Fiasco (35 page)

"How can I believe such a thing? You're saying that they wanted to use our help to commit suicide?"

"I say nothing—the facts speak. Their action, I admit, looks like madness. But a recreation of the cataclysm reveals its logic. We began the selenoclasm at the moment when the Sun was rising over Heparia and setting on Norstralia. The ballistic missiles directed at our sidereals were launched from the part of Heparia that was still behind the terminator—in other words, in the night. They required five hours to reach the perilune and strike our rockets. To keep us from destroying the missiles in time, the Quintans put them into an elliptical orbit, an orbit from which they could drop to the Moon some twelve minutes before the selenoclasm. There is no way around it: their missiles ambushed ours, moving along the ellipse segment farthest from Quinta and closest to the Moon. All attacked our sidereals, which were unshielded because we had not believed such a counteraction possible. I myself thought at first that the catastrophe had been caused by their miscalculation. But an analysis of the sequence of events rules out the chance of error."

"No. I cannot understand it," said Arago. "Although … one moment … does this mean that one side attempted to direct the blow at the other?"

"Even that would not have been so bad," said Steergard. "As far as a general headquarters is concerned during a war, any maneuver that discomfits the enemy is worthwhile and appropriate. But because they could not know the power of our sidereals or the time the selenoclasm would begin, or the initial velocity of the splitting masses of the Moon, they had to take into account that the dispersion pattern of the rocks might include their own territory as well. The Reverend Father is surprised? He doesn't believe me? The
physica de motibus coelestis
is the star witness in this case. Look at the situation from the point of view of the generals of a hundred-year war.

"A cosmic intruder appears above the battlefield with an olive branch. It wishes to establish friendly relations with the civilization; instead of responding attack for attack, it shows restraint, it remains peaceable. It will not attack? Then it must be made to! Does the population of the planet learn what really happened? Massacred, how can it doubt what its governments tell it: that the intruder is a ruthless, infinitely cruel aggressor? Did not the intruder level cities? And bomb
all
the continents, shattering the Moon
for that purpose?
Their own casualties? Blamed on the intruder. If we share the guilt, it is from undue innocence, because we did not foresee such a turn of events. Retreat, after what has happened, would leave the planet in the belief that our expedition was an attempt at murderous invasion. Therefore we do not withdraw, Reverend Father. The stake of the game was high to begin with. They have raised it, forcing us to play on…"

"Contact at any price?" asked the Dominican in white.

"At the highest that we can pay. Since I have ruffled our apostolic delegate with the announcement that the time for democracy—voting, shilly-shallying—is past, I think I should explain why, as I assume sole command and therefore sole responsibility for ourselves and for them, I will be taking this game to its conclusion. Shall I explain?"

"Please do."

Steergard went over to one of the wall lockers in his cabin, opened it, and said, as he looked for something in the pigeonholes:

"The thought of a nonlocal war expanded into space occurred to me after the catching of the wrecks behind Juno. And not to me alone. On the principle of
primum non nocere,
I kept it to myself, so as not to infect the crew with defeatism. It's known from the history of ancient voyages—Columbus's, the polar expeditions—how easily an isolated group of the best people can become a threat to themselves through the influence of an individual, particularly if that individual is one on whom they count, as if he were made of even better stuff than they. Therefore, I discussed this worst possibility only with DEUS. Here are the recordings of those discussions."

From a small, padded container that resembled a jeweler's box for precious stones he took a few memory crystals and inserted one in the slot of the reproducer.

His voice filled the room.

"How are we to establish contact with Quinta if there are blocs there locked in battle for years?"

"Provide the limits of the
n
-decisional space. It is strategically incalculable without starting parameters."

"Assume two, then three antagonists of approximately equal military potential, with the certain destruction of all in the case of heated escalation."

"The data remain insufficient."

"Give a minimax evaluation in a non-numerical approximation."

"The value in approximations is also indeterminate."

"Nevertheless, give me a stochastically weighted cluster of alternatives."

"That demands additional assumptions. They will be arbitrary and unsubstantiated."

"I know. Go ahead."

"For two antagonists on opposite continents, send two transmitters—in the atmospheric window of the infrared—with point-sharp collimation. Both should have antiradar camouflage and be self-directed at the planet's radio stations. This tactic takes for granted a thing that is open to question—because the antagonists may be not on opposite continents but in mutual possession of the same territory, both horizontally and vertically."

"In what way?"

"If, for example, they have entered the atomic phase with great saber-rattling, and each side takes aim at the enemy's population, making it hostage, threatening attack or reprisal. They fortify the means of thrust and parry, and, when saturation occurs, move underground. Their territories may be located far beneath the earth, like interlocking mines at many depths and levels. The same can happen above the atmosphere."

"Does an expansion of that type make contact impossible?"

"It rules out the tactic proposed, because with such an arrangement contact will not have separate addressees."

"Assume that there is no such subterranean colonization, with each side undermining the other."

"Where is the boundary to be drawn between the antagonists?"

"The meridian in the center of the ocean."

"That is simplest, but completely arbitrary."

"Go ahead."

"Very well. Assume the sending of probes, the signal emission—the delivery of the mail. And that they have received the codes transmitted and have mastered them. This assumption gives me a minimax fork. Send both sides the identical request for contact, either with a guarantee of neutrality which is genuine, or with a guarantee of exclusive support which is false."

"You mean, tell each side that we are addressing the other at the same time, or else assure it that we are approaching only it for contact?"

"Yes."

"Give the risk weighting of the branches."

"Honesty yields better chances if the message goes to the wrong address, and poorer chances if the message goes to the wrong address. Falsehood yields more chances if the address is right, and fewer if the address is right."

"That's a contradiction."

"Yes. The game-space is not quantifiable minimaximally."

"Show the reason for the contradiction."

"A bloc, assured of the exclusivity of contact with us, will be inclined to react positively—on condition that it can itself verify that exclusivity, independently of our communication. If, on the other hand, it learns that the other bloc has intercepted our message, or that—worse—we are playing a two-faced game, the chances of contact will fall to zero. One can also have a negative probability of contact."

"Negative?"

"A refusal is zero. I would assign a negative value to an answer that misinforms us."

"The setting of a trap?"

"Entirely possible. Here the forks branch factorially. A trap can be set by one side, or by both separately, or by both in a limited, temporary alliance—reasoning that if they call a temporary truce and cooperate to destroy us, or discourage us from contact, they will be running less risk than if they compete for exclusivity of contact with the
Hermes
."

"And what about their agreeing to a parallel, separate contact?"

"In that variant lies a fundamental contradiction. In order to achieve such parallelism, you must as sender guarantee to both sides our neutrality—convincingly. That is, you must give your word that you will keep your word. But an assertion, when reflexive, cannot assert itself. This is a typical antinomy."

"Where did you obtain the weightings for the decisional branches?"

"From your premise that on the planet there are only two players in mutual check. And that they hold to the rule of minimax. The prize of the game for them is the preservation of the
status quo ante fuit
, and for us, contact by breaking the impasse."

"Specifically?"

"It's trivial. I assume two empires, A and B. The optimal variant fork for us: both A and B enter into contact with us, each believing that it holds a monopoly. If either one is not sure of its privileged state—its exclusivity—it will suspect the monopoly. Whereupon, according to the rule of minimax, it will propose to the other a coalition against us, because it does not know the chances of entering into a coalition
with
us. That is obvious. Knowing their own history, they therefore know their rules of mutual conflict. But the rules of mutual conflict that pertain to us are unknown to them. If we make an offer of alliance to either A or B, it will be suspect.
Primo:
such an offer made by us to both adversaries is absurd.
Secundo:
if we choose one side only, we will be supporting it and will thereby antagonize the opposite side, gaining nothing for ourselves but participation in the ongoing struggle. Such a strategy of contact could be adopted only by a civilization of idiots. It is improbable even on the metagalactic scale."

"Yes. They can temporarily unite against us. What sort of game then results?"

"A game with indeterminate rules. The rules arise or change according to the course of the play. Therefore it is not known if the reward function will contain positive values. The game, probably, is zero-sum, because none of the players—ourselves included—stands to gain. All will suffer loss."

"The risk can't be brought to zero? Where is the minimum?"

"I have insufficient data."

"Go ahead without the data."

"The relieving of frustration in the face of insoluble problems does not lie within the domain of my computing ability. Do not ask the impossible, Captain. The branching tree of heuristics is not God's Tree of Knowledge."

In the silence after these words from DEUS, Steergard put a second crystal in the reproducer, explaining that this was from a dialogue with DEUS immediately following the selenoclasm. Again they heard the voice of the machine.

"Previously the risk was only incalculable. Now it has reached the power of a transfinite set; it is innumerable. Minimax holds, but now only for retreat."

"Could they be made to capitulate?"

"Theoretically, yes. For example, by the progressive elimination of their military technosphere."

"By destroying every instrument of war in the space around Zeta?"

"Yes."

"What are the chances of contact, with such an operation?"

"Minimal, making the most optimistic assumptions: that our deployment of the sidereals will be carried out without a hitch; that the Quintans will remain passive observers as their autoarmament sphere is peeled away, layer by layer; and that, stripped of these layers, they will fall into an armament stagnation. In the category of game theory, this would be a miracle—on the order of winning first prize in a lottery without ever having purchased a ticket."

"Present the variants of disarming their technosphere without miracles."

"The curve will have at least two saddles. Either they oppose us, offensively or defensively, or the pacification-destruction of the nonliving war zone heats up the conflict that is still smoldering on the planet, and thereby we push them into total war."

"Is it possible to destroy their cosmic war zone partially without disturbing the equilibrium of forces on the planet?"

"It is possible. To that end, one would have to destroy each orbital weapon after first learning the side to which it belonged. This reduces the military-cosmic potential of all the adversaries in the same degree, so as to preserve the dynamic balance between their forces. Two things are assumed: 1) that we will know the range in which they control their weapons in space—that is, the radius of their command effectiveness—and 2) that we will identify the combat systems beyond that radius and destroy them, and that, following the destruction of this autoarmament periphery, we will then be able to strip the civilization of the forces under its control inside the sphere.
In abstracto
it should be possible to strip it, so to speak, to the skin. But if we commit errors in identification—as to who controls what in the inner sphere—we will heat up the conflict on the planet, because we will be weakening one side to the advantage of another. And thus we push the antagonists from the precarious equilibrium of their arms race into total war. Captain, you remove me and yourself from reality. You wish success?"

"Of course."

"What is it to be, this success of yours? Contact? But in the above model the concept of success is indeterminate. It depends not merely on whether or not the
Hermes
will be able to overcome, besides the war sphere, the entire production of military devices that may be hurled continually into space. We will be waging an indirect war, attacking not the Quintans but their weapons. How can we be sure that, bringing new technologies into the battle, they will not master the resources that sustain us—the sidereal?"

"Assume that they will not."

"Very well. Besides the external factors—the ambiguous technological sets, the minimax decisions and calculations consistent with optimizing logic—the reactions of the Quintans will be determined, also, by irrational factors of which we are completely ignorant. We do know, however, what importance precisely these factors have had in terrestrial History."