Restricted Data

William A. Shurcliff is one of my favorite Manhattan Project dramatis personae. I’ve written about him before on here, some time back. In a nutshell, Shurcliff was a physicist who worked as a technical advisor to Vannevar Bush in the Office of Scientific Research and Development, and was connected to the bomb project only peripherally. In fact, his value to Bush was that he wasn’t really steeped in the work to make the bomb: he was a trusted, technically-competent outsider. So he was the person they called, for example, when they needed a censor for atomic patents, because he could be “read in” on the secrets but wasn’t otherwise in a position to have conflicted interests. Among his other roles on the bomb project was to be the copyeditor of the Smyth Report, and he later was the “official historian” for Operation Crossroads.

A painting of William A. Shurcliff from 1948 by his father-in-law, the American artist Charles Hopkinson.

What I love about the Shurcliff one finds in the Manhattan Project files is that he shows up in the most unusual, unsought places, and he loved to write unsolicited memos. I imagine him sitting around, thinking about some core problem related to the social and political future of atomic energy, and writing his thoughts out in a methodical fashion and sending them to the top. Occasionally there is evidence that these memos were read and circulated, though none were ever obviously used as the basis of policy going forward. Still, what’s really wonderful about someone like Shurcliff is that he wasn’t being exposed to all of the other scientists on the project, so he had a relatively independent outlook. This makes him a nice “barometer” for what kinds of thoughts were thinkable at the time, outside of the standard range of positions that the scientists took on the issues in front of them.

One of the issues that Shurcliff chimed in on was the prospects of long-term scientific secrecy. Late in the project (i.e. late 1944 and early 1945), the scientists at the University of Chicago had largely finished up their portion of the work (helping getting the Hanford reactors designed and running), and had more extra time for contemplation of long-term issues than those who were at Los Alamos. So they did things like write the Franck Report and other studies into the long-term prospects of nuclear energy, secrecy, the use of the bomb, and so on. A repeating theme in all of these reports is that long-term, postwar nuclear secrecy would not work. It is a position you will be familiar with from discussions today: secrecy would not prevent foreign nations (or “enemies” more broadly) from getting the bomb,  it would inhibit and slow future American work, and the worst thing imaginable would be a “secret arms race” between nations.

Vannevar Bush and James Conant, despite being key people behind the secrecy procedures of the Manhattan Project (which started well before the Army got involved), thoroughly embraced the anti-secrecy line. As Bush put it to President Truman in September 1945: “A secret race on atomic bombs can lead to a very unhappy world.”  In fact, almost every discussion I’ve found of postwar secrecy made during the Manhattan Project takes more or less this sort of position.

Shurcliff, however, approached it differently. I’m not sure how he picked up that these thing were “in the air,” though he was in limited doses exposed to the Chicago scientists while doing his patent work. In December 1944, he wrote a seven-page memo to Richard Tolman, another OSRD scientist who worked as a personal technical advisor to General Groves (among other things), with the lengthy subject heading of: “Analysis of the theses: (A) Maintaining secrecy on the details of the present weapon will not insure security. (B) Secrecy will come from ‘keeping ahead.'”¹

Click on the image above to view the full memo. Shurcliff’s memo was itself classified “Secret — Limited” which basically meant that only the very top-top level of administrators and advisors were allowed to read it. The “Top Secret” classification was only just starting to be used in this period, and probably only would have been used here if the memo had any insight onto when the United States would have a bomb ready to use.

Keeping to his form, Shurcliff’s memo is highly-structured and carefully argued. He starts it off with a statement of his motivations and his conclusions:

Explanation: Some analysis of these theses appears called for since they lie at the heart of the general secrecy policy which, in turn, is fundamental to the entire postwar policy. These theses have been endorsed by many persons heard by the [Interim] Committee.² The writer knows of no one-who has disagreed with these theses.

Conclusions: While it can be said that the theses are “more true than false,” it is apparent that they are seriously inadequate and to an appreciable extent misleading, since:

With regard to Thesis A, maintaining secrecy will make for security for a good many years at least — especially with respect to the many smaller countries incapable of developing nucleonics weapons independently.³ To place one’s faith in secrecy may be rash, but appreciably to dispense with secrecy may be even more rash.

With regard to Thesis B, even “keeping ahead” may prove futile when even “obsolete” nucleonics weapons can be employed by an enemy to wipe out our major centers, including nucleonics centers, in a single hour before declaration of war.

If you’re a regular reader of this blog, you’re probably recoiling from Shurcliff’s pro-secrecy arguments. They are pretty far distant from the “there is no secret” mantra of the postwar atomic scientists, but they are not bad arguments. Shurcliff’s approach is eminently pragmatic, not ideological. His memo is one about  technology transfer between nations, with an eye beyond seeing things as just a competition between two powers. Of course, he says, you can’t maintain such secrets indefinitely. But if you can maintain them for a few decades, that’s not nothing — time is a valuable commodity. 

Shurcliff also augments his analysis with the practical experience of technical espionage. Shurcliff’s main job at the OSRD was to be a liaison with other branches regarding information seized about enemy technology. So if the Allied soldiers found reports about, say, German tanks, they would send them to Shurcliff, and he’d figure out which of the OSRD divisions could make the most use of it. So unlike the scientists at Chicago, he actually knew a little bit about how difficult it was to construct technology based solely on knowledge alone:

Parenthetical note: The writer recalls many instances during 1943 and 1944 where, despite a wealth of fragmentary information from cooperative enemy prisoners, neutrals, and allied agents, the really significant technical engineering data on enemy devices remained wanting until uncomfortably late dates. Examples are: (a) technical characteristics of German infra-red search receivers and image tubes; (b) control frequencies for the German HS-293 glider bomb; (c) launching means, fuels, and radio control means of the German V-1 flying bomb. In-all these cases the serious gaps in our knowledge were not filled until reasonably intact specimens of the weapons in question had been captured. The abundance of such situations is believed to show that there is a good chance that appreciable amounts of highly-technical engineering data on secret devices may be kept out of enemy hands for years — perhaps decades.

Shurcliff’s estimates on the possibility of real espionage were, in the end, more optimistic than the reality. Neither he nor anyone else suspected that Los Alamos was full of a number of relatively high-level spies, and that direct design information on the bombs would be so immediately and thoroughly compromised. But it is worth noting that Shurcliff’s above discussion about the difficulty of reconstructing a physical technology from design information alone is, in fact, shown to be reasonably on the mark when we look at the history of the Soviet program. Even though the Soviets did have very detailed design information on the atomic bomb, it still took a tremendous effort to turn that into an actual bomb, and it has become much more clear over the years that information was not the primary determinant of when the Soviets developed their first nuclear weapons.

William Shurcliff, 50 years later.

Lastly, Shurcliff’s views on “staying ahead” feel remarkably relevant to our modern day, as well. Nukes, he argues, are not weapons were there is such a significant difference between the “best” and the “second-best.” Getting hit with an “obsolete” weapon is still going to be a disastrous thing. Does it matter that the North Korean’s largest test was 10 kilotons, whereas the largest bomb in the US arsenal is megaton-range? To most people, probably not — 10 kilotons will still ruin your day.

Shurcliff ends his memo with a set of “Concluding Remarks”:

We are entering an age (starting, say, in 1960) in which even inferior arms (e.g., 1950 nucleonics bombs) any be used suddenly to cripple and perhaps conquer the most advanced country. The coming age may be further characterised (in the following over-simplified and over-dramatic terms:!) thus:

An age in which surprise aggression can laugh at military defense;
An age in which nucleonics is the grand currency of military negotiations;
An age in which our scientists will no longer be able to contribute to the defense of the country;
An age in which international physical compulsion is possible, but in which international physical conflict is impossible;
An age in which international conflicts can only be moral conflicts;
An age in which the line separating international disagreement between two countries from sudden devastation of one of them may become vanishingly thin;
An age in which “balance of power” and “threat” are merely historical terms.

If the last war was a chemists’ war and  the present war is a physicists’  war, the next war may be an “administrator’s war” — a war whose outcome may be determined by the mere formulation and concealment of the administrative decision as to whether and when to strike.

What a conclusion!

So what became of Shurcliff’s analysis? He sent it to Tolman, who forwarded it to Bush, and Bush in turn forwarded it to Harvey H. Bundy, an assistant to Secretary of War Stimson (and father of McGeorge Bundy), with the following note attached:

Here is the pessimistic viewpoint, and I think you ought to read it. I would add 1) while scientific interchange is inevitable, transmission of details of weapons is not. 2) A sudden strike will not prevent a riposte, if stores of weapons are well protected underground. The case as between two nations with hidden and ample supplies is of most interest, as it will be the case probably, and is not here treated.

I doubt Shurcliff ever knew that his memo had been forwarded up the chain like this — the secrecy, ironically, meant that he rarely had any indication of what was going on other than his own little corner of things. And perhaps even more ironically, that never kept him from speculating and dreaming about the possibilities of the future.

I don’t think anything more came of his memo. But I do treasure it, not because I necessarily agree with it — though I do find it better rooted in the realities of technology and epistemology than many of the statements of the anti-secrecy scientists of the time — but because it is a little indication of the fact that there were some nuclear physicists in 1944 who could find ways to defend secrecy (a rare thing!), and also find ways to see, arguably with some clarity, the shape of things to come.

1. William A. Shurcliff to Richard C. Tolman (8 December 1944), Harrison-Bundy Files Relating to the Development of the Atomic Bomb, 1942-1946, microfilm publication M1108 (Washington, D.C.: National Archives and Records Administration, 1980), Roll 6, Target 4, Folder 75, “Interim Committee — Publicity.” [↩]
2. The Interim Committee was the main administrative body planning for what to do once the bomb was a matter of public record, i.e. after it had been used on Japan. [↩]
3. “Nucleonics” was at this time being floated as a new name for the entire field of nuclear technology, in analogy to “electronics.” It didn’t take off. [↩]