by Greg Egan
Kem was, in effect, mapping out dozens of possible futures for the Splinter, attempting to distinguish the safe paths from the hazardous. Shuffling the numbers in a single case was a straightforward process, but it was impossible to prepare for every combination of circumstances that might arise along their journey. Roi was almost certain that they had found the true geometry, but they still could not predict how the strength of the wind would change as they moved out from the Hub; its speed was dictated by the curvature of space-time, but its density was not. Anticipating the Wanderer's behavior was even more difficult. Though it followed a comprehensible orbit for short stretches of time, its motion was subject to unpredictable changes, and only some of these could be clearly linked to its visible eruptions.
Nobody could understand the Wanderer's nature. It appeared to be a ball of wind and light, but what could hold such a thing together? Nothing in the long run, apparently, since the wind and light were spilling out ever more violently the closer it came to the Hub. Whether the old stories of the Splinter's origins were true or not, the Wanderer was fragmenting in a very different way; instead of being torn brutally in half in one cataclysmic moment, it was forever being stripped of small portions. If the Splinter's mythical parent really had divided, that would have eased its plight for a very long time, halving the greatest of the weights until some ancient Jolt, or some generation-spanning drift toward the Hub had eventually increased them. The Wanderer's losses seemed to make no difference, as if each small excision only gave it a chance to offer up something more, like a fast-growing crop eager to be pruned.
Haf and the others kept bringing food. Sometimes Roi caught herself trying to guess who among them were her children, but even when that urge passed she was surprised at the strength of her feelings toward them all. Her sense of duty had always been directed toward her team-mates; of course she had never been indifferent toward hatchlings, and would have aided any child she found in need, but the idea that the well-being of the next generation was as important as the completion of her next shift had always been a remote one, with little emotional force and even less need to be acted upon. Eggs hatched themselves, and the hatchlings found teachers; this didn't require any attention from her. The clearest lesson on the matter she had received from her own teachers had concerned the need to practice contraception with sufficient diligence to avoid playing her part in bringing on a famine.
Now the sight of Haf, Pel, and Tio brought a warmth to her mind that was as strong as the buzz of cooperation. The hope she felt at the prospect of navigating the Splinter to safety was still, in part, the same kind of longing for a successful shift that she had known all her life, but that familiar emotion was increasingly overlaid with a compelling sense of what it would mean for the ultimate beneficiaries. The thought of her own premature death, of Gul's, of Ruz's, of all her team-mates', was dismaying, and more than enough to drive her, but the extraordinary idea that they could carry the hatchlings into a transformed world where this danger would finally lie completely behind them was imbued with both more urgency, and more prospective joy, than anything else she had ever contemplated.
Leh, who watched for light-messages from the junub edge, came to Roi with a written transcript. Ruz's team had measured a small increase in the Splinter's orbital period. It was tiny, but it stood out above the usual variations due to uncertainty in their observations and the imperfections of their clocks.
Roi waited three more shifts for the next report to arrive, before letting herself believe it. The second set of timings confirmed the earlier result: the Splinter was moving, drifting outward very slowly.
She sent the news on to Neth and Bard, then asked Kem to tell the other theorists. In no time at all there was a riot of delighted chirps coming to her from the surrounding tunnels.
When Kem returned, Tan was with her.
"It's good news," he said, "but I'm worried by how slowly we're moving. It doesn't give us much flexibility if we find ourselves in a dangerous situation."
Roi concurred. "Bard and Neth understand that. They'll make the new tunnels their priority now."
Kem said, "If the Wanderer continues to behave as it has been, I believe that with three tunnels we'd have enough control over our ascent to pass through the Wanderer's orbit on the opposite side of the Hub. The problem will be if that orbit shrinks rapidly without warning."
"There's something else we might need to consider," Tan said. "One of my recruits, Nis, came to me two shifts ago with a new idea about the Wanderer. I don't know how seriously we should take it, but he's working on the details, trying to make it more precise."
"What's the general idea?" Roi asked.
"The strength of rock is what holds the Splinter together," Tan said. "But the Wanderer doesn't seem to be made of rock. So why doesn't it simply fall apart from the usual weights? Spinning in different ways from the Splinter won't do the trick. Without strength of some kind, it ought to be smeared all along its orbit by now."
"So what kind of strength can it have?" Kem asked.
"Geometry," Tan replied. "The same thing that keeps us close to the Hub holds the Wanderer together. But now the Hub and the Wanderer's geometry are fighting each other for the Wanderer's wind and light."
Roi allowed herself a moment of pure exhilaration. It was a beautiful, audacious proposition. Who ever said the Hub was the only object in the void that could be wrapped in curvature? It might be the strongest thing in sight, powerful enough to twist the Wanderer's path around it, but that didn't mean the Wanderer itself was a mere passive follower of geometry, as the Splinter seemed to be. Why couldn't it carry its own curvature? The Hub had the glory of the entire Incandescence, but the Wanderer, at least for now, held on to its own portion of wind and light. The two were alike, one smaller, one greater.
Unfortunately, if this elegant solution to the mystery was true, it changed everything. If the Wanderer was wrapped in space-time shaped as it was around the Hub, then in turn the geometry around the Hub itself could not be described perfectly by the rotationally symmetrical solution that she and Tan had found. The presence of the Wanderer was like a dent hammered into a smoothly curved sheet of metal; it could be ignored from afar, but the closer you came to it, the more significant it would be.
Kem appeared dazed. "Two Hubs fighting? We need the geometry for two Hubs fighting?"
Tan said, "It's just one possibility. It needs to be thought through in a lot more detail."
"But how will we know?" Kem demanded. "When will we know?" Her careful preparations had always been in jeopardy from the Wanderer's erratic behavior, but this new unknown threatened to render half the calculations meaningless.
Roi said, "We need to observe something in motion close to the Wanderer. That's the only way to understand its geometry." If they waited for it to have a measurable effect on the Splinter's own motion, all their questions would be answered too late.
"What is there close it?" Kem asked forlornly. "Just its wind and light erupting chaotically every now and then. How can we make sense of that?"
Roi fought against a sense of panic. It seemed almost fortunate now that their progress outward was less rapid than they'd hoped; if it had been faster, she might have had to call a halt until they were certain this problem could be solved.
"What moves close to the Wanderer?" Tan mused. "Nothing stays, but many things pass by."
Roi said, "This is no time for riddles."
Tan chirped amusement. "After her victory over me, I would have thought Kem might find the solution first."
"Victory?" Roi had lost track of all the minor squabbles and disputes among the team; once things became clear, she had trouble remembering who had first thought of each idea.
"Light travels at Neth's speed," Tan replied, "or as close to it as makes no difference. As fast as that is, though, its natural paths still respond to geometry."
Roi thought she knew what he was hinting at now, but it was Kem who put it into words.
&n
bsp; "The Wanderer moves against the background of lights," Kem said. "We need to observe and record the positions of a small, closely spaced group of lights very carefully. We do this once when the Wanderer is among them, then once again when it's far away. If there's a change in the angles between them, it will tell us that the geometry has changed. It will tell us that the Wanderer is like the Hub, wrapped in curvature of its own."
As the Splinter spiraled slowly outward, Roi traveled with Kem and Nis toward the junub edge. This was too complex a matter to explain to Ruz through messages, whether sent in writing or by flashes of light. There was nothing to keep her at the control post; they would not be dodging and weaving around the Wanderer for many shifts yet.
The light-messengers remained at their posts, though, ready to get word swiftly to Tan and the others if anything suddenly changed. As Roi greeted each one along the way, it struck her that the whole Splinter was almost like a single work team, now. In a sense that had always been true: couriers had worked with depot organizers, susk herders had worked with cuticle-wrights, and even those teams that had not directly cooperated had shared a common goal: the welfare of the Splinter as a whole. There was no denying, though, that since the Jolt the old borders had melted, and the old system had been twisted into a new shape, far richer and more contorted than even the Hub's strange geometry.
When the delegation of theorists reached Ruz, he called his best recruits together and they listened to Kem's proposal.
"There's one problem," Ruz said. "We've already made the kind of measurements you've described, and to the limits of our ability we've found no changes."
Kem said, "Can you check the data? If you weren't expecting this effect, you might have dismissed it as just part of a background of observational error."
"Perhaps." Ruz sent someone to carry out that task.
Cho said, "There's a way we might be able to improve the accuracy of our measurements. It involves curving the metal plates, bending and polishing them to systematically distort what we see."
"Distort?" Roi was skeptical. "Isn't that just going to make the errors greater?"
"If we didn't know the shape of the plate well enough, then of course it would," Cho conceded. "But if we can calibrate the shape with enough confidence, we might be able to exploit a system of such plates to magnify the angles we're able to discern."
Ruz seemed displeased; Roi wondered if Cho had proposed this system before, and Ruz had ruled it out as too complex and uncertain. She tried not to let her loyalty to Ruz sway her; she had to be ready to judge the ideas of every team member on their merits.
"Can you explain in more detail?" she said.
Cho believed he'd found a simple geometrical principle governing the paths that light took when it struck metal: the angle it made with a line perpendicular to the surface when it hit the metal was the same as the angle when it departed. For a flat plate, the consequences were straightforward, and they lay behind the contraption the void-watchers were now using to observe the lights from the safety of the tunnel below the crack.
For a curved surface, Cho's principle had more complex ramifications. He had templates to show that if the shape followed a certain curve, light from distant objects could be brought together sharply on a certain plane; using the opposite side of the same shape, the light could be made to appear as if it was coming from a plane behind the metal.
By combining elements of these two kinds, Cho believed he could construct a system that, while bringing the light down from the surface, modified the geometry of its paths in such a way as to make it seem as if the observer was closer to the distant lights. With one of his designs, all the angles would become larger by a factor of twelve.
Roi looked to Ruz, to hear his objections.
"The principles appear sound," he said. "But we can't be sure that we can shape the metal to the necessary precision. And how can we test it, when our one passage to the surface is taken up completely with the system we need, and that we know is reliable? How can we disrupt all our observations to gamble on this?"
"I understand."
Roi struggled to weigh up the risks. They desperately needed to understand the nature of the Wanderer. If Nis's idea was right then the geometry that wrapped it would influence the Splinter's path, and while as yet she had no idea how to combine the Hub and the Wanderer's geometry into a single shape, the sooner they learned exactly what the geometry close to the Wanderer was, the better their chances of making sense of that complex interaction before it was too late.
At the same time, the present system was certainly running smoothly for the void-watchers. If they disassembled Cho's first invention to make room for his second, they would have to resort to observers clambering back and forth from the surface every junub dark phase. They would lose observations, or even people, in the rush.
She could always ask Bard to close the tunnel, to buy them more time. The Wanderer had its own schedule, though. If they kept delaying their ascent, the Wanderer would come to the Splinter, she was sure of that. To pass it by in its present large orbit was one thing; to be wedged in by it, this close to the Hub, would almost certainly be fatal.
"How thoroughly have you explored the area around here?" she asked Ruz.
Ruz knew exactly what she was asking. "There are no other cracks," he said.
"When we first came here," Roi recalled, "the one we were looking for, the one from Zak's map, was closed. It was sheer luck that we found an open one. But how much work would it take to reopen the old one?"
Ruz's posture shifted slightly, growing defensive, as if she had accused him of neglecting his duty. "I don't have enough people to do that job, to open it up."
Roi said, "Stop thinking the old way, my friend. Everyone is your team-mate now. We don't have to lure them away from their colleagues and recruit them, one by one. We just have to explain the need, and the urgency. We just have to make sense."
23
"You can have what you asked for," Zey told Rakesh. "You can take a part of me to study."
She had just finished her shift and had come out from the depot. The other workers were still milling around, talking among themselves before heading into the tunnels to sleep.
Rakesh felt no need to ask if she was sure; everything they'd spoken about for the last dozen shifts had been for the sake of informing her decision. He did feel he owed her a small moment of drama, though, so rather than admitting that, thanks to their long proximity, his own avatar was already plastered in her cells and he had no need to collect a sample, he reached out with one claw and gently scratched a soft part of her nearest leg.
Nanomachines inside his avatar swarmed over the cells, dissecting some destructively, infiltrating others to watch their components in action. The DNA sequences were only part of the analysis; they would be meaningless without the full context of cellular biochemistry.
Parantham spoke to him, back in the cabin. "You might have done this when I first suggested it, instead of elevating your own need for customary formalities over the real ethical issues." Rakesh ignored her.
He took the nanomachines' data and ran coarse-grained simulations of morphogenesis, precise enough to give a clear picture of the way the Arkdwellers' bodies were shaped generically, and to map out the strongest genetic and environmental influences on each individual, but not so precise that the simulation itself would experience anything.
The generic map of the Arkdwellers' brain that the simulation produced made visible what Rakesh had long suspected: their ability to form and manipulate abstract symbols was powerful enough to grant all of them general intelligence as a birthright. Though the data came solely from Zey's DNA, there were far too many genes involved for her to have mutant variants of all of them; the generic map encompassed the whole species. As for every human born since the Stone Age, as for the ancestors of every member of the Amalgam, there was nothing the universe was capable of doing that the Arkdwellers were not capable of comprehending. They were not mere
clever-looking animals, with some hard-wired repertoire of impressive but inextensible skills. With sufficient motivation and freedom from distractions — and perhaps a modest boost in longevity — they could have grasped anything. Apart from the subjectivities of art and language, where everyone needed tweaking to cross the species barriers, there was nothing in the Amalgam's million-year-old storehouse of knowledge that would have been beyond their reach.
That was the ability, the potential in every one of them. There was, however, no drive to realize it: no curiosity, no joy in discovery, no restlessness, no dissatisfaction. The Arkdwellers needed their full intellectual toolkit in order to master the complex tasks allotted to them in the present social order, so it was not a useless vestigial genetic fossil. It was the living, breathing embodiment of matter's ability to comprehend itself, but it was tamed and caged in a manner that Rakesh had never encountered before. It could rise to the occasion to overcome a limited range of mundane setbacks and challenges, but it would never soar.
Rakesh was not surprised by any of this, inasmuch as it applied to the Arkdwellers as a whole. It fitted his observations of their behavior perfectly. He did not yet understand Zey, though. Her team-mates could not be too different from her genetically, but he'd expected her to carry two copies of some rare, recessive gene that could explain why she alone was compelled to make full use of her intellectual abilities. If that had been the case, though, the coarse-grained simulation would have had no way of knowing what the ordinary version of the gene was, and erasing Zey's atypical urges from the generic map.
This proved that whatever had made her different could not have been entirely genetically determined. The simulation had smoothed out the possible environmental influences on brain development into a plausible average, but in doing so it had clearly missed something that had made all the difference to Zey.
Rakesh probed the data more deeply, looking for genes that might have been triggered only rarely, rather than being rare themselves. He simulated the chemistry of the developing embryo in more detail, looking for possible surges in morphogens, and the wave of changes they might bring.
