Science Fiction Project - Free Culture
Analog - All editorials - John Wood Campbell
* * Back * *


One of the major faults I find with the "scientific approach" to problems is the powerful tendency to hold that that-which-is-known necessarily embraces all-that-is-possible. Stated in that form, of course, any scientist would immediately deny it; it's normally stated in reverse form - i. e., "nothing known can produce such effects, so it is clearly a hoax, misobservation, or fraud".
I've been interested for some years in watching the case of Krebiozen, a cancer-treatment that has been vigorously attacked by the AMA - as have all cancer treatments other than their accepted, standard procedures of radiation, surgery, and caustics. There's been a running battle for years between the doctors who have used the stuff and believe they have clear evidence it works, and the AMA people who have not used it and insist it doesn't work.
For a long time, the Krebiozen faction refused, or claimed to be unable to supply a purified sample of their material for AMA-sponsored analysis; they demanded that the AMA make what amounts to a biological assay test - i. e., run a standard double-blind test of the effectiveness of the remedy in actual cancer cases. In a double-blind test, neither the doctor nor the patient knows which individuals are getting the test-drug, and which are getting blank solutions; only a central computer has the number-correlations that finally match results with identification. This is the one type of test that assures that subjective factors will not influence either the patient's reactions, nor the doctor's evaluations.
The new drug laws - resulting in part from the thalidomide furore - finally made it possible for the AMA, working through the Federal Drug Administration, to force the Krebiozen faction to supply a concentrate of their material for chemical analysis.
Chemical analysis has many powerful tools these days; infrared spectroscopy shortly permitted identification. The infrared spectrum of the Krebiozen sample was shown to match, one-to-one, the infrared spectrum of a well-known protein component of muscle tissue - creatine. The effects of creatine being well known - none whatever - it was at once clear that Krebiozen could not have any useful effects, as the AMA had long maintained.
There is, it seems to me, just one slight hitch in that simple-minded conclusion.
I am prepared to supply an extremely effective herbicide which I can positively guarantee can be shown by the most careful chemical analysis to consist of extremely pure water. "Conductivity water" in fact - water so extremely purified that it does not conduct electricity. There will be less than 0.001 parts per million of impurity in it. And chemical analytical techniques haven't even started to get good enough to reach that level of analysis!
Which seems sort of contradictory, in view of the fact that certain impurities in water, present in concentration of about 10^-15 are legally defined as making the water "impure". And I am not talking about radioactives, either - perfectly ordinary chemical compounds of stable elements.
My herbicide belongs to the same type-class; a concentration of 10^-18 or so would be quite adequate to ruin a field of growing plants. Readily proven by biological assay... but some ten orders of magnitude beyond the reach of chemical analysis!

How? Well, starting with conductivity water, I need add only a very minute trace of a known crystalline material - tobacco mosaic virus. The resultant solution, sprayed on young tobacco plants, could do quite a job, couldn't it?
And the legal definition of impure water has to do with the permissible concentration of bacteria coli in the water. Anyone want to try to spot that quantity of complex protein by chemical analysis?
So Krebiozen contains creatine? Well, well... And what else does it contain? Probably some hydrogen oxide too, I betcha. Since it's extracted from horse serum, it's quite apt to contain a couple of oddments of metabolic processes. Horses being noted for their quantity of muscle tissue, the presence of a muscle tissue extract of no significance isn't too startling, really. And since the mighty powers of modern chemical analysis can't find anything else present, that proves that creatine is all that's there.
Look friends... I have a bottle of a nice clear solution that should improve the situation for chemists who think like that. They're free to analyze it to their hearts content... if only they'll drink it after they've "proven" it has nothing of any significance in it. Let's see now... we could load it with botulinus spores... or concentrated polio virus... or even anthrax spores; then we could let him boil it for an hour before swallowing it, and still the damn fool would have personal experience with the fact that the limits of his knowledge and talents are not the limits of reality.
I have no personal interest in Krebiozen; I do have an acute personal, as well as citizenship interest in the honesty of thinking of science and medicine.
Anyone in a field of medical-biological work who considers, even temporarily, that chemical analysis is adequate for the determination of an unknown remedy is inexcusably incompetent, dishonest, or muddleheaded beyond toleration.
To consider that spectrum identification is an adequate tool for such work is even further in the direction of fantastic - appalling! - irresponsibility. Obviously spectrum identification requires that the spectrum be already known. This is a way to analyze an unknown material?
One microgram of cobaltamine - vitamin B-12 - is considered a reasonably adequate vitamin supplement for that material. If an adult consumes one kilogram of food, plus an additional kilogram of water in a day, the concentration of that vital cobaltamine is 0.005 Parts Per million in his diet. You will, maybe, find this with an infrared spectrum?
Cancer, it is currently believed, arises from cells whose growth-regulating mechanisms have gone wrong - somehow the DNA-RNA information has been altered disastrously.

It is currently believed that virus particles are more or less loose DNA fragments, wrapped up in a protein capsule. It's quite conceivable that the ultimate answer to cancer would be a highly specific virus that contained precisely the DNA codons that the cancer cells lacked - and attacked them, without, of course, bothering normal cells at all (one class of thing you cannot learn is that which you already know. The cells that already have the codon information on self-restraint wouldn't "learn" anything from such a virus; the cancer cells would).
If someone prepared such a virus, and submitted a highly active and adequate sample to the present AMA-FDA groups, it's evident that it would be reported as "fully identified as a solution of sodium chloride in water". Virus particles in normal saline solution, concentrated to about 10^-12 would be an extremely powerful solution... biologically!
Remember that it is inevitably necessary that I use analogies which you, the reader, already know the answers to. If I handed you an example, instead of an analogy, it would of course be meaningless! I can, for instance, hand you an example-for-a-1935-scientist, but it will be an analogy for you. Suppose I could borrow a time-transistor somehow, and slip back to the Bell Telephone Laboratories, in 1935, and hand them a collection of a few modern solid-state devices. Say a silicon diode rectifier, half the size of a golf ball, rated at 150 amperes, 400 volts peak inverse (that would drive the power-supply division into a flying frenzy!). And a grown-crystal audio-amplifier unit, with grown-in solid-state resistors, capacitors and transistors, half the size of a pea. A simple little semi-microscopic germanium diode detector, too, and perhaps be very generous and supply a couple of the new silicon carbide lasers ("You put in the juice here and here, and the coherent beam of blue light comes out here'').
Let us now stand back and watch the chemists try analyzing that stuff. That silicon rectifier, now... they'll find it's a single crystal of pure silicon. They haven't got a technique good enough to come close to guessing how pure (none of their reagents, or the water they're using to analyze it, are pure enough anyway; the techniques for getting commercially usable quantities of conductivity water weren't developed until transistor work forced them into use). And since they can't come close to the purity of the silicon, they can't possibly detect the doping impurity that makes it work. They won't do it with a spectroscope, either - partly because they don't know how to get a spectroscope clean enough to do any good! The "background noise" of contaminants in their reagents, their equipment, and the atmosphere they work in would conceal the doping impurity.
A large part of the work the Bell Labs did in the years after they did invent the transistor was concerned with developing techniques for getting clean tools, clean reagents, clean equipment, which made possible the modern transistor. It wasn't just the concept of the transistor they developed; Bell Labs had to develop a whole new chemical and industrial concept to make production possible.
"Zone refining'' was one of those - a technique whereby already ultra-purified germanium, silicon, or other material could be super-ultra-refined.

Back around 1940, the people working with copper oxide rectifiers at Bell Labs and other electronics industry research centers, knew that copper from a certain area in Chile made the best copper oxide rectifiers. Montana copper wasn't as good, nor was African or Mexican.
Yeah - sure - everybody knows that copper is an element, and an element is an element, and where it comes from has nothing to do with the matter. And in the days before knowledge of the "doping" behavior of semiconductive materials was available, who knew that a small-fractional-part-per-million impurity could make a huge difference?
The labs had tried analyzing the Chilean copper; they were perfectly sure that there was some impurity present. But no technique known to science as of 1940 was able to detect it.
After the information was of no practical significance - copper oxide rectifiers having been entirely displaced by silicon and germanium diodes - techniques developed in transistor research made it easy to determine the impurity. Zone refining, for instance, can sweep all the impurities on a bar of germanium - or copper - down to one end, thus concentrating them neatly for analysis. But by then, of course, it had become a completely academic question...
However, if someone says something like that business about only Chilean copper being good for the device he's invented... he almost certainly wins himself, right there, the "Strictly Crackpot!" label.
On the other hand, when the AMA and FDA proudly announce that they have completely identified a previously unknown remedy because they've identified the infrared spectrum of one component... that, sir, is Science at Work!
Pardon me while I go back to Magic.
The magicians used to try something out before they decided whether it worked or not.

January 1964

"Creatine has no known physiological effects"

The FDA's recent blast at Krebiozen as being "nothing but creatine" is of interest in view of the following item - a letter which appeared in Science (17 November, 1964, p. 1113).

Pre-1962 Creatine sought.

We are seeking help in finding 50 grams of Eastman (Distillation Products Corporation) creatine, catalog No. 951, purchased prior to 1962. We have been informed that the source of the creatine sold by Eastman has changed, and we find a different physiological response from that formerly found in rats fed creatine. Eastman has been unable to locate a supply of the earlier product for us.

W. R. Todd, Department of Biochemistry, University of Oregon Medical School, Portland.

Whether there has been an impurity in the old which is not in the new, or vice versa, seems to be undetermined at this point.
But of course, both lots were sold as being "nothing but creatine".
The rats appear to have considered otherwise.

September 1965