and crowds with illusions and fancies, there's three doing the same in the duke's court.
• For every village witch whipping up pox cures, there's surely a dozen dedicated to healing the nobility.
• For every wandering sorcerer with a lightning spell in their pocket, expect to see a score of them in the king's retinue.
• For every lonely wizarding tower, there's a whole magical academy that's for the children of the ruling military aristocracy and their cronies.

What does the process of state investment in magic look like? Well, we would expect the kinds of BBBs we're familiar with from history. People in power have historically been...

• ...willing to seize and indoctrinate children, so if children are detectably born with magic, that's what happens.
• ...willing to sponsor and regulate education where it suits them, so if magic is a studied ability, that's what happens.
• ...willing to search for what they want and pay exorbitantly to get it, so if magic is a subtly learned knack, that's what happens.
• ...willing to integrate with religious institutions to get an advantage, so if magic is a divine gift, that's what happens.
• ...happy to funnel resources disproportionately to their own children, so if magic is sparked by special circumstances, that's what happens.
• ...able to change laws, sculpt societies, and mandate cultural changes so that people have particular characteristics and feel a duty towards the state, so if magic is only granted to the 'worthy', that's what happens.

Or, remembering that people are often innovative and ruthless and cunning, we would expect the state to find new BBBs better-tailored to fit whatever process it is by which some people end up as magic-users. In practise, I would expect there to be combinations of different kinds of pressure.

Benefits Positive incentives

• "Cleric of war? Best opportunities are in the queen's army. Cleric of the sea? Everyone knows the best opportunities are in the queen's fleet. Cleric of ironfounding? The queen's workshops, easy."
• "The wizard guild's the de facto leader of the city. If you reach a position in the hall, there's a gold chain and a monthly feast and districting power. And it only gets better from there."
• "The duke's amassed so many mages that the ducal court is a place of plenty. You can get a levitating pillow and a month's conjured wine for just what he pays you for a week of work. It'd be worth it even without the food and housing, let alone the sheer prestige."
• "Marriage prospects? I guess, but I tell you they get a LOT better if you've done your customary stint of service to the emperor."

Blackmail Soft or hard coercion

• "Tax on mages outside the king's service is thirty shillings annually. I hear it's going up to forty next year."
• "Oh, you're a mage on your requisite annual pilgrimage to the imperial capital to provide a day's magic to the crown? Me too! I'm thinking of moving close to the city. There's plenty of work to be had there and of course the pilgrimage would be minutes, not weeks. Oh, you're thinking of moving closer too?"
• "We're from the enforcer mage agency. We know where you live. Here's your invitation to attend the earl's staff for two years. The pay's good. When you get back your family will be right here, safe and sound."
• "Join the royal magical academy without a noble sponsor? You must be out of your mind. Get lost."

Brainwashing Cultural programming

• "If you're a mage, you can't call yourself an upstanding citizen of Somewhereistan unless you put in your customary crown service. People would spit on you in the street if they knew you dodged service!"
• "When they found out I had magic as a child, they bought me from my parents and sent me to the local baroness to learn spells, of course. Tutelage, opportunities, and fealty to a powerful woman? I shudder to think of how my life would be if they hadn't found me."
• "Ares has called for a holy war! All men-at-arms, mages, and anybody who's ever strung a bow is marching under the temple banner! Join us or turn your back on the gods!"
• "Legionnaires serve for twenty years, but you can choose to switch to civic magic after twelve in the magical legion. Anyway, if your number gets called, then smile, boy - that's our whole people's future on your shoulders."

PART SIX: UNFORTUNATE COROLLARIES

The problem I've described has some additional emergent issues for a typical TTRPG fantasy setting.

Corollary 1. If you're buying magic items, you're doing so from the state, at the state's discretion (although grey markets and black markets for magic would likely exist, because pre-industrial states tend to be corrupt and inefficient). The same is analogously true for access to other kinds of magic (buying spells, casters for hire, spellswords, magical advice or prophecy, etc): you're limited to that which is permitted by the state. Access to other people's magic is only possible when the state gets more benefit (typically wealth) from selling that magic than it does by using it for itself. All of this is of course a huge issue for the flavour of a setting (but, ironically, it can actually help with certain other problems in TTRPG design).

Corollary 2. If there are dungeons, zones of adventure, outlaws, mysterious outsiders, tribal demihumans, ruins, and so on, they don't have magic items, because the state has too much of an interest in extracting those magic items via a state-organised expeditionary force. The only exception is when the entity is so incredibly powerful and dangerous that it's not worth the risk of the expedition, in which case what chance do the PCs have? A horde of orcs with some magical artefacts implies the state is too weak to find and kill them, which implies that horde of orcs is a burgeoning state itself.

Corollary 3. If a PC wants to be a magic-user, they are presumably under the state's control, and all that entails. Alternatively, they are an outlaw or pariah or being actively headhunted, and all THAT entails. Alternatively, this is a world where magic is too powerful to be controlled by states, and what you have in the group is effectively a demigod - and all THAT entails.

Corollary 4. Again, magic is a very useful military technology. How much does it cost for the state (i.e., some local knight or noble) to train a magic-user compared to a man-at-arms? In both cases, you're training a specialist, probably from a young age, probably for a decade or more, probably with associated equipment costs. Raising a man-at-arms was EXPENSIVE. Magic changes the nature of warfare so much that in typical TTRPG settings, I'd rather have a level one magic-user than a dozen men-at-arms. In which case even in the (surprising, to me) case where raising and equipping a war mage is TEN TIMES the cost of raising and equipping a man-at-arms, you're going to see retinues of forty war mages plus ten men-at-arms rather than ten war mages plus forty men-at-arms. This is probably contrary to most people's fantasy aesthetic.

Corollary 5. In most TTRPG settings, magic has opportunity costs and specialisation is possible. If you rain down purple fire on the enemy, that's one less use of Enchant Sponge Cake or whatever. For some kinds of magic-user, having access to Enchant Sponge Cake means not having access to a different spell today; for other kinds, it means locking in Enchant Sponge Cake as an option for the foreseeable future. So the objection "there aren't enough mages around to replace men-at-arms and serve en masse as bureaucrats in the imperial city" is actually just another way of saying, "because there's not enough magic to go around, there is an absolutely massive pressure from the state upon mages to specialise in those things that the state wants most: military magic, economic magic, and magic that can do stuff for the nobility that they can't get through mundane means".

CONCLUSION:

I wish I didn't believe this argument was valid. It makes world-building difficult!

Don't despair, though. In the next post I'm going to be examining possible solutions.