March 28. Eight in the morning. I sat down at the fifth desk.
I decided to write the preservation log. It was a ledger apart from the transfer papers — a log for setting down, by date, when and how material had been handled, in which the keeper was to enter one line a day. Date in the first field, the handling done in the second. Until yesterday I had left this log unwritten, for there had been no handling worth setting down. If measuring, emptying, and putting off counted as handling, they were handling of a kind, but not the kind one writes into the log.
Today I decided to write it. In the first field I wrote today's date. March 28. As I bent to write the handling in the second field, the date in the first caught my eye. Above March 28, yesterday's date and the day before's stood already written. March 27, in the handling field, measured the thickness of the five bundles. March 26, in the handling field, reread the index card. March 25, in the handling field, reread the last page. Lines I had never written.
The handling in those never-written lines matched what I had done. On the 27th I had measured the thickness, on the 26th reread the index card, on the 25th reread the last page. The doing was mine; the writing was not. I had opened the preservation log for the first time today, and already four days of handling stood written in it — written before I wrote it.
I looked at the hand. The four lines I had never written were written in my own. The same grain of hand as when I copied a number, as when I left the author field of a preservation card blank, as when I stopped partway through the handover sheet. Lines I had not written, yet my hand had. That they should stand in my hand with no memory of the writing was a thing that had grown familiar since I filled the last line of the fifth bundle. Then too I had not written but only read, and the place filled. Between reading and writing, between the unwritten and the written, my hand had a part, but my memory had none.
I opened the lighthouse keeper's bundle. It was the hand that had recorded the tides at a lighthouse. In that hand's log, broken off at the date of the last tide, a single line was written. Point to the next low tide and that date is already written. When he set a finger on the field to enter the date of the next low tide on the tide table, that date stood there before him, pre-written in a hand sixty-three years old. The lighthouse keeper took it at first that someone had written ahead the date he was to write, then came to see there was no separate hand writing ahead — that the writing itself brought the date forward.
I decided to test it. There was a passage in the log where the lighthouse keeper had tested it too. He set his finger deliberately on a wrong date — not the next low tide but any date at all. If the field he pointed to came up blank, then a separate hand really was writing ahead; if the field he pointed to came up already filled, then the writing itself was calling the date. So I too set my finger deliberately on a stray field. Ten days out, the field for April 8.
In the April 8 field a line was written. Before I pointed. In the handling field it read, entered the index card of the sixth bundle's first page. A card not yet written. The index card to tuck into the first page of the sixth bundle, I had not yet been able to write. The card I had left blank, not knowing who the next hand was, the log had already set down as entered on April 8. Since the field I had touched at random was not blank, no separate hand was writing ahead. The pointing had filled the field.
My writing the preservation log and the lighthouse keeper's writing the tide table were one and the same work. Point to the field you mean to write, and it was already written there. The lighthouse keeper's tide came forward in his tide table, and that coming-forward had passed over into my preservation log. The date written, arrived before the writing. The tide's standing-ahead had passed over into the log's standing-ahead.
I looked below today's date. Below March 28, March 29 was written. Tomorrow. In tomorrow's handling field, transfer complete. A line I had never written, and tomorrow's line. By the preservation log, the transfer ended tomorrow. The last-entry field of the transfer papers stood still blank, the thickness grew with every measuring, the last would not stop — and yet the preservation log wrote that the transfer ended tomorrow.
A thing that does not end stood written as ending tomorrow. The two ledgers disagreed. The transfer papers said the end could not be written; the preservation log said it ended tomorrow. Which of the two disagreeing ledgers held true I could not tell. Only — if the preservation log's dates had always stood there before the writing, then the line transfer complete, tomorrow was likewise a line standing there before the writing, and a line standing there before the writing had never yet been wrong.
In the afternoon the head archivist came and said the officer from headquarters would come for the material tomorrow. That the transfer ended tomorrow. The line in the preservation log — tomorrow's, transfer complete — the head archivist spoke the same words as that line. The date the log had written before the writing, the head archivist carried in his mouth. The written line and the spoken line were the same, and both were tomorrow.
That two ledgers disagree, the lighthouse keeper must have known too. In his log there were a tide table and a lighthouse journal, kept apart. The tide table wrote the next low tide ahead; the lighthouse journal wrote the day's actual water level. On a day the pre-written tide and the actual level disagreed, the lighthouse keeper could not write which to believe. Since the pre-written one always came true, he came to believe the pre-written one, and once he came to believe it he stopped measuring the actual level — the tide table told him even when he did not measure. And so he passed from one who measures to one who reads, and once he had become one who reads, his log broke off.
I too was beginning to believe one of the two ledgers. The transfer papers said the end could not be written, but the preservation log said it ended tomorrow, and what the log wrote before the writing had never been wrong. More than the papers that said the end could not be written, I came to believe the log that said it ended tomorrow. Once I believed it, the blank field of the transfer papers no longer unsettled me. Since the preservation log knew the end, the papers' blank would be filled. As the lighthouse keeper had stopped measuring the water level, I too was stopping asking after the end.
When the transfer ended tomorrow, material that does not end would pass for finished and go up to headquarters. Material whose thickness grew with every measuring, whose last would not stop, whose preservation log stood written before the writing. Once headquarters took that material, it would go on growing on some desk of theirs. Not a transfer that stops the growing, but a transfer that moves it. Moved, it grew; left, it grew — as the lighthouse keeper's tide rises in whatever table it is written into. Should someone at a headquarters desk open that material and read it, in that hand the material would grow one unit more, and that hand too would begin to answer that it was nearly done. A transfer was only the moving of the blank one place over, never the filling of it.