Homeric Atlas

Methodology

How this atlas decides what to draw, what to colour, and what to refuse to place. Everything below is a rule the dataset is validated against in CI, not an aspiration.

The four identification classes

Classes are assigned by hand and defended in each artifact. They describe the status of an identification, not how interesting a place is.

Accepted
A real site with mainstream scholarly and archaeological consensus.
Plausible
A real place proposed by mainstream scholarship, contested or unproven.
Speculative
An identification resting on a minority reading, a single proponent or later tradition.
Mythical
The poem places it outside the mappable world. No external tradition overrides that.

The 0–12 confidence rubric

Four criteria, three points each. The score is their sum and the band is derived from the score, so every badge on the site can be re-derived from the four numbers in the JSON.

0 = nothing · 1 = weak · 2 = partial · 3 = strong

#CriterionThe question it answers
1TextualHow specific is Homer about this location?
2ArchaeologicalIs there excavated Bronze Age material consistent with the text?
3ScholarlyHow far does current classical scholarship agree?
4GeographicDo the distances, bearings and landmarks fit?
max12

Bands

Thresholds are fixed and published:

  • High10–12 — all four criteria substantially satisfied.
  • Medium6–9 — a real case with a real gap in it.
  • Low3–5 — one criterion carries the claim.
  • Minimal0–2 — the claim exists; the evidence effectively does not.

The score is not a probability

A 9/12 does not mean a 75% chance that Odysseus lived there. It is an editorial judgement instrument: four questions, answered explicitly, so a reader can disagree with a specific answer instead of with a number that appeared from nowhere. Nothing in this atlas is a statistical estimate.

Rules the data must obey

  • Coordinates hang off a hypothesis, never off a place. A place with competing identifications has competing coordinates, and the map draws all of them.
  • A place marked unlocated may carry no coordinates at all. The schema forbids it, so nothing can quietly acquire a pin later.
  • An identification outside the Mediterranean frame is listed as off-map, never clamped to the border — that would draw a location nobody proposed.
  • An 'accepted' identification requires an archaeology or scholarship source. Tradition alone cannot promote a claim to accepted.
  • Every hypothesis and every competing theory cites at least one source. A claim with no source does not enter the dataset.
  • 'Mythical' is a statement about the poem's own geography, and no later tradition overrides it: Ogygia stays mythical however confidently a guidebook points at Gozo.

How sources are cited

Primary passages are cited by work, book and line, so any edition can be checked against them. Scholarship is cited by author, title and year. URLs appear only where a stable public one exists; a missing link means the citation was not verifiable online, not that it is weaker.

Primary textArchaeologyScholarshipTradition

Limits, stated plainly

  • The classes and the scores are one author's judgement, applied consistently and published in full so they can be argued with.
  • Coverage is not exhaustive. The atlas holds the places and episodes with a locational question worth asking, not every toponym in the poems.
  • The atlas takes no position on whether the Trojan War happened. Locating a city and dating an event are different questions, and only the first one is mapped here.
  • Some coordinates are approximate to a village, a bay or an island rather than a site. Each hypothesis says which.

Reusable data

GeoJSON carries one feature per identification, with its class, its rubric score and its canonical URL in the properties — so a map client can style by confidence without asking us what a number means.

CC BY 4.0 — reuse freely, credit the author, link the canonical URL.