Chapter 37
Hangzhou, August 1998
The batch came in with certificates and failed the entrance inspection: the wheel fasteners, the nuts that hold a hub on, hardness below specification across the lot. It was counterfeit material — melted and dressed as something better, which in 1998 was not an event but a weather. The purchasing department had already promised the batch to a coach builder whose line would stop without it. There was a meeting, in the room with the schedule board, with the delivery dates written in two colors: what we had promised, and what we had made. In the end the instruction was to release it, and the release required the inspector’s stamp on the inspection record — the three-check system, the countersigns, the personal chop in the box beside the form. I was the inspection engineer on the shift. The form came to me with the material status already argued through; my chop was the last box in the column, the one that made the paperwork a fact.
I had been four years in the quality section. The line pressure was real; the schedule was real; the man who made the decision was not me, and the chop on the form was.
A coach went off a mountain road that autumn. The report that came back through the quality section said twenty-six dead. The fasteners were in the wheel assembly; the metallurgy was as the entrance inspection had found it in June; my stamp was on the release.
I did not learn of it the way men in films learn of things — not at once, not from a single source, but in the slow arithmetic of a plant: the corridor going quiet when I entered it, two men from the works office standing at the quality section’s end with a folder held the way you hold something you would rather not have brought, the director’s door open for once in the direction of my desk. The report itself came by internal routing, in a gray sleeve, with the covering note unsigned and the date typed. I read it standing up. Then I read it sitting down. I left at the end of shift. Nobody noticed. Then I did the thing I have done with every unmanageable number since, which is to write it in a column of its own and put the column where the other columns could not see it.
I have kept the clipping. It is folded in the desk drawer at home, under a warranty card and a hotel receipt. Twenty-six — that is the number the report used, and it is the number I have used, for twenty-six years, because he never let them become anything else, and I am not going to give them faces now in order to be seen giving them faces. In the years since, I have watched a hundred tests: the sled, the barrier, the freeze-frame at contact. The dummy always lives. That is the design of the dummy. There was no sled in 1998. There was a road, and a number, and my stamp.
What I did in the weeks after is the part I have rehearsed least, because it is the part with the least to say. I was interviewed, as the inspection engineer of record, by a panel of three from the works and one from outside. They asked about the certificates, the entrance test, the release order, the chain of instruction. I answered each question with the truth that fits inside a question: the certificates were presented as valid; the hardness failed; the instruction to release had come from above me; my chop was on the form. Each answer was accurate. None of them volunteered anything, and volunteering is the thing I did not do, and there is no form on this earth with a box for it. The panel wrote what the panel could defend, and the file closed with a director’s countersign at the bottom of the report. I read that report twice, the way I read everything, and I have never said his sentence aloud to anyone in all the years since, and the not-saying has no date on it because it went on without stopping.
In the weeks between the release order and the report there was the particular silence of a plant that knows something without being told: the canteen going quiet when I came in with my tray, two inspectors from my own section finding reasons to be at the other end of the floor, the schedule board still carrying the coach builder’s dates in the column of what we had promised, which nobody crossed out, because crossing it out would have been a decision and the whole plant had by then agreed, without a meeting, to stop deciding anything. I went to work. I stamped forms. The forms were correct. That is the part no one writes down: that the machinery of a wrong decision keeps running on ordinary time for weeks before the number arrives, and during those weeks it is indistinguishable, on the floor, from the machinery of a right one.