A Query:

Me to Kai: How many relational dimensions?

There isn’t a single “correct” number, because “relational dimensions” are basically the minimal set of independent axes you need to predict how a relationship behaves in the world you care about.

But in practice, you can treat it like this:

• 2D is the floor (you + me, closer/farther; aligned/misaligned).
• ~8–12D is where relationships start to look real (because you can represent trust, power, time, obligation, etc. without collapsing them into mush).
• 20D+ happens fast once you include context-specific axes (money, health, kids, work hierarchy, cultural tikanga, legal duty, etc.).

Here’s a usable “core set” of 12 relational dimensions that shows up almost everywhere:

1. Proximity / Contact (how often, how close, access)
2. Alignment (shared goals/values vs conflict)
3. Trust / Reliability (keeps promises, predictable)
4. Care / Benevolence (wants your good, not just utility)
5. Power / Leverage (who can impose costs, who can exit)
6. Consent / Boundary Respect (stop-wins, coercion, pressure)
7. Reciprocity / Balance (give/take symmetry, exploitation risk)
8. Truth / Transparency (honesty, hidden ledgers, deception)
9. Obligation / Duty (roles, commitments, explicit contracts)
10. Time / History / Momentum (track record, decay, repair)
11. Third-party Entanglement (family, teams, institutions, publics)
12. Safety / Harm Potential (emotional, physical, financial risk)

If you want the very short answer: 12 gets you most of the way to “human-real.”

If you want a tighter “minimum viable” set for quick modelling, use 8D: proximity, alignment, trust, care, power, consent, reciprocity, time.

And if you’re doing Stanspace properly, the key move is: treat “new relational dimensions” as new independent axes only when collapsing them causes prediction failures. That’s how you keep it crisp instead of turning into an everything-bagel.