🏺 Linear B: how an architect showed the Mycenaeans wrote Greek

In 1952 Michael Ventris, an architect, showed that the clay tablets of Knossos and Pylos record an early form of Greek. He built on Alice Kober's analysis of word endings, worked with John Chadwick, and in 1953 a tablet listing tripods and jars from Pylos supplied the test.

Tags: decipherment, mycenaean-greek, linear-b, writing-systems, aegean

In 1900 Arthur Evans began digging at Knossos on Crete and found thousands of clay tablets written in a script he named Linear B. In 1939 Carl Blegen found a second archive at Pylos in the western Peloponnese, on the first day of his excavation. The script has about 90 syllabic signs plus logograms, which are pictures of the thing being counted: horses, chariot wheels, tripods, jars of wine and oil. For half a century nobody could say what language the syllables spelled. Evans took it for granted that it was the unknown language of the Minoans.

Kober's endings

Alice Kober, a classicist at Brooklyn College, published her key paper in 1946, "Inflection in Linear Class B". She worked without assigning a single sound value. She noticed groups of three related words, later called Kober's triplets, that shared a stem and differed only in their final signs. From these she argued that the language had inflectional endings. She also showed that some signs had to share a consonant while differing in vowel. Her 'bridging' signs became the raw material for a grid of consonants and vowels. Kober died in May 1950, aged 43, two years before the decipherment.

Ventris's grid

Michael Ventris was 14 when he heard Evans lecture in London in 1936. He trained as an architect and served in the war as an RAF navigator, not as a codebreaker. He believed for years that the language was related to Etruscan, and he said so in print in 1940, at 18. In 1951 Emmett Bennett published a standard sign list for the Pylos tablets. Using it, Ventris filled in a grid of signs sorted by shared consonant and shared vowel. He then guessed that certain words found only on the Cretan tablets were Cretan place names: ko-no-so Knossos, a-mi-ni-so Amnisos, tu-ri-so Tulissos. Once he put those values into the grid, other words came out as recognisable Greek: po-me 'shepherd', ko-wo 'boy'. Kober's triplets turned out to be place names alongside their masculine and feminine adjectives, such as a-mi-ni-so next to a-mi-ni-si-jo and a-mi-ni-si-ja.

Ventris circulated Work Note 20 on 1 June 1952 under the title "Are the Knossos and Pylos tablets written in Greek?" On 1 July he announced in a BBC radio talk that they were. John Chadwick, a Cambridge philologist, heard the broadcast and wrote to him. Their joint paper, "Evidence for Greek Dialect in the Mycenaean Archives", appeared in the Journal of Hellenic Studies in 1953.

Ventris's first readings
Linear BTransliterationMeaning
𐀒𐀜𐀰ko-no-soKnossos (place name)
𐀀𐀖𐀛𐀰a-mi-ni-soAmnisos (place name)
𐀶𐀪𐀰tu-ri-soTulissos (place name)
𐀀𐀖𐀛𐀯𐀍a-mi-ni-si-joof Amnisos (masculine)
𐀀𐀖𐀛𐀯𐀊a-mi-ni-si-jaof Amnisos (feminine)
𐀡𐀕po-meshepherd
𐀒𐀺ko-woboy

The tripod tablet

Blegen had dug up tablet PY Ta 641 at Pylos in June 1952, but it was not cleaned well enough to read until 1953. The tablet is about 15 cm long. Read with Ventris's values, the first word is ti-ri-po-de, Greek tripode 'two tripods', and the entry ends with a tripod logogram and the numeral 2. Further along, di-pa me-zo-e qe-to-ro-we 'a larger vessel, four-handled' is drawn with four handles, ti-ri-o-we-e 'three-handled' with three, and a-no-we 'handleless' with none. On 16 May 1953 Blegen sent Ventris a copy and wrote: "All this seems too good to be true. Is coincidence excluded?" The Scottish classicist Arthur Beattie claimed that the match was chance, and later that Ventris had seen the tablet in advance. Blegen replied that the tablet had been kept in a locked box until late July 1952, after Ventris had announced his result, and stayed illegible until the next year. The objection found little support.

Four entries on PY Ta 641
On the tabletTransliterationMeaning
𐀴𐀪𐀡𐀆 … 𐃠 𐄈ti-ri-po-de … *201VAS 2two tripods
𐀇𐀞 𐀕𐀿𐀁 𐀤𐀵𐀫𐀸 𐃡 𐄇di-pa me-zo-e qe-to-ro-we *202VAS 1one larger vessel, four-handled
𐀇𐀞𐀁 𐀕𐀿𐀁 𐀴𐀪𐀃𐀸𐀁 𐃡 𐄈di-pa-e me-zo-e ti-ri-o-we-e *202VAS 2two larger vessels, three-handled
𐀇𐀞 𐀕𐀹𐀍 𐀀𐀜𐀸 𐃡 𐄇di-pa me-wi-jo a-no-we *202VAS 1one smaller vessel, handleless

What the tablets say, and what they do not

The Greek is some five centuries older than Homer. It keeps the labiovelar kʷ, as in i-qo ikkʷos 'horse', later hippos, and the w sound that later Greek lost, as in wa-na-ka wanaks 'king', later anax. The texts themselves are inventories: rations, livestock, land holdings, offerings. None of them is literary. Ventris and Chadwick's Documents in Mycenaean Greek came out in 1956, a few weeks after Ventris died in a car crash at 34. Linear A, the older Cretan script from which Linear B borrowed most of its signs, can be sounded out with the same values. The words that result match no known language, so the Minoan language Evans was looking for is still unread.

⚖️ Compare 'horse' in Mycenaean, Ancient and Modern Greek
LanguageFamilySpeakersHorseWaterOne
Mycenaean GreekIndo-European (Hellenic, pre-classical)Extinct (~1450-1200 BCE)𐀂𐀦/ikkʷos/𐀄𐀈/hudɔːr/𐀁𐀕/hemei/
Ancient GreekIndo-European (Hellenic)Extinctἵππος/híppos/ὕδωρ/hýdɔːr/εἷς/hêːs/
GreekHellenic~13Mάλογο/ˈaloɣo/νερό/neˈɾo/ένα/ˈena/
💬 qe-to-ro- 'four', as in the tripod tablet — Mycenaean Greek
Mycenaean GreekAncient Greek
🌍 Go to the Palace of Nestor at Pylos Open the interactive map

Sources

More on this

See also: Trivia · Word Map — 168 core words · Han Map — Han character readings

Open the interactive map: Word Map · Han Map · Name Map · Time Map · Word Order · Tree

Part of LangMap — a linguistic visualization project. This is a static, crawlable summary; the interactive maps offer pronunciation audio, filters, and a globe view.

Made by Heuron Inc. · GitHub · cho@heuron.com · LangMap home