# n56 — the derivation of a: what fixes the monitoring rate's scaling
# (memo written before the toy; honesty grade declared inline)

## The question
n55 reduced the theory's PPN gamma to one exponent:
gamma = (1-2a)/(1+2a), a = d ln gamma_drain / d ln k. GR iff a = 0.
Is a = 0 derivable, or a wish?

## The argument, in three steps

1. WHAT IS MONITORED IS BOOKKEEPING, NOT ENERGY. The theory's defining
   axiom (n39, measured): the vacuum reads only violations of its own
   sum rule, v = A dk + c — a dimensionless bookkeeping object built
   from coupling SHIFTS and consumption counts. It carries no factor of
   the local scale k. The minimal Lindblad monitor is L ~ v with a rate
   set by the monitor, not by what is monitored.

2. THE MONITORED SECTOR SITS AT ZERO FREQUENCY, AT EVERY k. The Gauss
   sector is the exact zero mode of W (measured to 1e-32, n46b), and it
   stays the zero mode under any rescaling of k — the conservation law
   protects it. A monitor enforcing a STATIC constraint samples its
   environment at DC; substrate stiffness moves the phonon frequencies
   (sqrt-k), never the constraint sector's frequency (exactly zero).
   There is no frequency for k to enter through.

3. WHAT a =/= 0 WOULD REQUIRE. A k-dependent rate needs the monitor to
   weigh violations by the local scale: L ~ f(k_local) v. Nothing in
   the axioms sources such a dressing; it is an extra dimensionful
   structure inserted into the one sector the theory defines to be pure
   bookkeeping. Occam does not merely disfavor it — the theory as
   stated does not contain it.

## Honesty grade, stated plainly
This is a MINIMALITY derivation, not a theorem: a = 0 follows from the
monitoring axiom read minimally (the vacuum audits its ledger in ledger
units). A determined model-builder can write L = sqrt(k_local) v by
hand and get a = 1/2, gamma = 0. What we can say exactly:
- the theory's OWN axioms provide no source for the dressing;
- the choice is OPERATIONALLY EXPOSED: a is measurable. A k-dressed
  monitor makes the n48 decoherence knee sag with the local potential
  (delta-knee/knee = a * delta-k/k); the bare monitor keeps it rigid.
  gamma_PPN = 1 and knee rigidity are the same parameter measured in
  two places — the solar system already constrains it to |a| < 5.8e-6.
- an earlier hope for a stronger forcing ("k-dressing = state-reading =
  BMV-null branch") FAILS at rest and is recorded as failed: at v = 0
  the dressing multiplies zero; the branches decohere identically. The
  dressing is invisible at rest and visible only in the knee — which is
  exactly why the knee is the right observable.

## Quantitative note on the handcuff (correcting the memorandum's zeal)
The knee-sag observable exists but is weaker than Cassini in practice:
solar-system potentials give delta-k/k ~ Phi/c^2 ~ 1e-8, so knee
rigidity tests a at the 1e-6 level only with 1e-14 fractional knee
metrology — beyond foreseeable optomechanics. Cassini is and will
remain the binding constraint on a; the knee's real role is the
DISCOVERY channel (existence: knee vs floor vs nothing), not the
precision channel. The memorandum's "handcuff" is conceptually exact
and practically one-directional. Recorded.

## n56 toy (illustration, not proof — declared)
The n48 machinery with two monitor variants: L = v (bare) and
L = sqrt(k_local) v (dressed). Gates: (i) gamma_eff vs uniform k:
exponent 0.00 vs 0.50 (arithmetic made visible); (ii) the knee
position: rigid vs sagging by a * delta-k/k; (iii) protection at rest
IDENTICAL in both variants (the failed forcing argument, shown failed
honestly). The toy demonstrates the observable; the case for a = 0
rests on step 1-3 above.