Methodology in Personal Technical Archives

A personal technical archive is, at its best, a slow correction to the failures of memory. It is not a publication; it is not a reference work; it is not a substitute for the larger literatures it sits alongside. It is, more modestly, a record of working observations made over time, kept in a form that allows them to be retrieved later.

What the archive is for

The archive serves three purposes. First, it allows me to refer to my own prior reasoning. A working note made in 2019 about a particular method may, in 2024, be either confirmed or contradicted by subsequent experience; either way, the prior note is useful, because the contrast between past and present reasoning is itself informative.

Second, the archive supports correspondence. When a colleague asks about a topic on which I have previously worked, the archive provides a stable location to which I can point them. The question I am asked is rarely the question I have previously answered, but the archived material is generally close enough to the asked question to be of use, and it relieves me of the obligation to reconstruct the reasoning from memory.

Third, the archive supports my own returning attention. A methodological note I made years ago and have since half-forgotten is discoverable, in the archive, by the same indexes and tags that I used when filing it. The archive is, in this sense, an extended memory prosthesis of a particularly limited kind.

What the archive is not for

The archive is not for publication in the conventional sense. The material here is not peer-reviewed, is not held to publication standards, and would be embarrassing to many of the literatures it touches. A reader who arrives at the archive expecting the conventions of an academic publication will find those conventions absent.

The archive is not, similarly, a reference work. There is no attempt at coverage, comprehensiveness, or balance. The selection of material reflects only my own working interests, and the omissions are generally larger than the inclusions.

The archive is not a record of finished thought. Many entries are provisional, and remain so. A reader who would prefer settled answers to provisional observations should look elsewhere.

Operational discipline

The practice of maintaining the archive is, in operational terms, unremarkable. New entries are added when the underlying observation has stabilised sufficiently to be worth recording; this is a judgement call, and I am sure I make it incorrectly in either direction regularly. Each entry is annotated according to the small fixed schema described elsewhere. Each entry has a stable opaque reference identifier. The mapping from identifier to entry is preserved indefinitely.

The discipline that this requires, in time and attention, is small but non-zero. It is, in my own accounting, repaid by the times when I have wanted to refer to a prior observation and have been able to find it.