Mirror Images

Christopher Andrew

  • World of Secrets: The Uses and Limits of Intelligence by Walter Laqueur
    Weidenfeld, 404 pp, £25.00, November 1985, ISBN 0 297 78745 4

While writing World of Secrets, Walter Laqueur had discussions with the present and all surviving past directors of the Central Intelligence Agency save one, as well as with other senior Intelligence official and their main customers. Such goings-on would be unthinkable, or at the very least unavowable, in Britain. At the height of the Westland saga the Prime Minister’s press secretary, Bernard Ingham, found time to denounce publicly those journalists who had dared to print the name of the head of MI6, Christopher Curwen. The heads of the CIA and the KGB, William Casey and Victor Chebrikov, are of course public figures. But that, as Mr Ingham would say, is not the point. What the point is remains obscure.

In 1932 the novelist Compton Mackenzie was tried for a number of breaches of the Official Secrets Act which included revealing the name of the first chief of MI6, Sir Mansfield Cumming. Mackenzie protested that Cumming had been dead for nine years, and the judge seems to have regarded that as a mitigating factor. But the prosecution explained that he had revealed details about Cumming of potential interest to an enemy, including the fact that he had a wooden leg and a chin like ‘the cutwater of a battleship’. Mackenzie was fined £100 with costs. Cumming’s diaries are by tradition kept in the office of the present chief at Century House, Lambeth. It may be that, like some of his predecessors, Mr Curwen (who can expect his knighthood within the year) uses the diaries to entertain visitors with their colourful details of Cumming’s pre-First World War disguises. These disguises, however, are not to be taken lightly. Members of Sir Mansfield’s family have for some time been seeking access to his diaries. The Foreign Office, which is responsible for MI6, has explained that the Cumming diaries remain so vital to national security that not a single sentence in them, even as far back as 1909, can be declassified. Indeed the Foreign Office goes further. It concludes that it may never be possible to reveal either details of Sir Mansfield’s pre-1914 disguises or any other part of his diaries.

At present therefore the ‘World of Secrets’ is better explored, as Mr Laqueur explores it, from Washington than from Whitehall. Even Washington, however, has been slow to come to terms with its intelligence community. Henry Stimson took office as US Secretary of State in 1929 in the firm belief that ‘gentlemen do not read each other’s mail,’ and closed down the State Department’s SIGINT (signals intelligence) agency. The success of the Japanese surprise attack at Pearl Harbour in 1941 reflected the Administration’s continuing inability to make good use of the SIGINT it received. Even the Second World War left the Administration still confused about what the role of its intelligence community should be. Perhaps the oddest ceremony in the history of the post-war White House was that devised by President Truman to mark the founding in 1946 of the CIG (forerunner of the CIA established in the following year). The President presented his guests with black cloaks, black hats and wooden daggers, then called his chief of staff forward and stuck a large black moustache firmly on his upper lip.

Under its first three directors (DCIs) the CIA remained uninfluential. The DCI was not even invited to the first meeting of the National Security Council during the Korean War. It was nonetheless the Korean War which put the CIA firmly on the Washington map. In General Walter Bedell Smith, appointed DCI in October 1950, the Agency acquired its first influential chief. It was said of ‘Beetle’ that he was the most even-tempered man in the world: he was always angry. He raised the CIA budget from 4.7 million dollars in 1950 to 82 million in 1953, and secured a sevenfold increase in its foreign stations. The Cold and Korean Wars between them, however, drove the CIA to the neo-Marxist conclusion that its main purpose might be less to interpret the world than to change it. The comparatively painless removal of unfriendly regimes in Iran and Guatemala in 1953-4 confirmed the change of emphasis from the collection and collation of intelligence – the essential and primary function of all intelligence agencies – to ‘covert action’. By the end of Allen Dulles’s term as DCI (1953-61), the ascendancy of covert action over intelligence analysis was so great that the CIA’s own analysts were not consulted in planning the invasion of Cuba at the Bay of Pigs. Since Watergate, the emphasis has shifted firmly back to intelligence collection and analysis. President Reagan’s own liking for covert action seems to be at least partially restrained by the Congressional Select Committees on Intelligence which emerged from the post-Watergate investigations.

‘After much research and discussion with leading consumers of intelligence’, Mr Laqueur has concluded that ‘far from being an invisible government, far even from wielding great influence in the councils of state, intelligence has frequently been disregarded or ignored by decision-makers.’ Despite President Johnson’s and, still more, President Nixon’s interest in covert action, neither of them – in Mr Laqueur’s view – thought very highly of the intelligence community. If secret intelligence has not had a greater influence on policy, it is at least partly because policymakers – presidents in particular – have not known how to use it. The intelligence community has usually found it difficult to gain the ear of the President. As a result, it entered the 1970s as a series of overlapping agencies, each competing for Presidential attention, and believing, in the words of Richard Helms (DCI 1966-73), that it had ‘to have a publication that arrives in the White House every morning’. Because of presidents’ short attention spans and crowded time-tables, the ‘publications’ have increasingly resembled tabloid newspapers rather than the quality press. Helms found difficulty in holding the undivided attention of President Johnson for more than 90 seconds. The CIA today probably has similar problems with President Reagan. Some of Reagan’s comments at press conferences on intelligence matters have looked uncommonly like those of a president who has read only the first paragraph of his briefings. Last November he speculated publicly on the possibility that the Soviet redefector, Vitali Yurchenko, might have been a Soviet plant. No doubt the first paragraph of the President’s briefing referred to that speculation. But the second paragraph must surely have demolished it.

The information explosion, whose covert counterpart has been an intelligence explosion (based, however, more on signals and photographic intelligence than on human intelligence), has posed formidable problems for all intelligence consumers. Even the United States foreign affairs bureaucracy, though its love of good literature must surely exceed the President’s, has less time for reading than ever before. Two-thirds of its long working days are spent in meetings or on the telephone. American diplomats have less time and inclination than ever before to write as well as to read. The steady decline in the quality of diplomatic reporting, particularly of long-term trends, is, Mr Laqueur notes, depressingly ‘obvious to students of diplomatic history’. Future historians are unlikely to find in the State Department archives of the 1980s anything to compare with George Kennan’s famous cables from Moscow of 1946.

The priority attached within the American intelligence community to long-range ‘estimates’ (theoretically its most important activity) has been steadily undermined by consumers increasingly afflicted by the obsession of the American media with instant news and instant views. The former head of the Bureau of Intelligence Research, Thomas Hughes, concluded gloomily a decade ago that, in the struggle for policy-makers’ attention, the ‘bakers’ (long-term analysts) had been defeated by the ‘butchers’ (current intelligence specialists). Long-term analysis nowadays probably has greater prestige in the KGB than in the CIA. Since the Second World War, there has been a steady tendency for the amount of raw intelligence in the United States to get larger and larger and for White House briefs to get shorter and shorter. Under President Reagan foreign policy and intelligence analysts have more information to digest than ever before and probably less space than ever before to present their conclusions.

Mr Laqueur’s survey of intelligence failures is, however, telling evidence of the importance of long-term analysis. The historical record provides no easy solution to the traditional problems of intelligence assessment, now exacerbated by the information explosion. But it does provide indispensable guidance on the errors to avoid. Those who do not know the past are doomed to repeat it – and in the ‘world of secrets’ they frequently have. The line of policy-makers with strong convictions which they are unwilling to have disturbed by unwelcome intelligence is long and dismal. The Church Committee of the Senate concluded in 1976 that realistic CIA ‘estimates’ on the Vietnam War had a negligible influence on administrations unprepared to entertain pessimistic forecasts: ‘Vietnam may have been a policy failure. It was not an intelligence failure.’ The CIA has also frequently found itself caught between hawks and doves in the Administration resentful of any intelligence which disturbed their own views of Soviet policy. According to Mr Laqueur, Henry Kissinger sometimes put pressure on the CIA to provide more hawkish intelligence. Senator Fulbright, however, as chairman of the Senate Foreign Relations Committee, was liable to laugh pessimistic intelligence about Russian intentions out of court.

Contrary to widespread popular belief in Europe, American administrations over the last twenty years have been more frequently too reluctant than too eager to believe the worst about Soviet intentions. When Richard Helms gave a last-minute warning of the danger of a Soviet invasion of Czechoslovakia in 1968, he was dismissed as alarmist by an administration committed to arms-control negotiations. In 1979, technical intelligence provided ample evidence of the Soviet build-up on Afghan borders: but the Carter Administration stuck to its conviction that the Soviet Union, while willing to use Cuban proxies in Africa, was no longer willing to attempt direct armed intervention itself. Even the CIA was slow to credit mounting intelligence during the 1970s that Russia was aiming not at missile parity but at actual superiority. Not until President Ford established the so-called ‘B Team’ of outside analysts under the Harvard historian Richard Pipes did the extent of earlier underestimates become clear.

Part of the United States problem in interpreting Soviet intentions arose from the old intelligence problem of ‘mirror-imaging’ against which more historically-minded administrations might have been more on their guard. American assumptions of the desirability for both sides of stable deterrence were projected onto the Soviet leadership and read into Soviet military doctrine. The response to advocacy by the Chief of the Soviet General Staff, Marshal Sokolovsky, in 1962, of a strategic nuclear force larger than the Americans was to try to explain it away. The prevailing interpretation – that Sokolovsky was a traditionalist fighting a losing battle against Soviet ‘modernists’ who shared American assumptions – was a classic example of mirror-imaging based on the conflict over nuclear strategy between US military chiefs and defence analysts.

In Vietnam in the 1960s and Iran in the late 1970s the USA once again found difficulty in grasping strategies and motivations alien to their own. Kennedy’s and Johnson’s advisers ignored the clear lesson of the French defeat in Indochina (amply documented in the writings of Bernard Fall) that bombing could not prevent infiltration from the North. Instead of bringing North Vietnam to its knees, the saturation bombing of ‘Operation Rolling Thunder’ simply encouraged Hanoi to engage its regular forces in the South. American leaders steadfastly declined to allow their basic military assumptions in Vietnam to be challenged either by the experience of the French or by evidence from their own intelligence services. Mirror-imagers also cannot cope with the NIH (Not Invented Here) factor. Mr Laqueur gives the ‘recent illustration’ of a Polish engineer who defected with intelligence concerning a large-diameter Soviet jet. When calculations showed that the jet’s thrust would exceed any in the USA, the report was dismissed as impossible.

No policy-maker or analyst can, of course, avoid starting from certain assumptions. The best course, Mr Laqueur implies, is to try to follow a middle way between the ‘Pollyanna’ (wishful thinking) and ‘Cassandra’ (expecting the worst) syndromes. ‘Pollyanna’ holds the key to some of the classic examples of intelligence surprise. In 1941 the idea that small, backward, Asian Japan would dare take on the might of the United States seemed so outlandish that neither FDR nor his high command paid adequate attention to signs of impending attack. Almost forty years later Iran provided an even more devastating example of the Pollyanna syndrome. By 1977 there were both intelligence and other reports of rampant corruption, SAVAK (secret police) brutality, growing religious unrest and economic discontent. Yet in August 1978, only five months before the Shah’s fall, the CIA concluded confidently: ‘Iran is not in a revolutionary or even a pre-revolutionary situation.’ The Defense Intelligence Agency predicted in September 1978: ‘The Shah is expected to remain actively in power over the next ten years.’ Washington wanted so badly to believe the Shah was secure that, despite mounting intelligence to the contrary, it succeeded in doing just that.

Mr Laqueur has much less to say about Cassandra than about Pollyanna. A prime historical reason for Cassandra has been shortage of intelligence about a powerful opponent and the fears bred by ignorance. At the outbreak of the Second World War Whitehall had so little reliable intelligence about the capacity of the Luftwaffe that it believed it capable of lauching immediately a devastating Blitz on the capital. Within fifteen minutes of Chamberlain’s speech on the morning of 3 September 1939 announcing that the nation was at war, the staffs of Whitehall ministries were sheltering in their basements awaiting an air-raid which the Germans were incapable of launching until they had expanded the Luftwaffe and captured forward bases in France and the Low Countries in the following year. Satellite reconnaissance of the kind we now take for granted would have dispelled much of Whitehall’s ignorance in 1939, just as it would have provided unmistakable warning of the surprise attacks launched by Germany on Russia and by Japan on the United States in 1941.

One of the most dangerous contemporary versions of the Cassandra syndrome is conspiracy theory. Arguably the greatest damage done by the first generation of Soviet moles was the destructive climate of suspicion which they left behind in the British and North American intelligence communities. The discovery of the real moles inevitably left some of the investigators fearful that even more important moles remained undiscovered. Sir Roger Hollis, the director-general of MI5, David Murphy, the CIA chief of Soviet Bloc Intelligence, and James Bennett, head of Canadian counter-espionage, were all denounced as likely Soviet agents by conspiracy theorists within their own services. In his recently published memoirs,[*] President Carter’s DCI, Stansfield Turner, records that ‘British intelligence remains very self-conscious about its past.’ In 1979 Jim Callaghan asked Turner: ‘Do you trust us again?’ Turner ‘gave him an answer that I hope was somewhat reassuring, but still not too positive. I did have reservations.’

Mr Laqueur’s analysis of the uses and limits of intelligence is essential reading for those interested in the World of Secrets. Like all studies of its kind, however, it suffers from one insuperable handicap. The unclassified evidence available on contemporary intelligence is not merely very fragmentary but also in some respects actually misleading. Despite all the post-Watergate revelations about the American intelligence community, there is even more that we do not know, and, on the assumption that intelligence services are desirable at all, much that we should not know. As Laqueur acknowledges, intelligence failures usually become known well ahead of intelligence successes. By the time Philby fled to Moscow in 1963, much of the damage done by the moles and atom spies was public knowledge. But the successes of British code-breakers against Soviet codes forty years before had yet to be revealed. We still know almost nothing about the further successes against post-war Soviet codes which assisted the mole-hunt. The recent failures of Western intelligence analysed by Mr Laqueur have, almost certainly, been balanced by considerable successes which have yet to be disclosed. The defection last year of the KGB station chief in London and MI6 double agent, Oleg Gordievsky, illustrates the danger of assuming that there are no Penkovskys now working for the West within the Soviet intelligence community.

I suspect that in general Mr Laqueur makes too little allowance for the influence of SIGINT. NSA (America’s GCHQ) takes a larger share of the USA’s intelligence budget and personnel than the CIA, but receives very much less attention in World of Secrets – chiefly, no doubt, because its secrets are more jealously guarded than those of the CIA. In the Congressional hearings of the 1970s which exposed the CIA to unprecedented publicity, the NSA figured hardly at all. Laqueur acknowledges that SIGINT supplies perhaps 80 per cent of all raw intelligence collected but suggests that it produces ‘less and less important and useful information’. It is certainly unlikely that the code and cipher systems of any major power are nowadays as vulnerable as were those of Soviet Russia in the decade after the Bolshevik Revolution or of Germany during the Second World War. But Laqueur’s judgment may none the less one day require revision. After the Falklands conflict the director of GCHQ, Sir Brian Tovey, passed on to his staff the warm congratulations of the Government. ‘Never,’ he told them, ‘has so much praise been accorded.’ It seems likely that in the Cuban missile crisis of 1962 SIGINT also played an important and so far undisclosed role. Mr Laqueur himself acknowledges the possibility that in declassified reports of the missile crisis the role of human and photographic intelligence may have been exaggerated in order to disguise the role of SIGINT. As DCI, Stansfield Turner found ‘intense competition and little co-operation’ between CIA and NSA. NSA, he claims, has a ‘penchant for withholding information from the rest of the community so as to be able to give it directly to the President or the National Security Council. Scooping the rest of the Community is the game; the NSA plays it well and the overall intelligence effort suffers.’

The suspicion that Mr Laqueur may not have given sufficient weight to SIGINT is strengthened by his brief analysis of its historical development. ‘The dominant form of intelligence up to World War I was,’ he concludes, ‘human intelligence’ – and the French were ‘arguably the best in the world’. In fact, probably the best French intelligence source during the two most serious pre-war Franco-German crises (those of 1905 and 1911 over Morocco) were decrypted German telegrams. And Tsarist SIGINT was probably even better than the French.

The problems of assessing modern intelligence without adequate evidence on SIGINT are at their most acute in the case of the Soviet Union. ‘The overall record of Soviet intelligence achievements,’ Mr Laqueur assures us, ‘has been excellent.’ He tells us nothing, however, about Soviet SIGINT. (The one index reference to that subject refers instead to Western SIGINT about the USSR.) This is not surprising, since next to nothing has yet emerged about it. But if Mr Laqueur knows next to nothing about Soviet SIGINT and little about Soviet technical collection, how is he able to certify the overall excellence of Soviet intelligence? In fact, Russian SIGINT may actually have declined, relative to the West, since Tsarist times. Before the Revolution Tsarist SIGINT probably led the world. Given the Western lead in computer technology, it would be surprising if Soviet SIGINT still does.

Mr Laqueur seeks to analyse the ‘World of Secrets’. His problem is that most of the world has yet to surrender sufficient of its secrets to make possible a truly global analysis of the current ‘uses and limits of intelligence’. Apart from the USA, the only ‘open societies’ whose intelligence communities he considers (all in the space of thirty pages) are the British, German and Israeli. He has some interesting observations about Israeli intelligence but what he says about Britain is marred by a series of irritating, if usually minor, errors. His account of the origins of British military intelligence is erroneous; there was no ‘post of director of military and naval intelligence’; Room 40, the First World War SIGINT agency, is demoted to Room 39; Sir Vernon Kell was not head of MI5 throughout the Second World War (he was sacked in 1940), and, since he was the first holder of his office, it is odd to read that ‘he was a military man like all his predecessors’; aerial reconnaissance began in the First World War, not ‘in the mid-1930s’; Geoffrey Prime is renamed ‘Price’; Sir Percy Sillitoe was head of MI5, not MI6; Sir Stewart Menzies was not the last ‘military man’ to head MI6 (he was succeeded by Major-General Sir ‘Sinbad’ Sinclair).

Because source material on the current ‘uses and limits of intelligence’ is so much richer in the United States than elsewhere, there is inevitably a danger – which Laqueur recognises – that an analysis of the ‘World of Secrets’ will become simply the American experience writ large. ‘Mirror-imagery’ is a case in point. This is clearly not uniquely an American problem. The Foreign Office, like the State Department, appears to have been reluctant to credit intelligence pointing to the invasion of Czechoslovakia in 1968. ‘Mirror-imaging’ also helps to explain how the Israelis were taken by surprise in the Yom Kippur War. But it may (or may not) be that European intelligence services, though their resources do not compare with those of the United States, find it easier to take account of cultural differences of perception. Because of its sheer size and power, the United States suffers from what Mr Laqueur recognises as the ‘parochialism of bigness’, and is unusually prone to project its own standards and values on the outside world.

Perhaps the most basic lesson of this valuable book is that, despite the complex technology on which it nowadays depends, intelligence remains not a science but an art. American policy-makers have too often expected it to be a science. As Mr Laqueur observes: ‘It is ironic that whereas no reasonable person expects the Department of Health to stamp out disease altogether, or the Department of Agriculture to provide for good harvests every year, the various organs of intelligence are expected to be right every time.’ The ‘uses and limits of intelligence’, both considerable, can be understood only in the light of lengthy past experience. The two post-war British intelligence chiefs with the highest reputation, Sir Dick White and Sir Maurice Oldfield, both very nearly became university lecturers in history. It may be a good omen that the present head of MI6 is also a historian.

[*] Secrecy and Democracy: The CIA in Transition by Stansfield Turner. Sidgwick, 304 pp., £12.95, 13 March, 0 283 99330 8.