Martin E. Hellman
Work on Cryptography

I trace my interest in cryptography to three main sources:

• David Kahn's best-selling book The Codebreakers, which appeared in 1967. Kahn was also the banquet speaker at my first IEEE International Symposium on Information Theory, in January 1969.
• From 1968-1969 I worked at IBM's Watson Research Center in Yorktown Heights, NY. One of my colleagues was Horst Feistel, who had been brought in from classified government work to seed IBM's research in cryptography. IBM's work culminated in the Data Encryption Standard (DES) in 1975. While my work at IBM was not in cryptography, I had a number of discussions with Feistel that opened my eyes to previously unforeseen possibilities. The fact that IBM was investing in the development of cryptography for commercial applications also indicated the need and value of such work.
• In 1970, while I was an Assistant Professor at MIT, Peter Elias gave me a copy of Claude Shannon's little-known 1949 paper that related cryptography and information theory, my chosen area of research. I saw that information theory owed much of its birth to war-time, classified research on cryptography that Shannon had done at Bell Labs, and that much of what I had learned in my doctoral studies was directly applicable to cryptography. Maybe I could make a contribution after all.

• My ideas were embryonic and not likely to be judged worthy of financial support.
• In an area that was normally heavily classified, I feared that government support would be problematic at best, and impossible at worst.

In the early 70's, whenever I mentioned my budding interest in cryptography to colleagues, the almost universal response was that I was foolish to even consider the area. The concerns were two-fold. First, with NSA (the National Security Agency, America's primary cryptographic entity) having a multi-billion dollar budget, how could I hope to discover something that was not already known? Second, if I did anything good, they would classify it. I countered the first accusation of foolishness by noting that it didn't matter what was known in the classified literature since

• that knowledge was not available to meet the growing commercial needs for encryption, and
• credit went to the first to publish, not the first to discover and keep the work secret.

I didn't have a good answer to my colleagues' second concern, about my work possibly being suppressed by the government, and figured I would cross that bridge when I came to it. But I also have come to see that my counter-arguments were less logical than I thought at the time. As I noted in the discussion of my Work on War and Peace, the muse of the fools seems to whisper in my ear frequently or I seem to pay her more attention than most. In hindsight, I can see that I was drawn to cryptography in a way that defies rational explanation, and that my seemingly rational arguments were just rationalizations for what I was going to do, whether it made sense or not. My colleagues arguments were based on valid concerns, both of which reared their heads.

As Whit Diffie, Ralph Merkle and I, followed by Ron Rivest, Adi Shamir and Len Adleman (RSA) at MIT, and then others began to publish papers that NSA would have classified top secret, various forces within the government, and especially the intelligence community, agitated for such papers to be "born classified", even though we had developed our ideas without any benefit of the classified literature. There was even a threat that we could be prosecuted under the International Traffic in Arms Regulations (ITAR), since technical literature on cryptography was considered an implement of war by the ITAR.

On the advice of Stanford's general counsel, I even presented two papers at a 1977 symposium at Cornell University, instead of my usual practice of having the student co-authors do the presentations. The attorney told me that if the ITAR were interpreted broadly enough to include our papers, he believed they were unconstitutional. But a court case could drag on for years, severely hindering a new Ph.D.'s career (especially if the attorney's belief was not shared by the jury), whereas I was already a tenured professor.

I presented these thoughts to Ralph Merkle and Steve Pohlig, the students in question, but left the final decision to them. Initially they wanted to take the risk and give the papers, but eventually concern from their parents won out. Fortunately, the presentations went off without incident, though it was dramatic having Ralph and Steve stand mute by the podium, so they would get the recognition they deserved, as I gave the papers.

My colleagues' second concern did not rear its head until December 1997, when GCHQ, the British equivalent of NSA, issued a statement claiming that they had invented public key cryptography several years before we did. To my surprise, some people who should have known better, deviated from the well established rule that credit goes to the first to publish. For example, the journal Cryptologia published GCHQ's claim without any disclaimer to this effect and without noting that there was no way to verify the claim. And a colleague at another university broadcast an email giving the URL for the GCHQ claim and encouraging people to look there for what he called "the true story on the invention of public key cryptography." The researchers at GCHQ and NSA who toil in secret to protect our freedoms deserve our thanks. But it is more difficult to decide what credit they deserve in this instance. For what I believe is a reasoned approach to GCHQ's claim, I recommend Bruce Schneier's May 1998 newsletter.

For a more detailed discussion of some of the above topics, as well as many others, I recommend Simon Singh's Codebook and Steven Levy's Crypto. While David Kahn's The Codebreakers was written before public key cryptography was invented, it is an excellent history of cryptography prior to those events. All three books were written for popular audiences and do not require advanced mathematics.

Return to home page.