August 9. The commencement report is due tomorrow.
Two things go into the report. When it begins and how much there is. The beginning can be written as today. How much there is has to be counted.
In the morning I saw that it is work that begins with counting. For six months I have learned that counting is opening and opening is increasing. To the one who learned it came the work of counting and reporting.
At nine I ran the search first. One hundred and fourteen. Last week, on August 2, it was ninety-three, so twenty-one in seven days. Three a day.
From ten until twelve I opened the six items one at a time.
Twenty-nine places in the catalogue program. Forty-four in the release-register program. Six each in the defaults of the four terminals, so twenty-four. Nine each in the two reel drives, so eighteen. The index program had the most at one hundred and sixty-four.
The sum was two hundred and seventy-nine.
While counting I learned two things. First, the year places are not conspicuous. What comes up on the screen is the material's name and number, and the year rides behind them. Figures like twenty-nine and forty-four were not counted off a screen but off a printed configuration list. Second, three of the six were made elsewhere. The catalogue and the release register are head office's, and the index was fitted separately by the archive last year. Where the making differs, so does the way of laying out the year place. Both are two digits, and one holds them as characters and the other as a number. The one held as a number goes from ninety-nine to a hundred; the one held as characters goes from ninety-nine to zero zero.
The side that becomes a hundred is wrong, and it goes forward. Only the side that becomes zero zero goes backward.
The cards I counted separately. Twelve drawers, about six hundred to a drawer; counting one drawer in fact gave six hundred and twelve. Multiplied by twelve, I set it at something over seven thousand three hundred and forty. A card has one year place to a card.
At eleven I sealed it.
Coming back from lunch, one thing was on my mind. Two hundred and seventy-nine is the figure counted in the morning. Counting increases. If it increases, then in the afternoon it is not two hundred and seventy-nine.
At half past one I counted again. I opened the six in the same order.
Two hundred and seventy-nine.
I counted once more. It was three o'clock. The same.
Standing there, I was relieved for the first time. There is a thing in this room that does not increase when counted. In six months.
Having written it, I went to strike out the word contrary and left it. It is not contrary. When I counted view counts in July, opening was the same act as raising. A board rises when opened. A year place does not rise when opened. Where the manner of counting differs, so does the way the counted value moves. It may be that what rose by three a day for six months rose not because of opening but because of days.
Having had the thought, I did the arithmetic.
There are one hundred and thirteen days until the last day of November. To correct two hundred and seventy-nine places at only thirty a day takes less than ten days. Seven thousand three hundred and forty cards at three hundred a day is twenty-five days. Together thirty-five days. Thirty-five out of one hundred and thirteen.
It was work with an end.
In six months I have not been given work with an end in this room. Counting increased it, correcting increased it, deleting increased it. Work that increases has no end. The arithmetic I was given today came out right, front and back, for the first time. The front is larger than the back and seventy-eight days are left over.
I thought for a moment what work with an end is, in this room. When the thirty-five days are filled, this work is gone. Once it is gone I go back to being a person who only reads. So to say it has an end is to say there is an after, and what the after is, the form does not say.
That the arithmetic comes out right ought to be comfortable and was not. In six months no comfortable arithmetic has come out in this room, so when a comfortable one came out I began by looking for where I had counted wrong.
At four I counted once more to find where I had counted wrong.
Two hundred and eighty-two.
Three.
The first three counts were all the same and the fourth carried three. By the clock it is four in the afternoon; by the day it is inside today. Where within a day the rule of three a day attaches, I have not once asked in six months. The search was only ever run at nine in the morning. Three attaching every twenty-four hours and three attaching when the day turns are different things, and, looking only at nine, I have never divided them.
Where the three attached I could see. The year-storage items in the index program had gone from one hundred and sixty-four to one hundred and sixty-seven. I opened the three items.
All three were places holding a material's registration year and all three had values in them. 97, 97, 98.
They are not places I made. This morning, when I counted one hundred and sixty-four, those places were not there. Places that had not been came into being, and coming into being they came with values attached. It was not empty places that appeared but filled ones.
1997 and 1998. Before I took the post.
I looked at the three until five. At half past four the man from the third floor came down for a signature and asked what I was staring at so hard. The years of materials that do not exist, I said. He said the index was always untidy. Things get left out and things get left over, and twelve years of several people's hands would do that. He was right. For the first time in six months an explanation is standing beside me, and this one fails only to cover that the places were not there this morning. Following the material numbers, none of the three has any material answering to it in the catalogue. They are places with a year and no material. An index points at materials, and these three point at nothing.
One hundred and thirteen days to the end of November, and three a day is three hundred and thirty-nine more. Two hundred and seventy-nine plus three hundred and thirty-nine is six hundred and eighteen. At thirty a day it is twenty days' work. The front is still larger than the back. The arithmetic still comes out.
Only, something was missing from that arithmetic. Correcting is also opening.
At six I closed the log and looked at the envelope. It is the commencement report sealed at eleven this morning. Two hundred and seventy-nine is written inside it. If I post it tomorrow, the figure the company receives is this morning's figure.
When the figure counted and written down changes while one counts, which figure goes in the report? The figure of the day it was counted is right. What the company wants to know is not today's figure but how much work this is, and for that a rough figure will do. For the first time in six months I am doing work where a rough figure will do.
I should not have opened the envelope. I opened it.
I looked at the quantity box on the report. Two hundred and eighty-two.
It was my handwriting, with no mark of having been written over.