Trust Mapping Social Arrangements

In this article, I’m going to explore a variety of applications for the mechanism of Social Collateral, which I best introduce in “Building Social Trust”.

When I explore these various arrangements, I am going to use (and probably evolve) a visual language for describing these arrangements. As I share more examples, I’ll show how a variety of social arrangements can be diagrammed with this trust map syntax, which will give us a set of tools for analyzing other problems.

The Basics

Circles will represent “agents” (like people) in the system. The assets listed in the circles belong to those agents. Arrows in this graph will represent extending credit of the labeled asset to the other agent.

For example, here’s Alice giving Bob a key to her house:

When you trust another person with a capability, they naturally have the ability to extend that capability to another person. This isn’t a philosophy, this is just a description of a natural phenomena.

Bob can share the key to Alice’s house with Carol, and that’s all fine and well. What happens if carol trashes the house? Alice will have to fix it up, and she will probably change the locks.

Cycles Empowering Strangers

In this kind of Trust Graph, you can look for “cycles” to represent reciprocal relationships. Reciprocal relationships are healthy and good, and they tend to allow strangers to cooperate even if they don’t otherwise trust each other.

The most simple kind of reciprocal trust graph is collateral between two people:

By giving Bob the capability to keep her watch (note that in both their minds, the watch still belongs to Alice), she has established enough trust that Bob is comfortable lending her $50.

You can look for these cycles in other trust graphs to find other foundations of trust.

In our example above, where Bob lent Alice’s house key to Carol, Carol could be more inclined to trust Alice with something else, since the house key serves as a kind of social collateral.

This isn’t a perfect example, because for all Carol knows, while Alice is out with her car, she could change the locks on her house, leaving her carless. Although I guess she could still say “I know where you live!” to Alice, which is itself a sort of threatening and violent but very real tangible kind of threat she could make, having been extended the capability of visiting, presumably extended as the house address along with the key.

Smarter Contracts

With blockchain smart contracts, assets can be controlled by computer code, and this can be useful for systems of social collateral.

Imagine for example if Alice’s home were run with a smart-lock that would not let her lock out Carol for as long as she hadn’t returned the car. That’s a fairly horrifying example, since the house represents putting so much at stake, so let’s take it down a big notch, and think about some lower stakes, and see what we can do with that:

In this simple scenario, since we have strong trust flowing from Alice to Carol, Carol has a sort of recourse she could take if Alice were to betray her. Let’s say, for example, Carol asked Alice to hold her ice cream cone for a second while she tied her shoe:

And let’s say Alice absconded with the cone!

Since Carol has a line of credit from Alice through her friend Bob, she could withdraw what she thought the incident was worth (cone value, distress) from Bob, and send him a message about what happened.

At this point, Bob has two basic choices:

  • Agree with Carol’s complaint, and draw some amount of his credit limit from Alice (Maybe plus a fee, for the hassle of even hearing about it).
  • Disagree with Carol’s claim, maybe revoking the rest of her credit limit.

In either case, when Carol handed her ice cream cone to Alice, she basically had a sort of any-incident Alice insurance up to $50 (The maximal flow from Alice to Carol).

Just to emphasize a few things here:

  • Carol got insured for what she thought the damages were, period.
  • The only “judge” in this scenario is Bob, a mutual friend of the two parties. Far from “an impartial third party” that the traditional justice system uses, here we get into a world where the judge wants to keep everyone else happy, because they have friendships at stake, and that’s worth more than an ice cream cone, damnit!

Higher Stakes

If these credit limits were extended via blockchain smart contracts, and Carol was wary of funny business, she could even withdraw, or lock up the funds before handing over the cone, and could even wait for Bob to pull some of Alice’s credit limit up front, too, ensuring Alice can’t “change the rules” mid game (hey, it’s like the house key problem again!).

Multiple Paths

This same kind of collateral flow can work with more friend paths, it just adds overhead to the process.

If Bob broke a promise to Alice up to $40, Alice could start the withdrawl/complaint process from her social network, but since each member of that network basically has to “review the case”, and in turn adjust their own credit lines/relationships, there is a critical need for:

  • Lightweight financial infrastructure (ahem, blockchains).
  • Intuitive, nay viral, user experience.

A Brief Note on User Experience

Even if blockchains give us perfectly lightweight, cheap, free financial networks with our own credit lines and fancy policies, without usable interfaces, this is the stuff of theory.

In 2004, Ryan Fugger had the idea of using these sorts of “Trust lines” for peer to peer lending, but eventually decided the social aspect was hard, and so Ripple pivoted to optimizing inter-bank protocols instead:

That lending concept (and the name Trustlines) is currently being carried on by the Trustlines network.

Let’s Map Some Real World Examples!


If the arrows in these diagrams are “unconditional permissions to spend”, then they can even be issued when the account is empty. If Bob trusts Alice completely, he might accept her IOU at full price, and be patient to ever redeem it. (HODL on to your friends)

Of course, not all IOUs are worth the same. (Dumb & Dumber)

Fractional Reserve Banking

I love raising this example because it irritates some people who believe Bitcoin was designed to eliminate fractional reserve banking. I would argue it is better thought of as a tool for providing alternatives to untrustworthy, centralized fractional bankers.

With the very simple construct of a smart contract, you’re able to permit multiple people to spend from the same account. While you haven’t “created more dollars” per se, you have multiplied the number of people capable of spending that money, so if they are people you trust to spend in your interest, this can mean efficient coordination. No more arguing over checks at restaurants at the very least, and at the very most, you’ve increased the number of options available to your friends.

While “rational actors” (read: selfish strangers) might conclude that this is a race to withdraw the cash immediately from Alice, within a trusting friend group, these lines of credit have local social implications (that social collateral term comes to mind). If we assume that Bob and Carol both love Alice very much, and want what’s best for her, they will only withdraw as a matter of necessity, and if Alice truly believes in the lines she’s extended, she is likely to agree with expenses that Bob and Carol decide to use her funds on.

Charities and Kickstarter

Sometimes people just want to trust their money to a third party to do something they want to see done in the world! You can call it centralized, or you can call it a benevolent dictator, but sometimes group delegation is simply the tool for the job!

The square here represents how the recipient could be an organization, a company, or an individual. It’s a black box, and people could choose to trust this actor for any reason. In the case of GoFundMe, people often fund because they believe the agent will credibly do something good with it. On Kickstarter, there’s usually an (unenforceable) expectation of a product later on. On Charity Navigator, people might contribute to a group just because the site lists it with high ratings (hey, they delegated their own delegation to the site! It seems like brands may be related to trust… 😉).

Note that in traditional charity and crowdfunding, funds are given entirely and immediately. In a scenario where these are lines of credit instead of money themselves, the donors actually don’t lose the money until the project is successfully funded, or actually uses the funds, making for a little bit of extra liquidity for donors, a bit like the fractional reserve example.

Mutual/Hedge Funds

Putting aside the government/regulatory aspect for a moment, a mutual/hedge fund would look a bit like this, with many investors trusting an entity that trusts many other things.

ROSCAs and Gambling

A ROSCA or (Rotating Credit and Savings Association) is a sort of social investing group seen around the world where a number of people or entrepreneurs form an exclusive group to pool funds, pay dues, and take turns getting a loan.

The benefit of this arrangement comes from the fact that many ventures require more capital than a single person might have on hand, but many people might have more than they need to live day-to-day, so by taking turns holding “the pot”, as long as the group’s investments are sound, the members get opportunities to do things they couldn’t afford otherwise.

From a trust mapping perspective, this looks similar to a game of roulette, where several people put money into a single “game”, where one person will win. The big difference is that ROSCAs are largely built on a foundation of having highly exclusive memberships, and by ensuring that each member gets their turn at having a loan.

Another way to think of the gambling relationship is that a casino is a predatory ROSCA manager with open doors, high fees, and no promises or social trust. This sounds pretty critical, but I think casinos nonetheless provide a similar social service to its customers, who intuitively care more about gaining a lot than losing a little.


One thing I haven’t discussed is how this trust mapping can be used to represent other, governmental social organizations.

While a Kickstarter that doesn’t deliver can leave its users little recourse beyond forum posts, stocks and other regulated securities grant shareholders special rights, as enforced by their government, which also fits into a trust diagram:

You might think the company doesn’t give the SEC the capability to regulate, but the company chooses where to incorporate, and whether to register with the SEC, and its members choose where to live, so I would argue that in a way, the ability for the SEC to regulate comes from the combination of government power (partly tax dollars) and the cooperation of those under its jurisdiction.

Once we’ve trust mapped the stock arrangement, we can see that while people already willingly invest in crowd-funding opportunities with little or no guarantee, regulation can bring some extra consumer confidence, proportional to the amount of additional honesty the government is buying you here. Mind you, you also have to pay for that service through taxes.

I think this is significant, because in the context I’ve presented (which may seem backwards to someone accustomed to the stocks world), the government is merely playing a role that a strong social network might otherwise serve as. While someone might pitch this segment as “presenting peer to peer alternatives to government regulation”, I would argue that the government was always playing a social role, and we can dissect and debate how effective we think it has been at this.

You could get more granular, too, if you wanted to get pragmatic, philosophical, or cynical about where the power to regulate comes from:

Here you can see that part of the trust graph when you invest in a stock is the potential recourse to submit a lawsuit, initiating regulations, issuing a summons, and potentially inflicting violence! No wonder we can invest with confidence.

Once we look at how the social recourse for betrayal is already convoluted through multiple layers of bureaucratic organizations, those user experience problems seem a little bit smaller, don’t they?

Again, you might ask, does the company really give the police the ability to inflict violence? I guess it doesn’t have to. Sometimes other things happen. 😅

Enough for Now

While anything of value can serve as collateral when needing to trust strangers, social collateral can provide tools for analyzing complex social arrangements and identifying where trust might be asymmetrical, unfair, or resorting to violence where social bonds might have been able to serve the same purpose.

While using social bonds to replace the roles of violence in a huge, global society requires a combination of technology, usability, and new social norms, by identifying these patterns that are possible, I hope to help inspire people to build out the infrastructure and customs we would need to fully leverage these social tools.

Bonus References

I hope this has been an enjoyable walk through a way of modeling trust relationships.

While I was writing it, I learned about the field of Promise Theory, which maps promises in a social network, which looks quite similar! Just another field to add to the reading pile! I found out about it through this article that uses Promise Theory to formally analyze a blockchain! Check it out!

Also, while writing this, I learned (through the Trustlines Telegram chat) about a coop in Finland dedicated to these problems, too!