Psychohistory: Was Hari Seldon pulling our leg?

Note: This essay is based entirely on the first Foundation trilogy. The later books modify our picture of both psychohistory and certain key events in galactic history. But the original trilogy stood on its own for more than thirty years, and it is entirely fair to judge the concept of psychohistory presented in it, unmodified by Asimov's afterthoughts, exactly as it confronted readers for decades.

I also assume you've read the books, and am therefore quite cavalier about spoilers.
--Mark Rosenfelder

Does psychohistory work?

Discussions of psychohistory usually turn into debates on the role of the individual: can one person significantly affect the course of history or not? We all have our pet cases proving one point of view or the other. I have a strong opinion on the issue-- but I'm going to suppress it.

Instead, I'm going to ask another question: does psychohistory work in Asimov's world? Is Seldon's theory correct, and does his plan work, and actually account for the Foundation's success?

The answer might seem to be trivially true-- the Foundation does succeed, doesn't it, and Asimov has set it up so psychohistory is the reason why-- or trivially false, because of the Mule.

Asimov is actually rather more subtle than that, however. He never tells us that psychohistory works. He shows us actions taking place across the galaxy; he shows what First and Second Foundationers think is going on; but it is really up to us to decide if we're convinced or not.

The macro argument: Too many Mules

Everybody knows that the Mule is a problem for psychohistory. What isn't generally appreciated is a) how big a problem he is; and b) that there were two of him.

Seldon claims to be able to predict, on a gross level, 30,000 years in the future; and in detail, at least 1000 years. (Now, don't molest me with caveats about probability. Seldon knew those caveats as well as we do; and yet his whole argument rests on this figure-- there is no reason for the Plan at all if the 30,000-year dark age is not a near-certainty.)

Yet the plan derails within three centuries due to the appearance of a single mutant. The evidence-- from the books, mind you-- is that psychohistory can't accurately handle a hundredth of the time Seldon's argument needs it to.

Oh, but the Mule is a freak of probability. Sure; and if you play poker with me and I get a royal flush in the first hand you'll chalk it up to your own dumb luck. It could be luck. It could also be a basic flaw in the method. I could be cheating at cards. The Galaxy could be ignoring what Seldon tells it to do.

We're never told the probability of the Mule's mutation; but remember that the Galaxy has a quintillion people-- a billion billion little breeders. That's an awful lot of chances of hitting the jackpot. And that's to say nothing of non-biological threats to the plan-- new technologies that have as great an impact on galactic society as atomic power did, for instance.

How was psychohistory tested, by the way? A theory that can't be tested isn't science; and we're assured that psychohistory is science. The only conceivable answer is that Seldon tested his theories on the past-- he used it to explain the course of Imperial history up to his own time.

One immediately has to wonder how valid is the extrapolation to the galactic dark ages, inasmuch as the Empire has lasted for 12,000 years. What good data on dark ages and non-Imperial states of the Galaxy can remain after all that time?

The usual analogy is with thermodynamics; but a better analogy would be with a thermodynamics developed by Jovians, whose experience is with liquids and gases only. How confident would we be that their theories would correctly account for the behavior of solids?

The Mule could still just be incredibly bad luck. He could be; but surely two Mules would be an almost fatal blow to the theory? Wouldn't (say) two Mules in three centuries represent a frequency of psychohistory-derailment sufficient to make the theory completely untrustworthy, or at least reduce its scope to centuries rather than milennia?

Sure, but that didn't happen. Only it did. There were two galaxy-altering, psychohistory-derailing individuals in the span of just three centuries. One was the Mule; the other was Hari Seldon.

Seldon himself makes mincemeat of his own theory; for he shows in his very person that, no matter how much the actions of quadrillions of human beings can be systematized and predicted, history can still be altered forever by the action of a single human being-- and not in minor matters; we're talking about a 30,000 year dark age occurring or not.

Some boldly answer that the Galaxy might well have been ready for psychohistory-- if it hadn't been Seldon it would have been someone else. But we are told many times that psychohistory can't handle the complexity of a world that has psychohistorians in it. (That means the Second Foundation can't predict its own future actions, but let's leave that alone.) If the galaxy is ready for psychohistory, psychohistory will develop somewhere else a generation after Seldon-- and Seldon's predictive power becomes not 300 years but 30.

In a nutshell: If a psychohistorian (perhaps an alien observer) existed 1000 years before Seldon, then either he predicts the apperarance of psychohistory or not. If he doesn't, then his predictions are absurdly off. If he does, then his predictions are derailed anyway, because once psychohistory exists history becomes unpredictable.

The Second Foundation will take care of it. Sure it will. However, that's not psychohistory succeeding-- that's direct action. We'll get back to that.

The micro argument: What actually happened

If we look at the stories themselves, we see them explained by the characters as consistent with psychohistory. But their explanations may not be correct.

Consider the first Crisis: Hardin chases Anacreon off Terminus by appealing to Anacreon's three neighbors. This is, we're told, the only possible solution-- which it may well be. There is no discussion, however, of what might have gone wrong. And much might have.

Hardin speaks as if inaction, in awed contemplation of the Seldon Plan, was a virtue. The evidence from the stories, however, is that he acts when necessary. He wages political campaigns. He executes a putsch against the Board. He visits Anacreon's enemies. He comes up with the religious angle. He agrees to fix the Imperial cruiser. He visits Anacreon at the time of the second Crisis.

If we were told nothing about psychohistory, the first book, at least, would work just as well. The only evidence that psychohistory has any real power at all is in a few unimportant details-- e.g. Seldon first appears at precisely the crux of both an internal and an external crisis. But if he hadn't, so what? Hardin was plotting his coup anyway; given the political incompetence of the Board, it wouldn't have been much harder.

What we see psychohistory actually providing, of course, is confidence. We're told over and over again that the Foundation believes, almost to the point of carelessness, in its own destiny; and that even its enemies come to believe it, and to fear Seldon. This is not an unappreciable effect-- but it proves nothing about the actual effectiveness of psychohistory. To have this effect, psychohistory need not be true; it need only be believed.

The counter-interpretation: Psychohistory as fraud

A brilliant psychologist and political scientist foresees the imminent, violent fall of his own state. It's too late to prevent the catastrophe, nor is it desirable; the empire is arteriosclerotic and authoritarian, and must be remade, not rescued. A political entity capable of this remaking must be established.

Obviously the real nature of this enterprise must be kept from the imperial authorities, who can hardly be expected to succor their successors. The movement is therefore disguised as an educational institution, and established in a far region, where the central authorities will soon forget about it.

The secret is kept also from the bulk of his followers. In this he is only crazy as a fox. First, this stratagem prevents accidental revelations to the secret police, at least for the years the police can still pay it attention. More importantly, perhaps, the encyclopedia project attracts mostly scientists, not the politicos, power-seekers and malcontents normally drawn to a political movement.

Nor is the educational purpose entirely fraudulent, of course. The new culture is indeed intended to preserve existing science, and the new society's chief edge over its inevitable competitors will be technological.

The remote location of the foundation is also according to plan. Fogbound or status-hungry scientists will stay home; and the rough conditions will encourage a level of innovation both necessary for the success of the project and impossible to attain in the imperial center.

The power of ideology, the psychologist realizes, is too important to forego. The colonists' situation will itself provide half of it-- nationalism and a sense of superiority borne of their technological edge. He reinforces this with an edited version of his own beliefs and methods. The colonists will be chiefly sustained by a belief in their own exalted destiny, and in the preternatural foresight of their founder.

The political science he gives them is vague to the point of nothingness. Its purpose is not to provide detailed guidance, but to inspire. And as such it's self-reinforcing: every success of the colonists will be attributed to his own foresight. Just to make sure, he excludes psychology and political science from the sciences preserved in the colony. The doctrines he leaves are thus not subject to second-guessing.

Recognizing his own limitations, he also provides for the continuance of his own political analysis and for oversight over the developing society. A second group is founded, based at first in the capital where the greatest threat to the project lies. Its agents will also live among the colonists, of course, where events can be prodded along, and the destiny religion can be nurtured, for instance by staging posthumous 'appearances' by the psychologist-founder.

The primary purpose of this second group is political intervention on behalf of the colony. This is rather dangerous, much more so than merely establishing a colony, and certain safeguards must be provided.

First, the physical powers of the controllers is limited. To provide a balance of power with the scientific colony, their knowledge of science and technology is severely limited. And to prevent them from establishing a state of their own, which might rival the colony's, they are housed first in the Empire, and later in the colony itself.

Second, they also are given an ideology-- in effect, the psychologist's own, a representation of his own political methods and understanding, and his plan for the future.

Ideologies normally develop, and end up little resembling the founder's intentions. To hinder this, the psychologist codifies his thinking in the trappings of mathematics, and makes the acquisition and correct interpretation of these ciphers the path of advancement in the group. (Analogies with the Chinese Imperial examination system, or the American educational establishment, should be obvious.)

Trained in no other form of mathematics, the controllers think they are applying the results of science, when they are in fact only consulting his political oeuvre in a mythologized and circuitous fashion. Additions are permitted, but only anonymously, so that the only authority in the movement is his own. As a final check, the database containing his encoded political guidelines is maintained in specialized computer the controllers have no technological competence to alter or to reprogram.

Sometime in the far future, historians in the new civilization he has set in motion will examine their past, and begin to examine the sources of their success. They will discover the little deceptions he played on them-- the encyclopedia, the hokey appearances, the database, the destiny religion. He thinks they will forgive him. None of these deceptions, after all, are really serious. They will have the encyclopedia, after all, and the database of his political thought. And the ideology of political determinism-- well, it told them that they would avert a dark age and build a new civilization, and they did so, didn't they?

Asimov didn't do that, did he?

This reading of the trilogy won't appeal to everyone, of course. And I'll cheerfully admit it's more or less a retcon-- an interpretation that makes sense of a beloved story which, taken on its own terms, has more than its share of holes.

However, I wouldn't completely put it past Asimov. We're told in other sources that he never really bought psychohistory himself. He just used the idea to tell a series of good yarns.

It's not always realized that this was his modus operandi with the robot stories, as well: take an interesting idea and undermine it. The tension in the psychohistory stories is always Was Hari Seldon right? Will the present crisis be too much for Seldon's dead hand? Similarly, the robot stories are not celebrations of the Three Laws, but a series of elegant attacks on them.

Notes on the stories

What follows is a set of miscellaneous thoughts that occurred to me as I was rereading the series in preparation for this essay.


The Psychohistorians (FE 1)

Trantor's population is given as 40 billion, and its land area as 75 million mi2,-- that's 533/mi2-- which is rather unimpressive, given New York's 24,287/mi2. The third book says 400 billion instead-- still less than New York.

The empire is said to have a quintillion souls on 25 million inhabited words. That's 40 billion per world, which seems excessive.

Despite what apologists for psychohistory sometimes claim, Seldon very cheerfully makes predictions not only about individuals but about individual events. He has a 1.7% chance of being executed; Dornick has a 77.2% chance of acquittal. This makes hash of claims about the statistical nature of psychohistory-- thermodynamics can't predict the motion of individual molecules-- but then people aren't really molecules. Something must be done to paper over the reader's realization that individuals do in fact matter.

When psychohistory is translated into words, it becomes poli sci-- and nothing novel or striking, like Marvin Harris or Jane Jacobs; very conventional stuff.

The Psychohistorians (FE 50)

Terminus was discovered just 500 yrs before its settling?? After a 12,000 year galactic empire?

The technology throughout the trilogy is disappointing. Hyperspace, antigravity, and atomic wastebaskets, yes; but people scrawl on paper; taxis have drivers; people eat potatoes and smoke tobacco; metals (especially gold) are hot stuff. There's no power source fancier than atomic power; and when it's lost (in just fifty years??), Anacreon is back to "oil and coal"! (It's awfully fortunate that oil rigging and coal mining technology has been maintained in readiness during all those milennia...)

This story has no action at all; everything in it from exposition to threat to resolution is handled by talk. And most of the characters are dull; only the Anacreon and Imperial ambassadors are colorful. Hardin himself is a haranguer, worse than anything in Heinlein, scarcely more tolerable because he's right. Fortunately Asimov would learn as he went along.

The Mayors (FE 80)

The hierarchy of prefects, provinces, sectors, quadrants is missing a few levels. The Galaxy in these books doesn't seem like it has millions of worlds-- it seems like it has hundreds.

"As long as more than one course of action is possible, the crisis has not been reached. We must let things drift as long as we possibly can." Hardin has basically internalized fatalism in the form of subservience to a dead authority-- ironically, exactly what he accused the scientists and aristocrats of doing in the last story.

"It was only afterward that I thought I saw a pattern in events; but I've done my level best not to act on that knowledge." But to purposely dull one's own political instincts is not to be uninfluenced by psychohistory.

This is a good deal more satisfying story-- although its end can be predicted a long way off, as soon as Hardin agrees so casually to repair the Imperial cruiser. Of course he's going to gaff it.

The Traders (FE 135)

From a 1990s perspective, a rather nasty story! The Foundation, intent on galactic empire, uses any means necessary to crack even markets that don't want to be cracked. In its essentials it's the last story retold-- scientific tricks are used to overawe gullible provincials. But the goal here is not to safeguard the Foundation's existence or even its destiny, but to reduce an insignificant state whose principled stand agains the modern is doing (and can do) the Foundation no harm.

The Merchant Princes (FE 155)

I first read Foundation at about twelve, and it made a deep impression on me. Rereading it, I can see the source of some of my own juvenilia-- stories of interplanetary and parliamentary intrigue, conveyed entirely by talk, with a charismatic but inscrutable figure in the center.

As politics, this story is smug and unconvincing. That the housewives of Korell will grumble because their atomic knives are failing flies in the face of all we know of the behavior of civilian populations during wartime. People were miserable during the Blitz, but they didn't go over to the Germans because of it.

Asimov assumes that a superior economy can maintain its dominion indefinitely-- a hypothesis that looked more plausible in 1951 than in 1997. Why has no other Periphery state imitated the Foundation's success? Read Jane Jacobs to see how it could happen. You don't start by creating personal force-fields, of course. You start by making spare parts; the money you save not importing them goes into local development and more sophisticated imports. Once your economy makes enough spare parts, you start assembling simple machines. The small scale of these operations leads to widespread innovation and knowledge diffusion. In under a century you'll be making atomic cruisers.

Foundation and Empire

The General (~ FE 195)

Asimov's analysis-- "Cleon is strong because he permits no strong subjects"-- is simply ahistorical. The very name Bel Riose seems to be based on Belisarius, the strong Byzantine general who ended up being sacked by his strong emperor, Justinian. But Belisarius and his successor Narses didn't find "wars of conquest impossible"; they reconquered Italy, North Africa, Dalmatia, and southern Spain.

It's a particular pleasure to point this out because Ducem Barr, with his smart-mouthed evocation of Seldon's dead hand, is such a pill here. I was rooting for Riose.

A critic complains that Barr and Devers' jaunt to Trantor is completely irrelevant to the story, and perhaps indicates an authorial wrong turn detected, if at all,too late. Probably true, though Asimov's storytelling has so improved by this point that the flaw is noticed only in retrospect.

The Mule (~ FE 294)

One of the most effective of the stories (once you get past the creaky initial exposition). The story works for the same reason the Time Vault scene does: Hari Seldon is, for once, on the ropes.

Second Foundation

Search by the Mule (~ FE 299)

By the spaceways! The stars forbid! Galaxy! All those science-fictional oaths never convince me. This is a culture that's been spacegoing for 12,000 years; why would it swear by such banalities? It's as if we swore by airplanes or supertankers. People swear by what they hold sacred (so the occasional swearing by Seldon does make sense).

I rather liked Bail Channis. There's an inherent fascination in a character who could serve the Mule without being Converted.

The Mule gives up too quickly. Why should he take the First Speaker's word for it that he's finished?

Search by the Foundation (FE 376)

The transcriber as described might work for Asimov's own writing-- he rarely wrote a second draft-- but it wouldn't work for most of us; we need to edit. And evidently you can't-- when the essay is ruined by the transcriber being left on, Arkady has to start over. Nice UI.

The first chapter says Arkady was born in FE 348, and is 14 years old; that means it's 376. But chapter 16 says it's FE 348. Sloppy.

The Second Foundation is really chilling. My sympathies were once more with the (First) Foundation, the poor devils. Who wants a ruling class of mind-altering superhumans? Sounds horrible.

The hubris of the First Foundation is at least conventional-- they're Americans writ large, and their instincts are democratic. But the Second Foundation has the thinking of a Politburo; its aim is explicitly to establish itself as a ruling elite over inferiors. They also suffer no appeal-- you can't face them, can't disagree with them, can't overthrow them; and indeed, they can not only repress dissent, but reach into your mind and remove that dissent, and leave you feeling happy about it, like poor Winston at the end of 1984.

When Robert Heinlein conceived of similar creatures, he was decent enough to see them as implacable enemies of mankind, for they really strike at the very root of what it means to be human or to be free. The totalitarian can take your life; but the Second Foundationer can take your mind as well.

[back to Metaverse]