October 12. In the morning I wrote my four digits on paper and matched them against the screen.
It is the first time since January that I have written the four digits on paper. There is no occasion to. Putting them in a box on a form, the hand writes and the eye does not look. Today I wrote them in order to look.
I matched them place by place. All four were the same.
At nine I ran the search. Three hundred and six.
Having written agree, I had to write down why it makes no sense.
The archive code was changed last year. When the departments were reorganised in the winding-up after the merger, the four digits now in use came into being. Before that the number of places was different, I am told, and being told is not grounds. I went down to the stacks for grounds.
I looked at the recipient's box of the 1997 paper register.
It was five places.
Five places is not four. Two years ago this room's code was five places and now it is four. So four digits cannot be in a 1997 register.
What cannot be is on the screen in five lines.
At half past ten I looked out the transfer guidance. It is among what I was handed over in January. It says that when registering old material one is not to assign new codes but to carry over what is written on the original. No conversion table was made. There is no instruction to make one and I have never made one.
Then the five lines on the screen carrying a new code may be something put in by someone at transfer. If the guidance was breached, that is how it would be.
One thing blocks it.
The recipient's box on the same screen is five places.
To put a changed code inside one material, one must pick box by box. Leave the recipient and change only the recorder. And among the recorders, leave the two hundred and fourteen that hold names and change only the last five. There is nobody in this room to do that picking.
At half past eleven I checked again what the four digits point to.
The first two places are the department and the last two are the post. The post means the duty. This room has one post. One records keeper is attached to the computer room and that one is me. So these four digits point to a department and at the same time to one person.
Had it pointed only to a department it would have been less bad. The department existed two years ago too. That the last two places narrow to one person is today's matter.
From twelve until one I did cards. Some three hundred are left in drawer seven.
Lengthening two digits into four by hand, I thought again about this morning's matter. What I do is lengthen places. 97 is two places and 1997 is four. When places are lengthened, one value points to one year only.
The code went the other way. Five places shortened to four. Shortened, the values run short, so somewhere they overlap. Five places two years ago and four places now are different systems, so I have held that there is nowhere for them to overlap. With nowhere to overlap the same number cannot come out, and it has come out, so there is somewhere.
Where it overlaps cannot be seen by counting. The allocation tables of the two systems do not survive. The old system was destroyed in the winding-up, I am told. Being told is not grounds, and what is not there cannot be looked for.
At one I telephoned head office. The third time in three days.
I asked when the four-place system came into being. Last July, he said. I asked what it means if four digits appear in a record from before that. He said a transfer error. Codes were newly assigned at transfer, so send in the instance.
I said the transfer was mine and that I carried over the originals per the guidance. He was silent a moment.
Then somebody else has touched it, he said.
There is one account in this room that can register, I said. He said in that case to look at the access record for who came in on that account.
The access record I pulled whole four days ago. Nine hundred and twelve rows. From two until four I looked in it for lines that had touched the 1997 register.
Lines that queried it are three. Twice on 9 September and 8 October. All three read it and did not write it.
There was no line that entered anything.
So those five lines were not entered by coming in on this room's account. The way the man gave me is blocked too.
The party entering is unknown, I wrote. It is a thing I have written many times in eight months.
I counted the places in eight months where my four digits have appeared in this room. Four. In June as the display name on the board, in August in the recipient's box of the handover document, in August in the third of the three index places that came into being, and today in five lines.
The first three are all inside this room. I put them in or my account did. Today's is where a paper of two years ago is.
At half past four the third-floor man came down and took a signature. On his way out he said a call had come upstairs looking for me, and they gave the four digits, so nobody knew who it was. Give the four digits and it is me, I said. He laughed and went up.
After he went up I wrote it down. What calls me in this room is four digits. Four digits on paper and four digits on the screen. A name is for hearing on the telephone or being called in the corridor.
What is in the last five lines of the 1997 register is not a name. The two hundred and fourteen before them are names. A name is the word that calls a person and four digits are the word that calls a post. The five lines did not enter a person; they entered a post.
Enter a post and the line is right whoever sits in it.
Enter a name and that person must have been there. A name attaches to one person, and if that person was not there the line is wrong. A post is otherwise. A post is empty, then filled, then empty again. A line that enters a post is always right while the post exists.
The two hundred and fourteen are names, so they say a person was there two years ago. The five are a post, so they say only that the post was there. Two years ago this room's post was called by a five-place code, so it is not that post. Then the post those five lines entered is a post that did not exist two years ago.
A line that enters a post which did not exist is right after that post comes into being.
It is the second time in seven days that I have written that I can find no grounds for correction.
For eight months in this room I have seen figures that do not agree many times. Not agreeing, there is somewhere to correct. Today's four digits in the five lines agree. They agree place by place, they agree by the system, they agree by the form. They agree and they cannot be.
Where what cannot be is right, there is nowhere to correct.
At half past five I opened those five lines for the last time. I had not yet looked at the entry-time box.
The register keeps the date and the entry time apart. The paper has no time box and the screen has one.
The times of the five lines were all before dawn.
Two fifty-one. Two fifty-nine. Three o'clock exactly. Three oh four.
They were the same as the four I saw in the access record four days ago.
The fifth line's time was three eleven. That one was not in the access record.