The Failures of Eld Science

This time there were no robes, no hoods, no masks. Stu­dents were ex­pected to be­come friends, and al­lies. And ev­ery­one knew why you were in the class­room. It would have been pointless to pre­tend you weren’t in the Con­spir­acy.

Their sen­sei was Jeffreys­sai, who might have been the best of his era, in his era. His stu­dents were ei­ther the most promis­ing learn­ers, or those whom the beisut­sukai saw poli­ti­cal ad­van­tage in mold­ing.

Bren­nan fell into the lat­ter cat­e­gory, and knew it. Nor had he hes­i­tated to use his Mistress’s name to open doors. You used ev­ery av­enue available to you, in seek­ing knowl­edge; that was re­spected here.

“—for over thirty years,” Jeffreys­sai said. “Not one of them saw it; not Ein­stein, not Schröd­inger, not even von Neu­mann.” He turned away from his sketcher, and to­ward the class­room. “I pose to you to the ques­tion: How did they fail?”

The stu­dents ex­changed quick glances, a calcu­lus of mu­tual risk be­tween the wary and the merely baf­fled. Jeffreys­sai was known to play games.

Fi­nally Hiriwa-called-the-Black leaned for­ward, jan­gling slightly as her equa­tion-carved bracelets shifted on her an­kles. “By your years given, sen­sei, this was two hun­dred and fifty years af­ter New­ton. Surely, the sci­en­tists of that era must have grokked the con­cept of a uni­ver­sal law.”

“Know­ing the uni­ver­sal law of grav­ity,” said the stu­dent Taji, from a nearby seat, “is not the same as un­der­stand­ing the con­cept of a uni­ver­sal law.” He was one of the promis­ing ones, as was Hiriwa.

Hiriwa frowned. “No… it was said that New­ton had been praised for dis­cov­er­ing the first uni­ver­sal. Even in his own era. So it was known.” Hiriwa paused. “But New­ton him­self would have been gone. Was there a re­li­gious in­junc­tion against propos­ing fur­ther uni­ver­sals? Did they re­frain out of re­spect for New­ton, or were they wait­ing for his ghost to speak? I am not clear on how Eld sci­ence was mo­ti­vated—”

“No,” mur­mured Taji, a laugh in his voice, “you re­ally, re­ally aren’t.”

Jeffreys­sai’s ex­pres­sion was kindly. “Hiriwa, it wasn’t re­li­gion, and it wasn’t lead in the drink­ing wa­ter, and they didn’t all have Alzheimers, and they weren’t sit­ting around all day read­ing we­b­comics. For­get the cat­a­logue of hor­rors out of an­cient times. Just think in terms of cog­ni­tive er­rors. What could Eld sci­ence have been think­ing wrong?”

Hiriwa sat back with a sigh. “Sen­sei, I truly can­not imag­ine a snafu that would do that.”

“It wouldn’t be just one mis­take,” Taji cor­rected her. “As the say­ing goes: Mis­takes don’t travel alone; they hunt in packs.”

“But the en­tire hu­man species?” said Hiriwa. “Thirty years?”

“It wasn’t the en­tire hu­man species, Hiriwa,” said Styr­lyn. He was one of the older-look­ing stu­dents, wear­ing a short beard speck­led in grey. “Maybe one in a hun­dred thou­sand could have writ­ten out Schröd­inger’s Equa­tion from mem­ory. So that would have been their first and pri­mary er­ror—failure to con­cen­trate their forces.”

Spare us the pro­pa­ganda!” Jeffreys­sai’s gaze was sud­denly fierce. “You are not here to pros­ely­tize for the Co­op­er­a­tive Con­spir­acy, my lord poli­ti­cian! Bend not the truth to make your points! I be­lieve your Con­spir­acy has a phrase: ‘Com­par­a­tive ad­van­tage.’ Do you re­ally think that it would have helped to call in the whole hu­man species, as it ex­isted at that time, to de­bate quan­tum physics?”

Styr­lyn didn’t flinch. “Per­haps not, sen­sei,” he said. “But if you are to com­pare that era to this one, it is a con­sid­er­a­tion.”

Jeffreys­sai moved his hand flatly through the air; the maybe-ges­ture he used to dis­miss an ar­gu­ment that was true but not rele­vant. “It is not what I would call a pri­mary mis­take. The puz­zle should not have re­quired a billion physi­cists to solve.”

“I can think of more spe­cific an­cient hor­rors,” said Taji. “Spend­ing all day writ­ing grant pro­pos­als. Teach­ing un­der­grad­u­ates who would rather be some­where else. Need­ing to pub­lish thirty pa­pers a year to get tenure...”

“But we are not speak­ing of only the lower-sta­tus sci­en­tists,” said Yin; she wore a slightly teas­ing grin. “It was said of Schröd­inger that he re­tired to a villa for a month, with his mis­tress to provide in­spira­tion, and emerged with his epony­mous equa­tion. We con­sider it a fa­mous his­tor­i­cal suc­cess of our method­ol­ogy. Some Eld physi­cists did un­der­stand how to fo­cus their men­tal en­er­gies; and would have been se­nior enough to do so, had they chose.”

“True,” Taji said. “In the end, ad­minis­tra­tive bur­dens are only a generic ob­sta­cle. Like­wise such an­swers as, ‘They were not trained in prob­a­bil­ity the­ory, and did not know of cog­ni­tive bi­ases.’ Our sen­sei seems to de­sire some more spe­cific re­ply.”

Jeffreys­sai lifted an eye­brow en­courag­ingly. “Don’t dis­miss your line of thought so quickly, Taji; it be­gins to be rele­vant. What kind of sys­tem would cre­ate ad­minis­tra­tive bur­dens on its own peo­ple?”

“A sys­tem that failed to sup­port its peo­ple ad­e­quately,” said Styr­lyn. “One that failed to value their work.”

“Ah,” said Jeffreys­sai. “But there is a stu­dent who has not yet spo­ken. Bren­nan?

Bren­nan didn’t jump. He de­liber­ately waited just long enough to show he wasn’t scared, and then said, “Lack of prag­matic mo­ti­va­tion, sen­sei.”

Jeffreys­sai smiled slightly. “Ex­pand.”

What kind of sys­tem would cre­ate ad­minis­tra­tive bur­dens on its own peo­ple?, their sen­sei had asked them. The other stu­dents were pur­su­ing their own lines of thought. Bren­nan, hang­ing back, had more at­ten­tion to spare for his teacher’s few hints. Be­ing the be­gin­ner wasn’t always a dis­ad­van­tage—and he had been taught, long be­fore the Bayesi­ans took him in, to take ev­ery available ad­van­tage.

“The Man­hat­tan Pro­ject,” Bren­nan said, “was launched with a spe­cific tech­nolog­i­cal end in sight: a weapon of great power, in time of war. But the er­ror that Eld Science com­mit­ted with re­spect to quan­tum physics had no im­me­di­ate con­se­quences for their tech­nol­ogy. They were con­fused, but they had no des­per­ate need for an an­swer. Other­wise the sur­round­ing sys­tem would have re­moved all bur­dens from their effort to solve it. Surely the Man­hat­tan Pro­ject must have done so—Taji? Do you know?”

Taji looked thought­ful. “Not all bur­dens—but I’m pretty sure they weren’t writ­ing grant pro­pos­als in the mid­dle of their work.”

“So,” Jeffreys­sai said. He ad­vanced a few steps, stood di­rectly in front of Bren­nan’s desk. “You think Eld sci­en­tists sim­ply weren’t try­ing hard enough. Be­cause their art had no mil­i­tary ap­pli­ca­tions? A rather com­pet­i­tive point of view, I should think.”

“Not nec­es­sar­ily,” Bren­nan said calmly. “Prag­ma­tism is a virtue of ra­tio­nal­ity also. A de­sired use for a bet­ter quan­tum the­ory, would have helped the Eld sci­en­tists in many ways be­yond just mo­ti­vat­ing them. It would have given shape to their cu­ri­os­ity, and told them what con­sti­tuted suc­cess or failure.”

Jeffreys­sai chuck­led slightly. “Don’t guess so hard what I might pre­fer to hear, Com­peti­tor. Your first state­ment came closer to my hid­den mark; your oh-so-Bayesian dis­claimer fell wide… The fac­tor I had in mind, Bren­nan, was that Eld sci­en­tists thought it was ac­cept­able to take thirty years to solve a prob­lem. Their en­tire so­cial pro­cess of sci­ence was based on get­ting to the truth even­tu­ally. A wrong the­ory got dis­carded even­tu­ally—once the next gen­er­a­tion of stu­dents grew up fa­mil­iar with the re­place­ment. Work ex­pands to fill the time al­lot­ted, as the say­ing goes. But peo­ple can think im­por­tant thoughts in far less than thirty years, if they ex­pect speed of them­selves.” Jeffreys­sai sud­denly slammed down a hand on the arm of Bren­nan’s chair. “How long do you have to dodge a thrown knife?

“Very lit­tle time, sen­sei!”

Less than a sec­ond! Two op­po­nents are at­tack­ing you! How long do you have to guess who’s more dan­ger­ous?

“Less than a sec­ond, sen­sei!”

The two op­po­nents have split up and are at­tack­ing two of your girlfriends! How long do you have to de­cide which one you truly love?

“Less than a sec­ond, sen­sei!′

A new ar­gu­ment shows your pre­cious the­ory is flawed! How long does it take you to change your mind?

“Less than a sec­ond, sen­sei!”


Sweat was form­ing on Bren­nan’s back, but he stopped and ac­tu­ally thought about it—


“No sen­sei! I’m not finished think­ing sen­sei! An an­swer would be pre­ma­ture! Sen­sei!

Very good! Con­tinue! But don’t take thirty years!

Bren­nan breathed deeply, re­form­ing his thoughts. He fi­nally said, “Real­is­ti­cally, sen­sei, the best-case sce­nario is that I would see the prob­lem im­me­di­ately; use the dis­ci­pline of sus­pend­ing judg­ment; try to re-ac­cu­mu­late all the ev­i­dence be­fore con­tin­u­ing; and de­pend­ing on how emo­tion­ally at­tached I had been to the the­ory, use the crisis-of-be­lief tech­nique to en­sure I could gen­uinely go ei­ther way. So at least five min­utes and per­haps up to an hour.”

Good! You ac­tu­ally thought about it that time! Think about it ev­ery time! Break pat­terns! In the days of Eld Science, Bren­nan, it was not un­com­mon for a grant agency to spend six months re­view­ing a pro­posal. They per­mit­ted them­selves the time! You are be­ing graded on your speed, Bren­nan! The ques­tion is not whether you get there even­tu­ally! Any­one can find the truth in five thou­sand years! You need to move faster!

Yes, sen­sei!

“Now, Bren­nan, have you just learned some­thing new?”

“Yes, sen­sei!”

“How long did it take you to learn this new thing?”

An ar­bi­trary choice there… “Less than a minute, sen­sei, from the bound­ary that seems most ob­vi­ous.”

“Less than a minute,” Jeffreys­sai re­peated. “So, Bren­nan, how long do you think it should take to solve a ma­jor sci­en­tific prob­lem, if you are not wast­ing any time?”

Now there was a trapped ques­tion if Bren­nan had ever heard one. There was no way to guess what time pe­riod Jeffreys­sai had in mind—what the sen­sei would con­sider too long, or too short. Which meant that the only way out was to just try for the gen­uine truth; this would offer him the defense of hon­esty, lit­tle defense though it was. “One year, sen­sei?”

“Do you think it could be done in one month, Bren­nan? In a case, let us stipu­late, where in prin­ci­ple you already have enough ex­per­i­men­tal ev­i­dence to de­ter­mine an an­swer, but not so much ex­per­i­men­tal ev­i­dence that you can af­ford to make er­rors in in­ter­pret­ing it.”

Again, no way to guess which an­swer Jeffreys­sai might want… “One month seems like an un­re­al­is­ti­cally short time to me, sen­sei.”

“A short time?” Jeffreys­sai said in­cre­d­u­lously. “How many min­utes in thirty days? Hiriwa?”

“43200, sen­sei,” she an­swered. “If you as­sume six­teen-hour wak­ing pe­ri­ods and daily sleep, then 28800 min­utes.”

“As­sume, Bren­nan, that it takes five whole min­utes to think an origi­nal thought, rather than learn­ing it from some­one else. Does even a ma­jor sci­en­tific prob­lem re­quire 5760 dis­tinct in­sights?”

“I con­fess, sen­sei,” Bren­nan said slowly, “that I have never thought of it that way be­fore… but do you tell me that is truly a re­al­is­tic level of pro­duc­tivity?”

“No,” said Jeffreys­sai, “but nei­ther is it re­al­is­tic to think that a sin­gle prob­lem re­quires 5760 in­sights. And yes, it has been done.”

Jeffreys­sai stepped back, and smiled benev­olently. Every stu­dent in the room stiffened; they knew that smile. “Though none of you hit the par­tic­u­lar an­swer that I had in mind, nonethe­less your an­swers were as rea­son­able as mine. Ex­cept Styr­lyn’s, I’m afraid. Even Hiriwa’s an­swer was not en­tirely wrong: the task of propos­ing new the­o­ries was once con­sid­ered a sa­cred duty re­served for those of high sta­tus, there be­ing a limited sup­ply of prob­lems in cir­cu­la­tion, at that time. But Bren­nan’s an­swer is par­tic­u­larly in­ter­est­ing, and I am minded to test his the­ory of mo­ti­va­tion.”

Oh, hell, Bren­nan said silently to him­self. Jeffreys­sai was ges­tur­ing for Bren­nan to stand up be­fore the class.

When Bre­nann had risen, Jeffreys­sai neatly seated him­self in Bren­nan’s chair.

“Bren­nan-sen­sei,” Jeffreys­sai said, “you have five min­utes to think of some­thing stun­ningly brilli­ant to say about the failure of Eld sci­ence on quan­tum physics. As for the rest of us, our job will be to gaze at you ex­pec­tantly. I can only imag­ine how em­bar­rass­ing it will be, should you fail to think of any­thing good.”

Bas­tard. Bren­nan didn’t say it aloud. Taji’s face showed a cer­tain amount of sym­pa­thy; Styr­lyn held him­self aloof from the game; but Yin was look­ing at him with sar­donic in­ter­est. Worse, Hiriwa was gaz­ing at him ex­pec­tantly, as­sum­ing that he would rise to the challenge. And Jeffreys­sai was gawk­ing wide-eyed, wait­ing for the guru’s words of wis­dom. Screw you, sen­sei.

Bren­nan didn’t panic. It was very, very, very far from be­ing the scariest situ­a­tion he’d ever faced. He took a mo­ment to de­cide how to think; then thought.

At four min­utes and thirty sec­onds, Bren­nan spoke. (There was an art to such things; as long as you were do­ing it any­way, you might as well make it look easy.)

“A woman of wis­dom,” Bren­nan said, “once told me that it is wis­est to re­gard our past selves as fools be­yond re­demp­tion—to see the peo­ple we once were as idiots en­tire. I do not nec­es­sar­ily say this my­self; but it is what she said to me, and there is more than a grain of truth in it. As long as we are mak­ing ex­cuses for the past, try­ing to make it look bet­ter, re­spect­ing it, we can­not make a clean break. It oc­curs to me that the rule may be no differ­ent for hu­man civ­i­liza­tions. So I tried look­ing back and con­sid­er­ing the Eld sci­en­tists as sim­ple fools.”

“Which they were not,” Jeffreys­sai said.

“Which they were not,” Bren­nan con­tinued. “In terms of raw in­tel­li­gence, they un­doubt­edly ex­ceeded me. But it oc­curred to me that a difficulty in see­ing what Eld sci­en­tists did wrong, might have been in re­spect­ing the an­cient and leg­endary names too highly. And that did in­deed pro­duce an in­sight.”

“Enough in­tro­duc­tion, Bren­nan,” said Jeffreys­sai. “If you found an in­sight, state it.”

“Eld sci­en­tists were not trained...” Bren­nan paused. “No, un­trained is not the con­cept. They were trained for the wrong task. At that time, there were no Con­spir­a­cies, no se­cret truths; as soon as Eld sci­en­tists solved a ma­jor prob­lem, they pub­lished the solu­tion to the world and each other. Truly scary and con­fus­ing open prob­lems would have been in ex­tremely rare sup­ply, and used up the mo­ment they were solved. So it would not have been pos­si­ble to train Eld re­searchers to bring or­der out of sci­en­tific chaos. They would have been trained for some­thing else—I’m not sure what—”

“Trained to ma­nipu­late what­ever sci­ence had already been dis­cov­ered,” said Taji. “It was a difficult enough task for Eld teach­ers to train their stu­dents to use ex­ist­ing knowl­edge, or fol­low already-known method­olo­gies; that was all Eld sci­ence teach­ers as­pired to im­part.”

Bren­nan nod­ded. “Which is a very differ­ent mat­ter from cre­at­ing new sci­ence of their own. The Eld sci­en­tists faced with prob­lems of quan­tum the­ory, might never have faced that kind of fear be­fore—the dis­may of not know­ing. The Eld sci­en­tists might have seized on un­satis­fac­tory an­swers pre­ma­turely, be­cause they were ac­cus­tomed to work­ing with a neat, agreed-upon body of knowl­edge.”

Good, Bren­nan,” mur­mured Jeffreys­sai.

“But above all,” Bren­nan con­tinued, “an Eld sci­en­tist couldn’t have prac­ticed the ac­tual prob­lem the quan­tum sci­en­tists faced—that of re­solv­ing a ma­jor con­fu­sion. It was some­thing you did once per life­time if you were lucky, and as Hiriwa ob­served, New­ton would no longer have been around. So while the Eld physi­cists who messed up quan­tum the­ory were not un­in­tel­li­gent, they were, in a strong sense, am­a­teurs—ad-lib­bing the whole pro­cess of paradigm shift.”

“And no prob­a­bil­ity the­ory,” Hiriwa noted. “So any­one who did suc­ceed at the prob­lem would have no idea what they’d just done. They wouldn’t be able to com­mu­ni­cate it to any­one else, ex­cept vaguely.”

“Yes,” Styr­lyn said. “And it was only a hand­ful of peo­ple who could tackle the prob­lem at all, with no train­ing in do­ing so; those are the physi­cists whose names have passed down to us. A hand­ful of peo­ple, mak­ing a hand­ful of dis­cov­er­ies each. It would not have been enough to sus­tain a com­mu­nity. Each Eld sci­en­tist tack­ling a new paradigm shift would have needed to re­dis­cover the rules from scratch.”

Jeffreys­sai rose from Bre­nann’s desk. “Ac­cept­able, Bren­nan; you sur­prise me, in fact. I shall have to give fur­ther thought to this method of yours.” Jeffreys­sai went to the class­room door, then looked back. “How­ever, I did have in mind at least one other ma­jor flaw of Eld sci­ence, which none of you sug­gested. I ex­pect to re­ceive a list of pos­si­ble flaws to­mor­row. I ex­pect the flaw I have in mind to be on the list. You have 480 min­utes, ex­clud­ing sleep time. I see five of you here. The challenge does not re­quire more than 480 in­sights to solve, nor more than 96 in­sights in se­ries.”

And Jeffreys­sai left the room.