Method
How this archive treats evidence
Every entry in the Codex names its sources. Not "ancient texts" or "legend has it", but the book, the chapter, the collector and the year: Prose Edda Gylfaginning, 1 Samuel 17, Argonautica Book 4, Barandiarán working in the field in the Basque country, Sébillot's questionnaire of 1883.
The catalogue is split in two, and the split is the point.
Myth and Folklore holds traditions that are documented. Someone wrote them down, and we say who and when. Where the manuscripts disagree with each other, the entry says so rather than quietly choosing the version that reads better.
Bones and Shadows holds claims. Modern sightings, disputed finds, stories with no named witness and no photograph. These entries exist because the claims exist and people go looking for them. They are labelled unverified, and they stay labelled unverified no matter how good the story is.
Some entries here are deliberately short. Where a tradition cannot be reliably attributed to the community it belongs to, we say so, because writing a fuller account would have meant inventing one. An empty space is information too.
Nothing here is presented as hidden history. Where the evidence stops, the entry stops.
Corrections
4 August 2026, the daeva post
A long post published on X on 4 August traced the word for giant from Proto-Indo-European through Avestan into Persian, Georgian, South Slavic and Albanian. Four things in it were stated more firmly than the evidence allows.
The post said the usual reading is that a reform demonised the gods of a rival cult. That sentence merges two different scholarly explanations of the Avestan inversion into one, and they are not the same argument.
The post said the reversal itself is not in doubt. Encyclopaedia Iranica lists whether the daivas were originally gods among the questions still open, so that was too strong.
The post glossed the Proto-Indo-European root as the shining one. The standard gloss is the celestial one, and Slavic divo and diviti se sit on the same root without meaning god at all.
The post named Albanian div and dif in its opening line. We have not been able to source that to a reference work we can cite, and we are not repeating it until we can.
The Georgian point stands. Georgian is Kartvelian, the word did not descend into it, and it was carried in. The borrowing route into South Slavic stands too, and is better supported than the post allowed.