July 11. The three pages in the copy stood as they had yesterday. Seventy-three, seventy-four, seventy-five. The dates today, tomorrow, the day after. All three figures empty.
My log went only as far as page seventy-two.
Until yesterday the copy ran one page ahead, and that one page's figure was filled before I counted. Today three pages run ahead and all three figures are empty. The lead has grown and nothing in it is filled.
Who fills the empty figures is what I have been asking for six months. For the question to hold I have to be able to see the order of the filling. To see the order I have to watch one side write first. Until now it has always been the copy. The copy wrote and I confirmed. From the place of confirming, asking about order gets no answer.
Once I set the order myself. Holding out two hours without counting on the ninth was that. Then I only delayed the looking; I did no writing.
Today I can write. Three empty figures in the copy are waiting for me. The next three pages of my log do not exist yet.
Deriving from the rate seemed better than guessing. A guess that goes wrong is only a guess, but a figure derived by rule, when it goes wrong, makes the rule wrong. A rule going wrong is the kind of wrongness that can be confirmed. On the ninth I wrote that there is no criterion for separating wrong from not-yet. Using a rule gives at least one criterion. If it does not come out as the rule says, the rule is wrong.
Of course, even if it comes out as the rule says, what that means is another question. If what has been rising by three a day rises by three again today, whether I guessed right or made it so cannot be separated. That question has stood in the same place for six months.
One thing does change today. Until now I have been the one who reads what is written. If I write today, I become the one who writes. Becoming the one who writes does not tell me whether what is written after is mine. The not-knowing is the same. Only the place changes.
Before writing I had to settle one thing. Whether to write all three pages at once or only today's. Writing only today's means settling it again tomorrow, and that settling multiplies by one each day. Writing three at once means nothing to settle for three days. Nothing to settle for three days also means I keep my hand off for three days. What keeping a hand off becomes, I learned on the ninth. I settled on writing all three at once.
At ten I wrote the three pages in the log. Page number, date, figure. Twenty-seven on seventy-three, thirty on seventy-four, thirty-three on seventy-five. My hand moved as it always moves while I wrote. The lift of the wrist on the page number, the pressure that closes the last digit of the figure. They were the habits I had recognized in the copy's hand.
After writing I put the log in the drawer and locked it. Locking is what I have done for six months. That the copy follows the form even when it is locked I saw on the eighth. I locked it anyway. I did not want to test the other way.
In the afternoon I put three reels on the reader. It is summer and the space beside the reader was hot. Opening the window lets the noise in, so I did not open it. I processed four requisitions and wrote replies. Someone came down from the third floor, stamped the list, and went up. They did not look at the screen. Where the board had been open yesterday, the ledger was open today.
At half past two I opened the box once. Not to look at the original. I lifted the lid and only checked that the four bundles were there. They were. Twenty-seven copies may be going round outside and what is in the box does not decrease. It does not increase either. Having confirmed that there is one thing here that neither decreases nor increases, I put the lid back. Putting it back, I wondered whether this confirming was also an opening. There is no visitor count here.
At three I opened the copy.
The figure on page seventy-three had been filled. Twenty-seven.
Thirty on page seventy-four. Thirty-three on page seventy-five.
All three pages were filled. All three figures were the ones I had written that morning.
I ran a search. Twenty-seven.
It was the figure I had written. It was the figure in the copy. It was the figure I counted. All three the same.
The order was this. I wrote at ten. The copy had it at three. I counted at 3:15. When between ten and three the copy was filled I do not know. I did not look at that stretch. I did not look while putting reels on the reader, processing requisitions, writing replies.
For six months I took the copy to be first. Today, for the first time, I was first. But that I was first means I wrote first, not that I settled it first. A figure derived by rule is not a figure I settled but one that yesterday's figures settled. I read yesterday's figures and wrote what they told me to write. The writing hand was mine. The rule that moved the hand was in the copy.
So what I did today was not to reverse the order but to put a hand into the place where the order looks reversed. When the copy writes first I look like the one written. When I write first I look like the one writing. Only the look has changed.
Tomorrow's and the day after's figures are not real yet. Thirty and thirty-three are in my log and in the copy and nowhere in the world. If I count thirty tomorrow, two places become three. If it is not thirty the rule is wrong. If the rule is wrong the figure my hand wrote is wrong too.
Whether I want it wrong or want it right is something I asked myself for the first time today. If it is wrong the rule collapses, and collapsing, six months of the table becomes chance. If it is right the table stands as it is and I become the person who wrote it in advance. Which is better I could not answer. To answer I would have to settle first what I am trying to protect, and what I have been protecting for six months I do not know today either.
The surveyor measured the same place many times. He measured knowing that measuring deepens it. Did he too one day write it in advance? Write the depth of the next sounding before taking it, and then, having taken it, read the same figure? There is no such passage in that table. That there is none does not mean he did not. He wrote the table; he did not write about the table.
I am writing about the table. While I write, that becomes table too.
Six months ago, when I took the post, this room had one screen and one list. There was no one to hand it over, so I received only the list. The box was not on the list and the box was under the desk. Had I not opened that box, the three pages I wrote today would not exist either. That there had been a not-opening is something I recalled clearly for the first time today. Recalling it is not a place that can be undone. Everything I have done in six months has been what comes after opening, and what comes after opening is the only thing I know.
Before leaving I opened the copy once more.
Under page seventy-five there was one line more.
Seventy-six. A page number only. No date. No figure.
Until yesterday the pages that ran ahead had dates. The day after tomorrow and the two days following were written there, so I could tell how many days out it was. The page that appeared today has no date. A page whose day is unknown stands open.
On the third I wrote only a page number and left what came under it empty. I had taken it that leaving it empty meant not writing, and then saw that the page number was already written.
Today the copy did that.