The Ariana Projectundeciphered · readings are proposals, not translationsAll textsWordsWho were the Minoans?The data

Who were the Minoans?

A Bronze Age civilisation on Crete, at its height roughly 1900–1450 BCE. We do not know what they called themselves. "Minoan" is Arthur Evans's coinage, after the legendary king Minos; their neighbours called them Keftiu (Egyptian), Kaptara (Akkadian) and Caphtor (Hebrew).

Four things can be said with some confidence, and one cannot be said at all.

They were local

Genome-wide data has settled the origin question. Minoans derive 62–86% of their ancestry from the first Neolithic farmers who brought agriculture from western Anatolia into the Aegean, plus 9–17% from populations related to those of the Caucasus and Iran. They were not colonists from Egypt or Libya, as Evans supposed.

The sampled Minoans carry no steppe ancestry. Mycenaeans — otherwise genetically very similar — do. That is the single genetic difference between the two peoples, and it matters below.

A stranger finding: on Bronze Age Crete, first-cousin marriage appears to have been systematic. Of more than a thousand published ancient genomes worldwide, no comparable system has turned up anywhere else. It may have kept inherited land from fragmenting.

Caveats: sample sizes are small — 19 genomes in the 2017 study, 102 across the whole Aegean and several millennia by 2023 — and Mediterranean heat is hard on DNA. "Minoan" labels a material culture, not a bounded people.

Their language is unknown, and probably has no relatives

Linear A is undeciphered. We can pronounce much of it, by borrowing sound values from Linear B, without understanding it.

Two families have been argued for repeatedly, and neither has held:

  • Indo-European. Duhoux's structural work finds heavy prefixing and suffixing, a poor fit. And since steppe ancestry tracks Indo-European spread imperfectly but really, its absence in Minoan samples — beside its presence in Greek-writing Mycenaeans — removes a prop rather than supplying one.
  • Semitic. Pursued from Gordon and Best to the present without result. The dominant ancestry signal is Anatolian farmer, not Levantine, so the language would have had to arrive without the population.

Borrowed words are not a language family. ku-ni-su beside Akkadian kunāšu (emmer) and sa-sa-me beside Semitic šamaššammu (sesame) both sit in grain lists at Hagia Triada — commodities arriving with their names attached, which is ordinary trade contact.

The current consensus treats Minoan as an isolate: no known relatives, living or dead, and therefore no comparative framework to derive meanings from.

The one real hope is Egyptian. Three 18th-Dynasty sources — a scribal tablet of Keftiu personal names, a list of Cretan place names, and the London Medical Papyrus, which carries spells labelled as being in the Keftiu tongue — may record the Minoan language in a script we can read. Whether "Keftiu" always means Crete, and whether the spells are Minoan at all, are both contested.

They traded across the eastern Mediterranean

Direction Languages of the partners
Egypt Egyptian
Levant Amorite, Canaanite, later Ugaritic — with Akkadian as the commercial lingua franca
Cyprus unknown; written in Cypro-Minoan, a script derived from Linear A
Anatolia Luwian, Hittite; Hurrian to the south-east
Aegean and mainland pre-Greek languages, then Mycenaean Greek

The Mari tablets (c. 1780–1760 BCE) record tin moving from Mari through Aleppo, Qatna and Hazor to a Kaptarite at Ugarit — and record an interpreter being paid to handle the exchange. Someone was bridging a language gap, and it was worth writing down.

Linear A itself turns up on 48 inscriptions at 12 findspots beyond Crete: heavily in the Cyclades (Akrotiri, Ayia Irini, Phylakopi), at Miletus and Troy in Anatolia, at Mycenae and Tiryns, and once as far as Tel Haror in the Negev.

They wrote on clay that was never meant to last

Minoan administrative clay was sun-dried, not fired. A sun-dried tablet is a temporary object; wet it and it is mud again. It survives only where a building burned and baked it by accident.

So the corpus is a record of disasters. 1,167 of its 1,884 inscribed faces — 62% — come from Hagia Triada alone, one complex destroyed by fire around 1450 BCE.

Most of it is not tablets:

count what it is
nodules 913 clay squeezed round string or a fastening, sealed, often one sign
tablets 489 small page-shaped accounts
roundels 194 discs sealed round the rim — receipts, one impression per party
stone vessels 63 libation tables, carved, mostly repeating one religious formula

The durable objects are the exception, and they are not archives: stone vessels, inscribed pottery, metal (rings, pins, the Arkalochori axe), cave graffiti, and the eight inscriptions of the ink class (Zc), written in ink.

That ink is the clue to what is missing. People who write in ink on a cup can write on hide, papyrus or wood — and none of that survives in Crete's climate. Some clay nodules even preserve, on their undersides, the impressions of the folded documents they once sealed. The seals outlived the records.

The one thing that cannot be said

What the Minoans thought, said or believed.

The medium preserved accounting — lists, totals, commodity marks. A decipherment needs connected language, with syntax to work on. The Minoans almost certainly had that; it was on the perishable material the nodules were wrapped around.

What is left is about 1,600 clay objects that count things, plus 63 stone vessels repeating a single formula — and that formula carries nearly the whole weight of every translation proposed to date.

Sources

Corpus figures are from this repository's database and can be reproduced: python3 -m lineara stats, python3 -m lineara show 'HT 31'.