Some Remarks: Essays and Other Writing

by Neal Stephenson

  A few engineers pad through the place. The setup is instantly recognizable; you can see the same thing anywhere nerds are performing the kinds of technical hacks that keep modern governments alive. The Manhattan Project, Bletchley Park, the National Security Agency, and, I would guess, Saddam Hussein’s weapons labs are all built on the same plan: a big space ringed by anxious, ignorant, heavily armed men, looking outward. Inside that perimeter, a surprisingly small number of hackers wander around through untidy offices making the world run.

  If you turn your back on the equipment through which the world’s bits are swirling, open one of the windows, wind up, and throw a stone pretty hard, you can just about bonk that used book peddler on the head. Because this place, soon to be the most important data nexus on the planet, happens to be constructed virtually on top of the ruins of the Great Library of Alexandria.

  THE LALLA ROOKH

  When William Thomson became Lord Kelvin and entered the second phase of his life—tooling around on his yacht, the Lalla Rookh—he appeared to lose interest in telegraphy and got sidetracked into topics that, on first reading, seem unrelated to his earlier interests—disappointingly mundane. One of these was depth sounding, and the other was the nautical compass.

  At the time, depths were sounded by heaving a lead-weighted rope over the side of the ship and letting it pay out until it hit bottom. So far, so easy, but hauling thousands of meters of soggy rope, plus a lead weight, back onto the ship required the efforts of several sailors and took a long time. The U.S. Navy ameliorated the problem by rigging it so that the weight could be detached and simply discarded on the bottom, but this only replaced one problem with another one in that a separate weight had to be carried for each sounding. Either way, the job was a mess and could be done only rarely. This probably explains why ships were constantly running aground in those days, leading to a relentless, ongoing massacre of crew and passengers compared to which today’s problem of bombs and airliners is like a Sunday stroll through Disney World.

  In keeping with his general practice of using subtlety where moronic brute force had failed, Kelvin replaced the soggy rope with a piano wire, which in turn enabled him to replace the heavy weight with a much smaller one. This idea might seem obvious to us now, but it was apparently quite the brainstorm. The tension in the wire was so light that a single sailor could reel it in by turning a spoked wooden wheel.

  The first time Kelvin tried this, the wheel began to groan after a while and finally imploded. Dental hygienists, or people who floss the way they do (using extravagantly long pieces of floss and wrapping the used part around a fingertip), will already know why. The first turn of floss exerts only light pressure on the finger, but the second turn doubles it, and so on, until, as you are coming to the end of the process, your fingertip has turned a gangrenous purple. In the same way, the tension on Kelvin’s piano wire, though small enough to be managed by one man, became enormous after a few hundred turns. No reasonable wheel could endure such stress.

  Chagrined and embarrassed, Kelvin invented a stress-relief mechanism. On one side of it the wire was tight, on the other side it was slack and could be taken up by the wheel without compressing the hub. Once this was out of the way, the challenge became how to translate the length of piano wire that had been paid out into an accurate depth reading. One could never assume that the wire ran straight down to the bottom. Usually the vessel was moving, so the lead weight would trail behind it. Furthermore, a line stretched between two points in this way forms a curve known to mathematicians as a catenary, and of course the curve is longer than a straight line between the same two points. Kelvin had to figure out what sorts of catenary curves his piano wire would assume under various conditions of vessel speed and ocean depth—an essentially tedious problem that seems well beneath the abilities of the father of thermodynamics.

  In any case, he figured it out and patented everything. Once again he made a ton of money. At the same time, he revolutionized the field of bathymetry and probably saved a large number of lives by making it easier for mariners to take frequent depth soundings. At the same time, he invented a vastly improved form of ship’s compass which was as big an improvement over the older models as his depth-sounding equipment was over the soggy rope. Attentive readers will not be surprised to learn that he patented this device and made a ton of money from it.

  Kelvin had revolutionized the art of finding one’s way on the ocean, both in the vertical (depth) dimension and in the horizontal (compass) dimensions. He had made several fortunes in the process and spent a great deal of his intellectual gifts on pursuits that, I thought at first, could hardly have been less relevant to his earlier work on undersea cables. But that was my problem, not his. I didn’t figure out what he was up to until very close to the ragged end of my hacker tourism binge.

  SLACK

  The first time a cable-savvy person uses the word slack in your presence, you’ll be tempted to assume he is using it in the loose, figurative way—as a layperson uses it. After the eightieth or ninetieth time, and after the cable guy has spent a while talking about the seemingly paradoxical notion of slack control and extolling the sophistication of his ship’s slack control systems and his computer’s slack numerical-simulation software, you begin to understand that slack plays as pivotal a role in a cable lay as, say, thrust does in a moon mission.

  He who masters slack in all of its fiendish complexity stands astride the cable world like a colossus; he who is clueless about slack either snaps his cable in the middle of the ocean or piles it in a snarl on the ocean floor—which is precisely what early 19th-century cable layers spent most of their time doing.

  The basic problem of slack is akin to a famous question underlying the mathematical field of fractals: How long is the coastline of Great Britain? If I take a wall map of the isle and measure it with a ruler and multiply by the map’s scale, I’ll get one figure. If I do the same thing using a set of large-scale ordnance survey maps, I’ll get a much higher figure because those maps will show zigs and zags in the coastline that are polished to straight lines on the wall map. But if I went all the way around the coast with a tape measure, I’d pick up even smaller variations and get an even larger number. If I did it with calipers, the number would be larger still. This process can be repeated more or less indefinitely, and so it is impossible to answer the original question straightforwardly. The length of the coastline of Great Britain must be defined in terms of fractal geometry.

  A cross-section of the seafloor has the same property. The route between the landing station at Songkhla, Thailand, and the one at Lan Tao Island, Hong Kong, might have a certain length when measured on a map, say 2,500 kilometers. But if you attach a 2,500-kilometer cable to Songkhla and, wearing a diving suit, begin manually unrolling it across the seafloor, you will run out of cable before you reach the public beach at Tong Fuk. The reason is that the cable follows the bumpy topography of the seafloor, which ends up being a longer distance than it would be if the seafloor were mirror-flat.

  Over long (intercontinental) distances, the difference averages out to about 1 percent, so you might need a 2,525-kilometer cable to go from Songkhla to Lan Tao. The extra 1 percent is slack, in the sense that if you grabbed the ends and pulled the cable infinitely tight (bar tight, as they say in the business), it would theoretically straighten out and you would have an extra 25 kilometers. This slack is ideally molded into the contour of the seafloor as tightly as a shadow, running straight and true along the surveyed course. As little slack as possible is employed, partly because cable costs a lot of money (for the FLAG cable, $16,000 to $28,000 per kilometer, depending on the amount of armoring) and partly because loose coils are just asking for trouble from trawlers and other hazards. In fact, there is so little slack (in the layperson’s sense of the word) in a well-laid cable that it cannot be grappled and hauled to the surface without snapping it.

  This raises two questions, one simple and one nauseatingly difficult and complex. First, h
ow does one repair a cable if it’s too tight to haul up?

  The answer is that it must first be pulled slightly off the seafloor by a detrenching grapnel, which is a device, meant to be towed behind a ship, that rolls across the bottom of the ocean on two fat tractor tires. Centered between those tires is a stout, wicked-looking, C-shaped hook, curving forward at the bottom like a stinger. It carves its way through the muck and eventually gets under the cable and lifts it up and holds it steady just above the seafloor. At this point its tow rope is released and buoyed off.

  The ship now deploys another towed device called a cutter, which, seen from above, is shaped like a manta ray. On the top and bottom surfaces it carries V-shaped blades. As the ship makes another pass over the detrenching grapnel, one of these blades catches the cable and severs it.

  It is now possible to get hold of the cut ends, using other grapnels. A cable repair ship carries many different kinds of grapnels and other hardware, and keeping track of them and their names (like “long prong Sam”) is sort of like taking a course in exotic marine zoology. One of the ends is hauled up on board ship, and a new length of cable is spliced onto it solely to provide excess slack. Only now can both ends of the cable be brought aboard the ship at the same time and the final splice made.

  But now the cable has way too much slack. It can’t just be dumped overboard, because it would form an untidy heap on the bottom, easily snagged. Worse, its precise location would not be known, which is suicide from a legal point of view. As long as a cable’s position is precisely known and marked on charts, avoiding it is the responsibility of every mariner who comes that way. If it’s out of place, any snags are the responsibility of the cable’s owners.

  So the loose loop of cable must be carefully lowered to the bottom on the end of a rope and arranged into a sideways bight that lies alongside the original route of the cable something like an oxbow lake beside a river channel. The geometry of this bight is carefully recorded with sidescan sonar so that the information can be forwarded to the people who update the world’s nautical charts.

  One problem: now you have a rope between your ship’s winch and the recently laid cable. It looks like an old-fashioned, hairy, organic jute rope, but it has a core of steel. It is a badass rope, extremely strong and heavy and expensive. You could cut it off and drop it, but this would waste money and leave a wild rope trailing across the seafloor, inviting more snags.

  So at this point you deploy your submersible remotely operated vehicle (ROV) on the end of an umbilical. It rolls across the seabed on its tank tracks, finds the rope, and cuts it with its terrifying hydraulic guillotine.

  Sad to say, that was the answer to the easy question. The hard one goes like this: You are the master of a cable ship just off Songkhla, and you have taken on 2,525 kilometers of cable which you are about to lay along the 2500-kilometer route between there and Tong Fuk Beach on Lan Tao Island. You have the 1 percent of slack required. But 1 percent is just an average figure for the whole route. In some places the seafloor is rugged and may need 5 percent slack; in others it is perfectly flat and the cable may be laid straight as a rod. Here’s the question: How do you ensure that the extra 25 kilometers ends up where it’s supposed to?

  Remember that you are on a ship moving up and down on the waves and that you will be stretching the cable out across a distance of several kilometers between the ship and the contact point on the ocean floor, sometimes through undersea currents. If you get it wrong, you’ll get suspensions in the cable, which will eventually develop into faults, or you’ll get loops, which will be snagged by trawlers. Worse yet, you might actually snap the cable. All of these, and many more entertaining things, happened during the colorful early years of the cable business.

  The answer has to do with slack control. And most of what is known about slack control is known by Cable & Wireless Marine. AT&T presumably knows about slack control too, but Cable & Wireless Marine has twice as many ships and dominates the deep-sea cable-laying industry. The Japanese can lay cable in shallow water and can repair it anywhere. But the reality is that when you want to slam a few thousand kilometers of state-of-the-art optical fiber across a major ocean, you call Cable & Wireless Marine, based in England. That is pretty much what FLAG did several years ago.

  IN WHICH THE HACKER TOURIST TREKS TO LAND’S END, THE HAUNT OF DRUIDS, PIRATES, AND TELEGRAPHERS. AN IDYLLIC HIKE TO THE TINY CORNISH TOWN OF PORTHCURNO. MORE FLAGON HOISTING AT THE CABLE STATION. LORD KELVIN’S HANDIWORK EXAMINED AND EXPLAINED. EARLY BITS. THE SURVEYORS OF THE OCEANS IN CHELMSFORD, AND HOW COMPUTERS PLAY AN ESSENTIAL PART IN THEIR WORK. ALEXANDER GRAHAM BELL, THE SECOND SUPREME NINJA HACKER MAGE LORD, AND HIS MISGUIDED ANALOG DETOUR. LEGACY OF KELVIN, BELL, AND FLAG TO THE WIRED WORLD.

  50˚ 3.965' N, 5˚ 42.745' W Land’s End, Cornwall, England

  As anyone can see from a map of England, Cornwall is a good jumping-off place for cables across the Atlantic, whether they are laid westward to the Americas or southward to Spain or the Azores. A cable from this corner of the island needs to traverse neither the English Channel nor the Irish Sea, both of which are shallow and fraught with shipping. Cornwall also possesses the other necessary prerequisite of a cable landing site in that it is an ancient haunt of pirates and smugglers and is littered with ceremonial ruins left behind by shadowy occult figures. The cable station here is called Porthcurno.

  Not knowing exactly where Porthcurno is (it is variously marked on maps, if marked at all), the hacker tourist can find it by starting at Land’s End, which is unambiguously located (go to England; walk west until the land ends). He can then walk counterclockwise around the coastline. The old fractal question of “How long is the coastline of Great Britain” thus becomes more than a purely abstract exercise. The answer is that in Cornwall it is much longer than it looks, because the fractal dimension of the place is high—Cornwall is bumpy. All of the English people I talked to before getting here told me that the place was rugged and wild and beautiful, but I snidely assumed that they meant “by the standards of England.” As it turns out, Cornwall is rugged and wild and beautiful even by the standards of, say, Northern California. In America we assume that any place where humans have lived for more than a generation has been pretty thoroughly screwed up, so it is startling to come to a place where 2,000-year-old ruins are all over the place and find that it is still virtually a wilderness.

  From Land’s End you can reach Porthcurno in two or three hours, depending on how much time you spend gawking at views, clambering up and down cliffs, exploring caves, and taking dips at small perfect beaches that can be found wedged into clefts in the rock.

  Cables almost never land in industrial zones, first because such areas are heavily traveled and frequently dredged, second because of pure geography. Industry likes rivers, which bring currents, which are bad for cables. Cities like flat land. But flat land above the tide line implies a correspondingly gentle slope below the water, meaning that the cable will pass for a greater distance through the treacherous shallows. Three to thirty meters is the range of depth where most of the ocean dynamics are and where cable must be armored. But in wild places like Porthcurno or Lan Tao Island, rivers are few and small, and the land bursts almost vertically from the sea. The same geography, of course, favors pirates and smugglers.

  On the other hand, what looks to a pirate like an accessible port of entry can be a remote refuge to a landlubber. Cornwall, like Wales, is one of the places where peculiar and unpopular Britishers have long gone to seek refuge—it was the last part of England to become English. And when Kublai Khan was storming China, the last Mongol emperor fled southward until he reached—you guessed it—Lan Tao Island, where he and his dynasty died.

  But all becomes clear when you clamber over yet another headland and discover Porthcurno, a perfect beach of pale sand sloping gently out of clear turquoise water and giving way to a cozy valley that, a few miles inland, rises to the level of the inland plateau. To the hacker tourist, it comes as no surprise
to learn that much of that valley has been owned by Cable & Wireless, or its predecessors, for more than a century. To anyone else, the only obvious hint that this place has anything to do with cables comes from the rusty yellow signs that stand above the beach proclaiming “Telephone Cable” as a feeble effort to dissuade mariners from using the bay for anchor practice.

  It was here that the long-range submarine cable business, after any number of early-round knockdowns, finally dragged its bloody self up off the mat and really began to kick ass.

  By the year 1870, Kelvin and others had finally worked the bugs out of the technology. A three-master anchored off this beach in that year and landed a cable that eventually ran to Lisbon, Gibraltar, Malta, Alexandria, Cairo, Suez, Aden (now part of Yemen), Bombay, over land to the east coast of India, then on to Penang, Malacca, Singapore, Batavia (later Jakarta), and finally to Darwin, Australia. It was Australia’s first direct link to Great Britain and, hardly by coincidence, also connected every British outpost of importance in between. It was the spinal cord of the Empire.

  The company that laid the first part of it was called the Falmouth, Gibraltar and Malta Telegraph Company, which is odd because the cable never went to Falmouth—a major port some 50 kilometers from Porthcurno. Enough anchors had hooked cables, even by that point, that “major port” and “submarine cable station” were seen to be incompatible, so the landing site was moved to Porthcurno.That was just the beginning: the company (later called the Eastern Cable Company, after all the segments between Porthcurno and Darwin merged) was every bit as conscious of the importance of redundancy as today’s Internet architects—probably more so, given the unreliability of early cables. They ran another cable from Porthcurno to the Azores and then to Ascension Island, where it forked: one side headed to South America while the other went to Cape Town and then across the Indian Ocean. Subsequent transatlantic cables terminated at Porthcurno as well.