Everything is sourced or it does not go up
Every factual claim on this site should be traceable to a named source, and most pages carry their bibliography at the foot of the page. The full library is at the references page.
Where we could not verify something, we cut it rather than hedging it. A softened claim with no source is still a claim with no source.
Three tiers of certainty, and we mark them
Confirmed — published in a peer-reviewed journal or an excavation report. We state it plainly.
Reported — announced by a ministry briefing, an excavation director in an interview, or a news agency, but not yet published. We say so, and we name where it came from. Much of the Taş Tepeler material reaches English this way, sometimes years before formal publication. It is worth reporting; it is not the same thing.
Contested — the evidence exists and the reading of it is disputed. We give the competing interpretations and say which archaeologists hold them, rather than picking a winner and presenting it as settled.
Description and interpretation are kept apart
What was found, and where, is one kind of statement. What it meant is another, and in this period it is almost always contested.
A T-shaped pillar with arms, hands and a belt is a description. A T-shaped pillar representing an ancestor is an interpretation — a reasonable one, widely held, and still an interpretation. We try never to let the second borrow the confidence of the first.
Superlatives
"Oldest", "first" and "world's first" are the currency of coverage in this subject and they are usually doing more work than the evidence supports.
Where we use them, they should be attributed and bounded: oldest known monumental architecture of its kind, on current dating. Where the underlying claim is narrower than the headline — Özdoğan's Sayburç claim is for the earliest known narrative scene, not the earliest story — we use the narrower one.
How the writing is produced
Research, drafting and site-building are AI-assisted, using Claude, with a human editor responsible for everything that publishes. This is stated on the About page too.
The review that matters is source verification: citations are checked against the actual publication, not accepted from a summary. That has caught real errors, including a citation drafted with the right authors and the wrong journal and year.
What we will not do
We will not present a tour-selling claim as an archaeological one.
We will not use a photograph of one thing to illustrate another. Pages here have no image rather than a stand-in, and where a photograph is contextual rather than the object itself, the caption says so.
We will not quietly change a claim. See corrections.
Last reviewed 21 July 2026.