Jo Marchant's book Decoding the Heavens: Solving the mystery of the world's first computer was first published in November 2008 by William Heinemann (UK) and in February 2009 by Da Capo Press (US). It purports to tell the stories of the people who uncovered the Antikythera mechanism, a remarkable device that was in the cargo of a 2000 year-old shipwreck discovered off the island of Antikythera in 1900.
Pages 171 to 193 are concerned with the collaboration of Mr Michael Wright and Dr Allan G. Bromley, an Australian computer scientist and historian of computing. Allan died in 2002. His wife and some of his friends believe that many of the facts in this account are wrong and that Jo Marchant's description of Allan's character is a cruel smear. This commentary shows some of the evidence for our belief.
The first version of this commentary was sent to the UK publisher in mid June 2009 in the hope that some action would be taken to correct the record. To date (4 August 2009) the only reaction has been an acknowlegement of receipt. We intend to continue updating this page if more information becomes available.
Our commentary shows that the chronology of Jo Marchant's account is wrong and hence disproves much of her interpretation of the events. There is no way to prove conclusively that her character description is even more wrong: it might seem that this is just a difference of perspective. However, the reader is entitled to know that Marchant did not contact anyone in Australia during research for her book. This failure is possibly the reason she is wrong about such a simple detail as the date Allan died. This failure certainly cost her the recollections of clockmaker Frank Percival, who she mentions in passing as one of the builders of a physical reconstruction of the mechanism. This failure also means that Marchant did not test her interpretation of the accounts of Allan's character she received from her sources in Britain. The result is a distressing distortion.
Decoding the Heavens has problems, some serious and some silly, in other sections as well. For example, the description of the sun's annual movement through the zodiac (p. 49) is literally backwards; if the sun moved as Marchant describes the order of sun signs would be from Aries to Pisces, Aquarius, Capricorn, and so on, instead of the other way around. The Antikythera Mechanism Research Project website contains a commentary which details this and other mistakes and misinterpretations of the science and misattributions of credit.
You might think that this page is an over-reaction, that the book is just presenting a side of Allan that we never saw. You might be right, though when so many details are factually incorrect you might also wonder how reliable any of the interpretation or inference can be.
The book's index contains a reference for Bromley, Allan G.: character of. It points to this statement:
To Bromley, knowledge really was power. Throughout his life he had defined himself by what he knew and others didn't. It was what drove him to become the world's foremost expert on Charles Babbage, and it was what drove him to return to Athens year after year to study the Antikythera mechanism. Knowledge wasn't to be shared, it was to be held on to like a currency, to be bartered at a later date...
Here are some different views of Allan. They come from people who, unlike Jo Marchant, knew him.
We welcome questions or comments, especially if you knew Allan Bromley or have information about his work on the Antikythera mechanism.
These comments refer to Jo Marchant's account of the collaboration between Michael Wright and Allan Bromley, which makes up the second and first parts of chapters 6 and 7 of Decoding the Heavens. They are based mainly on a first reading of unpublished correspondence between Wright and Bromley from 1984 to 1996, now held by Anne Bromley. The author did not attempt to contact Anne Bromley during the preparation of the book. Had she done so this material would have been made available to her.
The correspondence is in the form of letters, many hand-written. From Wright and Bromley's joint and individual publications, from the recollections of people in Sydney, and from the content of the letters dated after 1996 there is good reason to think that Marchant's assertion (p.188) that that communication between the two dried up is wrong. The email archives from the the University of Sydney Department of Computer Science are the obvious place to look for evidence of this but retrieving them could take a while and we feel that the paper correspondence provides sufficient evidence that Marchant's account is seriously inaccurate in other ways as well.
Some of these comments draw on conversations in May and June 2009 with Frank Percival, the clockmaker who worked closely with Bromley from 1985 and 1987 to build a reconstruction of the Antikythera mechanism and continued to contribute from his specialist knowledge until Bromley died. Michael Wright stayed with Percival for a couple of nights during his visit to Australia in 2000, and has remained in touch with him. Percival is now in his eighties and might have been reluctant to talk with a stranger about his significant part in the studies of the Antikythera mechanism and his long friendship with Bromley, but his recollections could have been another source of information and verification of Wright's memories.
Marchant has said that she did not contact anyone in Australia for her book because there are so many different points of view about the story of the Antikythera mechanism that she decided to write from the perspectives of each of the main protagonists in turn to tell things as they saw it. We feel this explanation is inadequate. Apart from the author's own, Wright's perspective is by far the most prominent in the book and it seems to us that normal journalistic practice would require at least an attempt to verify his memory and her understanding of events from 25 years ago. If Bromley was alive Marchant would certainly have sought his input. Her decision to ignore the best sources of information about his perspective on the story has left her reporting of Wright's perspective open to questions and of Bromley's role in the overall story seriously flawed.
Sandra Oates, Anne Bromley
18 June 2009
Page 170: The two men [Michael Wright and Derek de Solla Price] had briefly met, when Price had called at the Science Museum to see the newly discovered Byzantine sundial during his visit to London in 1983. But that was before Wright started studying the Antikythera mechanism in earnest, and within a fortnight of that meeting Price was dead.
and
Page 185: For the last five years he [Wright] had spent every spare moment thinking, planning, preparing for his work on the Antikythera mechanism.
Price died in September 1983. A letter from Bromley dated June 1984 mentions that he is still getting over the flight home and goes on to discuss calculations of gear trains, the al-Biruni device (later called the London Sundial Clock and apparently the Byzantine sundial), and the Antikythera mechanism. Thus Wright and Bromley started studying the mechanism “in earnest” at about the same time and their studies were collaborative from the beginning.
The correspondence also shows that they got into misunderstandings about credit and publications and worked them out as they went along. This is hardly unusual.
[Bromley] was convinced that [Babbage's] designs would work, but he had been looking at them from the point of view of a computer scientist. He understood the logic and the theory behind them, but he wanted to know more about how the parts of the mechanism would have been made and put together...
These passages set up a false contrast between Wright as practical worker and Bromley as dilettante theorist that becomes overt later in the chapter.
Just like Wright and Price, and indeed like thousands of other children up to the 1970s and 80s, Bromley's interest in mechanics and Meccano began in childhood and was life long.
Bromley built his collections of mechanical computers to learn and to demonstrate the multiplicity of techniques involved. They were the reason he got started on Babbage and the knowledge he had developed in working with them was part of the basis for his confidence that Babbage's design for Difference Engine No. 2 could be built.
With that background, it's meaningless to say that Bromley was looking at Babbage's designs “from the point of view of a computer scientist”. He went to Wright for details of nineteenth century techniques, not because he didn't know about “making geared mechanisms.”
Wright would have been one of the few people who could understand what Bromley was on about. Maurice Wilkes was another, witness the correspondence he edited after Bromley died. Bromley formed strong friendships with both of them.
Page 176: ...Instantly, Bromley was mentally converting [the Antikythera mechanism's] gear trains into circuit diagrams...
The author gives no indication of the source of this insight into Bromley's instant mental conversion, which on the face of it is highly unlikely. Bromley had profound knowledge and a long-standing interest in mechanical computing. Why would his first reaction to an ancient and unique mechanical calculating device be to convert it to electronics?
The correspondence shows that in fact his first reaction was to analyse the gear trains and consider the astronomy.
Page 176: ... and a new plan began to form in his mind. He would be the man to solve the Antikythera mechanism.
Marchant gives no hints about what Bromley's plan might have been, and neither does the correspondence of the next 10 years. In the second half of 1984 Bromley and Wright exchanged at least eight letters, each asking for and receiving comments on draft papers on various subjects. From 1985 to 1995 they discussed strategy for research on the mechanism and many details of their individual and joint work on it. They also discussed other matters of technical interest, and professional and personal problems. The letters indicate a respectful friendship between two mildly eccentric people who had many interests in common and who were working, in spite of obstacles, on a joint project both found deeply interesting.
Marchant appears to have invented this “plan to be the man” as part of her dramatic notion that the Antikythera mechanism took possession of researchers (page 2, 216 and elsewhere), like some kind of Hellenistic Sauron's ring.
Page 177: ...Bromley had a theory that the device couldn't have been driven by the slow-moving big wheel...
This was an observation rather than a theory and Bromley didn't claim it. In his first Antikythera paper, published in Centaurus in 1986, Bromley thanked Wright for pointing it out to him and stated his understanding that E. C. Zeeman had made it in public. He acknowledged Zeeman again in later papers.
Frank Percival laughed when he heard this and said that it was Wright who had terrible trouble getting triangular teeth to mesh properly in his own work on old clocks. According to Percival, when Bromley left the newly completed model with Wright for a while Wright pulled it apart to see how they had got it to run so smoothly and then couldn't get it back together.
Bromley and Percival worked on the reconstruction for more than two years. On the practical side, the work included making gear cutters and experimenting with methods of dividing the wheels, all using hand techniques that could have been or that they knew to have been available to a metalworker in Hellenistic Greece. It's incorrect to say that they had to round the teeth off at the edges, except for those on the very smallest wheels. For most of them the technique was to adjust the depth of the cuts.
The “proper reconstruction” was completed in 1987. The work of developing it and the completed model itself were a significant factor in Bromley's obtaining permission to study the original mechanism.
Bromley and Wright discussed this reconstruction at their usual detail and length while it was in progress.
The word “meanwhile” makes it difficult to know when Marchant suggests Wright's loss of faith occurred, though the next paragraph implies that it was before Christmas 1989. Wright's disbelief in Price's work is a key element in Marchant's portrayal of Wright, but the correspondence suggests that she overstates the significance Wright gave it and how early it developed.
In October 1984 Wright commented in response to one of Bromley's early draft papers on the gearing of the mechanism, “It seems to me that the real point of 'Gears from the Greeks' is the message, 'they could do it and they did!'. Isn't Price's reconstruction almost incidental?” This and further correspondence in that year strongly suggest that at that stage Bromley was more interested in the details of Price's reconstruction than Wright was.
In May 1989 Bromley applied to the University of Sydney for special study leave. His supporting statements include a summary of work to date on the Antikythera Mechanism, in which he describes how “Reluctantly, I reached the conclusion that all of Price's observations would have to be repeated as the number of errors was substantial enough that none of his observations could be taken entirely on trust.”
In November 1993, five years after the events of this section and shortly before the last trip Bromley and Wright made to Athens to complete their radiographs, Bromley wrote from England to Percival, “After a lot of hard talking I have convinced Michael that we can be certain of almost none of Price's reconstruction '- basically the front dials with parapegma, the 127 tooth gear (D2?) in the Metonic train, and that's it.”
Bromley might have been wrong to think that Wright's conviction was so recent, but if Marchant had known of the letter she could at least have asked Wright about it.
Wright's jaw dropped. All this time he had been dreaming about studying the mechanism and now Bromley – his friend, the man he had confided in – had stolen his idea.
Bromley's first trip to Athens was in Christmas 1988, and Wright knew all about it. He wrote some funny letters responding to Bromley's numerous postcards and their cryptic observations of the Antikythera mechanism and the Tower of the Winds.
Throughout 1989 Wright and Bromley wrote frequently, sometimes daily, about their thoughts on the mechanism and about plans for study in Athens. In May Bromley sent Wright a copy of his own research proposal with the comment “Feel free to use any of it that is of aid in raising grants to go to Athens. I want to see you there!”
On 23 October Wright reported that he had applied for a grant from the Royal Society for his own trip to Athens but that he thought he was unlikely to get leave from the Science Museum. A letter from Bromley dated 1 November 1989 told Wright the exact dates he would be in Athens and London and provided two Athens addresses.
Bromley's arrival from Athens was in no way a surprise to Wright.
The first written mention we have found of studying the mechanism in Athens is in a letter from Bromley dated 2 November 1984, and discussions about work in Athens continued at intervals for the next four years. Wright and Bromley discussed what they could expect to achieve, how to demonstrate their qualifications and seriousness as scholars, and possibilities for finance.
Studying the mechanism directly was never Wright's idea alone and Bromley did not steal it.
Page 178: Wright... swallowed his pride and asked Bromley to take him along as his assistant.
The idea that Bromley could casually take an assistant along to study these unique items, with virtually no notice to the authorities and on his own first study trip to Athens, is unlikely on the face of it and it's not what happened.
The author might have been confused by the fact that in 1990 Bromley was able to partly fund Wright's stay in Athens by including him as an associate researcher on a study grant from Sydney University. In a letter to Wright Bromley carefully explains that the grants are only available to members of the University and this is the only way he could obtain the funds. In other words, he went to considerable trouble to help Wright get to Athens with him.
This is possible, though the implication that Bromley was goofing off while Wright did all the work is just a slur. But it's also possible that Bromley was unwell because of Glandular Fever, which he contracted around this time and to which he and his doctors attributed much of his ill health throughout the 1990s. There's no way to tell now. It's only worth pointing out because it's another stroke in the ugly portrait of Bromley that Marchant is painting.
Did the museum have only one scientist? Seems unlikely. According to the Antikythera Mechanism Research Project website “Eleni Mangou is chemist and Head of the Chemical and Physical Research Department of the National Archaeological Museum in Athens since 1979.”
The author doesn't seem to know anything about Bromley and Wright's experiments with stereoscopic viewing. These preceded the tomography, and provided significant new detail about the mechanism (including some new inscriptions) as well as for the design of the tomography equipment.
Wright did indeed build the equipment, but he and Bromley collaborated on its design through frequent letters and occasional phone calls. Percival, whose continuing contributions to the work Marchant ignores, made a test model of the mechanism so that Wright could test his equipment before the trip to Athens. Bromley and Wright both “read up on tomography”, they both experimented with photographic techniques, and years later Wright asked Bromley to review the draft of his paper about it.
This is ridiculous. What university would ever “provide some extra funds” at no notice to an absent academic, however sweet his talk? If Bromley indeed found some extra funds it was by re-allocating money within his research grants, and he had to account for it. This makes him sound like a con artist.
... Bromley coming during his university's summer vacation and Wright using his precious holiday time.
Here we go again. Marchant implies that Bromley sat around drinking while Wright slaved in darkness developing the plates, that Wright used up his precious holiday time while Bromley just filled in time while the university was on vacation. Taking the exposures was also a delicate job requiring heavy concentration and steady hands. Bromley also used precious holiday time for at least some of these trips, and in some years was unable to take his unused holidays because he had courses to teach. They both worked very hard. They both enjoyed themselves enormously.
Page 185: It was February 1994...
Bromley's student (Bernard Gardner) finished his honours project in November 1993, so the date of this alleged announcement is incorrect. Moreover the announcement itself was unnecessary. In February 1991 Wright commented on the “great news” of a student who could consider problems of resolution and storage of information and that building a scanner sounded like fun. So Wright knew and agreed that the radiographs would be scanned in Sydney and that one of Bromley's students would be doing it. There was no bombshell.
The correspondence shows that Bromley had done the same thing, much of it in correspondence and conversation with Wright.
Bromley had built two reconstructions, learned new skills, studied a new language and patiently coaxed details out of the stubborn fragments in collaboration with his colleague and friend, Michael Wright.
Two months later, in April 1994, Wright faxed Bromley a proof of a report he had given on linear tomography and asked for comments. So much for exhaustion and defeat.
In December 1995 Wright's alleged sense of betrayal did not prevent him from writing that “you are most welcome to stay with us”. Bromley stayed home, but responded with a chatty letter, written on Christmas Eve, in which he described a talk he had recently given about the mechanism. “I felt the speaker glossed over some details – like what the heck it was and what the hell it did.”
Page 187: The years following Wright's last trip to Athens with Bromley...
and
Page 188: Meanwhile any correspondence from Bromley had dried up.
These excerpts are from the introductory pages to Chapter 7. The chronology of this section is fuzzy, but the implication is that it's about events in Wright's life that occurred between February 1994 and the receipt of a letter from Bromley's wife in 1999 or 2000. The section also strongly implies that there was no meaningful communication between Bromley and Wright during this period. Both implications are false.
Wright and Bromley talked over a number of personal matters, including the breakup of Wright's first marriage, the beginnings of his relationship with his second wife, and his difficulties at work, in letters dating from the late 80s to the end of the paper correspondence in 1996.
The paper record contains further letters or email printouts from 1997 or 1998. From the content the record is incomplete, but it is sufficient to show that there was no breakdown in the relationship and the break in the correspondence does not appear to have caused undue distress to either party.
He knew it at the end of 1995. In the Christmas Eve 1995 letter, in addition to the remark quoted previously, Bromley wrote: “I think we are missing the wood for the trees and the mass of information in the X-Rays has become an excuse for me to not think about [the mechanism]. I think instead about how to digitise and image process {yuck!} the X-Rays. Typical busy work. But where to instead?”
“Everyone around him” in the 90s knew that Bromley often felt inexplicably tired and frequently unwell. He saw doctors, and was told that he was suffering after effects of Glandular Fever (Epstein-Barr virus). This seemed a sufficient explanation.
Anne Bromley does not recall writing any such letter.
The wording seems to support Marchant's claim that Bromley had been trying to hide his illness. In fact the diagnosis of Hodgkin's lymphoma was not made until 1998 and Bromley immediately told anyone he thought might be even faintly interested.
It's understandable that Wright was exhausted and nervous when he reached Sydney, but he didn't just turn up on the doorstep. Anne Bromley met him at the airport and brought him back to her home.
There was no denial about his illness and Bromley did not attempt to escape the probabilities of its outcome.
This was a private conversation and only Wright can now comment on the accuracy of the report. But the alleged dream is out of character. Bromley was not much interested in personal fame; if he had been he would never have gone into such an obscure field of study (the history of mechanical computing) and once there he would have published much more than he did. He was interested in understanding things. Frank Percival comments that with the Antikythera mechanism Bromley was “like a kid.”
This is grotesque, as dozens – perhaps hundreds – of students, colleagues, and friends could have attested. It's particularly disturbing that this section is not written as Michael Wright's view; it is written in the voice of the author, with all the weight of an apparent authority. Yet Marchant has stated that she made no attempt to test this description through Sydney University, where Bromley was based for his entire professional life, or with anyone in Australia.
The prominence of this baseless description is indicated by the presence of an index entry, presumably the work of an independent indexer, for “Bromley, Allan G.; character of”. It should read “character assassination of”.
Bromley freely gave from his knowledge, some considerably more than was expected. More importantly he consistently and generously supported people in pursuing the knowledge they wanted to find. This included many hours of unpaid work editing grant applications and coaching younger colleagues, as he did Michael Wright, through the processes of research proposals and publications. Outside the academic world he was a popular speaker to organisations such as the Sydney Clockmakers Society and the Australian Computer Museum, and at a personal level provided encouragement and practical advice to many people who were unsure of their own abilities. Encouragement often meant encouraging people to talk about their enthusiasms, including those whose enthusiasms were uncommon and who were very sensitive to signs of boredom. The arrogant knowledge hog Marchant describes could never had been the generous listener that Bromley was.
"That," he said finally, "is for me to know and you to find out."
According to Bromley's sister, this was one of their mother's favourite expressions. It's at least possible that Bromley was joking, though it would be easier to judge that if we knew the question and the inquisitor.
"You've done this great body of work," she coaxed. "Let Michael go and make sense of it all, so that your efforts can be recognised."
Anne Bromley does not recall saying this or anything like it. Although this is presented as a direct quotation, the author never attempted to verify it with the alleged speaker.
Bromley died on 16 August 2002.
This obituary appeared in the Horological Journal (November 2002). The correct date of Bromley's death is prominent in its first paragraph.
The correspondence between Wright and Bromley raises many questions about Marchant's account of the history of their collaboration. However it confirms this sentence in Wright's obituary, which comes immediately after the one Marchant quotes: His other great gifts to me were to teach me to order my thoughts and to trust my own judgement. Those are not the gifts of someone who pursued knowledge for the sake of power. They are gifts that build intellectual independence and the capacity to express original ideas.
It's unfortunate that the “perfectly glowing accounts that appeared elsewhere” apparently raised no questions in Marchant's mind.
Connectives Home The Connectives The Books A Booklover's Web About Connectives