Saturday, July 9, 2016

Book review – ‘Code Warriors: NSA's Codebreakers and the Secret Intelligence War Against the Soviet Union’

Stephen Budiansky, author of Battle of Wits: The Complete Story of Codebreaking in World War IIand Blackett's War: The Men Who Defeated the Nazi U-Boats and Brought Science to the Art of Warfare has published a new book, this time dealing with the Cold War operations of the NSA and the efforts to solve Soviet high level cryptosystems.

Code Warriors: NSA's Codebreakers and the Secret Intelligence War Against the Soviet Union’ is a history of the National Security Agency with an emphasis on the work done on Soviet cryptosystems.


The book starts in 1943, when US codebreakers were solving several important Axis cryptosystems (such as the German Enigma and the Japanese PURPLE cipher machines). At the time Soviet diplomatic traffic was being collected but it was only sorted not actively attacked. During the year a small group was formed to study this material and make an attempt at solution.

In theory the Soviet authorities used codebooks enciphered with one time pad which meant that their messages should have been unbreakable. However the ‘Russian problem’ group was able to make a stunning discovery. It turned out that there were cases of additive pad reuse, which meant that some messages could be decoded.

This was the start of the famous VENONA project and thanks to the decoded messages of the Soviet intelligence agencies it was possible to identify a large number of Soviet agents and communist sympathizers.

NSA and GCHQ continued to solve important Soviet systems in the period 1945-48, such as the cipher machines Coleridge, Longfellow, Pagoda and a modified version of the Hagelin B-211. They were also able to intercept the Soviet civilian network thanks to German equipment, captured in the last days of the war.

Operations came to a standstill in 1948 when after being warned by one of their agents the Soviets introduced new secure cipher procedures. From then on NSA would continue its efforts against Soviet high level cryptosystems but with little to no success and this despite devoting most of its resources to the Soviet problem.

The author looks into the efforts of the NSA to solve Soviet high level cryptosystems, the investments in new technologies such as high speed computers, the crisis resulting from repeated failures and the huge resources devoted to the Soviet problem (at the expense of other targets). In the end the failure to solve Soviet ciphers using the ‘standard’ methods meant that more resources had to be directed to ELINT satellites, ‘bugs’ and traffic analysis. It was only in the late 1970’s that a combination of new supercomputers (built by the Cray corporation) and mathematical research (from the Institute for Defense Analyses) that allowed the NSA to solve Soviet high level ciphers.

Overall the book covers NSA operations from WWII till the end of the Cold War and looks into all aspects of the agency’s work, their codebreaking successes, the relationship with the CIA, their investment into high speed computers, operations in Korea and Vietnam and even the organizational and security problems of running an organization of such size.

There are also five appendixes with short explanations of enciphered codebooks, the Soviet cipher teleprinter (from TICOM sources), cryptanalysis of the Hagelin machines, Turing’s deciban method and Friedman’s Index of Coincidence.

Considering the information presented in the book it is a valuable contribution to Cold War cryptologic history.


The author was kind enough to answer some of my questions. 

1). You’ve written several books on signals intelligence and codebreaking. How did you become interested in this subject and how did you go from writing ‘Battle of Wits: The Complete Story of Codebreaking in World War II’ to ‘Blackett's War: The Men Who Defeated the Nazi U-Boats and Brought Science to the Art of Warfare’ and now ‘Code Warriors: NSA's Codebreakers and the Secret Intelligence War Against the Soviet Union’?

I think what first drew me to the subject, and what has been the common theme of all of these books of mine that you mention, is my abiding interest in the intersection of science and warfare. Since World War II, science has dominated warfare, and it’s simply impossible to understand modern military history without the technical story of scientific developments that have shaped conflicts, weapons, and strategy and tactics. It also involves an often fascinating intersection of very different cultures and personalities.

2). There are only a few books that cover the Cold War operations of the NSA. What new information have you discovered for your latest book that set it apart from previous efforts? How hard is it to research cryptologic history compared to social, economic or standard military history? 

NSA has been regularly releasing and declassifying documents from the post-World War II period. It’s admittedly slim pickings, and a very frustrating process. I did file several Mandatory Declassification Review requests with NSA in the course of my research and actually received several of the important reports I was seeking to have declassified—only to find that NSA’s declassifiers had redacted so much from the documents that they needn’t have bothered even pretending they were releasing anything: so much was chopped out that there was literally nothing of significant historical information left by the time they were done. 

And we’re talking about information that is as much as 70 years old, involving for example Soviet rotor cipher machines from 1947. You can buy a Soviet “Fialka” machine on the collectors’ market these days, but NSA still refuses to acknowledge that such a thing even exists, much less anything about its cryptanalysis.

That said, there are a few significant things that have come out recently, notably in the William Friedman Papers released (or partially released I should say — these too were subjected to the usual heavy-handed redaction censorship) last year.

But the main thing I tried to do which I think is different from earlier books was to pull the clues from disparate sources together, provide essential context, and try as much as possible to synthesize the technical story of cryptanalysis in the Cold War, as I did for World War II in my earlier book “Battle of Wits.” There’s a lot that one can figure out from context, correlating sources, and applying a basic knowledge of cryptology

3). During the Cold War the NSA’s budget and manpower rivaled those of the largest companies in the world. Do you think that this investment paid off for the US government? How can one evaluate the operations of an agency that works in secret?

I’ve repeatedly argued that NSA would be much better off if they were more open. It’s very difficult to get them to reveal their successes, and the result is exactly the problem you note: Why should the American public continue to support these agencies and their activities if they can’t know what they’re getting for the investment? The public tends to hear about NSA when there’s a failure or a scandal. 

That said, I do think NSA’s greatest success in the Cold War was preventing us all from being blown to bits in World War III. That’s a negative argument which you can't prove of course. But until the advent of real-time photoreconnaissance satellites in the 1970s and 80s, SIGINT was the primary source of early warning of Soviet military activity and in particular was the only real source of strategic warning of Soviet preparations that would precede a nuclear attack. The reassurance that NSA’s surveillance gave US leaders that the Soviets could not launch a first strike without us having significant warning greatly reduced the hair trigger of the Cold War nuclear standoff. And we specifically know, as I note in the book, that during some key crises in the Cold War—such as the Suez Crisis in 1956—the information NSA provided was crucial in convincing US leaders that Soviet threats of military intervention were a bluff, not backed up by any actual movement or mobilization of its forces, which greatly helped to defuse those crises. It’s not hard to imagine an escalation that could quickly have gotten out of hand had we been in the dark and left to guess what the Soviets were up to.

4). What are your thoughts on the recent Snowden revelations regarding the NSA interception of US civilian communications?

I think they show NSA to be very much a creature of its history. The problems that the Snowden revelations point to regarding NSA’s efforts to “get everything,” to exaggerate the effectiveness of its bulk collection activities, its willingness to press a maximal and at times highly dubious view of its legal authorities, and to misrepresent the truth when confronted with embarrassing facts, are strikingly similar to the mindset and institutional culture of the agency (and its predecessors) going back as early as World War II.

5). What areas of intelligence history do you find most interesting and what topics do you plan to research for future books?

I feel in a way that intelligence history is at a real crisis point. We’ve beaten to death World War II—I mean how many books about breaking the Enigma or Operation Zig Zag do we need? — and the completely broken system of official declassification of post–World War II documents has left intelligence historians with precious little to work on. Until some fundamental change occurs—and it has to come from the top, because the entire FOIA and declassification process is the *problem*, not the solution—I think I’m going to head for another field entirely, if I decide to write another book. I’m a bit worn out from reading documents with every other word crossed out!

4 comments:

  1. Great interview - what do you say Christos as a fellow ww2 researcher asking for more declass of dox?

    ReplyDelete
    Replies
    1. The NSA definitely takes too long to release documents. I don’t understand why they keep reviewing documents from the 1940’s. It’s a waste of their resources. That material should be released without review.

      Delete
  2. Q&A with Stephen Budiansky


    And we’re talking about information that is as much as 70 years old, involving for example Soviet rotor cipher machines from 1947. You can buy a Soviet “Fialka” machine on the collectors’ market these days, but NSA still refuses to acknowledge that such a thing even exists, much less anything about its cryptanalysis.

    Es kann keine Fialka sein, diese wurde anfang der 1960 in Serie produziert.
    Ab 1967 in der NVA eingesetzt. Bzw. Warschauer Vertrag.
    NSA and GCHQ continued to solve important Soviet systems in the period 1945-48, such as the cipher machines Coleridge, Longfellow, Pagoda and a modified version of the Hagelin B-211. They were also able to intercept the Soviet civilian network thanks to German equipment, captured in the last days of the war.

    Wie waren die russischen Namen der Chiffriermaschinen?

    ReplyDelete
    Replies
    1. 'Wie waren die russischen Namen der Chiffriermaschinen?'

      The Soviet B-211 was called K-37 - Crystal (Кристалл). For the other cipher machines i don't know the designations.

      Delete