However he stuck it out, was awarded a diploma, and in September 1949 joined the Ministry of Education as part of a development group of architects dedicated to the design of new schools, who were located in breeze-block offices with regulation ‘battleship grey’ linoleum floors in a central government building in Curzon Street (later well known to Londoners for being occupied by the secret service). This work was more satisfying than the planning course, because it had a definite focus for new thinking and design: the need to put into practical effect the large expansion of state education legislated by the Butler Act of 1944. School teaching was becoming less rigid and more liberal, and the era of open-plan classrooms, movable desks, learning by doing, bright colours and a generally less ‘Victorian’ educational environment, was dawning. A fellow architect and graduate of the AA, who joined the group at exactly the same time, Dargan Bullivant, remembers quite an intense, even intellectual atmosphere in the office, influenced by two slightly more senior figures, Mary Crowley and David Medd, a childless, husband-and-wife team reminiscent of the formidable Fabian team of Beatrice and Sidney Webb. Bullivant soon realized that Ventris was ‘an intellectual aristocrat’, though he found him to be easy company, never a show-off: ‘he was not a natural expositor, and if you didn’t follow him the first time, you would not feel you could ask him again. But he didn’t try to humiliate you with his cleverness.’
From surviving drawings of a school by Ventris, he clearly worked to his usual clean-lined, fastidious standards. Still, it is hard to see how designing schools could have excited or challenged his mind for long – especially as he had an aversion for ‘sharawaggi’ in design, had not much enjoyed his own school days, and was not very interested in children. Soon, his mind was moving on. A younger architect in the office who became a friend, Edward Samuel, remembered that during lunch breaks Ventris would work on Linear A and Linear B (Samuel would help him look for similarities of sign in the text on different tablets), and he would make phone calls to his stockbroker; once, he told Samuel after finishing a call that he had just made more money in that one call than his entire year’s salary. He was also deeply involved, outside the office, in assembling a ‘Guide to Modern Architecture in Western Europe 1900–1950’, naturally including all the buildings he admired in Copenhagen and in Sweden, which was based on a survey questionnaire of his own design circulated among architects in Britain and on the Continent. And he was looking for other architectural work: for example, a Festival of Britain Information Office in Leicester Square, which he and Oliver Cox designed together.
His small spring-backed, loose-leaf design book for this modest project still exists, dated October 1949. Interleaved among the sketches of the room, its furniture and various gadgets, are lists of Linear B sign groups with possible Etruscan parallels. There can be no more eloquent witness to the way in which design and decipherment ran along side by side in Ventris’s complex mind.
A month or so later, while working at the Ministry of Education, he took his next major step towards the decipherment. His own introductory words describe it best. ‘At the end of 1949 I sent out copies of the following questionnaire to a number of scholars who have been working in recent years on the problems set by the language and writing of the prehistoric Aegean, suggesting that we might make New Year 1950 the occasion for an informal exchange of views, reviewing the position reached at the end of the half-century’ – i.e. 50 years after Evans first discovered the Minoan scripts. ‘The answers to the questionnaire are collected together in this progress report, which I have had duplicated and circulated privately at my own expense: a venture which enables me to add some notes of my own which are too disjointed for more formal publication.… I have, by general consent, typed the whole bulletin in English, translating a proportion of the contributions which were in French, German, Italian or Swedish; and must therefore take responsibility for any conclusions awkwardly or imperfectly expressed.’
The ‘venture’ – which Ventris also called ‘an interesting experiment in international cooperation’ and which soon became known as the Mid-Century Report – was wholly typical of him. No professional classicist would have conceived it, and if one had, he or she would almost certainly have lacked the linguistic skills to translate the replies, which came from all over Europe and from the United States. It was ‘group working’ applied to a completely different field, the decipherment of the Minoan scripts, in a disinterested effort to break the scholarly impasse.
The questions were numerous, penetrating and detailed. For example:
1 What kind of language is represented in the Linear B inscriptions, and to what other known languages is it related?
8 Do you feel that certain distinctions should be made between particular Linear B signs which have tended to be confused in the signaries so far published?
10 If the signs are syllabic, what kind of syllables do they represent?
12 Is the Cypriot syllabary directly descended from the Minoan scripts?
17 To what extent are prefixes, or compounds, or suffixes, the typical mechanism for forming Minoan names?
20 Which sign groups in Linear B (and in Linear A if you consider the language identical) appear to be VOCABULARY WORDS (nouns, adjectives, pronouns, numerals, prepositions, adverbs, conjunctions) rather than proper names?
Several scholars wrote highly detailed responses, as did Ventris himself, and every scholar actively working on the Minoan scripts replied, except for two. The first was the Prague scholar Bedřich Hrozný, whose recent published decipherments of Linear B, the Indus Valley script and several other undeciphered scripts had been roundly rejected by scholars, despite his respected work during the first world war in deciphering Hittite cuneiform. (‘From this occupational disease of decoders we may all wish to be preserved’, wrote Ventris, after Hrozný’s death.) The second was Alice Kober. She wrote testily: ‘I have no intention of answering the questionnaire. In my opinion it represents a step in the wrong direction and is a complete waste of time.’ By then she had only three months to live.
There was little consensus, however. Perhaps the most striking fact is that not a single scholar, including of course Ventris, ventured to suggest that Linear B’s language might be Greek. Some thought the language might be Indo-European, others non-Indo-European, while Bennett and Myres (and of course Kober) refrained from speculation.
Ventris concluded his twenty-page response by stating: ‘I have good hopes that a sufficient number of people working on these lines will before long enable a satisfactory solution to be found. To them I offer my best wishes, being forced by pressure of other work to make this my last small contribution to the problem.’ For the rest of 1950, he went back to full-time architecture.
But, as ever, Linear B would not leave him be. During that year, the pace of his correspondence with Bennett quickened. In July, they met for the first time, in London, when Bennett visited Myres on his way to Greece. He showed Ventris some of the Pylos material, and they formed the beginnings of a friendship. In November, after reading an excellent article by Bennett (the one on the numerical systems of Linear A and B) in the American Journal of Archaeology, Ventris wrote that he was so excited he had to jump out of bed at 3 a.m. in order to jot down his (lengthy) response. Publication of the Pylos tablets was now scheduled for the spring of 1951, and surely the long-awaited Knossos tablets could not be much further off. The prospects for success were brightening. Soon, Ventris decided he must quit his job, draw upon his private income, and sit at home in Highpoint to concentrate exclusively on deciphering Linear B. Characteristically, he did not explain his reasons to any of his colleagues at the Ministry of Education.
A sketch of 47 Highpoint by Oliver Cox.
He would never really return to architecture. So maybe this is the moment, at the parting of his parallel careers, to summarize exactly what the two had in common. First of all, there is the problem-solving nature of architecture and decipherment: both architects and decipherers, at the
most fundamental level, are looking for solutions to problems that satisfy a given set of constraints – whether these be that a design must fit a certain space or that a language must be Indo-European. Then there is the functional nature of both buildings and writing systems: both are designed by human beings to be used; neither is art for art’s sake. As for the methods adopted by Ventris to solve design and decipherment problems, there can be no doubt that he applied group working to both architecture and Linear B; and that the way he would record all his ‘working’ on Linear B (including errors), in a series of twenty Work Notes, was a direct evolution of the detailed design books he kept as an architect. Finally, uniting fundamental similarities with methods, there is the way in which both architecture and decipherment require the mental manipulation of large amounts of visual and written data. Perhaps Chadwick – in one of his rare references to Ventris’s work as an architect – put the link between one career and the other best when he wrote: ‘The architect’s eye sees in a building not a mere facade, a jumble of ornamental and structural features; it looks beneath the appearance and distinguishes the significant parts of the pattern, the structural elements and framework of the building. So too Ventris was able to discern among the bewildering variety of the mysterious signs, patterns and regularities which betrayed the underlying structure. It is this quality, the power of seeing order in apparent confusion, that has marked the work of all great men.’
5
Into the Minoan Labyrinth
‘To wait for a bilingual to help us solve our problem is to cry for the moon.’
Michael Ventris, Work Note 15, September 1951
And so, at the beginning of 1951, we reach the most intense phase of Ventris’s work on Linear B, which would culminate a year and half later in his announcement of the decipherment. Writing after the event, in 1953, he set down the basics of his approach in words so concise and masterly they are worth quoting, before we try to understand the details of how he did it.
Any decipherment, he wrote, ‘needs to be planned in three phases: an exhaustive analysis of the signs, words and contexts in all the available inscriptions, designed to extract every possible clue as to the spelling system, meaning and language structure; an experimental substitution of phonetic values to give possible words and inflections in a known or postulated language; and a decisive check, preferably with the aid of virgin material, to ensure that the apparent results are not due to fantasy, coincidence or circular reasoning.… Prerequisites are that the material should be large enough for the analysis to yield usable results, and (in the case of an unreadable script without bilinguals or identifiable proper names) that the concealed language should be related to one which we already know.’ With Linear B, the analysis phase lasted up to June 1952, when the substitution phase took over and revealed that the language was probably Greek, which was then confirmed in the check phase after May 1953, when sensational virgin material from mainland Greece became available.
What makes Ventris’s decipherment so remarkable is, of course, that before it took place the script was completely unreadable; there were few reliable clues to the concealed language (Ventris’s guess, an Etruscan-related language, was entirely wrong); there were no bilinguals or identifiable proper names (no Rosetta stone or inscribed obelisks, containing predictable names like Ptolemy, Cleopatra and Alexander); and finally that, until 1952 when the Knossos tablets were published, Ventris had access to a rather limited quantity of material (unlike in the decipherment of Egyptian hieroglyphs, Babylonian cuneiform and Mayan glyphs).
‘The main thing’, Ventris told Bennett in mid-1951, ‘is to discuss the data objectively, at this first stage, without looking forward to conclusions which one has reached by more experimental means.… You’re less likely to put your foot wrong here than I am.’ The nearly two hundred pages of his twenty Work Notes bear out his self-criticism. Densely typed (by Ventris himself, like all his work) and profusely illustrated with his own Linear B signs, they overflow with highly ingenious linguistic analysis, hypothesis and experiment, jumping back and forth between these modes of attack and between a bewildering range of Linear B inscriptions (buttressed with references to inscriptions from other languages and neighbouring civilizations). And naturally the notes show only what Ventris chose to commit to paper, not the mental gymnastics and intuitive leaps that led to the words on the page; these can only be conjectured. ‘If the secret of the Linear B labyrinth was at some points penetrable, he was the man to penetrate it, thanks to his scouting and his probing’, a scholar wrote much later. But there is unfortunately no thread like Ariadne’s running through the labyrinth. John Chadwick admitted as much, and in The Decipherment of Linear B made little attempt to present the decipherment in the same order as the Work Notes. Even Ventris, despite his best efforts, was unable to produce a coherent narrative of his method (like Champollion, it is interesting to note); and he went so far as to advise Chadwick that Work Note 19 – written a mere two months before his breakthrough – was in fact ‘complete nonsense’. So we must, I fear, be humbly content with a summary of Ventris’s key techniques and conclusions, based on his own simplified account supplemented by his Work Notes and letters to Myres and Bennett, followed by a fuller explanation of what was unquestionably his moment of revelation.
To kick off, let us look at the ‘grid’, mentioned in the last chapter in connection with Alice Kober’s analyses of the unknown vowels and consonants in the syllables of Linear B. Not only do Ventris grids exist in paper form in three of the Work Notes, Ventris also constructed a physical grid which he could play with whenever he felt like trying out a new move. An architect friend, Michael Grice, who used to visit 47 Highpoint, remembered seeing an array of labels with Linear B signs written on them hanging from hooks in a wooden board fixed to a wall of the flat. Although Ventris characteristically did not talk about his toy, Grice was struck by how similar it was in concept to the way in which architects think about solving problems. (Sadly, unlike Crick and Watson’s almost equally primitive metal ‘double helix’ model of DNA, now in the Science Museum in London, Ventris’s model is not to be found in any ‘museum of decipherment’, because it no longer exists.)
The very first Work Note, dated 28 January 1951, contains a grid, with vowels down the columns and consonants across the rows. Ventris has experimentally placed a range of signs on it, according to theinterrelationships he had worked out from the few published tablets; some of these signs would turn out to be in the correct position. He has also guessed some phonetic values on the basis of Etruscan analogies; these values were mostly incorrect, as might be expected.
Ventris’s phonetic ‘grid’ from Work Note 1, January 1951. (Ignore the labels such as ‘ag’, ‘id’ and ‘ol’ which are not proposed phonetic values.)
As the Work Notes progressed, the grid steadily filled up with Linear B signs like words in a crossword, while Ventris dropped or refined his guesses at the phonetic values of the vowels and consonants. Although the positions of the signs continued to be revised, the grid endured – and assisted Ventris to predict, if only very tentatively, new interrelationships between signs, as when a crossword is gradually completed. According to Ventris and Chadwick, writing after the breakthrough: ‘The problem of decipherment is in this way reduced to the correct distribution of five vowels and twelve consonants to the columns of the grid; and since a proposed reading of only two or three words may, by a “chain-reaction”, predetermine rigid values for almost the entire syllabary, a very severe discipline is imposed on the earliest stages of a decipherment. If the initial moves are wrong, it should be quite impossible to force any part of the texts into showing the slightest conformity with the vocabulary or grammar of a known language; even though that might be quite easy if one were free to juggle with the values of 88 mutually unconnected signs.’ While this was true in principle, in practice the Linear B grid never played as decisive a role in the decipherment as their statement implies.
The first
substantial progress was possible when Ventris received The Pylos Tablets: A Preliminary Transcription from Bennett in March 1951. At this time, even six years after the end of the war, Britain still had rationing and a relatively austere regime; gifts from the United States were especially welcome, though carefully vetted by customs. An amused Bennett recalled decades later that he received a letter of thanks from Ventris explaining how when he went to pick up the packet containing Bennett’s small book, a suspicious London postal official asked him: ‘I see the contents are listed as PYLOS TABLETS. Now, just what ailments are pylos tablets supposed to alleviate?’
In Work Note 7, dated 1 May, his first note on the Pylos tablets, Ventris recorded a series of important ‘assumptions and objectives’ behind his approach:
1 Until the Pylos material has been fully analysed, or until the Knossos material is published, discussion is best confined to the Pylos tablets, and the Knossos evidence only brought in where a definite analogy is helpful on a specific problem.
2 Conclusions drawn from an initial analysis of the Pylos inscriptions must be regarded as tentative until the whole of the Knossos material has been published, and, in the case of fragmentary evidence, until the transcriptions have been collated with the originals.
3 In spite of slight differences in script and in accounting method which suggest a different date and place of origin, the Knossos and Pylos tablets are written in the same language, which is not Greek.…
The Man Who Deciphered Linear B Page 7