THE DRIFT / NORMALIZATION OF DEVIANCE
Almost nothing in this collection arrived by decision. The taxonomy is published years before the removal. The shell forms over decades. The banner is believed before it is convenient. The connecting mechanism is drift.
Diane Vaughan, in her 1996 study of the Challenger launch decision, named the process normalization of deviance: a departure from a standard occurs, no adverse outcome follows, and the departure is thereafter treated as acceptable. The revised baseline becomes the reference point for the next decision. Repeated, the process produces an organisation whose operating standard has moved a long way from its stated standard, staffed by people who each approved only a small movement.
Vaughan's finding, and the reason the study is cited here rather than a simpler account, is that the process required no villain, no concealment and no incompetence. The decisions were made through the proper channels by qualified people applying the evidence available, and the evidence available had itself been shaped by the previous departures.
WHY DRIFT IS THE ENGINE. Each of the other faces requires a first instance that would be visible if performed all at once. Nobody accepts the full taxonomy on day one; nobody signs the whole ledger. Drift solves this by never presenting the whole. It presents only the increment, and the increment is always defensible, because it is always small and because the last one was survived.
The reference point moves with the practice, so at no moment does anyone experience themselves as having crossed a line. The line is behind them, and it moved.
THE TEST. Comparison against the present is useless, because the present is the drifted position. The only usable comparison is against a fixed external reference: a written standard from before the drift began, an outside auditor with no history in the organisation, or a person encountering the practice for the first time. The last is the cheapest instrument available and the most commonly ignored, because the newcomer's objection is heard as inexperience.
THE RELATION TO THIS ARCHIVE. This collection is written at a fixed point in time by someone who will drift. The records are dated and versioned for that reason. If a future edition of the Archive finds these standards naive, the correct question is not whether the standards were naive, but which direction the reference point has moved and whether anything outside the Archive was consulted about it.