This Little Britain Page 17
Such has been the mystique around Enigma that few people realize how limited its function was. It was hand not machine operated, which made it slow and labour intensive, capable of a maximum transmission of just five hundred characters. Although it was used extensively by Germany’s navy and air force, the army high command made little use of the device, preferring instead the non-Morse-based Geheimschreiber, or ‘secret writer’. The Geheimschreiber was completely automated, permitting messages of any length to be sent. Unlike Enigma, where British intelligence knew the physical make-up of the machine, the Geheimschreiber was a completely unknown quantity. The only route into it was the enciphered intercepts themselves.
For a long time, frustratingly little progress was made. Then, in August 1941, two signals were received both beginning with the same twelve-letter indicator, HQIBPEXEZMUG. This moment of slackness on the part of some German signaller provided the crucial breakthrough that the British had been needing: the opportunity to crack open the code for the first time. It wasn’t that the signals contained useful military intelligence—they didn’t—but some brilliant detective work by Bletchley’s finest finally laid open the the logical structure of the Geheimschreiber itself. In effect, they used the nature of the encoding to work out the physical design of the machine that had generated it. They had figured out what they were up against. The next task was to beat it.
Easier said than done. Bletchley now knew that their opponent was a far more complex beast than Enigma. Enigma, poor dear, had managed only a few thousand billion combinations. The Geheimschreiber used a few thousand million billion. Against this monster, the old Bombe technology was of no help. There continued to be a trickle of decrypts, which came about largely because of human error on the German side. As the Germans continued to tighten their communications security, Bletchley Park risked losing even the limited gains it had made so far.
A new approach was called for. That approach relied on the oldest trick in the cryptanalyst’s book, the statistical analysis of letter frequency. In English, E is the most common letter, followed by T, A, O, and so forth. In the particular language of Germany military transmissions, following decryption, the most common characters were 9, E, N, 5 and 8. (The numbers in this sequence all had specific meaning: 9 representing the space key, for example.) The trouble was that in the encrypted text these very marked variations were all evened out. There was random variation in the encrypted text; nothing more. The actual insight that led to beating the Geheimschreiber for good is so complex that it’s (thankfully!) well beyond the scope of this book to describe it.* In shorthand, though, Bletchley Park came to understand that the final random-looking stream of characters was created from the combination of two non-random streams. If a moment of enemy sloppiness allowed the British to get a handle on the pin settings that generated one of those streams, then a brute-force statistical analysis might well be able to do the rest.
But just how brutish would that force need to be? Even half cracked, the Geheimschreiber was reluctant to give up its secrets. A two-thousand-character message might need a total of 17.8 million operations to crack it. Each message would need to be separately decrypted. Plainly enough, the task called for mechanization—but mechanization how?
Two approaches were mooted. The first—the Robinson, later renamed the Heath Robinson—involved a hybrid electromechanical device involving punched paper tape. The second, an all-electronic device, was advocated by Tommy Flowers, a senior engineer at the Post Office. Flowers’ proposal, involving as it did a massive use of electronic valves, was thought unlikely to be reliable. The decision was therefore taken to go with the Heath Robinson, which was duly built. The Heath Robinson (looking rather like two bedsteads festooned with toilet roll) worked, but it was slow, temperamental and acutely vulnerable to human error. More was needed.
Fortunately for all concerned, Tommy Flowers was as stubborn as he was creative. He commented, ‘They [Bletchley Park] didn’t commission me [to build a computer, but] they said if you feel like it, that’s up to you. So we said, we’ll do it.’ And do it they did. The Post Office gave Flowers authority, manpower, cash and freedom.* Work started (at Dollis Hill of all places) in February 1943. The first prototype was ready in December of that year. The first machine was delivered in January 1944. By February, it was already largely responsible for cracking the enemy codes. This machine, Colossus, was the world’s first electronic computer. In its first version it could handle an astonishing 5,000 characters per second—and it was only constrained to this speed because the paper tape ripped apart if they ran it any faster. The second Colossus ramped its speed up to 25,000 characters per second. The paper still had the same tearing speed, but the computer got around that with parallel processing and wider paper.
This was an extraordinary degree of computing power. The German security apparatus never conceived of facing such an adversary. The Geheimschreiber, with its 160 thousand million billion combinations, was undone. What’s more, the Colossus and its successors were no mere one-trick ponies. From the start, Flowers had built logical flexibility into the design of the machine. If German code-setting procedures altered (as they did), the Colossus could simply adapt to deal with them. The machine proved perfectly reliable, just as Flowers had predicted. The D-Day landings and the subsequent war in Europe were conducted with an unprecedented understanding of the enemy’s intentions, an advantage of immense strategic value to the Allies.
The war over, the British were understandably reluctant to reveal their secrets. Colossus was dismantled. Those who knew about it were sworn to secrecy. As far as anyone knew, Colossus had never existed at all.* All the same, all technologies have their moment, and the post-war world was clearly ripe, scientifically and technically, for computing. In America, large and well-funded research projects began to grind systematically towards the rediscovery of the computer. Remarkably, however, their ill-funded, make-do-and-mend British counterparts were at least as well advanced, and probably more so. The first serious work on computer programming? British. The first stored program computer? British. The first genuinely flexible programmable computer? British. The first computers to be built and delivered by commercial manufacturers? British. The first office computer? British. The American historian Kenneth Flamm summarized this fertile period, saying simply, ‘The first modern electronic computers, by any definition, were built in Britain.’
All this comes as heartening confirmation of the resourcefulness and creativity of the British boffin. Yet it will not have escaped the reader’s notice that twenty-first-century Britain is not exactly well endowed with computer companies. The great hardware manufacturers—IBM, Compaq, Dell, HP, Apple—have all been American. The two principal processor manufacturers—Intel and AMD—are both American. The most important software companies—Microsoft, Lotus, Sun, Oracle—have all been American, as have the major providers of IT services—IBM, EDS and Accenture. The most important dot.commers—Google, Amazon, Yahoo!, eBay—are as British as pumpkin pie, cheerleaders and drive-by shootings.
The truth is that the great age of British invention ran for two centuries from 1750 to 1950. In the first half of that time, British invention was effervescent, rampant, world-beating. Increasingly thereafter it became stymied, not by any lack of brains or creativity, but by the inadequacy of the British companies that should have nourished it. It was probably inevitable that computing’s centre of gravity should shift to the United States. After the war, the American commercial lead was so immense that American buyers would form by far the most important market for all the fancy new equipment being built. Britain’s make-do-and-mend computers of the early electronic era were masterpieces of innovation, but they didn’t necessarily form the most obvious basis for lasting commercial success. But, to paraphrase Oscar Wilde, to lose leadership in one technological sphere may be regarded as a misfortune, to lose leadership almost everywhere looks like carelessness. There are still technological areas in which Britain has (or rather shares)
world leadership. We’re top notch in pharmaceuticals, and certain military and aerospace areas. We’re strong in some niche areas such as Formula One racing cars. But in any field that requires mass manufacturing, we’re nowhere. How could it be otherwise, when we have no mass manufacturers left? From 1850 onwards, invention and innovation had less and less to do with eccentric geniuses cobbling something together in a garden shed. Success in modern technology required organization, discipline and the industrialization of research. When invention left the potting shed and entered the research lab, British inventors were doomed to be left behind.*
An investment banking friend of mine once heard a senior executive from the German company Mannesmann talking about their recent acquisition of a British maker of forklift trucks—the last such independent manufacturer in Britain. The Mannesmann executive said that every single one of the major innovations to have been made in the forklift industry since the war had been made in Britain, but went on to comment that ‘follow-through is lousy’. That epitaph should be carved in stone, and placed over the gates of every failing, failed or extinct British manufacturer.
British inventors have often been among the best in the world. British manufacturing, alas, has been a forking disaster.
* It’s almost beyond the wit of man to do so. Paul Gannon’s Colossus is an admirably clear and readable history, but even so it’s forced into using sentences like: ‘The delta stream of an enciphered text stream is the same as that which would be produced by creating the delta of the plain, the delta of the Chi and the delta of the extended-Psi and then combining them.’ Got that?
* Not unlimited cash, though. Flowers often paid for essential parts out of his own pocket.
* One by-product of which is that Tommy Flowers has never obtained due recognition. Since he built the first computer and won the Second World War (with a bit of help from his friends), isn’t it time that changed?
* Of course, there’s more to creativity than just inventing things. While Britain’s commitment to R&D is notoriously poor, its creative industries (design, publishing, broadcasting etc) are almost twice as larg in relation to national income as those of its nearest competitor.
ECONOMY
WHOSE LAND?
If you were to sneak out on to the tarmac at Heathrow airport and look around at the vast agglomeration of stuff that lies all around—planes, terminal buildings, runways, baggage trucks, boarding ramps, catering units, engineering works and the rest—you wouldn’t be able to find a single object that wasn’t owned. Every blade of grass has an owner, every rivet, every gallon of fuel. What’s more, those assets exist in an invisible web of legal relationships. The airline leases a plane, pays for landing rights, sells tickets, and so on. If any of these rights—whether of ownership or contract—are breached, then the injured party can go to court and get compensation.
Because this system is all we’ve ever known, it seems as natural as the air we breathe, but of course it’s no more natural than the jet engine or the airline meal. I was reminded of this most starkly myself when (during my dark past as an investment banker) I worked on the privatization of CSA, the Czech national airline, not long after the fall of communist rule. To privatize a company, you have to know what you’re selling. In Prague, this was no simple matter. Take a really basic asset: the runway. Who owned it? Did it belong to the airline? Or the airport? Or the civil aviation authority? Or the transport ministry? Or central government itself? In Britain, these questions would have had a precise answer. In Prague, it made no difference whom you asked; the standard answer took the form of a shrug followed by that all-encompassing communist answer, ‘the state’. Getting the airline ready to be sold involved a massive allocation of assets and rights—in essence, inventing ownership from scratch. The final inventory of assets was a computer print-out so massive it took two hands to lift it.
That experience was a salutary reminder of the importance of property rights to any capitalist system. If you want to go into business, you have to be pretty damn sure that you’ve got a right to the assets you’re using. If not, you’ll work hard, take risks, accumulate profit—then stand to have the whole lot taken away from you. The first and most important building block of any capitalist system is that invisible network of legal rights, in particular the right to own.
It isn’t just ex-communist societies that have no firm idea of individual property rights. Take a classic peasant society—tsarist Russia, for example, or pre-communist China. One family would typically farm a piece of land from generation to generation. Those generations tended to live and work together, labourers on a common project. Such a system was almost completely insulated from the market. Labour came from family members, not hired hands. The bulk of all production was consumed at home. In such a system, money did feature, but not much. It wasn’t central.
Unsurprisingly, such systems didn’t think of ownership the way we do. The ‘head of the household’—a man, nearly always—would be the notional owner. But suppose that man was a drunkard? Or bone idle? Or terminally stupid? In Russia, such a man could, in effect, be fired and replaced by a more competent brother or son. The head of the household wasn’t really an owner at all; he was more like the custodian of an asset owned in some eternal and elemental way by the family itself. If someone wanted to sell the land or leave it in a will to someone outside the family, he couldn’t do it—or at least his ability to do so was very heavily restricted.
Pretty obviously, the system had its advantages. Drunken fathers couldn’t drink their children into landlessness. But it had disadvantages too. The system resisted acquisitiveness. It resisted innovation. It was anti-capitalist.
In eastern Europe, the system lasted into the nineteenth century. In western Europe, the system was less sharply defined and faded sooner, but it was present nonetheless. And England? Pretty obviously the country made a transition at some point from a subsistence farming culture to a capitalist one—but when and how?
Fifty years ago, the standard view went a bit like this. At the time of the Norman Conquest, England was a peasant society, much like tsarist Russia. Some time around the late Tudor/early Stuart period our peasants began to flirt with capitalist modes of production. A couple of hundred years later and—boom!—proper capitalist farming emerged and with it the Industrial Revolution, technology and, in due course, the jet engine and the airline meal. On this view, property rights developed much as you’d expect them to: very important later on; not nearly so much so earlier.
Then historians began to chip away at the evidence, and the pesky thing about evidence is that it has a nasty habit of overturning even the neatest of ideas. Luckily for researchers, England is exceptional in the wealth of its archive sources. Perhaps only Japan has archives as rich and continuous as our own,* and those archives began to reveal some curious things. One historian looked at the pattern of land transfer in Leighton Buzzard from 1464 to 1508. On the traditional view, you’d have expected a sluggish land market, with not many transfers, nearly all of which would have been within the family. That traditional view was simply wrong. Not only was the land market exceptionally active—over nine hundred transfers within a forty-four-year period—but three-quarters of land transfers did not involve family members. Just 10 per cent of all transfers involved an owner passing goods on to his family when he died.
Further studies were done, the historical telescope focusing on the ever more distant past. The same thing emerged: an unending flow of property transactions, a large core of which had nothing to do with family. Further surprises surfaced. Classic peasant societies are structured round the extended family, yet this wasn’t the case in England as far back as the late Middle Ages, and quite possibly earlier still. Not only were English families largely nuclear, but sons and daughters went out to work from a young age. The surprise here isn’t that kids were put to work (it would have been a miracle if they hadn’t), the surprise is that they worked in the labour market for cash, rather than simply cont
ributing their labour to the family farm. Kids weren’t simply kids, they were economic agents. Perhaps more strikingly, even unmarried women—virtually invisible in most traditional systems—were perfectly capable of buying, selling and bequeathing assets. Courts registered transactions. Money changed hands. As far back as 1200, England seems to have been a small but perfectly formed capitalist society. (Perfectly formed, that is, with the not-so-small proviso that feudal obligations severely restricted the freedom of many labourers to seek work beyond the manor. Those capitalist energies would really start to get full expression only with the upheaval in the labour market that followed the Black Death.)
Although the historical record simply falls silent before 1200 or so, there’s no reason to think that this date marked any kind of turning point—indeed, rather the reverse. The Anglo-Saxon system arrived in England from Germany. As the French writer Montesquieu commented, ‘In perusing the admirable treatise of Tacitus [a Roman historian] On the Manners of the Germans we find it is from that nation that the English have borrowed their idea of political government’—and, he went on to add, their laws of property and inheritance—‘This beautiful system was invented first in the woods.’ As far as we know, Montesquieu had it pretty much spot on.
There’s lots one could say about all this, but a few points stand out.
First, it’s easy for us to think of the British as really becoming exceptional only during the rapid thrust towards modernity—say in the hundred years before the Industrial Revolution and the hundred years after. That view is simply wrong. There’s no doubt that the most obvious fruits of our exceptionalism emerged then, but the soil had been prepared long, long before. Why did Britain become the first industrial society in the world? In part, because it had long been the world’s truest capitalist society: the one with the most developed labour and product markets; the one with the most developed system of property rights.