The Man Who Deciphered Linear B Read online

  The throne room of the palace of Knossos, reconstructed according to Evans’s ideas.

  (From A. Evans, Palace of Minos at Knossos, vol. IV, 1935.)

  Today, with the passing of the modern movement in architecture and the general critical reaction against functionalism and central planning, Highpoint has lost much of its appeal. But even in 1936, some of its admirers were critical. Although Ventris lived there for a long period, until 1953 – and ‘cracked’ Linear B there – he showed no great enthusiasm for its design, and the house he eventually built for himself off Hampstead Heath does not echo the design of the Highpoint flat in any way. His mother Dorothea went so far as to call Highpoint ‘noisy and blatant’ in a letter to Marcel Breuer. But she loved the furniture that Breuer had designed for the flat during his brief stay in England in 1935–37 before emigrating to the United States: a chrome and laminate tea trolley, a combined radiogram/sideboard, a bookcase and sideboard with roller fronts, a glass-topped table with tubular metal legs (later used as a work desk for decipherment), a sofa, chairs with broad bentwood arms forming shelves underneath, and other items. ‘I count myself extraordinarily fortunate to have this little centre which you made…your interior with its shapes and colours and textures of which I never tire. Other people appreciate it too, but no one as much as Michael, who has such a firm affection for his room that I am sure he will never let me give up the flat.’ The following year, 1939, she again told him: ‘In spite of the fact that we may be going to be bombed to pieces we still have the same enthusiasm for your flat.… We have two new pictures, both by Picasso. One hangs in the dining room, and is flat Cubist, 1923, with very strong masses of brown, grey, white and terracotta against a background of lilac and yellow. The lilac and yellow are marvellous with your blue table.’ But she admitted that she had never had the ‘courage or inspiration’ to introduce into the interior the kind of small personal touches that she knew Breuer wanted, to make the flat feel like it was hers, rather than a showpiece.

  Highpoint, Ventris’s home from 1936 until 1953, where he deciphered Linear B.

  (Photo Dell & Wainwright. © Architectural Review.)

  Artistically minded and acutely sensitive though she was, Dorothea Ventris was no artist: one looks in vain for her name among the records of the exciting artistic life of 1930s Hampstead, though she knew personally Naum Gabo, Ben Nicholson, Henry Moore and other local artists, and keenly collected their work. Their strong personalities attracted her but, unlike Michael, she found them overwhelming. ‘Perhaps you realized in the end how completely I lack self-confidence,’ she told Breuer. ‘I know I must have seemed unfriendly and ungrateful. The fact is that I am simply terrified of people. I was in a panic that you would bring your friends here [to the flat] – capable, worldly people…who make me feel a fool. So I just faded out and you thought I was indifferent to your interests. Or being the kind, gentle person you are, you were merely chilled and puzzled.’

  Nevertheless, she could be very practical, especially where her son was concerned. Her devotion to him is transparent and she strongly encouraged him when he began to show an interest in studying architecture in his last year at Stowe, despite the likely difficulties of getting trained and finding a job as a modernist architect in 1940s Britain, and the undoubted heavy expense. She wrote to Patrick Hunter: ‘You will agree with me that Michael is not very sensitive to literature. He loves language, but his aesthetic taste is stronger in music, in the volumes (or spaces) of architecture and sculpture.’

  Influenced by Highpoint, Breuer’s designs and the constructivist sculpture of Gabo – who refused to distinguish between sculpture and architecture – Michael had started to subject the sublimely inappropriate neoclassical setting of Stowe to some functional analysis of his own. ‘Who has ever heard of a functionally built chapel?’ he told Gabo. ‘I’m not religious myself but I suppose other boys are…’. He toyed with a design for a public school, which he thought would be ‘quite interesting just as a technical study’ even though a public school was not ‘a progressive institution’ – a project which was a sort of precursor of his future, post-war design work for the Ministry of Education. Although nothing much came of all this, the mental effort helped to persuade him (and Gabo, who would become something of a father figure for Michael), during the first half of 1939, that he should abandon the study of classical literature and follow his modernist instincts.

  Breuer wrote from Harvard University in July, where he was now teaching, and gave solid advice: that Michael should study at the Architectural Association School in London, which was now more sympathetic to modernism than in the mid-1930s. ‘A fresh wind seems to blow there and I am sure it is more useful for professional studies than Oxford.’ But he agreed with Michael that in England an Oxford education had a special social significance. Other possibilities were Harvard and a technical college in Zurich.

  In September, Michael was more or less forced to decide. When the Germans invaded Poland, his mother’s Polish father lost all his land, property and income, with a disastrous effect on Dorothea’s income. She immediately wrote to Stowe withdrawing Michael (he had intended to stay at least one more term) and asking if the fees for the autumn term could be waived – a request that Roxburgh granted. He also, like Breuer, encouraged Michael to apply for admission to the Architectural Association School; this had remained open despite the outbreak of war, while moving from its central London quarters in Bedford Square to temporary premises outside London, at Hadley Common – luckily not far from Highgate. After taking a crash course in drawing at a local art school that autumn – a ‘rather gruelling’ experience, Michael told Gabo, since he had done no drawing at Stowe and was not naturally talented as an artist as he was as a linguist – he was accepted at the ‘AA’ and asked to begin his architectural training in January 1940.

  Although his mother was pleased, her thoughts were bleak, dominated by the war. She had already lost her brother in the first world war and her husband was dead; now her father was a refugee in London and her only son, 17 years old in 1939, faced the prospect of military call-up. To Roxburgh she wrote in September that she hoped Michael ‘would live to do you credit’. But she told Gabo in November: ‘I still shiver at any mention of air raids, but in my calmer moments I am convinced that the fear and horror which Picasso made conscious are a universal deterrent and that no nation will dare to start a campaign against open towns because of the reprisals.’ Within less than a year, Dorothea herself would be dead, leaving Michael entirely alone.

  2

  The War Years

  ‘One can remain sure that no Champollion is working quietly in a corner and preparing a full and startling revelation, as no one has access to sufficient reproductions.’

  Michael Ventris, as an RAF serviceman, in a letter to Sir John Myres, October 1942

  At Easter 1940, presumably during a break in his architectural training, Ventris sat down in his room at 47 Highpoint and typed a two-page letter to Sir Arthur Evans of extraordinary self-assurance for someone so young. It began: ‘I don’t know whether you remember my writing to you a few years ago about some theories I had on the elucidation of Minoan. Actually I was only 15 at the time, and I am afraid my theories were nonsense; but you were very kind and answered my letters. I was convinced that the key would prove to be in Sumerian,’ – that is, the earliest-known language of Mesopotamia – ‘but I am glad to say I have given these ideas up long ago.’ The rest of the letter went on to propose a new candidate language for ‘Minoan’: ‘a dialect closely related to Etruscan’.

  Six months later, after tearing up two drafts, Ventris had produced a lengthy, scholarly manuscript about his new theory. As the Germans began to bomb London in early September and the battle of the Atlantic got underway in earnest, he mailed this to the American Journal of Archaeology, the leading archaeological journal of the United States. Immediately, it was accepted and rushed into print by the editor in the last number for 1940 under the title ‘Intr
oducing the Minoan language’ with the wholly unknown author given simply as ‘M. G. F. Ventris, London’. He was just 18 years old.

  During the decipherment and after, a decade later, Ventris would largely disown this paper, as he had disowned his juvenile letters to Evans, but it is in fact of great interest. Not because its conclusions were correct – they were dead wrong – but because of the light it throws on the workings of its author’s unusual mind. The paper’s curious mixture of cold logic, scholarly caution, wide reading, wild assertions and imaginative leaps suggests why it would be Ventris, and no ordinary classical scholar, who would eventually decipher Linear B.

  At the very beginning, Ventris dismissed what he called ‘the fantasy’ of a number of scholars ‘which makes Minoan out as Greek’. He pointed out the flaws in their methods and noted that their various readings, supposedly all in Greek, were largely unrelated to each other. This was fair and reasonable, but then he let slip his main argument: ‘The theory that Minoan could be Greek is based of course on a deliberate disregard for historical plausibility.’ In other words, Ventris was consciously taking the same line as Evans, mentioned in the Introduction: that the civilization of the Minoans on Crete was known to be historically earlier and artistically different from that of the Greek mainland at Mycenae, and therefore the Minoan language had to be a pre-Greek language.

  In following the same line, Ventris was aware that Evans himself had come dangerously close to the Greek hypothesis for ‘Minoan’, in volume four of The Palace of Minos. There, in search of further clues to the decipherment of Linear B (after he had identified the word dividers, numerals and pictograms), Evans had turned east, to Cyprus. Here was another island on which an ancient script had been found, dating to about 800–200 B.C. But unlike Linear B, the classical Cypriot script had been deciphered (in 1871), because it appeared alongside the classical Greek alphabet in a number of ‘bilingual’ inscriptions:

  Greek/Cypriot inscription.

  (© British Museum, 2001)

  To digress for a moment, the spoken language represented in these ‘bilinguals’ is the same in the case of both scripts: Greek – a dialect of Greek in the Cypriot case. The historical reason for this, according to classicists of Evans’s day, was that Greek speakers fleeing the Trojan war had brought Greek to Cyprus. Since the sounds of the Greek alphabetic signs were known, the sounds of the Cypriot script could be deciphered and matched with their corresponding signs. But the Cypriot script turned out to be, not alphabetic, but syllabic, with 56 signs, one for each syllable; an inconvenient way to represent Greek sounds, if manageable. The Cypriot syllabary is a so-called ‘open’ syllabary, in which a syllabic sign stands not for a consonant C but for a consonant with an inherent vowel, CV. (In ‘closed’ syllabaries, a sign stands for CVC.) This means that when an ‘open’ syllabic sign is used to represent a final consonant in a word, the sign’s inherent vowel must be assumed to be silent, i.e. C(V). In classical Cypriot, therefore, the many words that finish with the syllabic sign , se, have a silent e and actually end in s, which agrees with a very common ending for words in classical Greek, ‘-s’ (e.g. logos, Dionysos). This was as expected in a dialect of Greek.

  Evans, however, was not looking to the Cypriot script for its Greek connections – rather the opposite, given his Minoan predilections. His hope was that the known sounds of the Cypriot script could help him to decipher the unknown sounds of the Linear B script, but without assuming that the language of Linear B was Greek, or even a dialect of Greek. His idea may seem somewhat perverse, but it was based on a theory of his that the Cypriot script was somehow derived from the ‘Minoan’ Linear B script, while the Cypriot language was not derived fromthe‘Minoan’ language. (One might recall the modern Turkish script, which Kemal Atatürk deliberately derived from the roman script, even though the Turkish language is not derived from any European language written in the roman script.) According to Evans, ‘Minoan’-speaking people, possibly traders to begin with, must have settled in Cyprus, bringing their script with them from Crete. That was why, he said, some of the Cypriot signs looked so similar to the Linear B signs, despite being up to a thousand or so years younger than Linear B.

  Here are the eight most similar signs and their syllabic phonetic values in Cypriot:

  Evans decided to test these values on a promising-looking tablet from Knossos:

  (Horse tablet from Chadwick, The Decipherment of Linear B, 1958)

  He noticed on the tablet six horse heads, two of which were incomplete. (The join in the tablet was made by John Chadwick years after Evans’s death, so Evans’s drawing below does not include the left-hand portion.) Of the four horse heads in the middle and on the right of the tablet, two had manes and two did not. The ones without manes, foals presumably, were preceded by the same pair of Linear B signs:

  (Tablet drawing published in A. Evans, Palace of Minos at Knossos)

  According to the Cypriot phonetic values, the two signs should read po-lo. What might ‘polo’ mean in the ‘Minoan’ language? Evans duly noted that it resembled the classical Greek word ‘pōlos’, young horse or foal, (and its dual form ‘pōlo’, two foals); in fact ‘foal’, the English word, comes from the same source as Greek ‘pōlos’. If the ‘Minoan’ language and the Greek language were related after all, ‘Minoan’ ‘polo’ could easily be the equivalent of classical Greek ‘pōlos’. The tablet would then mean:

  horses 2 polo foals

  polo foals 2

  horses 4

  Presumably, the word (‘polo’) had been added by the Minoan scribe to make it absolutely clear that the maneless pictogram was a foal and not an adult animal.

  But Evans rejected this plausible beginning, almost out of hand. For one thing, he noted that, unlike Cypriot words, very few Linear B sign groups ended in the sign (se), ‘-s’, which suggested that ‘Minoan’ and Greek were not related. A logical enough deduction, and one that would trouble all subsequent decipherers including Ventris. Less logical was that Evans simply could not accept that the Minoans spoke and wrote an archaic form of Greek, which they took with them to Cyprus. In Evans’s view, it was Minos and the Minoans, not the mainland Greeks, who ruled the roost: the ‘Minoan’ language could not possibly be Greek. He dismissed the similarity of the Linear B and Cypriot signs in the case of ‘polo’ as a mere coincidence of the kind that, in fairness to Evans, must be admitted to be only too misleadingly common in historical linguistics and decipherment. It is a general rule of decipherment that to allot sound values to unknown signs purely on their visual resemblance to known signs is a risky procedure, likely to be wrong.

  Of course, this left the big question wide open: what was the language spoken by the Minoans, if it was not Greek? Ventris decided that it must be related to one of the languages known to have been spokenaround the Aegean in the second millennium B.C., which included the three Anatolian languages Lydian, Carian and Lycian – and also Etruscan. Although Etruscan was spoken only in Italy during classical times and eventually disappeared altogether with the rise of Latin, the language was said by Herodotus originally to have been the language of an Anatolian people from Lydia who had migrated through the Aegean to Italy in pre-Greek times, presumably in the second millennium. Some archaeological support for this legend came from a 6th-century-B.C. stone inscription found on the island of Limnos, near Anatolia, in the 19th century, which appeared to be written in a language similar to Etruscan. (Modern scholars, it should be said, disagree with Herodotus and treat the Etruscans as indigenous to Italy; and they regard the Limnos inscription as a controversial enigma.)

  The principal reason why Ventris favoured Etruscan over the three Anatolian languages came from his acceptance of the Evans theory that the Cypriot syllabary provided a key to Linear B and its ‘Minoan’ language. He noted that the Cypriot syllabary did not represent the consonants b, g and d, and proposed that the same was true of the ‘Minoan’ language. He further noted the very same feature in Etruscan, but not in the A
natolian languages. Therefore, ‘Minoan’ must be closely related to Etruscan. Pursuing this theory, he then allotted Cypriot sound values to many Linear B sign groups which he believed to be names – on the assumption that similar-looking signs in Cypriot and Linear B had similar values – and, hey presto, some Etruscan-sounding names, such as Vilie (compare Etruscan Velia) popped out of the tablets from Knossos! ‘That we should find this typically Etruscan name used by the women of Minoan Crete…is a startling demonstration of the fundamental linguistic unity’, Ventris audaciously claimed.

  He was building a house of cards on shifting sands. Yet it is worth asking why he remained so committed to the Etruscan hypothesis – and so opposed to the Greek hypothesis – from 1940 right up to 1952. There were some solid reasons, such as Evans’s arguments against Greek, the undoubted existence of pre-Greek inscriptions in the Aegean area, the various words in the ancient Greek language which the Greeks themselves knew to be ‘barbarian’ imports, and the evidence of Homer, Herodotus and other commentators for the existence of pre-Greek (non-Indo-European) languages in the Aegean. But over and above these, Ventris seems to have set his heart on a non-classical, non-Indo-European linguistic solution to ‘Minoan’ because he disliked the dominance of the classical languages over western culture. After the war, he even wrote: ‘Our researches have been prejudiced by conscious or unconscious acceptance of the official Nazi doctrine, according to which all admirable civilizations are due to Nordic blood (or at least to the speakers of Indo-European languages), and which incidentally saw in the Etruscans an earlier prototype of all those non-Aryan vices of luxury, obscenity, cruelty, necromancy and usury for which the Jews were later to be scapegoats.’ At bottom, it would seem, Ventris’s commitment to the Etruscan solution – like that of Evans to ‘his’ Minoans – was emotional, not intellectual.