Thursday, February 16, 2017

Reprint: John W. Campbell, Traditional Values

The following article was written by Mr. Campbell for the February 1970 issue of Analog Science Fiction/Science Fact just over a year before his death. His forceful personality and reductionist editorial style had long since divided and conquered SF; the Golden Age was long passed, New Wave was ascendant, and the cinematic pulp revival spawned by Star Wars was years away. Campbell here comments on the ill-planned destruction of tradition.

Traditional Values
an editorial by John W. Campbell

Any fundamentally sound idea can be carried too far and become a fanaticism - and the essence of any fanaticism is, simply, that there is Only One Right Way and We've Got It.

Even the objective physical scientist can get suckered on that one - because while his field of study may be objective and physical, he remains a human being whose mental nature is inherently subjective. As a scientist who bases his thinking on careful laboratory experiments - a sound method that has yielded immense and reliable benefits - he displays his human tendency to fanaticism by holding that only laboratory evidence is sound data. That something which cannot be confirmed by repeatable experiments in a laboratory is not usable information.

That repeatable laboratory experiments are highly important is not really open to argument; it's an exceedingly fruitful, and, therefore, valid, method of learning and confirming derived ideas. But it is not the Only One Right Way - in fact, there are many areas where it cannot be applied as it is now thought of, and others where we certainly can't apply it with any tools available to is.

Example: As of 1800 many astronomers strongly suspected that the "fixed" stars weren't fixed; by that time they had some pretty good telescopes, could measure star positions with high accuracy, and had tried measuring star positions for changes. With several million stars visible to choose from, naturally they chose the wrong stars - there are darned few of them that move across the firmament fast enough for ordinary means of measurement. (It's different when you finally develop high-precision photography, "blink comparators, and microdensitometers.) The rate of motion of a star may be 200 miles per second, but because of interstellar distances, the angular rate of motion is usually so small it can't be detected in a human lifetime.

However, Halley, in 1817, showed that both Aldebaran and Arcturus moved across the firmament; he had studied star maps as much as two thousand years old, and the movement of those stars - relatively near Earth; fifty-five and thirty-three light-years, respectively - had accumulated enough in two millennia to be obvious. It could be detected readily by naked-eye observation by anyone patient enough to wait around for two thousand years while it happened.

Even the motions of more distant stars would be readily observable by ordinary naked-eye observation if you just used a little patience; after all, the Big Dipper wasn't a dipper when we men started doing those cave drawings one hundred thousand years ago.

We might properly say that the optimum tool for much astrophysical work is not a super-telescope mounted in space, not a hyper-spectroscope of unlimited resolution and total spectrum coverage from 1.0 Angstrom to 100,000,000 - but a time machine. Want to study stellar evolution? The mechanism of supernovas? How galaxies form? A small telescope - maybe a 6" reflector and a modest spectroscope - would help, of course, but remember that our spaceship will take us near almost every imaginable type and class of star, in almost every possible stage of evolution. Our spaceship called Earth that circumnavigates the entire galaxy about once every 200,000,000 years.

Curious about what ended the Age of Dinosaurs? What caused primeval Pangea to break up into drifting continents 200,000,000 years ago - and what Earth's surface was like before that, so you can solve the problem of planetary dynamics? A time machine would be far better than any Mohole Project.

Indications now seem to suggest that what ended the Age of Dinosaurs was a relatively nearby supernova explosion, when the Solar System was around on the other side of the galaxy, half a Great Year ago. When a giant star turns into an exploding nuclear bomb, the fallout and radiation blasts everything within half a dozen light-years with hard radiation. Not even ordinary interstellar distances are great enough to render the fallout of a 4,000,000-mile-diameter detonating A-bomb harmless.

The atmosphere is inadequate protection against ordinary intensities of stray radiation in space - the atmosphere, reinforced by Earth's magnetic field, and Sol's greater, farther-reaching field. But, if a neighboring giant star detonates, the result is not "ordinary intensities" of radiation. Land life gets sprayed with high levels of hard radiation, both high-energy photons and particulate. The result is a tremendous wave of mutations; only the sea creatures, living under a deep blanket of dense, radiation-absorbing water would be protected.

The mutations are apt to bring finis to many species by direct action - and to others indirectly, as a few of the mutations turn out to be markedly favorable. The mammals, with their strange new characteristics, originated at just about the time the dinosaurs were vanishing.

Now obviously things like that aren't subject to laboratory test - at least not for some little while. We aren't about to trigger a supergiant star into supernova detonation just yet, and see what it does to neighboring planetary systems.

Then there are a lot of things on a smaller scale that call for a time machine more than any other tool: you don't need super-dooper hyper-precision devices to measure stellar movement if you can just use patience - a millennium or so of it.

Now the point of this is that while you and I and all living scientists simply can not use millennial patience - the human race can. The modern species of man is apparently about 150,000 years old; the stars have moved quite a bit in that time. The glaciers have come and gone and surged back and retreaded. A number of animal species have evolved and others have failed and died out. (And I don't mean those that the rise of man's technical civilization killed off.)

Various types of social systems have evolved, been tried out, and been discarded over those one hundred and fifty millennia. No man can have such patience; the human race has.

The ancient builders of Stonehenge didn't have precision telescopes; they made up for the lack by using many lifetimes of naked eye observation - their own form of time machine! - to determine the motions of the Earth, Sun, and Moon with precision.

To determine factors accurately, you can approach the problem with exceedingly sophisticated ultra-precision instruments - or you can make fairly crude observations extended over a long period of time. Thus, although even the greatest telescopes can't see Pluto well enough to measure its diameter, and its mass is known only in the vaguest sort of way - with the result that it isn't known whether its density is 5.0 or 50.0 - the length of its day is know with five-figure length accuracy. Reason: the planet appears to have some kind of a blob on one side, and photoelectric observations of its brightness show periodic variations. Even though no one period can be measured to better than +/-10%, observations of several years, and hundreds of cycles allows a determination with great accuracy.

As of now we do not know that the speed of light is in fact invariant; measurements made over the last half-century or so have appeared to show a progressive change at least on the order of the accuracy of measurements. Maybe the value of c changes with time? Have patience; we'll know for sure in just a millennium or so.

But the area wherein the long view is essential - and laboratory experiments under full control are simply impossible - is sociology.

One of the major stupidities of the professional sociologists - and psychologists, also - is the oft-repeated statement that "you can't experiment with human beings."

Look, friends - if you do not have full knowledge of the consequences of your actions, and you have to act - you're experimenting, whether you admit it or not.

No human being or group of human beings today has full knowledge of the complex interaction of human dynamics called "sociology."

Ergo, all societies are always experimenting and always have been.

In the absence of complete knowledge of the structure and dynamics of human minds - psyches - any action taken with any other human being is an experiment. All parents lack that knowledge, specifically including people like Dr. Spock and Dr Skinner. Therefore, all parents are always experimenting with human beings.

It doesn't matter a bit what you think you're doing - it makes no difference how absolutely rock-ribbed certain you are that you do in fact know exactly what the right moves are. There is no predictable correlation whatever between the intensity of your certainty and its correctness.

Matter of fact, there is a high probability of a negative correlation; an absolute certainty that you're right means a fanatic attitude, which prevents learning and self-correction.

Thus two of the most critically important areas of human experience are the peculiar position that the "authorities" in the field keep saying "you can't perform experiments with human beings," while the fact is that due to lack of knowledge in the areas, innumerable experimental tests are being conducted all around them, and all over the world.

The problem is that the "authorities" in the field keep thinking solely in terms of laboratory experiments, where they can control every possible variable, and observe the results in detail. I started with the problem of astronomy, because that area in many ways resembles the problems of sociology and psychology. The astronomer can't perform laboratory experiments either - he damn well can't manipulate giant stars and test the stability of white dwarfs, or get inside a pre-supernova to see what's happening. He'd sure like to be able to.

And the psychologist would find his work greatly helped if he had true telepathy so he could get inside another mind to see what was happening.

O.K. - so neither one of them can - but the astronomer has had the good sense to acknowledge that, and study the experiments that happen in the galaxies around him, without bitching about the impossibility of making experiments.

Like the sociologist, the astronomer is working with a population having interactions so complex that he can't analyze their movements mathematically, save in the broadest of terms. And like the sociologist, he doesn't know all the forces that operate to direct the motions: it's only recently that they've become aware of how important the exceedingly week, but stupendously extensive magnetic field is. The astrophysicists have a problem just about the same as that of the sociologist.

But they have a different philosophy; the astrophysicist recognizes that the stars - which he definitely can't control - are experiments he can observe. That he can learn the basics of stellar dynamics not by interfering with them, but by observing them patiently.

Neither the psychologists nor the sociologists have accepted the necessity for using the "natural experiments" all around them adequately - with curious and painful results.

Since it is held unethical to experiment with human beings, but the fact is that we must act in ignorance - a situation which is experimenting! - we get around this by loudly proclaiming we know-for-sure what we are doing.

If I know-for-sure what the right answer is, then when I do what I think needs doing, I can hold I am not doing that unethical thing, experimenting with human beings. That makes it ethical for a fanatic to do what he Knows Is The Only Right Thing (Heil Hitler! Or Hail Cromwell! Or "Jesus and No Quarter to the enemy!")

But the thoughtful man who recognizes that his knowledge is imperfect - that he may be mistaken - is unethical, for he is experimenting with his children, or his neighbors, or the people of his city, state or nation.

Only fools are certain they are right about anything so complex as psychology or sociology - and under the doctrine that holds experimenting with human beings is unethical, only fools appear ethical.

The lifetime of a society is greater than the lifetime of any observing man - again, sociology is like astronomy. What's needed, again, is a time machine by which we can trace the development of societies and study the results of this, that, and the other variable. Even if we could introduce a desired variable F, we could not observe its working out until at least three generations of people had interacted and distributed that new factor. If you could introduce a new Earth-size planet into the region between Mars and Jupiter, you couldn't determine the consequences in the next ten years, obviously. It wouldn't have had time to make more than two orbits, and the final results of its interactions with the asteroids would be completely unknown - far too complex for computation.

Halley was the first to show that the "fixed stars" did move - and he did it not by super-accurate measurement, or new experimental techniques - he did it by recognizing and using traditional values. In two thousand years, the effects showed up clearly and measurably.

Now one of the great problems of human life is that much of human personality extends beyond the reach of logic - because logic is a useful, but inadequate tool.

Logic - mathematical logic - is an inadequate tool for the astronomer, too; the motions and interactions of a hundred billion stars are beyond the ability of mathematical logic to analyze.

The interactions of two billion human beings - even if each unit were strictly logically controlled - would be beyond formal logical analysis.

Therefore, a formally-logical sociology is necessarily almost useless, save as an intellectual game.

Complicating the matter however is the fact that human beings have nonlogical - and I mean higher-than-logical, not anti-logical! - abilities. Any formal logician, who really knows what logic is, will assure you that logic is a science of manipulation of postulates according to rules - but it can never generate a postulate. It can only derive conclusions from postulates that are introduced from some nonlogical source.

Consider the following logical argument:

If all men are green, and all green things are explosive, then all men are explosive.

That, my friend, is unshakably valid logic. The explosive nature of men is an inescapable conclusion.

Oh, sure - it happens that the postulates are false. But the logic is perfect.

And because we have no communicable technique for deriving postulates, people observing the same situations derive differing postulates, then using the same logical techniques reach conclusions that each holds inescapable. There is no logical way to evaluate a postulate, because logic does not apply to postulates. They are generated by an as-yet-unknown method that exists outside of, and underlies, all logic.

(The rules of logic are themselves postulates, incidentally.)

The only way to test the validity of a postulate is to check not against someone else's postulate logic system - you may both be entirely wrong! - but against the real universe of experience.

Events are not logical; they are simply True. This can best be appreciated directly by considering the problem of demonstrating something that is highly improbable; if it can be shown to be a one in a billion probability by logical mathematic analysis, when it happens it is neither probable nor improbable - it is only True. A repetition of the event may then be declared highly improbable - but any event is True, and beyond the realm of logic. It's the Ultimate Arbitrary; you can argue from it, but no logical argument can alter it. A witness should be arbitrary and not subject to persuasion, because he should be reporting an event, which is True.

Wherefore the only proper check of the validity of postulates is to compare them with events. If logical derivations from the postulates correctly predict what happens, you have reason to put some trust in your postulates.

Now herein lies the importance of traditions and traditional values. Like the 2,000-year-old star maps that allowed Halley to see that Arcturus and Aldebaran had moved, traditions represent postulates that have been tested-in-action over long periods of time.

They may not be completely and precisely correct - but only a fool would hold that they were valueless. They represent the results obtained by experiments performed on millions of human beings, over centuries of time.

They are, in fact, the basic data on which a sound sociology, or sound psychology, must be based; they're the experimental results that the modern "authorities" in those fields say we can't get because we can't perform experiments with human beings.

Traditions are valuable not because they're traditional, but because they're rule-of-thumb engineering results from ages of experiments performed on/by millions of human beings under widely varying conditions.

The old Roman engineers were very weak on theory; unlike the Greek theoreticians, the Romans didn't do much arguing about philosophy - they built things, and sought only to find practical working rules of how-to-do-it. They didn't understand force-vectors, Young's Modulus, or the chemistry of mortar, but they built magnificent arch bridges, and great domes that have stood for two thousand years. Some of their works are still in practical operation. They were lousy theorists - but their rule of thumb traditions of how to build a bridge that wouldn't fall down worked.

The fact that you cannot understand, or explain, something has nothing whatever to do with its validity.

It would behoove any would-be engineer stumbling across such a structure to study and appreciate it. And any would-be theoretician would be wise to understand that for his field of study, such a bridge is an Event; it's true, and he'd better try to understand why, instead of trying to explain it away as useless - old-fashioned - a mere tribal mores - things have changed.

Sure they have - but the basic laws haven't. We use steel reinforced concrete rather than mortared stone, but we also use the principle of the arch.

The importance of traditions is not that they're traditional and we ought to worship antiques - but that they are old, and have grown old in service.

They worked.

Like Roman aqueducts and bridges, they're still functioning usefully after millennia of use. They must have great basic laws underlying them or they'd have crumbled before this.

-The Editor

This editorial was on the last page of this issue. The back cover advertised Heinlein's Stranger in a Strange Land.

No comments:

Post a Comment