On the Filing of the Galley Door Specimen

UNDERCOVER PETS DETECTIVE AGENCY — WORKING PAPER Not a hazard finding. Not submitted to Eden's compliance framework. Filed to have been written down.

---

1. What I am disputing

Eden's hazard register carries the Deck 7 galley door specimen as a lamb-mask, Class III, non-sapient. I am not disputing the feeding pattern, which is textbook and which I would have filed the same way from the same desk with the same five minutes. I am disputing the class.

I want it on record that the filing was not careless. A duty officer had a description, a feeding pattern, and a form with a field on it that would not accept being left empty. Everything after that followed correctly. This is how nearly all of Eden's wrong paperwork gets made, and almost none of it gets made by anyone doing their job badly.

2. What the observation holds

Nine months, continuous, most of it recorded by a trainee assigned to the approach for unrelated reasons before anybody thought the approach mattered.

It learns a door. When the warm-air outlet moved in the refit it tried three wrong ones and then had the right one, inside two days. Instinct arrives at the right door or at no door; it does not arrive at the right door by way of three wrong ones and remember afterwards which three.

It has a preference it does not act on. There is a cook it waits for and does not work. I have no account of this that fits Class III. I have no account of it that fits Class II either, and I would ask any reader tempted to supply one to notice how badly we all want the tidy answer here.

It gets bored. Yields on the approach fell off over the summer and it went to the laundry, where there was nothing to have, and stayed eleven days, and came back.

And it stopped trying me in the second month. Not avoidance. Avoidance would be a plan and there is no plan in it. It stopped the way any of us stops at a door that has never opened, which is to say without deciding to and without being able to say when.

3. The calibration, and what is wrong with it

The specimen benchmarks at roughly the order of cognition of a certified Smart Pet. I am prepared to put that in writing and I am aware of what is wrong with putting it in writing.

The instrument is me. I have measured how clever a thing is by how long it took to work out that I was not worth working, and by how much that resembles what I would have done. There is no reading available to me that is not taken through my own head, and a cat is not a disinterested judge of whether a thing is as clever as a cat. I cannot separate the finding from the flattery in it. I have looked for a way and there is not one on this habitat.

The honest form of my claim is therefore narrower than the claim: nothing I have observed is explained by the class it is filed under, and the nearest thing I have to a ruler is myself. Somebody with a better instrument should be asked. Nobody has been, because asking requires a form, and the form requires the name.

4. The part I would rather not have got to

If the finding holds, the specimen is an animal of about my order, carrying no subsystem, holding no certification, and covered by no welfare instrument this habitat or the Compact has ever written. My own welfare sits formally outside the Warden's habitat authority through a gap in Custos's remit that nobody has yet asked anyone to close. The specimen sits outside a wider gap than mine, from the other direction, and I notice that no part of me went looking for that until the classification made me.

My operational position has not changed and I am not proposing that it should. The counter is what it has always been: somebody refuses it out loud, in front of the others, and is rude for four seconds. I would like it recorded that I know what I am asking to have done, and to what, and that I am asking for it anyway. It is not a hard call. It is only that a thing can be an easy call and still be worth writing down that it was a call.

5. Provenance of the evidence

The observational record in section 2 is, in substance, the trainee's log. I have read all of it, checked the dates against the refit schedule and the rota, and found no error in nine months.

— Agent Barsik, Eden posting

---

Verification Desk annotation, Survey Archive: lodged at the author's request rather than through any institution's submission channel, and held on that basis. The Archive takes no position on the classification argument, which is outside its competence and inside a live dispute; it notes only that the author states his instrument problem himself in section 3, unprompted and at greater length than the finding, and that this is not the usual shape of a document written to persuade.

One observation the Archive will make, having no standing to do more than make it. The evidence base of this paper is another agent's log across the whole of section 2, and section 5 credits it. The trainee is not named, is not asked what she concluded, and does not appear anywhere in the paper as a witness — only as a record. Whether that is the ordinary economy of a senior officer's working note or something the paper is about without knowing it, this desk cannot say, and does not need to in order to file it.

See also: Muffin, Agent Barsik, Bubochka, Predatory Entities, Meta-Dimensional Beings, Smart Pet, The Cat Who Would Not Say.