Nine Years in the Archive, and What It Did to My Sentences

People ask where the research in my books comes from, and expect an answer about libraries. The honest answer is that I spent nine years being paid to handle the raw material of other people's disasters, and I have been metabolising it ever since.

I catalogued for a regional maritime collection: crew lists, insurance claims, harbour master's logs, the correspondence of a shipping firm that failed in 1931. It is not romantic work. It is dusty, repetitive, and conducted under fluorescent light to a standard of description so rigorous it borders on the theological.

The discipline of the finding aid

An archival description has to be accurate, complete, and entirely free of interpretation. You may write: 'Claim form, damaged nets, 1924.' You may not write: 'Claim form documenting the ruin of a family.' The second sentence might be true. It is not yours to add.

Nine years of that does something to a person's prose. I came out of it with an almost superstitious distrust of the interpretive sentence — the one that tells the reader how to feel about the thing it has just described. I still cut them compulsively. My editor has a shorthand for it in the margins: 'archivist'.

Where it shows up

Directly, in Nightjar, which is essentially a novel about the ethics of accurate description. Less directly everywhere else. The house clearance in The Weight of Quiet Things is a finding aid in narrative form: object, condition, disposition. The survey in The Salt Archive is the same trick again, with instruments.

I did not plan this. I noticed it around book three, when a reviewer described my work as 'forensic' and I felt caught out.

What I miss

The certainty, mostly. In the archive the answer to a question either exists in the box or it does not, and either way you will know by Thursday. Fiction offers no such comfort. You can work for a year on a chapter and still not be able to say whether it is any good, and there is no box to check.

But I kept the accession log. Every working day I record what came in — a phrase, an image, an overheard thing on a bus — with a date and a reference number. Four hundred index cards on a wall, and my partner's standing offer to help me file them, which I decline every time. The disorder is doing something. I am fairly sure the disorder is doing something.