The Theological Anthropology Built Into the Protocol

• 1. Identity is self-sovereign: your key as your unavailability
• 2. Trust is relational (web of trust), not merely algorithmic ranking
• 3. Coordination without a centre
• The common thread

§1. Identity is self-sovereign: your key as your unavailability

Technically: your identity is the key pair. The npub is not registered with a provider, it is not granted and it cannot be revoked. No account that anyone can suspend; no “Sign in with” that inserts a third party between you and your self. The identity exists cryptographically before any platform recognises it — and regardless of whether it ever does.

Theologically, this hits the notion of unavailability precisely: your identity precedes the institution. In platform logic you are reduced to whatever Google, Meta or X administers about you, so your identity is an entry in someone else’s database, on loan to you, terminable at any moment. That is the inversion of creatureliness and of a humanity thought from God’s side: not withdrawn from disposal, but made totally and absolutely available and handed over, without the freedom to decide for oneself. Nostr reverses this. Your being is not derived from a platform’s recognition; it precedes it. Structurally this is the same movement as “the human being is a creature prior to any ascription by systems”: your dignity is inviolable and a gift, it is not for sale, and not a provisional status that could be withdrawn at any time.

Unfortunately this also builds fallibility into the system, because your sovereignty means that you can lose your key. No “forgot password” rescue, no support desk that restores you. That is inconvenient and at the same time honest. Availability without the risk of loss can only be had at the price of a third party holding control. Sovereignty and fallibility are two sides of the same coin. Theologically this is almost too fitting: a freedom that cannot rule out error and failure is the only real one.

And at the same time a natural limit applies, because the key is not your person. An nsec can be copied, delegated, sold, inherited, handed to a bot. The key protects your unavailability over against the platform, but it can also reduce the person to their key pair. That is precisely the temptation: to mistake cryptographic sovereignty for personal dignity. Creatureliness is more than control of a key.

§2. Trust is relational (web of trust), not merely algorithmic ranking

Technically: in the platform model a central algorithm decides who is visible, who is credible, who is “relevant”. Your influence over that ranking is limited, because you do not know its closed criteria and cannot affect them. In a web of trust, trust emerges from publicly traceable relationships: whom do the people I follow follow? Who is confirmed by people I trust? Trust arises along your relationship graph, not on a scoring server.

Theologically this is the difference between ascribed and attested credibility. Algorithmic ranking is a secular predestination: an invisible system has already decided who counts, and you only learn the outcome. The web of trust replaces this with something deeply relational: trust as testimony, passed on through a community. Structurally this is the logic of vocation: nobody calls themselves; authority arises relationally, through attestation within a community, not through self-optimisation in the rankings. And it is the logic of forgiveness and of love of neighbour, because someone vouches for you instead of leaving you at the mercy of a scoring system.

For religion-related education this strikes me as enormous: the credibility of content, of people, of code is not decided by SEO or engagement metrics (which structurally reward the loud, the polarising, the optimised) but by a network of people standing surety for one another. That is far closer to how communities of faith actually build trust: through invested time, relationship, shared practice. It carries better than any feed algorithm ever could.

The limit is sharp: a web of trust can tip over into exclusion. A network of trust is also a network of exclusion. Whoever is not attested, or does not present themselves as resonant, effectively does not exist. This reproduces existing structures of proximity: whoever is already inside becomes visible; the strange, the new, the other stays outside. That is the dark reverse side, and theologically it would be the perversion of community into cronyism, the opposite of hospitality towards the stranger. A web of trust without a deliberate opening towards the un-attested becomes a closed society. This is exactly where religion-related education has to offer resistance: at one decisive point the gospel always also shows itself as anti-web-of-trust, because it addresses precisely the one for whom nobody vouches. The technology does not carry this opening on its own; it has to be built in culturally, by the community.

§3. Coordination without a centre

Technically: there is no Nostr Inc., because relays are interchangeable (anyone can run one), clients can be swapped, and even the protocol is open. Coordination here emerges without a coordinating authority, solely through shared conventions (NIPs) and overlapping sets of relays, not through a centre that issues and enforces rules.

This is precisely the situation the church is actually in: purely institutional constitution as a self-evident framework carrying everyone is eroding. Declining membership, loss of trust and the end of the Volkskirche as a coordinating centre appear as a real and painful rupture of a framework once taken for granted, but not as the end of collaboration or community. And it is exactly there that the question shifts: the old question, “how does the institution hold this together?”, is replaced by a new one: what holds collaboration together when the shared institution no longer forms the foundation and carries it? Nostr is a technical answer to exactly this question.

The three classic answers:

• Institutional order (top-down, hierarchical): no longer carries - that’s the finding.
• Order through competition (proof of work, market, optimisation): buys cohesion with energy and competition - the Bitcoin/AI logic that creatureliness precisely does not want.
• Relational order (web of trust, shared practice, attestation): cohesion through relationship rather than through coercion or combustion.

relilab, edufeed and FOERBICO are in fact already attempts at relational order: networks without a dominating centre, held together by shared practice and trust rather than by institutional authority to instruct. Nostr gives this an infrastructure whose value decision is already in the protocol. That is the real point: you do not have to enforce relational order against the technology, because the technology is already built relationally.

The limit: the idea of “no centre” is admittedly something of a wishful dream, because in practice relays centralise (most traffic hangs on a few large ones), clients centralise (a handful of dominant apps set the de facto standards) and capital centralises (whoever pays for infrastructure has influence). Decentralisation looks more like a process variable than a state, because without active care it keeps drifting back towards the centre simply because centralisation appears more efficient. So it may well be that the old optimisation logic that made the platforms strong comes back through the back door. The carrying order will presumably have to be agreed anew and deliberately and at the same time it lives from a gift that reaches beyond the agreement: hospitality towards the one who did not take part in the negotiation. Otherwise the vacuum is filled by the next hyperscaler, the loudest relay operator or the agreement excludes again precisely the person that point 2 has just made room for.

§The common thread

The same movement runs through all three points: Nostr shifts order from disposed (platform, institution, algorithm, proof of work) to relational (key, attestation, shared practice). And the same limit runs through all three: relational order is not self-sustaining. It can tip into reduction (key = person), into exclusion (web of trust as cronyism), and into covert re-centralisation.

If we make ourselves aware of this, we may be able to use the relational infrastructure without sanctifying it, keeping open the fracture points at which the technology betrays its own anthropology. Then we would not be dealing with a tech adaptation for education, but with the construction of an ecosystem as educational infrastructure. One that knows the limits of technology and realises collaborative learning precisely because of that.