How Linear B Was Deciphered

1. The Labyrinth and the Silent Clay

In March 1900, British archaeologist Sir Arthur Evans broke ground on a low hill called Kephala on the northern coast of Crete. Within days, his team uncovered the sprawling, multi-storied ruins of a vast Bronze Age palace: Knossos, the legendary seat of King Minos and the mythical Minotaur's labyrinth.

Buried within the ash-streaked floors of the palace storerooms, Evans discovered something far more valuable than gold or frescoes: thousands of unbaked clay tablets incised with delicate, flowing inscriptions.

                  THE ACCIDENTAL SURVIVAL OF LINEAR B
                  
   Late Bronze Age Scribe               Palace Cataclysm (c. 1200 BCE)
   ┌────────────────────────────┐       ┌────────────────────────────┐
   │ Incises administrative     │       │ Enemy invasion or earth-   │
   │ tallies on raw, wet clay.  │ ====> │ quake burns palace down.   │
   │ Intended to be dissolved   │       │ Intense fire bakes tablets │
   │ and recycled after 1 year. │       │ into rock-hard ceramic!    │
   └────────────────────────────┘       └────────────────────────────┘
                 │                                    │
                 ▼                                    ▼
       Temporary Ephemeral Ledger          PERMANENT HISTORICAL ARCHIVE

These tablets had never been meant to survive. Bronze Age palace scribes treated clay tablets as temporary scrap paper. They inscribed inventory records, tax assessments, and livestock tallies on damp clay; after the annual audit, the tablets were routinely dumped into water tubs to be kneaded back into fresh clay.

The tablets survived solely because of catastrophe. Around 1200 BCE, an immense fire consumed the palaces of Knossos, Pylos, and Mycenae. The conflagration collapsed the timber roofs and burned at temperatures exceeding 800^∘C, accidentally firing the raw clay tablets into permanent terracotta ceramics before burying them under tons of rubble for over three millennia.

Evans cataloged the recovered inscriptions into three chronologically and stylistically distinct scripts:

1. Cretan Hieroglyphic (c. 2000–1650 BCE): Highly pictorial signs depicting human figures, tools, animals, and plants.
2. Linear A (c. 1800–1450 BCE): A cursive, abstract script written from left to right, representing the indigenous, non-Indo-European Minoan language (which remains undeciphered today).
3. Linear B (c. 1450–1200 BCE): A modified, more angular descendant of Linear A found in massive administrative archives at Knossos on Crete, and later on the Greek mainland at Pylos, Mycenae, Thebes, and Tiryns.

For more than four decades, Linear B resisted every attempt at decipherment. It stood as the most tantalizing linguistic enigma in classical archaeology.

---

2. The Cryptanalytic Challenge: A Script and Language Unknown

To appreciate the scale of the intellectual achievement, one must contrast Linear B with history's other famous ancient decipherments.

When Jean-François Champollion deciphered Egyptian hieroglyphs in 1822, he possessed the Rosetta Stone: a decree carved in three scripts—Hieroglyphic, Demotic, and Ancient Greek. Because Champollion could read the Greek text fluently, he possessed an exact translation of the inscription, enabling him to match royal cartouches (Ptolemy, Cleopatra) directly against known names.

Similarly, when Sir Henry Rawlinson and Edward Hincks cracked Mesopotamian cuneiform in the 1840s, they utilized the trilingual Behistun Inscription carved high into a Persian cliff face, containing identical texts in Old Persian, Elamite, and Babylonian.

                THE FOUR QUADRANTS OF ANCIENT DECIPHERMENT
                
                        KNOWN SCRIPT            UNKNOWN SCRIPT
                     ┌───────────────────────┬───────────────────────┐
       KNOWN         │ Ordinary Translation  │ LINEAR B (1950)       │
      LANGUAGE       │ (e.g., Reading Latin  │ (Unknown symbols,     │
                     │  in Roman alphabet)   │  but Greek language)  │
                     ├───────────────────────┼───────────────────────┤
      UNKNOWN        │ ETRUSCAN              │ LINEAR A, INDUS SCRIPT│
      LANGUAGE       │ (Letters are Greek,   │ (Unknown symbols,     │
                     │  but grammar unknown) │  unknown language)    │
                     └───────────────────────┴───────────────────────┘

For Linear B, there was no bilingual inscription.

There was no Greek, Egyptian, or Phoenician parallel translation. Epigraphers confronted the ultimate cryptanalytic nightmare: an unknown writing system expressing an unknown language.

The Blunder of Arthur Evans

Decipherment was further crippled by the dogmatic preconception of Arthur Evans himself. Evans was infatuated with his discovery of Minoan civilization. He insisted that the prehistoric inhabitants of Crete were an advanced, non-Greek maritime race completely distinct from the "barbaric" mainland Hellenes of the classical era.

Evans argued that Linear B recorded a purely Minoan tongue. He vehemently rejected any suggestion that the language could be an early dialect of Greek, claiming that Greeks had not entered Crete until centuries later.

Because Evans was the undisputed titan of Aegean archaeology, his theory intimidated classical philologists for forty years. Scholars attempted to translate Linear B using Basque, Etruscan, Hittite, Hebrew, and Egyptian—yielding nothing but arbitrary, contradictory nonsense.

Sign Count Analysis: Deducing the System Type

Before reading a single word, a cryptanalyst must determine the structural type of a writing system.

Writing systems throughout human history fall into three distinct structural categories, each dictated by a characteristic number of distinct signs:

               SIGN-COUNT DISTRIBUTION ACROSS WRITING SYSTEMS
               
   Type:             Alphabet               Syllabary             Logographic
   Number of Signs:  20 to 35               60 to 90              Hundreds to Thousands
   Examples:         Latin (26)             Japanese Kana (46)    Chinese (50,000+)
                     Greek (24)             Cherokee (85)         Egyptian Hieroglyphs (1,000+)
                     Phoenician (22)        LINEAR B (87)         Sumerian Cuneiform (1,000+)

Evans and his contemporaries cataloged approximately 87 distinct phonetic characters in Linear B, alongside numerical symbols and dozens of pictorial ideograms (such as drawings of horses, swords, tripods, jars, and grain).

The mathematical implication was inescapable:

• 87 signs is far too many for an alphabet. If Linear B were alphabetic (one sign per vowel or consonant), it would require an impossibly complex phonetic inventory of 87 distinct phonemes.
• 87 signs is far too few for a logographic or ideographic system. If each sign represented an entire word or concept (like Chinese characters or Sumerian cuneiform), 87 signs would not even suffice to write a basic child's vocabulary.

Linear B could only be one thing: a pure open-syllabary.

Each sign did not represent an isolated letter or an entire word. Each sign represented an open syllable, consisting of either a pure vowel (V, such as a, e, i) or a single consonant followed by a vowel (CV, such as ka, te, ro, mi).

---

3. Alice Kober and the Cryptographic Matrix

The breakthrough that transformed Linear B from an archaeological curiosity into a solvable mathematical cipher came from an American classicist at Brooklyn College: Alice Elizabeth Kober.

Kober recognized that all previous attempts had failed because scholars began by guessing phonetic values based on superficial visual resemblances between Linear B signs and other ancient alphabets. Kober insisted on a principle of absolute scientific rigor:

Do not try to pronounce the signs. Treat the script as an encrypted code. Analyze the internal statistical and morphological relationships of the symbols before assigning a single sound.

Working with limited funding during the 1940s—laboriously cutting index cards out of spare wastepaper and filing them in cigarette cartons during wartime paper shortages—Kober examined thousands of Linear B words across hundreds of tablet transcriptions.

                 THE KOBER INFLECTIONAL TRIPLET STRUCTURE
                 
         Case I (Nominative):     Sign A ── Sign B ── [ Sign C1 ]
         Case II (Genitive):      Sign A ── Sign B ── [ Sign C2 ]
         Case III (Dative):       Sign A ── Sign B ── [ Sign C3 ]
                                  └──────┬───────┘    └────┬────┘
                                   Constant Stem       Variable Inflection

The Kober Triplets

Kober noticed that certain groups of Linear B words occurred in recognizable "triplets"—sequences of words that appeared in identical contexts on inventory lists, sharing the exact same root symbols but showing systematic variations in their final characters.

She identified dozens of these patterns, which became known to history as Kober's Triplets.

Consider an abstract example of two related triplets observed by Kober across the Knossos tablets:

By analyzing the administrative lists, Kober determined that the difference between the word ending in C and the word ending in D was grammatical inflection (such as the shift between nominative subject and genitive possessive, or between masculine and feminine).

This morphological observation led to an algebraic deduction of staggering power.

In an open syllabary (CV), a word stem that ends in a consonant (for example, the consonant k) cannot write that consonant in isolation. If the grammatical case ending changes its vowel—say, from an −o in the nominative to an −a in the feminine—the scribe must replace the sign for ko with the sign for ka:

Kober realized that:

1. The language behind Linear B was an inflected language, exhibiting complex declensions of nouns and adjectives (characteristic of the Indo-European language family).
2. Even without knowing how C or D sounded, sign C and sign D must share the exact same consonant, but possess different vowels!
3. Conversely, if two different stems with different consonants adopted the same grammatical case ending, their final signs must share the exact same vowel, while possessing different consonants!

The Birth of the Grid

Kober began assembling an algebraic two-dimensional matrix.

On the vertical axis, she plotted unknown consonants (C₁, C₂, C₃…). On the horizontal axis, she plotted unknown vowels (V₁, V₂, V₃…).

Whenever a triplet demonstrated that two signs shared a consonant or a vowel, she placed those signs into the corresponding row or column:

                  ALICE KOBER'S STRUCTURAL SYLLABIC MATRIX
                  
                     Vowel 1 (V₁)    Vowel 2 (V₂)    Vowel 3 (V₃)
                   ┌───────────────┬───────────────┬───────────────┐
   Consonant 1 (C₁)│    Sign 08    │    Sign 12    │    Sign 32    │ <- Shared Consonant
                   ├───────────────┼───────────────┼───────────────┤
   Consonant 2 (C₂)│    Sign 52    │    Sign 70    │    Sign 44    │
                   ├───────────────┼───────────────┼───────────────┤
   Consonant 3 (C₃)│    Sign 05    │    Sign 37    │    Sign 55    │
                   └───────────────┴───────────────┴───────────────┘
                           ▲
                           └─ Shared Vowel Column

Kober had built the cryptographic scaffold of Linear B. If a scholar could discover the phonetic value of just one or two key cells, the values of adjacent signs would immediately snap into place across entire rows and columns through sheer geometric constraint.

Tragically, Alice Kober did not live to see the completion of her work. Suffering from cancer, she died in 1950 at the age of forty-three, leaving behind the mathematical blueprint that would unlock the Aegean Bronze Age.

---

4. Michael Ventris and the Syllabic Grid

Enter Michael Ventris.

Ventris was not an academic philologist by trade; he was a brilliant young English architect who had served as an RAF navigator and cryptanalyst during the Second World War. But Linear B had been his personal obsession since 1936, when, as a fourteen-year-old schoolboy visiting Burlington House in London, he had heard the elderly Sir Arthur Evans lecture on the undeciphered tablets of Knossos.

Ventris inherited Alice Kober’s structural approach. He recognized that trial-and-error phonetic guessing was futile. Instead, between 1950 and 1952, Ventris compiled and distributed a series of mimeographed "Work Notes" to a network of international scholars, systematically building an expanded, 60-character syllabic grid.

                  THE TWO-DIMENSIONAL PHONETIC MATRIX
                  
                     Vowel a         Vowel e         Vowel i         Vowel o         Vowel u
                   ┌───────────────┬───────────────┬───────────────┬───────────────┬───────────────┐
   Pure Vowel      │      *a       │      *e       │      *i       │      *o       │      *u       │
                   ├───────────────┼───────────────┼───────────────┼───────────────┼───────────────┤
   Consonant d-    │      da       │      de       │      di       │      do       │      du       │
                   ├───────────────┼───────────────┼───────────────┼───────────────┼───────────────┤
   Consonant k-    │      ka       │      ke       │      ki       │      ko       │      ku       │
                   ├───────────────┼───────────────┼───────────────┼───────────────┼───────────────┤
   Consonant m-    │      ma       │      me       │      mi       │      mo       │      mu       │
                   ├───────────────┼───────────────┼───────────────┼───────────────┼───────────────┤
   Consonant p-    │      pa       │      pe       │      pi       │      po       │      pu       │
                   ├───────────────┼───────────────┼───────────────┼───────────────┼───────────────┤
   Consonant t-    │      ta       │      te       │      ti       │      to       │      tu       │
                   └───────────────┴───────────────┴───────────────┴───────────────┴───────────────┘

Like Evans, Ventris initially assumed the underlying language was an extinct Aegean language akin to Etruscan. But as he refined the grid, he searched for a phonetic wedge—an external anchor that could attach absolute sound values to his relative coordinate grid.

The Toponymic Anchor: Cretan Place Names

Ventris made a brilliant geographical deduction.

Archaeologists knew that while kings and civilizations fall, geographical place names (toponyms) frequently survive for millennia. Mountains, rivers, and ancient towns often retain their original names even when conquered by new peoples.

Ventris noticed that a small group of words occurred with immense frequency on the tablets found at Knossos on Crete, but were completely absent from the tablets found at Pylos on the Greek mainland.

If these words appeared exclusively on Cretan tablets, they were likely the names of local Cretan cities and sanctuaries.

What Cretan cities existed in classical antiquity that might have preserved Bronze Age names?

• Knossos (the palace site itself)
• Amnisos (the ancient harbor town located five miles east of Knossos, mentioned in Homer's Odyssey)
• Tulissos (a prominent ancient settlement west of Knossos)
• Phaistos (the southern Cretan palace)
• Lyktos (an ancient inland city)

Ventris examined the sign patterns of these candidate place names.

One particular word consisted of three signs:

Ventris knew from previous comparative work with the already-deciphered classical Cypriot syllabary that Sign 08 frequently appeared at the start of words where a pure vowel was expected. In Cypriot, a very similar sign represented the pure vowel a.

If Sign 08 was a, could this three-sign word be the harbor town of Amnisos?

                  THE TOPONYMIC PHONETIC BREAKTHROUGH
                  
   Hypothesized City:      A-MNI-SO-S (Amnisos)        KO-NO-SO (Knossos)
   Linear B Symbols:       [ 08 ] - [ 12 ] - [ 32 ]    [ 52 ] - [ 08 ] - [ 32 ]
                               │         │        │          │        │        │
   Assigned Syllable:          a   -    mi   -   so         ko   -    no   -   so

Ventris tested the hypothetical values:

• If Sign 12 was mi, then its column in the grid was the −i vowel, and its row was the m− consonant.
• If Sign 32 was so, then its column was the −o vowel, and its row was the s− consonant.

Now look at the word for Knossos. In an open syllabary that cannot write a double consonant cluster like kn-, a scribe must insert a dummy "echo vowel": Ko-no-so.

Ventris found a three-sign word that ended with the exact same Sign 32 (so) and contained Sign 08 (no) in the middle:

Next, he looked at Tulissos:

The values interlocked with geometric perfection. The grid did not merely suggest the sounds—it demanded them.

---

5. The Shock of 1952: The Language Was Greek!

Once the place-name values were inserted into the grid, phonetic values cascaded across dozens of cells.

Ventris opened his transcription notebooks and began applying the new syllabic values to ordinary, everyday vocabulary items found across the Knossos and Pylos tablets.

He expected to find Etruscan or an unknown Mediterranean language.

Instead, as the syllables were pronounced aloud, the words emerging from the clay sounded shockingly, unmistakably familiar:

                  THE EMERGENCE OF ARCHAIC MYCENAEAN GREEK
                  
   Linear B Spelling       Mycenaean Pronunciation      Classical Greek          Meaning
   ═════════════════════════════════════════════════════════════════════════════════════════
   po-me                   poimēn                       ποιμήν (poimēn)          Shepherd
   wa-na-ka                wanax                        ἄναξ (anax)              High King / Lord
   ra-wa-ke-ta             lāwāgetās                    λααγέτας (lāgētās)       Commander of Army
   do-e-ro                 dohelos                      δοῦλος (doulos)          Slave / Servant
   i-je-re-u                hiereus                      ἱερεύς (hiereus)         Priest
   ko-wo                   korwos                       κοῦρος (kouros)          Boy / Son
   ko-wa                   korwā                        κόρη (korē)              Girl / Daughter
   ti-ri-po                tripos                       τρίπους (tripous)        Three-legged pot

The language of Linear B was not an alien Minoan tongue. It was Greek!

It was not classical Athenian Greek of Plato or Sophocles, nor was it the Ionic Greek of the Homeric epics. It was Mycenaean Greek—an archaic, ancestral dialect spoken five centuries before Homer, frozen in the ashes of the Bronze Age palatial administration.

On July 1, 1952, Michael Ventris broadcast an unheralded radio talk on the BBC Third Programme. He announced to the world that Linear B had been solved:

"During the last few weeks, I have come to the conclusion that the Knossos and Pylos tablets must, after all, be written in Greek—a difficult and archaic Greek, seeing that it is 500 years older than Homer and written in a rather abbreviated, clumsy code, but Greek nonetheless."

Among the listeners was John Chadwick, a thirty-two-year-old classical philologist at Cambridge University who had served as a naval cryptanalyst at Bletchley Park during the war. Chadwick immediately realized that Ventris was right.

Chadwick possessed what Ventris lacked: formal training in historical Indo-European linguistics. The two men formed an extraordinary partnership, spending the next four years demonstrating that every bizarre spelling convention in Linear B followed consistent, predictable phonetic rules.

---

6. The Pylos Tripod Tablet (Ta 641): Irrefutable Proof

Despite the elegance of Ventris and Chadwick’s work, skeptics across Europe remained unconvinced.

Eminent philologists argued that Ventris had simply tortured the 87 arbitrary signs until, through coincidence and selective anagramming, they yielded words that looked like Greek. After all, if you invent enough phonetic rules, you can make almost any unknown script look like almost any language.

What the decipherment needed was an empirical blind test: an unread, newly discovered tablet that had never been seen by Ventris during the construction of the grid.

In May 1953, that test arrived.

American archaeologist Carl Blegen, digging at the palace site of Pylos in the Peloponnese, excavated a fresh cache of clay tablets. Among them was an extraordinary, pristine inventory tablet designated Ta 641.

                  THE PYLOS TRIPOD TABLET (TA 641) PROOF
                  
   Linear B Inscription (Ventris Grid):     Ideogram Carved on Clay:
   ════════════════════════════════════     ════════════════════════
   ti-ri-po   (tripos = one tripod)         [ ⚱️ with 3 legs ]  x 1
   
   ti-ri-po-de (tripode = two tripods)      [ ⚱️ with 3 legs ]  x 2
   (Dual grammatical number!)
   
   ti-ri-po-we (with 3 ears/handles)        [ ⚱️ with 3 handles ]
   
   qe-to-ro-we (with 4 ears/handles)        [ ⚱️ with 4 handles ]
   
   an-o-we     (without ears/handles)       [ ⚱️ with NO handles ]

Blegen had in his hands a copy of Ventris's proposed phonetic grid, which had been mailed to him weeks earlier. Blegen sat down in his excavation tent and transliterated the unread tablet line by line.

The first line began with a word spelled with three signs:

According to Ventris, this was tripos—the Greek word for a three-legged cooking cauldron or tripod.

Directly following the word ti-ri-po, the ancient Mycenaean scribe had incised a small, unmistakable pictographic ideogram: a drawing of a three-legged pot, followed by the numerical tally mark for 1!

The second line read:

In classical Greek, nouns possess singular and plural numbers. But archaic Greek possessed a rare third grammatical number: the dual number, used exclusively for pairs of objects! The dual of tripos was tripode.

Directly following ti-ri-po-de, the scribe had drawn two three-legged pots, followed by the numerical tally mark for 2!

Further down the tablet, the inscriptions described the handles of the vessels:

• ti-ri-po-we = tri-ōwes (three-handled), accompanied by a pot drawing with three handles.
• qe-to-ro-we = kʷetro-ōwes (four-handled; archaic Indo-European kʷetur- for "four", classical tetra-), accompanied by a pot drawing with four handles.
• an-o-we = an-ōwes (earless / handleless; the Greek privative prefix an- + ous "ear"), accompanied by a pot with zero handles!

Blegen immediately fired off a historic letter to Michael Ventris:

"All this seems almost too good to be true. May I congratulate you on your solution of the problem?"

There was no longer any room for doubt. The probability of random signs coincidentally spelling the Greek words for "three-legged pot", "two three-legged pots (dual)", "three-handled", "four-handled", and "handleless" while directly matching corresponding line drawings was mathematically zero.

Linear B was decisively, triumphantly deciphered.

---

7. What the Tablets Revealed: The Palatial Bronze Age World

The decipherment of Linear B shattered and rebuilt classical history overnight.

Before 1952, historians believed that Greek culture began around 750 BCE with the introduction of the Phoenician alphabet and the composition of the Iliad and Odyssey. The earlier Mycenaean civilization was treated as a semi-mythical, pre-Hellenic world.

Linear B proved that the Mycenaeans were Greeks.

They worshipped the same Olympian gods, spoke the ancestor of the same language, and lived in an extraordinarily organized, bureaucratic palatial society:

1. The Divine Pantheon

Scholars were astonished to find the names of classical Greek deities incised on 3,200-year-old clay:

• Zeus (di-we) and Hera (e-ra)
• Poseidon (po-se-da-o), who was the supreme deity of Mycenaean Pylos, receiving sacrifices of bulls, sheep, and wine
• Athena (a-ta-na po-ti-ni-ja = "Mistress Athena")
• Ares (a-re) and Hermes (e-ma-a)
• Dionysus (di-wo-nu-so), whose appearance on a Pylos tablet disproved the long-standing academic theory that Dionysus was a late foreign import from Thrace or Phrygia.

2. A Bureaucracy of Bronze and Wool

The tablets contained no poetry, no literature, no personal letters, and no historical annals. They were the cold, hyper-detailed ledgers of a centralized command economy:

• Scribes tracked the distribution of bronze to regional smiths down to the fraction of an ounce.
• They audited the palatial sheep flocks—recording over 100,000 sheep in central Crete dedicated to industrial wool production.
• They allocated rations of barley, figs, and olives to workgroups of captive women and children (ko-wa and ko-wo).
• A famous series of Pylos tablets records "Watchers by the Sea"—emergency troop deployments and rowers dispatched to coastal watchposts as the palace braced for an imminent maritime attack that would soon destroy it forever.

                  THE TWO ERAS OF GREEK WRITING
                  
   BRONZE AGE: LINEAR B (1450 - 1200 BCE)       IRON AGE: THE GREEK ALPHABET (c. 750 BCE - Present)
   ┌────────────────────────────────────┐       ┌────────────────────────────────────┐
   │ Clumsy open syllabary (87 signs)   │       │ Precise phonetic alphabet (24 letters)│
   │ Exclusively palace scribes         │       │ Widespread civilian literacy       │
   │ Records inventories and taxes      │       │ Records poetry, law, history, logic│
   │ Extinct in 1200 BCE palatial fire  │       │ Survived to form modern literacy   │
   └────────────────────────────────────┘       └────────────────────────────────────┘

---

8. Why Linear B Vanished While the Alphabet Endured

Why did a writing system that governed the Aegean for three centuries vanish so completely that its very existence was forgotten for three thousand years?

The answer lies in the sociological architecture of literacy.

Linear B was never a democratic, popular tool. An 87-sign syllabary with dozens of ideograms and complex phonetic omissions is difficult to learn and cumbersome to write. Literacy was the closely guarded monopoly of a tiny caste of specialized palace scribes (Bureaucrats of the Throne Room). The ordinary farmers, shepherds, and warriors of Mycenaean Greece could neither read nor write.

Around 1200 BCE, the Late Bronze Age Collapse swept across the eastern Mediterranean. Within decades, the great Mycenaean palatial citadels—Pylos, Mycenae, Tiryns, Knossos—were burned and abandoned.

When the palaces burned, the economic machinery that sustained the scribes ceased to exist. Because writing was tied exclusively to palatial administration, when the administration collapsed, literacy vanished with it.

Greece plunged into four centuries of epigraphic silence—the Greek Dark Ages (c. 1200–800 BCE). During this period, the Greek language survived exclusively as an oral tradition, passed down through memory by wandering bards singing tales of the Trojan War.

When literacy returned to Greece around 750 BCE, it did not resurrect Linear B. Instead, Greek merchants adapted the revolutionary Phoenician alphabet (as explored in How the Alphabet Was Invented), adding explicit vowel letters to create the twenty-four letters of classical Greek.

Unlike Linear B, the alphabet was simple, democratic, and easily mastered by ordinary citizens—paving the way for the lyric poetry, philosophy, mathematics, and drama that would define Western civilization.

Yet beneath Homer’s hexameters and Aristotle’s logic lies that older, miraculous layer of human memory: the baked clay tablets of the Bronze Age, silent for three millennia until the cryptographic genius of Alice Kober and Michael Ventris taught them to speak Greek once more.