Epistemic Status: Pattern matching. The model completes a pattern in an intuitive way. Little originality besides the act of pattern completion itself.
As I was reading the recent discussion about networks of trust, I noticed a mismatch between models described therein and the social environment around me. In particular, I found that I would be happy to participate in the kinds of networks that came up as examples. That confused me, because in practice, I actively try to distance myself from some local communities. In fact, I've always thought that trust-based relationships were just a burden without any practical benefits for anyone involved. As it turns out, the reason for the mismatch is that the little bubble where my intuitions came from is actually different.
In the networks of trust that I can observe out there people don't actually seem to be interested in helping others where help is actually needed. Or at least, they don't act like it. And it isn't just a bug, either, because no one involved asks for, or expects, any helpful help. In the meantime, there is a complex culture of ritual incantations and arbitrary norms around it that everyone is careful to follow for fear of losing social status points. In other words, it smells of the exact same kind of bullshit that simulacrum level 3 is notorious for.
So I completed the pattern. The trust that is actually trust will be simulacrum level 1 (S1). The "trust" I observe out there will be S3. The pattern does complete in an obvious way for the other two levels, and the resulting model does explain my two datapoints well.
Level 1: The Handshake
The definition of S1 is "I help you, and expect you to help me; you help me, and expect me to help you."
At this level, you may offer a hand to your neighbor because you think you could be helpful.
In the world of trust-proxies, a company may deliver what the customer has paid for, because they think they offer something of value for a fair price.
A suftaja, as described by Henrik Karlsson, also belongs to this level. As a quick summary:
The level of trust [...] between merchants in the great Malay entrepôt Malacca, gateway to the spice islands of Indonesia, was legendary. The city had Swahili, Arab, Egyptian, Ethiopian, and Armenian quarters, as well as quarters for merchants from different regions of India, China, and Southeast Asia. Yet it was said that its merchants shunned enforceable contracts, preferring to seal transactions “with a handshake and a glance at heaven”.
Level 2: The Backstab
The definition of this level is "You help me, and expect me to help you; also, your naïveté is adorable."
At this level, you may ask for or accept help, signaling that you are interested in building a trust-based relationship, but then defect and get the better end of the deal.
In the corporate world, a company may refuse to deliver, or deliver something that predictably doesn't satisfy the customer. Only after they have the money, of course.
Level 3: The Contract
The definition of S3 is "We will pretend not to screw with each other."
At this level, you may invite your neighbors to a dinner party because last time it was your neighbors who invited you to their dinner party, and you are afraid you will be seen by the community as the kind of person who doesn't return favors if you don't.
In the corporate world, a company may keep the letter of some contract or terms of service just so that people don't think that they are pulling an S2 on their customers.
Notice that at S3 nobody cares about helping others anymore. The name of the game is plausible deniability and trustworthiness signaling. There are norms or legal frameworks that define what constitutes defection. So long as you maintain the standards for inviting guests to dinner parties, no one will hold it against you. Whether or not you or your guests hate dinner parties doesn't factor into that (unless specified by the norms). Whether or not a product ends up being useful doesn't matter any more than it's declared in the contract.
This doesn't of course mean that none of these transactions are helpful. Sometimes, ""help"" has the side effect of being "help". But only so long as that is required to maintain the appearance of trustworthiness.
Here, frameworks of trustworthiness become objects that can be manipulated. It's difficult to find good real-world examples of this. I suppose the stakes are too high and everyone is careful not to let this happen.
At this level, when you are sick and have a relative with the "I help sick relatives" trustworthiness signal, you may ask them to do arbitrary things for you. Even if both of you know some of that could definitely wait until you have recovered, they might find the price in the signaling world too high.
In a world where it isn't (yet) illegal, you may squeeze someone else's short position on a stock. They will be forced to close their position for the generous price of one zillion, or else lose their trustworthiness by refusing to fulfill their contractual obligations. Losing this kind of trustworthiness is not good, as it will reduce the willingness of the police to engage with them in the mutually beneficial way.
In the case of the simulacrum levels of trust, the equivalent of the symbol is an act of transaction, and the equivalent of it's meaning is the trust-based relationship. S4 is the level that needs not pretend that there is any trust behind the transaction. Indeed, both parties can be aware that one side is defecting. The incentive to maintain appearance of trustworthiness is strong enough to overwhelm the prospect of receiving no or even negative reward in the form of "help".