Kim Philby and a web of trust

Show me your Kim Philby and I’ll show you your concerns. Ever since his defection, British writing on the iconic spy has always reflected the anxieties of society at the time of writing, modulated by the latest lot of declassified documents. What else could it do, faced with such a character, a man who couldn’t have been more protean or more expert at letting his interlocutor project their own thinking on to him? In the 1960s, his role as the ultimate Etonian predominated. In the 1970s, sexuality took over. In the 1980s, to go with the revival of the cold war, it was the communism again. But what would it be now?

Anyone writing about Philby is standing on the shoulders of giants. The literature is enormous, and has regularly been transformed by the disclosure of more information. However, this viewpoint often mostly serves to show that the giants are skew-whiff on their pedestals.

Ben Macintyre’s A Spy Among Friends: Kim Philby and the Great Betrayal is the latest to have a go. He decides to frame the book around the notion of friendship and the relationship between Philby and his friend, defender within SIS, boss, agent handler, and interrogator, Nicholas Elliott. As a result, the book is in large measure a group biography, always a neat trick.

I’m not sure how well the friendship element of the book works, except to highlight that Elliott was almost comically decent to Philby and everyone else involved, visiting Aileen Philby daily in the psychiatric hospital while Kim, self-absorbed as always, barely seems to have noticed she was ill. But that’s OK. Macintyre brings up whole chunks of new story.

One especially rich chunk is the women. A Spy among Friends is among other things a succession of incredible women, like the chief of staff in SOE Section D who was also the chief organising officer of the Conservative Party – surely we need a brief life at least – and the Sunday Express war correspondent who recommended Philby to her, or the MI5 investigator Jane Archer whose abilities Philby rated so highly he insisted on getting her transferred to his SIS counterintelligence group to make sure she didn’t by chance get assigned to his case. Marks & Spencer executive Flora Solomon appears at a succession of critical junctures, helping him launch his career, politely declining the opportunity to join him in the KGB, and eventually providing crucial evidence against him – in part, out of anger at his appalling treatment of his wife, a close friend of hers. (She was also furious about his journalism, specifically his criticism of Israel.)

Peter Wright, by contrast, mentioned Litzi Philby/Kohlmann in passing and hardly says anything about any of the other women in the story, not even his colleague Archer. Neither does The Culture of Treason although it’s a long time since I read it. There’s your answer: the highly un-gay Philby was remembered as the Gay Traitor, not out of homophobia, but out of sexism. This shift of focus tends to diminish Philby as a character, to pigeonhole him as a posh chancer who left a trail of broken wives and unpaid bills behind him, dumping his self-inflicted chaos on – who else? – his friends. But this is a welcome cold bucket of realism.

Another important point is the politics of professionalism. The critics of SIS tended to make much of its habit of recruiting through the old-boy network and of operating in a highly informal fashion. They used the words professional and modern a lot and affected classlessness. But even as they were saying their piece, Michael Young was already writing the strongest possible rebuttal, The Rise of the Meritocracy.

In many ways, the post-defector recriminations were a debate about how the access to elite status would be regulated in future. On one side, the traditionalists believed that the best solution was a diffuse network of contacts who themselves could be trusted and a specific culture, as hard to define as it was unmistakable. On the other side, the modernisers (and I use the word deliberately) believed the best solution was a graduate profession with defined career progression, credentials, and tests.

Importantly – and here’s where Michael Young comes in – this doesn’t for a moment mean that the same sort of people wouldn’t be recruited. If the modalities of Nicholas Elliott’s recruitment couldn’t have been more stereotyped (a word put in by his old housemaster, an approach at the races), it was also true that he was an Etonian with a good degree, someone whose chances of passing through a credentials-based, supposedly modern and prejudice-free, system would have been between excellent and certain. The same was of course true of Philby, whose academic qualifications could not have been more stellar.

Another way of looking at this is from an information security perspective. The traditionalists didn’t have the vocabulary, and public-key cryptography hadn’t been invented yet, but they were essentially a web-of-trust. Trusted individuals vouched for others, whose trustworthiness they initially evaluated through diffuse markers of subculture. This had the advantage that it was much, much harder to fake being the right sort of chap than it is to bang on the effort, get the credential, and keep your nose clean. Any fool can fake conformity, but it takes a genius to fake weirdness convincingly. It was even harder again to gain the connections into the web of trust required.

(Macintyre is good on eccentricity as strength; he needs to be, because the book is just full of prize weirdos. Also, there is a hell of a lot of drinking, hardcore, vase full of Martinis, whisky before breakfast, falling out of windows, eight pints and gin chasers the night before a zero-dark-thirty call for a dangerous mission, drinking. Arguably, had anyone worried about the booze and drugs, they might have rumbled the entire spy ring in the early 1940s – but hardly anyone was remotely sober. It’s the 1950s as seen in this classic Jamie Kenny post Oh yes, and did you know Commander “Buster” Crabbe RN was in the habit of wearing his diving suit to make love? Apparently his marriage was falling apart at the time of his sticky end in Portsmouth harbour, and the suit wasn’t helping.)

The modernisers wanted something more like a public-key infrastructure. A succession of tests administered by a central source of truth would lead to the issue of credentials, which could of course be revoked.

As it turned out, the Cambridge spies reveal both the strengths and weaknesses of the web of trust. The system was very robust; it took multiple, individually brilliant, persistent, patient, cooperating, already partially trusted attackers to break into it. It was also resilient; it continued functioning despite the massive security breach, and many people in it did in fact detect the attackers. However, it was extremely difficult to restore after the disaster, as the continuing hunt for more spies after Philby demonstrates. Also, its very resilience made it difficult to know that it had been subverted, and easy to take refuge in denial.

The Americans adopted the second approach. Theoretically, a PKI is easier to administer and more scalable, but more brittle. Spies like Aldrich Ames and Jonathan Pollard had to fake it, adopt the conventional signs of conformity, but once they faked it, they were in. (It didn’t help that the Americans were so keen on lie-detectors, an example of security theatre marrying pseudo-science.)

But brittleness can be useful. When, for example, a SSL certificate authority has been compromised, it has (as far as we know) become obvious very quickly, and the problem has (as far as we know) been solved quickly, by revoking and re-issuing the certs. The diversity of CAs, and hence competition between them, is useful here. Even the Heartbleed fiasco is mostly remarkable for the swiftness with which it was fixed. It is often useful for things to break unequivocally and publicly.

And webs of trust have other problems. Any system that intends to determine if some people can be trusted or not is exclusionary; that is sort of the point. As a result, they tend to absorb and harden the prejudices that are ambient in society.

So, I show you my Kim Philby: the key concerns seem to be sexism, disenchantment with graduate career prospects, and the failure modes of distributed cryptosystems. I guess I was right.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.