Two farmers have a dispute, Tonia says. One says the other promised to pay for work; the other says that the work was shoddy and the agreement was to only pay if the work was good.

Sounds pretty sad.  Guess they won't be bargaining again next time.

That works in a small village but Tonia thinks it won't work if they have too many people for everyone to know everyone else's reputation.

People should stay to small villages and farm until they have enough food, then.

Yep, Tonia thinks that's probably just correct about Golarion humans. But maybe the dath ilani will be willing to add a ton of complexity in order to get to not live in villages.

Adding tons of complexity right away sounds like a bad idea!  What's the next chunk of complexity that enables somebody to solve a problem that current small villages cannot?

Well, can two villagers combine their plots and be understood to have shared rights to the plot that results. Can villagers exchange their plots for food.

What customer problems are these new features intended to solve?  We should make sure the customer has an exciting important problem to which some feature is the simplest solution, before we spend a lot of time implementing, testing, and debugging that feature in our next-gen village design.

Well, combining two plots smooths out your variance in harvests which is good because having insufficient food is more bad than having too much food is good, and one other person is few enough for social enforcement against slacking to mostly work and for people to mostly get the benefits of their labor. And selling your plot for food is something you might do if you were imminently starving because better to be in a bad situation later than right this minute.

Keltham supposes he's starting to see a bunch of value that could be unlocked by this system, in which, to write down some of the rules so far:

- Food can be tagged with a 'socially-constructed' imaginary arrow that points to a particular person, let's call them the 'owner'.  There's a socially respected rule (not a physical law) saying that only the 'owner' of food can eat it.

This rule of itself isn't obviously very useful on the face of it - if food just came out of nowhere, a rule like this would just prevent hungry people from eating food that's nearby?  Or at least, nobody else has explained yet why you'd want 'socially_constructed-ownership_tags' for food if it just came out of nowhere?

However, this rule becomes (visibly) useful when combined with a new one:

- Land can be tagged with a socially constructed imaginary 'pointer' to an owner.  There's a socially respected rule saying that all food which grows from that land then gets tagged with the same owner, and those food-tags are then respected as above.

This incentivizes people to invest increasing amounts of effort into farming their land, since they get more food and more reliable food in proportion to putting in more effort.  This seems probably more efficient than having a whole group guarding a lot of land and then also expending effort on monitoring how much effort all those people are putting into farming, monitoring the monitors, trying to have an incentive system that encourages the right amount of farming, etcetera.  Though obviously this whole tagging system is also going to have social costs?

Still, it's plausible the tagging system is more efficient.  It has a compact local physical relationship between how much people work and how much they eat, instead of a complicated socially-constructed one involving lots of monitoring and people with power over other people who might go corrupt.

Now some new rules are being proposed which seem to say things like:

- People who tag land or food can, as a voluntary action, taken under particular circumstances, announce that the 'pointer' should change to point to somebody else instead.
- You can announce today that some of the food your land produces later will, when it appears, get tagged with a 'pointer' to somebody else instead.
- There's a social construct whereby multiple 'atomic' actions and conditions of this kind can be packaged together into a single bound 'molecule' (he's using the same words as in chemistry) that only gets socially respected as a single piece after everybody involved has made the necessary announcements.  You can 'trade' a lot of food for a bit of land, but not in a way where you announce that the food-tag points to somebody else now, and then they can say haha and refuse to announce the change on their land-tag.  You say 'Change the ownership pointer on this food to them and that land to me' and the other says 'Change the ownership pointer on this land to them and that food to me' and the joint declaration only gets socially respected after all tagged parties have announced the same molecule-arrangement of the same 'atomic' actions.

The Taldane word for this would probably be contract, or something like that.  There's a single-syllable word 'contract' in Baseline but the more technical term would be 'multi-signature transaction'.

These new rules unlock potential actions like 'trade some of your land's future food for food you can eat right now' - as might be a good thing, if something went wrong with this year's labor, as sometimes happens.

Though, you know, you could also see how that sort of thing could potentially go wrong?  Especially if any of the people involved were reasoning less than perfectly?  Still, let's ignore that part for now and plunge ahead.

As of the most recent suggestions, somebody was proposing that two pieces of land could be owned by two people.  Keltham's not sure he could back them on this idea?  This sounds fraught.  If food has a tag pointing at two people, who gets to eat the food if the two of them disagree about who should eat it?  Do both of them have to agree on who eats the food, and announce that socially using multiple signatures, before it can be eaten?  If people end up in angry lasting disagreements, does the food rot?  That seems like a waste.

Keltham thinks the system should say that any socially constructed tag points to one place.

But you could, maybe, allow people to announce a 'multi-signature transaction' whereby the tags shift to point to a new social construct that isn't a person, and instead can be a complicated imaginary construct like FairBot or PrudentBot.  This new construct could say, for example, that both of the announcers will now own 'shares' in the land, and the food produced by the land will be tagged to the people who own the shares, in proportion to how many shares they own; and that these shares themselves now have ownership-tags and those tags can be moved around in 'multi-signature transactions'.

This new persistent complicated imaginary construct, to actually be socially effective, would have to be something that all the people who are supposed to respect the tags and the 'multi-signature transactions', would be able to understand.  If somebody can't understand what the new complicated imaginary social construct does, they can't figure out who's socially allowed to eat the food.  There's obviously going to have to be a socially constructed 'specification-language' for imaginary constructs like these, and some 'persistent contracts' in the 'specification-language' could potentially fail to 'validate', because they don't fulfill all the conditions that society demands from a contract.

For instance, if you try to announce a complicated contract where, under some circumstances, food ends up with no ownership tag and nobody is allowed to eat it and it rots, the people around might tell you that your contract fails to 'validate' and that they won't respect this contract even if it's announced.  They might then, if they feel like being helpful, suggest a 'debugged' version of your contract in which all the food always ends up with a well-formed ownership 'pointer', and say that they'd respect that contract instead.

Any farmer who tries to announce a contract in which they trade a cherry to themselves and end up owning all the food in existence, would similarly be told by other farmers that their contract fails to 'validate', because they haven't fulfilled the condition - in this whole imaginary system that everybody is maintaining inside their own imaginations - that you need 'signatures' from all the parties in a 'molecular transaction' who are currently deemed to own anything whose tag gets shifted around in the 'transaction'.

Likewise if you said that the food produced by land gets owned in proportion to how much work two people put into it, society would tell you that your contract doesn't 'validate' because society can't directly observe how much work was put in.  Society needs all the terms in a contract to be 'evaluable' by society in order for the contract to validate; any fact referred to inside an 'expression' needs to be one that society can observe, cheaply, by the time society needs to 'evaluate' the 'expression'.

(To be clear, Civilization doesn't literally think that contract specification languages were invented by farmers immediately after they invented ownership as a social construct - it's widely suspected that this wouldn't have occurred until hundreds of years later, after people got in some practice with simpler systems.  This would, however, probably have been the first use-case for the explicit idea of a programming language, invented thousands of years before anyone could build an actual computer.)

"That's not how actual farmers do things here, just so we're clear," says Tonia faintly.

Yeah, the Intelligence 10 thing is having really weird effects on their society.  His students are now being exposed to the sort of thinking that happens in a world where they’re normal average people instead of smart people.

How does this hypothetical society handle contract disputes or contract breaking.

Well, there's presumably something you have to do if an individual 'defector' from society goes around eating other people's food?  Even a tiny fraction of people like that can do a lot of damage, and of course, you'll have even more of them if you don't build any way of stopping them.  There's probably some kind of 'police investigator' whose job it is to hunt down food thieves, and then that person has to eat...

Well, they probably get a small fraction of food from a lot of farms?  And the 'police' don't protect any farms that don't contribute those little bits of food?  And thieves target those farms and only those farms that don't have police investigators.

If you need that mechanism anyways, the 'police' then have to decide what is thievery in the first place, and keep track of where the tags go.  So probably the 'police' would be the main ones where you'd want them to tell you that the contracts 'validate'.  Though, obviously, you'd also want the other people in your society to agree on not coming over and taking your food.  You wouldn't want any contracts which were complicated to the point where other members in society couldn't figure out that they shouldn't eat your food.

The notion of a contract breaking is a weird one?  For a contract to 'validate', the police should always be able to figure out what to do in any case of observable terms.  The contract might say that, if somebody doesn't do some deed - or rather, something the police can observe about that, like some contracted observer saying that somebody didn't do a deed - some land or food they own, then belongs to somebody else?  And the person who didn't do the deed, might lose reputation about that?  But contracts shouldn't 'break' so much as describe what happens in unpleasant contingencies as seen from a police perspective.

If the police can't figure out what a contract means - if there's a 'dispute' the contract doesn't imply how to resolve - then they need to do more careful validation next time!  But, sure, you could have some sufficiently respected group of five people, whose majority would decide what would happen in the hopefully very rare cases of contracts that validated and shouldn't have.

And if the police is corrupt everyone who is paying them, since they're doing that voluntarily, can just start paying someone else instead?

As long as they can collectively outfight the Hypothetical Corrupted Governance Military, yes.

This is easy while everyone's a farmer.  The police outfight individual thieves, all the farmers together win against the police.

If all this social complexity enables you to build more powerful weapons, and the police have those weapons, and nobody else has those weapons, then there is perhaps a bit of a problem.

If you count the rehearsal festivals for it, Civilization spends more on making sure Civilization can collectively outfight the Hypothetical Corrupted Governance Military, than Civilization spends on its actual military.  And even then there's some real and awful questions about whether, for example, the military plus the Keepers could take the rest of Civilization.

There's an order of Meta-Keepers whose job is specifically just 'keep an eye on the Keepers', and in real life that's probably worth anything.  The Keepers hopefully can't win against all the rest of Civilization plus its usual military; and the usual military is in fact specifically designed to resist Hypothetical Corrupted Keepers.  One similarly tries to make it hard for the Keepers collectively and the Military collectively to trust each other in a pact to take over together, especially when they were already betraying their oaths to Civilization.

It's hard to have a really simple reason why Governance couldn't possibly go corrupt, once your weaponry reaches the level of 'giant flaming craters'.  Civilization has thrown complicated solutions at this because they don't have simple solutions, and nobody is happy about that.  But 'just don't build overly powerful weapons' isn’t an answer, because how do you stop rogue groups from building those, possibly in secret, if there's such a prize to be won by being the only group with overly powerful weapons?

What is Keltham's estimate of the probability Governance secretly was corrupt, Meritxell asks.

Which failure mode?  What counts as “corrupt”?  There was probably at least one person somewhere in Governance who was corrupt?

The whole thing being secretly full of, Keltham doesn’t know, people getting paid much more than their theoretical salaries with secret mansions full of full-time sex workers?  That seems legit unlikely?

Keltham once tried to do his own totally separate checks on what Civilization could be hiding in the ways of massive secret projects.  It looked to him like at most 2% of the economy was being hidden off-the-books… at least in ways that didn’t require a really massive conspiracy with tons and tons of data being falsified and lots of people outside of Governance being in on it?  Which would be a lot easier for them if they actually had a good reason?

Keltham doesn’t know, here, the parts of Civilization he got to see were pretty nice to him!  Especially by Golarion standards!  He supposes he’d be ticked to find out that actually half the female population was masochists, but he was prevented from finding out that he was a sadist in order to leave more masochists for people in on the Conspiracy?

It seems legit hard to hide a massive Government Conspiracy in dath ilan, unless there's a good reason that gets the Keepers and lots of civilians who’d notice an anomaly in their section to go along with that?

Keltham is really having a hard time evaluating this probability!  If they’re hiding mansions they’re good at that!  It leaves him mainly with ‘priors’, plus his guess that a Dark Conspiracy doesn’t train people to think nonconformingly and be ready to defeat the Hypothetical Corrupted Governance Military.

Maybe 2%???

To Chelish ears that sounds kind of outrageously low.

Though it's hard to argue that dark conspiracies don't usually encourage you to think nonconformingly and be ready to defeat the government militarily.

"If there were a conspiracy," says Gregoria, "with the sort of people that exist in Golarion, it wouldn't be to divvy up all the actually-plentiful masochists unfairly and have mansions with full-time sex workers."

Keltham is sincerely confused about what the point of the Dark Dath Conspiracy would be!  Does Golarion wish to enlighten him?

"It would be to just have painful sex with people who don't like it, since there aren't any masochists," says Gregoria flatly. "And to not pay them, because why would you pay them. And to frame people who offended you for crimes, and to murder people and make it look like an accident. I don't think there's a Conspiracy in Cheliax but if there is, Her Infernal Majestrix is not hoarding masochists."

Obviously Abrogail isn't hoarding masochists!  There are masochists right here!  Not being hoarded!  They are, allegedly, plentiful enough here that nobody would need to hoard them!

And what is the actual point of this Dark Conspiracy?  Keltham can imagine that there are some 'criminal-sociopaths' in Governance with sexualities not quite like his, who can get off on inflicting pain on somebody who isn't responding sexually to that and to whom it doesn't mean anything emotionally except as something to hate.  He can imagine that if some dark group of dark conspirators had managed to secretly take over dath ilan anyways, the criminal-sociopath-sadists among them would be having sex with nonmasochists so long as they were at it.  He's not seeing that being the whole point of the Conspiracy.