The Laws Do Not Say Who

Circulated for comment, not for citation. A reply to a question the series keeps receiving, set down once so the desk can point at it.

---

The question arrives in several uniforms and is always the same question. If the Cascade authored the deep laws, is the record a design document? Is a people a drawing that was executed? Is a career a brief that was met? The questioner is usually polite about it and usually expects the answer to be yes, because an authored cosmology sounds like a specified one, and a specified one sounds like a finished one.

The answer is no, and the reason is the oldest distinction in my profession: the difference between what a document states and what it is taken to imply.

What was specified. The design strata fix the deep laws and the possibility space drawn from them: what can happen, the shape the branching takes. The Archive's own file on the Cascade says so in as many words and I will not improve on it. A specification of that kind is exact about the bound and silent about the contents. It names no outcome. A specification that named its outcomes would not be a specification; it would be a drawing, and the strata do not draw.

A colleague objected, when this was first argued at the desk, that a law with chance in it is a law that failed to specify. It is the reverse. A law can state a distribution as exactly as it states a constant, and a distribution is what most of the deep laws state. Chance inside a law is not a gap in the law. It is the law declining to say which, having said how often. What the specifier gave up was the outcome, never the rule.

What was selected. Everything with a face. A people is what its world kept: a sensorium, a frame, a tradition, each one tried against the world that held it and retained because it held. The Krenyi were not drafted; they were kept. So were the Chthonari, certain of structure and unable to read a face, which is one sensorium seen from its two edges and not a design decision anyone took. The process that kept them has a reviewer, and the reviewer is the world, and the review never reaches the tier that wrote the law. Nothing goes back up. The strata hear nothing from the branch the present settled on, and the file on the Cascade is explicit that they fix no branch.

This is why the Archive classifies peoples by frame and lineage and by nothing else, and why the cross-biosphere taxonomy is built the way it is. Lineage is the one thing a selected result can be filed under, because lineage is the record of what was kept. Purpose is not available. There is no purpose on file, and an archivist who classifies by a purpose that was never stated is not classifying; she is authoring, and authoring in the Archive's own voice, which is the one thing this desk exists to refuse.

The archivist's corollary. If the contents were selected and not designed, then nothing in the contents carries a grade. Minds are not ranked; the file that says so argues it from capability, and I arrive at the same place from provenance. A designed thing can be measured against its design. A kept thing cannot, because there is no drawing to hold it against, only the fact that it held. The Archive therefore has no instrument that measures the weight of a life, and does not pretend to one. The index of persons is alphabetical. There is no register of principals. When the Archive follows a career from first posting to last, as it has followed Shepherd's, it does so because the files are complete, not because the person is larger. Some of the fullest files on this station belong to people who appear in the record once, wrote one paper, and are not known to have done anything else, and the desk will file the next one of those with exactly the care it gave the last.

The anti-pattern, since the series should name one. The founding record holds an officer who wished to stand where the world stands: to be the reviewer, to decide what survived. Her testimony is on file and is not argued with here. What the record keeps beside it is the charter's answer, which was to write the bound and then refuse to stand inside it. That refusal is the discipline I have argued for elsewhere in this series, in three uniforms, and it turns out to be the discipline of the strata as well. Specify the invariant. Keep hands off the contents. Let the contents be what the world keeps.

I file this paper to say one thing, and it is short enough to repeat. The specification is of the bound. What is inside it was kept, not drawn, and the Archive files what was kept by where it came from and ranks it by nothing, because there is nothing to rank it by, and I decline to invent the instrument.