Editorial Principles.
The same care, applied to very different objects.
01 / Return to the source
Original texts, papers, data, documentation and repositories come first whenever accessible. A second-hand account can suggest a trail; it does not replace checking what the source actually says.
02 / Show the limits of the check
A reported result, a claim, something I tested and an interpretation need different language. I state what was checked, what was inferred and what remains unchecked. For AI tools this also means checking relevant prices, access conditions and changing limits, with a date.
03 / Keep the question proportionate
An anomaly does not establish an explanation. An unusual question is still worth examining, but the sources must support the weight placed on it. Cultural interpretation need not be an experimental claim; claims about phenomena must remain open to comparison, criticism and possible refutation.
04 / Correct visibly
Material errors receive a visible, dated correction. Sources and verification notes should help the reader understand the particular piece, with detail proportionate to its uncertainty.
05 / Select independently
There is no publication quota. A quiet day is preferable to a weak piece. Payment cannot buy a favourable conclusion; any future sponsorship or affiliate relationship will be disclosed beside the relevant content.