#On the Sealed Witness
Deposition 7-K/Pell is the sworn account of Andros Pell, Bell-Warden First Class of the Elder Tower of Thessaloniki, concerning the failure of his bell in A.S. 198 (1908 CE). It is filed under Seal Amber. It runs to fourteen pages. Of everything produced by anyone during the Silence of Thessaloniki, it is the most disciplined document, and its author learned his prose on a tower slate.
The witness needs little introduction here. Pell's competence is set out at length in his dossier. For this file it is enough to say that a First Class warden may strike dawn, dusk, chain-window and alarm without waiting for a clerk's permission slate, and that in Thessaloniki those strikes are law. The harbour opens on sound and seals on sound. A captain who passes beneath the chains without a peal sails uncounted, and no sailor from Marseille to the Bosphorus does that willingly. When the Elder bell stops speaking, the man at its rope becomes the only witness to the failure from inside the bell-chamber.
The Bureau of Doctrine sealed his account for that reason. The seal notice gives its other reasons plainly, and they are better read in its own words than in mine.
#On the First Six Pulls
By Pell's testimony, it began at dusk on the fourteenth of Ashmonth, A.S. 198. He climbed to the crown-chamber at the prescribed hour and did what a First Class warden does before touching the rope. He verified rope condition, verified counterweight tension, verified the crown-chamber lamp, and checked the schedule slate. Then he pulled.
The clapper struck bronze.
Careless men have stated this badly, so it must be stated with care. The strike occurred. Pell felt the blow travel through rope and wrist, shoulder and boot, jaw and tower-stone, as he had felt it every dusk of his service. Instruments later confirmed the physical impact. As machinery, the bell was struck exactly as a bell should be struck. As an instrument, it produced nothing. No note left the bronze. The air of the chamber carried none. The city below waited for the lower tone that seals the harbour and heard its own waiting.
Pell pulled again: impact, then absence. He pulled six times in that first sequence, and each pull delivered the same pair of facts, a blow the body received and a sound the air never got. After the sixth, Chainmaster Kosta ordered emergency runners dispatched, and the matter stopped belonging to one man on a rope.
#On What the Fourteen Pages Retain
Eleven of the fourteen pages chiefly describe sensation. This ratio has annoyed every investigator who wanted cause, culprit and a clean diagram. It does not annoy me. A liar runs to the cause. A witness stays where it hurt.
Page one is administrative: time, station, personnel, weather, schedule. Those entries stand in the file. This Codex does not reproduce them, and I will not invent a sea-state to make the page look furnished. Pages two through twelve belong to the body. They record heat in the hands, pressure at the jaw, vibration in the stair and rope recoil after each strike. They record that the air did not move after impact. They record a difference, and Pell insists on it, between the silence outside the tower and the silence inside the bell.
He sorts that silence four ways: empty, held, swallowed, returned. These are his words, written by a man determined that one kind of nothing should not be filed as another. They are not a classification of agents. Doctrine has adopted them as categories of nothing at all, and no reader should mistake a bell-warden's precision for a statement about what occupied the bronze. The Bureau of Rites found the taxonomy undisciplined. The Bureau of Bells found it useful. The seal fell on both opinions.
Page Nine survives as a readable extract, except for two passages.
DEPOSITION 7-K/PELL, PAGE NINE: “At the fourth pull I perceived resistance in the silence, as though the air between bronze and ear had ████████████████. At the fifth pull the rope returned warm. At the sixth I heard, beneath the absence, a note moving downward through water, not from the bell but from ████████████████. I did not answer.”
This Codex holds the extract as sealed. It holds nothing beneath the bars, and it will not guess.
Read without the blacked-out words, the extract is a sequence: resistance at the fourth pull, warmth in the returning rope at the fifth, and at the sixth a note perceived moving downward through water. It closes with four words of conduct.
Page eleven adds that the absence in the Elder "moved downward after each strike." Pell wrote this without access to Diver-Matron Sera's report. That report is separate testimony about the submerged chain-links, and it was in nobody's hands at the rope. Anyone with the grade to read both may lay the two accounts side by side. Neither may be bound into the other with the binder's awl.
#On the Rope and the Audit
The last two pages record what Pell did once the first sequence failed. Warning pulls were suspended. An emergency slate was sent. The Elder rope was secured. The crown-chamber was closed against civilian entry. A secondary mechanical audit followed, witnessed by two chainwrights and one Rites acolyte. The deposition does not give their names, and this Codex will not supply any.
The audit found no evidence of mechanical fracture, no evidence of clapper misalignment, no evidence of foreign matter, no evidence of sabotage.
I have seen that sentence quoted by men who wished it said more. It says what it says. An audit that finds no fault in the bronze has established that the bronze showed no fault by any measure the auditors possessed. It has established nothing about anything else. Engineers grieve over a finding of that kind. Theologians, who ought to know better, treat it as a permission slip for whatever they had already decided.
The safeguards governed the pulls that followed. At each scheduled hour Pell stood at the rope, waited for the countersign from below, and pulled under witness. Every slate entry took the same form: DUSK — STRUCK — NIL; DAWN — STRUCK — NIL. He never wrote failed. A failure implies an attempt that fell short. The bell had received the blow and withheld the note, and Pell recorded exactly that much.
#On the Dated Return
Pell gives two dates. The first silent pulls came at dusk on the fourteenth of Ashmonth, A.S. 198. The return of sound came on the morning of the seventeenth. The Codex dossier on Pell, drafted in this office, calls the interval seventy-two hours and names the return the fourth dawn. The event plate says the same. The dated anchors do not agree with those phrases, and I will not force them to. Pell's dates are Pell's testimony. The seventy-two hours are the city's phrase and the Bureau's, repeated until they acquired the dignity of a number. This office records both and reconciles neither. A sealed witness is not improved by a clerk with an abacus.
On the morning of the seventeenth Pell pulled, and the Elder rang.
It rang the Younger's note, the higher coastal tone that belongs to the other tower and to Bell-Warden Second Class Katerin Liss, whose account of the Younger stands in its own record. The bronze was the Elder's bronze and the rope was Pell's rope, but the voice was borrowed. He held the rope until the note died on schedule. Then he wrote DAWN — STRUCK — YOUNGER TONE PRESENT IN ELDER, and underlined nothing.
Operations resumed with that entry. Normality did not. The harbour opened and sealed again, and the bell that opened it spoke in another tower's voice. As of A.S. 201 it still does, and Pell is still at the rope under restricted observation.
A training précis circulated in A.S. 199 states that Pell "restored normal operations" on the fourth dawn.
Withdrawn. He restored operations. The Younger's note sounded in the Elder's bronze, and no competent reader of his slate could call that normal. The précis stands amended. The slate never required amendment.
#On What Seal Amber Allows
The seal notice is short and should be read whole: witness reliable; metaphorical content restricted; sensory observations retained; theological conclusions excised; distribution limited to Bell-Warden Grade and above.
Each clause does a separate job. Reliable is a finding about the man. Retained governs the body pages: the heat, the jaw, the recoil, the four silences. Restricted governs his comparisons. The most famous one escaped anyway: striking the bell, he said, was like shouting into a pillow held by someone who is not there. Rites has spent years wishing it back into the file. Excised governs whatever Pell concluded, theologically, once he stopped describing and started explaining. The excision does not record its subject. Readers who supply the Drowned Choir as the missing subject have supplied it themselves, and the file does not thank them.
The arrangement keeps what can be checked and removes what can be argued. The Bureau of Bells receives an exact record of a strike without a note. The Bureau of Doctrine keeps its determination, 198-K/7, free from competition with a bell-man's theology. That determination calls the event a test of fidelity and names no examiner. The deposition, stripped of conclusions, names none either. The two documents therefore agree, and agreement of that kind is the only kind Doctrine can certify.
The Creator, if He examined Thessaloniki, left His marks on a bell-warden's jaw and nowhere a clerk could stamp them.