# Reality Revision 1.48 — origami is an observation map, not a hidden-state proof

## Concrete addition

Matter Harmony Lab 0.5.0 adds an ordinary explicit-charge water response. A
neutral rigid three-site molecule with charges -0.834e, +0.417e, +0.417e has a
2.346971 D model dipole. The macroscopic reference is a single-Debye response
calibrated to the published 18.56 GHz primary relaxation of pure water at 25
degrees C:

    tau dP/dt + P = E,    tau = 8.575159 ps.

At the relaxation frequency the analytic amplitude ratio is 1/sqrt(2) and the
phase lag is 45 degrees. A separate numerical integration agrees to 3.529e-10.
These are ordinary dielectric mathematics and calibrated reference inputs, not
a prediction of water from hidden dimensions.

## Origami theorem and executed counterexample

Let the microscopic dipole state be x in R^(3N) and let the oscilloscope report

    y = A x,    A = (1/N)[I3 I3 ... I3].

For N=64, rank(A)=3 and its ambient kernel has dimension 189. The engine builds
two unit-dipole trajectories x_A(t) and x_B(t) whose transverse vectors cancel
pairwise. They differ microscopically by at least 0.688568 RMS at every stored
time, while A x_A(t) = A x_B(t) exactly in the generated arrays. Explicit
partial-charge dipoles reproduce the same equality to 4.441e-16.

Therefore a perfect polarization trace still does not reconstruct a unique
molecular movie. This is a written linear-algebra result plus a finite
constructed example. It preserves ambiguity instead of identifying an
unconstrained residual with a fifth dimension or theological cause.

## Limits

The two microscopic movies are controlled realizations, not molecular
dynamics. The model does not yet reproduce a held-out dielectric spectrum,
structure factor, diffusion coefficient, viscosity, magnetic response, or
quantum observable. The sampled-sinusoid alias remains: frequencies separated
by an integer multiple of the sampling rate agree at every recorded tick.

All 52 tests pass, and an independent implementation checks 92,160 atomic
coordinate scalars and the principal response/collision invariants. The next
test is a frozen multi-frequency complex-permittivity comparison on held-out
measurements, with a structural observable added whenever polarization alone
cannot distinguish candidate microscopic geometries.

Revision 1.14's idempotent complete deterministic closure and revision 1.15's
sampled non-crossing, angular-impulse, arbitrary-generator, and recurrence-
class corrections remain unchanged. The 2K continuation guarantee still
requires a noiseless homogeneous constant-coefficient recurrence of order at
most K from the initial state.