Gwen's Finest Hour?

 

[TC Comment: this is the last instalment of Gwen Tovey's memoirs. I hope you've enjoyed them. This story hasn't been published before. I'll explain at the end why I think it is important as well as entertaining.]

One morning early in 1969, when I came into the office, I found some rather startling signals from RAF stations in Germany (Butzweilerhof and Teufelsberg). They were reporting that they could hear none of the usual air-to-ground voice traffic from the fighters of the Tactical Air Forces (TAF) in Germany, which should have been in full swing by that time. Now, a stand-down of all aircraft at least 24 hours before an attack was one of the indicators in a report which I had compiled some 10 years earlier, about likely Sigint indicators of Soviet intentions to attack.

So I had to do something about this at once. I summoned all eight Soviet Intelligence coordinators and asked them to go and find out what was happening on their targets, with particular reference to anything that had been mentioned in the Indicator paper. Some 10 minutes later they reported back and it seemed that all aircraft — the fighters, the light bombers, the bombers of the Long Range Air Force (LRAF) and the Transport Commands were silent and there were no tracking reports on the air defence nets.

We put together a report, which specified all this, but with a comment which played the whole thing down by pointing out that no other indications, which would be expected if this meant an intention to attack, were discernible. Particularly, there was nothing in the communications of the LRAF (the bombers of which could carry atomic bombs), which resembled their preparations for major exercises. There was nothing on the Strategic Rocket Forces (SRF) links, there was nothing on naval links. In fact the Navy seemed to be preoccupied with an annual tin-pot little exercise in the Baltic. But nevertheless I issued an Alert [TC: a way in which all Five Eyes Sigint intercept stations could be warned that something might be happening] and hastily got someone to send a teleprint to NSA, saying what we were up to, just before issuing the FLASH report.

The next morning the front pages of newspapers were full of photographs of the usual Red Square scenes of long columns of tanks and guns, bands and missiles and hundreds of marching troops to celebrate — Oh Hell! — International Women's Day. So then, when I got into the office, I had to write the hardest signal of my life, apologising humbly and saying that all was now well and totally normal.

What was my horror then to find later that day, that in a couple of days' time the Senior Intelligence Officer from Strike Command proposed to come down and visit us. On the appointed day I sat shaking in my shoes waiting to be summoned to the Division Head's office, where the RAF officer would doubtless lambast me. However, suddenly the door of my office opened and in he strode with a great grin on his face, saying 'Let me shake you by the hand. That was the best exercise we ever had'. Now what he had done was to put the whole V Bomber force in the air and when he had done this, he was assailed by queries from station commanders saying 'What the hell is all this about?' And he said 'This just might this time be for real'.

I don't think an intelligence coordinator today would have received the totally blank silence I did, when I consulted the JIC Watch and the Foreign Office before sending off the original alarm, because the Anniversary is quite well known about today. But I think it will be a very long time before Whitehall grants us an extra day's holiday every time March 8th comes around.

[TC Comment: It's only since retiring that I realise that I would have taken some of the lessons this story contains for granted while still serving. Specifically:

a.       The GCHQ 'Soviet Intelligence Coordinators' would have been the equivalent of Senior Executive Officers: middle managers, with a wealth of knowledge and experience of their target, but six grades below the Director, and five grades higher than the most junior clerical staff. And yet they could put out a qualified Alert message to the whole of the western alliance saying that something big might be happening.

b.      In 1969 it was a woman who made the decision, and nobody questioned it, as she was intelligence coordinator for the Soviet Tactical Air Force and knew and understood her target. Of course GCHQ in 1969 wasn't an outpost of 21st century liberal values about diversity – it's only just appointed its first woman Director! – but however much harder Gwen had to fight to get on as far as the men she was working with, once there her knowledge and expertise were more important than the fact she was a woman.

c.       When GCHQ realised that it had issued a misleading report it took responsibility and explained its mistake. Getting something wrong isn't the end of the world if you can explain why you drew the conclusions that you drew and show how that is a knowledge gap that has been plugged. The best example of this I have seen is in HW 75, the intelligence reporting on the Soviet Bloc being released into The National Archives, in which a report on Soviet Field Post Numbers (FPNs) from the very early 1950s was corrected more than 20 years later because the location of one of the FPNs had finally been shown to be erroneous. What matters in intelligence reporting is accuracy so that those assessing it understand exactly how much weight can be put on it.]

Sigint Liaison with Finland

 

The UK now has a Defence Agreement with Finland and Sweden, and will soon be formally allied to both countries in NATO. It is not well known that for a short period in the Second World War the UK and Finland found themselves on the same side: our enemy's enemy was our friend and in spite of the alliance between Nazi Germany and the USSR, German occupation of Poland, Denmark and Norway, the Soviet occupation of the Baltic States and the part of Poland the Germans had agreed to let them have, and Swedish neutrality, the UK and Finland enjoyed a Sigint partnership for just over a year.

British Sigint had had relationships with the French, Italians and Americans during the First World War. These had all disappeared by the time GC&CS was formed on 1 November 1919. RAF Sigint did its own deal with the Estonian military Sigint authorities, and for most of the 1930s equipped and trained the Estonians in return for Estonian intercept logs of traffic of the Soviet Air Force, though this ended in 1939. In the 1930s, under the auspices of SIS a relationship was developed with the French to discuss Enigma, and from this grew the 'Enigma Relay', the exchange of information that led to Enigma becoming an exploitable cipher system.

This meant that GC&CS had become receptive to the idea that foreign Sigint services could be met as partners and early in 1940, this led to the idea that somebody should visit Finland to see whether the Finns were doing anything against the Soviet Union, the ally of Nazi Germany, that might benefit UK Sigint. At this point there was no dedicated UK effort against the Soviet military: an interservice cryptanalytical section would not be set up (in Wavendon, near Bletchley) until the summer of 1940.

Tiltman, the Head of the Military Section, went to Finland to meet the Finnish Sigint organisation, a meeting brokered by the SIS representative in Helsinki and the UK Military Attache. He arrived in Finland on March 1st 1940 and stayed for ten days most of which he spent with Major Hallamaa, the head of Finnish Army Sigint, at Kerava. Hallamaa showed Tiltman the whole of his organisation and expressed the willingness of the Finnish General Staff to cooperate fully in Sigint matters with the British.

The Finns were willing to exchange their knowledge of Russian codes and ciphers in return for the radios they were very short of. They needed 50 receivers, 6 DF sets and 6 high-powered transmitters to link their units in the field. The War Office in London confirmed that some of the equipment was available and the Finns handed over a forty page handbook on Russian Army and Naval communications.

The equipment was released on 15 April 1940, and the Finns were further informed that the UK would fund the Finnish purchase of further radio equipment in Sweden up to the value of £500. Some equipment arrived in Helsinki but when another GC&CS officer arrived in Finland to develop liaison on the Russian target he found that the Finns had only obtained 4 receivers, though a search while he was there produced three of the DF sets as well, and the Finns appeared satisfied that the British were acting in good faith. They agreed to provide Soviet intercept provided the British reciprocated: in the Finnish system the clerical effort involved would have to be justified to higher authority by a quid pro quo. They gave the UK the current encryption system being used by the Soviet Baltic Fleet as well as military systems which had become mainly readable. In return GC&CS offered its work on encryption systems used by the Soviet Navy Black Sea Fleet which was being taken in Palestine.

The exchange of intercepts was not straightforward. The air route between the UK and Finland was only open during the summer months, and the Finns were unwilling to open a new two-way radio link to the UK so the only way was to send the traffic by air to Stockholm from Helsinki in a diplomatic bag and to forward the traffic from there on commercial wireless circuits and vice versa. This would have worked if a Typex machine could have been sent to Stockholm and if sufficient Typex capacity had been installed at Sarafand to deal with the Black Sea traffic. However, there were significant problems in the supply of Typex, and by March 1941 Denniston argued that if only GC&CS could be allocated 10 lbs weight in the Stockholm daily bag the need for Typex would disappear, but neither Typex nor the 10 lbs of freight became available.

In spite of all the assurances it had given, the War Office did not actually send the rest of the radio equipment originally promised. It took until summer 1941 before this was sorted out and by then the situation had changed completely. The Germans declared war on Russia and a German mission was sent to Finland just as the final decision to send the missing sets was taken.

The last signal from Finland contained an assurance that they would not reveal to the Germans their Sigint cooperation with the UK and asked for an assurance that the British would not tell the Russians of Finnish cryptanalytical success against Russian systems. This assurance was given, and the Finns and the British each kept their promise (something perhaps easier for the British side than for the Finnish).

GC&CS developed its links with the US, and maintained its links with France and the exiled Poles until December 1942, but this marked the high water of non-Five Eyes cooperation. The Finnish experience had shown GCHQ that a buccaneering attitude to foreign relationships might bring results, but always at a potential cost of the partnership being prey to other considerations, and the dispersal of French and Polish cryptologists after the Nazi occupation of Vichy led to a new and cautious approach which would last for many years.

Sigint Liaison with the Russians

After the German invasion of Russia in June 1941 a military mission (No 30 UK Military Mission) was sent to Moscow and learned that the Russians had a Y Service. In August, 1941 a Y officer from the Middle East went to Moscow. He was given little information, and wasn't allowed to see anything of the Soviet organisation. However, as British relations with the USSR had been bad for many years this was considered unsurprising, and in October 1941 the War Office on its own initiative attached an Army officer (who had been working as a Sigint traffic analyst) in 30 Mission. At the same time the Admiralty negotiated permission to build a Naval Y Station in the far north aimed at providing early warning of any attack on the convoys planned to bring supplies to the Russians. On the Army and Air Force side it was hoped principally to obtain Russian-intercepted German traffic and some help if possible on Japanese systems. In fact, apart from the supposition that Russian Sigint would be less efficient than British there wasn't any clear policy, other than that allied intelligence liaison was a "good thing". 

The Admiralty received Soviet permission to erect a Y Station at Polyarnii, near Murmansk, and British permission to give the Soviets information about Auka (a low-grade Luftwaffe system), and, in May 1942, sent a naval officer from GC&CS to Murmansk to arrange for the exchange of Naval and TA intelligence. As a result of this some naval intercepted traffic reached GC&CS and there was also an interchange of DF bearings.

The War Office representative also made some progress. He was authorised in February 1942 to train the Russians to break the German police ciphers which were being used in rear areas on the Russian front, as well as showing them some traffic intercepted in the UK. (It was never the intention that British success against Enigma should be revealed.) He reported that this had produced a favourable reaction and he asked to be allowed to expand the exchange, giving the Russians identities recovered from callsign analysis of certain German stations in return for traffic intercepted in Russia. In June 1942 he was recalled for discussion of policy and went back to Moscow as representative, not of the War Office, but of the Y Board. But the mood in Moscow had changed: the Russians stalled and eventually in November the Head of the Soviet Army section dealing with Traffic Analysis suggested that cooperation should be discontinued.

Naval cooperation in the far north continued on a restricted basis, though some intercepted material inaudible elsewhere did reach the UK. This lasted until spring 1943 when, just as the UK staff were preparing to hand over another batch of Y material, the Russians suddenly entered the British site, sealed up the transmitters and took away the "meaconing" apparatus (a means of falsifying German Beacon transmissions.) At the same time it came to British knowledge that the German Y Service was reading some high echelon Russian cipher systems without much difficulty. Convoys were not actually in operation at this time, but were expected to resume and the closure of the station and Soviet insecurity greatly increased the hazards of running them.

In return for German Police material, the UK had given the Soviet military the Japanese Military Attaché code, the "Bird book" and index (the GC&CS process for breaking out German enciphered callsigns), the solution to German aerodrome serviceability reports, the process for recovering German police keys, Auka, and some ISOS keys. The Russians had also been made aware that the UK could read the Army tank code, Army medium grade field ciphers as well as the cipher used by one of the German sabotage units. The UK side decided that unless and until the Russians resumed cooperation, no further material should be sent. There was no point in keeping a representative in Moscow and the liaison officer, while retaining the title 'Russian Liaison Officer', was relocated to GC&CS. GC&CS had only one vote on the Y Board but it was a significant one. its attitude towards Russian cooperation was becoming more assertive and more reluctant. 'Jumbo' Travis wrote to Heads of Sections in May 1943 saying that he was prepared to consider an exchange, but only if there was a solid return "as I am not now a believer in Russian cooperation".

Although it had been decided that the Y Board should have no representative in Moscow, 'C' felt that an SIS liaison officer might usefully represent Sigint as well. One went to Moscow in the summer of 1943, just in time for several things to happen all at once. The Soviets asked for a full description of the German Enigma machine; they captured a copy of Auka themselves; and German Y Service traffic intercepted by the UK revealed that Russian ciphers were still being read by the Germans.

A meeting took place in summer 1943 after the British Ambassador made representations to Molotov, and it became clear that the Soviets' main interest was to find out what exactly what the British evidence of Russian cipher insecurity was. They were told that the information had come from a disaffected Austrian prisoner of war and from information from Japanese Military Attaché decrypts; they were also told that the UK side would give no further Sigint-derived information until the Russians delivered on their promises of the previous year. There were a few sporadic meetings and it became perfectly clear in spite of the SIS representative's attempts to put the best possible interpretation on what was going on, that the Russians were trying to get what they could from the British without giving anything away themselves. Telegrams continued to pass desultorily throughout 1944 and an Enigma machine and a small amount of very low level Y information was sent, though probably more for SIS's benefit than for Sigint's. Before the end of the war cooperation had ceased.

Why were the Soviets so reluctant to carry out an exchange of material? GCHQ believed post-war that the Soviet Sigint organisation had made the authorities in Moscow aware that the material being offered by the UK was low grade and that therefore either the UK was keeping its aces up its sleeve, or it didn’t have any capability worth wasting Soviet time on.

In my last post I told the story of how some information derived from Enigma decrypts had been published after having been passed by military censors. Commenting on Twitter @jock_bruce said that "in that story the point is that the folk clearing stuff for publication need to know what they’re actually protecting". I think that there is a similar issue here: those making decisions – the single services, and subsequently the Y Board – didn't understand that any half-way competent Sigint organisation would realise that valuable information was being withheld.

Sigint liaison worked with the French and the Poles when there was a full exchange of knowledge between the national Sigint organisations of each country, and subsequently worked with the US and Australia, Canada and New Zealand, because there was full sharing and integration of staff in each other's organisations. It couldn't work if the exchange was partial and low-level, and it certainly wouldn't work if non-Sigint personnel were representing Sigint.