The futility of scientific secrecy

"We are, I rather assume, going to have a whole series of crises as a result of increasing scientific knowledge that is adaptable to blowing the hell out of the world."  -- David Lilienthal, chairman of AEC, September 1945.

So why do government actions tend to spawn the opposite result of what was intended? We have already seen how forcing down interest rates may have raised unemployment (rather than lowering it), just as raising interest rates three decades ago had the counter-intuitive effect of reducing unemployment.

We now find that banning texting while driving results in a slight increase in accident rates, probably because texting drivers have to hold the their device below normal sightlines and scan for police, in addition to the task of driving.

The international community (by and large) seems determined to keep Iran from developing the technology required to build a nuclear weapon. At first glance this seems a laudable task. But how effective can it be? Not long ago Shimon Peres gave a speech on the futility of restricting scientific advancement. Knowledge can be disseminated by too many routes.

In 1945 American scientists faced a very similar situation. America had successfully tested and then used nuclear weapons against Japan, and while the scientists involved in the Manhattan Project were not permitted to disseminate the knowledge of what they had done to the world, they recognized that muzzling free scientific communication was doomed to failure.

This story is conveyed very nicely in the May 2012 edition of Physics Today, in this article, which is available free of charge.

A group of scientists, who had not been involved in the atomic bomb project, set themselves about duplicating the procedure, and by 1946, had basically succeeded. They attempted to publish their findings in a book. And that's when the trouble began. Among the things the scientists had done is conjectured possible methods of triggering nuclear detonation--which was the idea the AEC most urgently wished to keep secret. The AEC wanted to censor the book--but there was a problem. If they pointed out what they wanted removed, that would send a clear signal to the authors that their conjectures were probably correct. Furthermore, since these scientists had already given a series of public lectures, anyone who had attended the lectures could read the book and deduce which information the AEC viewed as most sensitive. So there really was no way to prevent the most sensitive information on nuclear weapons from being disseminated.

Eventually the AEC did force some material to be censored (the censored material has thoughtfully been made available here as a pdf).

(As an aside, here is my solution--let's say there are 40 ideas in the book, and six of them are sensitive. Toss a coin, or use some other random method to censor, say ten of the ideas, and allow the rest to pass, even if they turn out to be sensitive material.)

Nevertheless, every so often some physics graduate student would read the material that had been published, and from that deduce how to build a nuclear bomb. Assuming that Iranian scientists are at least as smart as American graduate students, they must already have the knowledge to build a bomb. Assassinating nuclear scientists is futile, and merely establishes a precedent for assassinating scientists engaged in scientific endeavours that might be inconvenient to your country. When American (or, perhaps, Israeli) scientists start dying mysteriously we will know that some form of international parity has been realized.