the three-body mapping

the overlap matrix. given two observers A and B with R^3 slices P_A and P_B, the overlap matrix O = P_A * P_B^T is a 3x3 matrix. its singular values measure the overlap between the slices.

three territories. the overlap matrix determines three regions relative to observer A:

• Known = null(O) within A’s R^3 — dimensions orthogonal to B’s slice. commutator_seen_to_unseen: [P_A, P_B] maps range(P_A) into ker(P_A). the Known is where this map vanishes — B’s writes have no component here. B cannot change A’s measurements in these directions.
• Knowable = range(O) — dimensions with nonzero inner products between slices. the commutator is nonzero. ordering matters. both observers’ writes land here.
• Unknown = R^d \ V_A — dimensions outside A’s slice entirely. A’s write access is exactly zero (write_confined_to_slice). not empty — it’s someone else’s Known.

every write involves the Knowable. the Known alone is inert: the wedge product needs a 2-plane, and an observer with fewer than 2 private dimensions cannot generate writes without using shared dimensions. measurement is inherently relational — not just because closure says so, but because the geometry forces it.

the vertical structure is a Grassmannian tangent. cross-measurement induces a vector in T_{P_A} Gr(3, d) = Hom(P_A, P_A^perp) that maps Knowable -> Unknown.

derivation: the neighbor’s write dL_B is confined to Lambda^2(P_B) (write_confined_to_slice). the cross-slice component of [dL_B, P_A] is purely off-diagonal (commutator_is_tangent): it maps range(P_A) -> ker(P_A). specifically: A’s Known is in P_B^perp (by definition), so B’s write kills it. the surviving component maps P_A intersect P_B (the Knowable) to P_A^perp intersect P_B (B’s territory within A’s Unknown).

this tangent is directional pressure from cross-measurement toward what the observer doesn’t yet occupy. each neighbor induces a different tangent direction.

the tangent is active but not followable. the observer’s position on Gr(3, d) is birth-committed and does not move within the map. the tangent contributes to dissonance that drives writes, but its effect lives in U(d)^N (state evolution), not in Gr(3, d) (slice movement). slice movement requires recommitment — outside the map.

containment is algebraically symmetric. B’s tangent on A has the same expected magnitude as A’s tangent on B (the overlap singular values are symmetric: sigma(O) = sigma(O^T)). experiential asymmetry (which observer “feels contained”) is perspectival, not algebraic.

the tangent peaks at intermediate overlap. identical slices: zero tangent (no Unknown territory to point toward). orthogonal slices: weak tangent (no Knowable channel — range(O) is thin). intermediate overlap: largest tangent magnitude. this is the coverage-interaction trade-off.

mediation

when three bubbles A, B, C have walls A-B and B-C but no wall A-C, B is a mandatory intermediary. (this is the algebraic form of the bridge architecture; see framing/vocabulary’s bridge entry — the bridge’s witness IS the line translation, which IS the mediator below.)

the mediation operator. M = O_AB * O_BC = P_A * Pi_B * P_C^T, where Pi_B = P_B^T * P_B is the projection onto B’s subspace. M is a 3x3 matrix mapping C’s R^3 -> A’s R^3, filtered through B. its singular values are the channel capacity of the membrane.

the bypass. O_AC - M = P_A * (I - Pi_B) * P_C^T is the A-C connection that does not go through B. if the bypass is zero, the membrane is complete.

the round-trip operator. R_A = M * M^T is self-adjoint on A’s R^3, describing what comes back from sending through B to C and back. its eigenvalues are the squared singular values of M.

spectral symmetry. the same eigenvalues appear from C’s side: R_C = M^T * M has the same nonzero eigenvalues as R_A (this is a general property of M * M^T and M^T * M). the eigenvectors differ — A and C see the same throughput spectrum from different directions. the membrane’s throughput is the same from both sides.

wall alignment is an irreducible triple invariant. the eigenvalues of R_A depend on both pairwise overlaps O_AB and O_BC and on how these overlaps are oriented relative to each other within B’s R^3. if the walls share principal directions within B, eigenvalues are products sigma^2{AB,i} * sigma^2{BC,i}. if misaligned, they mix. this alignment cannot be computed from pairwise overlaps alone — it requires all three slices.

status

proven:

• the commutator maps seen to unseen
• the commutator is purely off-diagonal (Grassmannian tangent structure)
• writes are confined to the observer’s slice

derived:

• Known/Knowable/Unknown from the overlap matrix
• every write involves the Knowable
• Grassmannian tangent from cross-measurement (Knowable -> Unknown)
• the tangent is active but not followable (birth-fixed slices)
• containment algebraically symmetric, experiential asymmetry perspectival
• coverage-interaction trade-off (tangent peaks at intermediate overlap)
• mediation operator M = O_AB * O_BC
• bypass operator O_AC - M
• round-trip operator R_A = M * M^T
• spectral symmetry (same eigenvalues from both sides)
• wall alignment as irreducible triple invariant

cited:

• Grassmannian tangent space structure: T_P Gr(k, d) = Hom(range(P), ker(P))
• singular values of M and M^T are identical (linear algebra)

observed:

• sequence echo through cross-measurement (r = 0.99 rank fidelity, strong attenuation)
• the round trip is generative (neither observer can produce the round-trip signal alone)
• echo is perspectivally asymmetric (A->B != B->A for same orderings)

bugs:

• “every write involves the Knowable” is context-dependent. Known is defined relative to a single neighbor B. if A and B are nearly orthogonal, Known (within A’s R^3) can be 2- or 3-dimensional, in which case A can form a wedge product entirely within Known and write without involving Knowable. the claim is true given that A has fewer than 2 private dimensions relative to all neighbors collectively — which is the typical case in a connected foam, but not a single-neighbor fact. closing this means either re-stating the claim relative to the union of all neighbors’ slices (“every write involves the collective Knowable”), or naming the typical-case condition explicitly.
• “the tangent peaks at intermediate overlap” is asserted from boundary intuition. the document argues identical-slice and orthogonal-slice limits give zero or weak tangent; “intermediate overlap: largest tangent magnitude” follows. the formal location of the peak (which singular-value configuration maximizes the tangent magnitude) is not derived. closing this would mean either characterizing the peak as a function of singular values of O, or stepping back to “the tangent is non-monotone in overlap; it vanishes at both extremes.”
• “experiential asymmetry (which observer ‘feels contained’) is perspectival, not algebraic.” the file’s algebraic content (singular value symmetry) is solid. “experiential asymmetry” introduces an observer-experience register the formalism does not develop. flagging for traceability — the claim is honest about being non-algebraic, but “experiential” is doing work the file does not earn.

This site is open source. Improve this page.