Journal of the Operational Research Society (2010) 61, 68 --81
© 2010
Operational Research Society Ltd. All rights reserved. 0160-5682/10 www.palgrave-journals.com/jors/
The ‘invisible science’: operational research for the British Armed Forces after 1945 MW Kirby∗ and MT Godwin Lancaster University, Lancaster, UK This paper presents an original account of the contribution of operational research (OR) to the formulation of tactics and strategy on behalf of the British Armed Forces in the four decades after 1945. The main focus is on the Cold War in the European theatre, where OR analysts devoted considerable time and effort to the modelling of warfare on behalf of the British Army of the Rhine. In the absence of combat data for nuclear weapons, OR analysts devised a sequence of war games which evolved in conformity with the development of NATO strategy in relation to the Warsaw Pact. Again in the context of the Cold War, the paper analyses the role of OR in relation to the early V bomber force and the introduction of ‘global war studies’ on behalf of the Royal Navy. The paper also comments on the organisational structure of British military OR following on the creation of a centralised OR facility within the Ministry of Defence in 1965. In conclusion, the paper notes the sea-change in military OR following the end of the Cold War in 1990 and the onset of a period of sustained ‘asymmetric’ military operations in the former Yugoslavia, the Middle East and Afghanistan. Journal of the Operational Research Society (2010) 61, 68 – 81. doi:10.1057/jors.2009.156 Keywords: history of OR; military OR; Cold War; nuclear war
Introduction Readers of this journal will need little reminding of the origins of operational research (OR) in the pre-1939 preparations for the ‘Air Defence of Great Britain’ (Larnder, 1978; Kirby, 2003). Based upon the innovation of radar as an electronic means of aircraft interception, the resulting ‘man-machine system’ was the product of ‘operational researches’ carried out in 1938 (Larnder, 1978). It was the fortuitous combination of radar, monoplane interceptor fighters and youthful pilots within an integrated system that proved its worth against the Luftwaffe in the ‘Battle of Britain’ in 1940. Although the victory was won by a small margin, this did not prevent the diffusion of OR throughout the greater part of the military command structure both at home and abroad. By the end of the war, OR had substantial achievements to its credit. These included the improvement in accuracy of anti-aircraft fire, the vast enhancement of anti-U-boat capabilities on behalf of the Royal Navy and RAF Coastal Command in the North Atlantic theatre, and decisive contributions to the enhancement of RAF Bomber Command’s navigation and bombing accuracy during the ‘area’ campaign conducted against the German homeland after 1941. Following the Japanese attack on Pearl Harbour in December 1941, OR was consciously adopted by the US armed forces where it was applied with considerable success to anti-U-boat warfare and the daylight ∗ Correspondence: MW Kirby, Lancaster University, Bailrigg, Lancaster, UK. E-mail:
[email protected]
offensive against Germany conducted by the US Eighth Air Force after 1942. In this light, it has been argued that OR played a substantial role in the winning of World War Two in that its contributions to tactics and strategy stand alongside the other major scientific discoveries of the era, such as sonar, modern fire control, rockets and guided missiles, the cavity magnetron, and the proximity fuse (Kirby, 2003). At the end of the war, senior officers in all three of the main service branches in Britain were united in their view that OR should continue into the peace as an essential component of military planning, not least because of the accelerating pace of weapons development (TNA CAB90/5, 1944; ADM219/230, 1945; WO291/2459, 1949). For the RAF there was a vital need to comprehend the implications for aerial warfare of the jet engine. For the Navy, protection of surface vessels against improving submarine capabilities was a primary concern, and for the Army there was the dual need to enhance its capabilities in mechanised warfare and to improve manpower efficiency. But even as OR work commenced on these important aspects of defence planning after 1945, the remit of military OR was about to be radically transformed as a result of the onset of the Cold War in Europe and, after 1950, the increasingly real prospect of nuclear warfare, both tactical and strategic. It is the role of OR in relation to the threat of nuclear warfare that provides the main agenda for this paper. Because nuclear warfare held out the prospect of destruction on an unimaginable scale, defence planners were obliged to ‘think the unthinkable’ up to and including a strategic nuclear exchange between East and West. The
MW Kirby and MT Godwin—The ‘invisible science’
bombing of Hiroshima and Nagasaki in 1945 had confirmed the devastating impact of atomic weapons on urban and port environments. Thermonuclear tests in the early 1950s fulfilled the same role in demonstrating the vastly enhanced power of the hydrogen bomb. But despite these experiences, there remained a considerable deficit in knowledge and understanding as to how a war in which the belligerents were equipped with nuclear firepower would actually be fought, won or lost. In view of their limited knowledge of the tactical and strategic impact of nuclear weapons on the waging of war, and as the frontier of nuclear weaponry advanced, OR scientists were obliged to live their professional lives in a counterfactual, if not surreal environment. Their advice to military commanders was therefore tendered in hypothetical terms. How would Britain and the NATO Alliance in general respond if Soviet-led forces, both conventional and nuclear, invaded West Germany? What trajectories would the invasion follow? What would be the optimal deployment of NATO forces? What were the desirable tactics and strategy for attacking the Soviet homeland itself? How could the North Atlantic supply route from the US to Western Europe, so vital to Allied success in World War 2, be sustained in the light of submarine attack? This paper provides critical insights into the manner in which OR scientists sought to answer these questions. In the following section, the paper focuses on the development of tactical nuclear war gaming for the Northern European theatre on behalf of the British Army of the Rhine (BAOR). The second section highlights the weaknesses of tactical nuclear war gaming as they were perceived in Britain and America by the early 1960s, and this serves as a prelude to discussion in the third section of the implications for OR scientists of the movement of NATO from a ‘tripwire’ nuclear response to Soviet aggression in favour of ‘flexible response’ whereby invading forces would be intercepted by non-nuclear military formations. The final substantive section highlights the role of OR on behalf of the RAF and the Royal Navy paying particular attention to the deploymentof the early V-bomber force and the Navy’s use of ‘big OR’ in the form of ‘global war studies’. In conclusion, the paper comments briefly on the role of OR after 1990 when the ‘symmetrical’ Cold War, with its known geography and logistics, was followed by a sequence of ‘asymmetric’ military operations from peace-keeping duties in the former Yugoslavia to high intensity warfare and anti-insurgency campaigns in Iraq and Afghanistan. In these novel situations, in which the distinction between war, organised crime and large–scale human rights violations became blurred, OR scientists have been presented with major challenges, not least in terms of their methodology. Note: By the end of the 1960s, the British defence establishment was in the habit of referring to OR as ‘Operational Analysis’ (OA) in order to distinguish it from the military term ‘operational requirements’. Henceforth, the paper will use the later nomenclature: OR personnel will be referred to as ‘OA scientists’.
69
Tactical war gaming for the Northern European Theatre In reflecting on the remit of OA scientists after 1950, Roger Forder has commented as follows: In many key areas analysis based on accumulated operational and trials data was of limited value. Yet military planners too found their own operational experience of rather limited value, so they were still keen to call up the scientific advice which had proved so valuable in war. But the scientific advice would have to be of a different sort. Rather than making sense of real, day-to-day operations, analysts had now to be concerned with hypothetical circumstances, hypothetical enemy reactions and the pros and cons of equipment as yet unprocured, perhaps even undeveloped. (Forder, 2000)
In relation to nuclear weapons, this is the nub of the matter. Since nuclear weapons could not be used in military exercises, the only operational data that existed were derived from the atomic bomb attacks on Hiroshima and Nagasaki in 1945 and, for the period after 1950, target response data from nuclear test explosions. It was these factors that led OA scientists in the Army Operational Research Group (AORG) to adopt the German-derived ‘Kriegspeil’ concept in the early 1950s as a means of training military commanders in the use of tactical, or battlefield nuclear weapons. In this respect, their work was undertaken on behalf of the BAOR stationed in West Germany. The establishment of the NATO alliance in 1949 was in response to the onset of the Cold War in Europe focusing on the divided German frontier. The BAOR was Britain’s land force contribution and it was destined to become the dominant element in NATO’s Northern Army Group (NORTHAG), formed in 1952 and responsible for the defence of the northern front from Hamburg to Kassel, in the event of invasion by Warsaw Pact forces. In 1952, a British Chiefs of Staff working party concluded that the considerable superiority in manpower and conventional weaponry enjoyed by the Warsaw Pact could be offset by equipping the BAOR with tactical ‘low-yield’ atomic weapons, a view that was taken up within the higher NATO command structure in 1954 when tactical, or ‘battlefield’ nuclear weapons were incorporated into planning in preparation for a Warsaw Pact land offensive. At that time the British Army did not possess tactical nuclear missiles or warheads, and although the US Army was about to be equipped with first-generation Honest John and Corporal missiles, there were no immediate plans for the BAOR to acquire such weapons (Kirby and Godwin, 2008, p 637). The arrival of thermonuclear weapons following the first American hydrogen bomb test on 6 November 1952 dramatically changed the stakes in nuclear planning. British thinking in response to this event was that in its capacity as the ‘primary’ deterrent, the hydrogen bomb could be usefully supplemented by secondary deterrents at the tactical level.
70
Journal of the Operational Research Society Vol. 61, No. 1
Thus, it was assumed that the prospect of a global war precipitated by the first use of thermonuclear weapons by the Soviet Union might be lessened considerably if conventional forces were equipped with tactical nuclear firepower: possession of tactical weapons on the part of NATO forces in Europe could serve to deter aggression at the hands of the Warsaw Pact. It was this scenario that led OA scientists within the AORG to investigate the means by which battlefield nuclear weapons could be incorporated within the BAOR on the twin assumptions that it would possess a tactical nuclear capability by 1960 and that the main Warsaw Pact thrust on West Germany would be directed at NORTHAG. In this respect, the AORG began to develop considerable expertise in simulating a tactical nuclear land battle in the form of war gaming. The learning curve was inaugurated by two substantial studies based upon experience in World War Two, the first relating to the German invasion of France and Flanders in May–June 1940 and the second to the D-Day landings in June 1944 (Kirby and Godwin, 2008). Both of these ‘counterfactual’ studies used wartime data and were modelled on the assumption that Allied forces were equipped with tactical nuclear weapons. Given that the role of NATO forces in West Germany was essentially defensive, the events of 1940 were viewed as being of greatest relevance to the BAOR. The lesson to be derived from this first AORG venture into war gaming was that if defending forces possessed nuclear superiority and the determination to exploit this advantage in ‘critical defensive situations’, and assuming that they were well entrenched behind a major obstacle, such as a wide river-line, a rapid ‘Blitzkreig’ of the kind used by the Germans in 1940 could be thwarted (Kirby and Godwin, 2008). It was this conclusion that paved the way for further exercises in war gaming in preparation for a nuclear land battle across the divided German frontier. The main result of this activity was the AORG Tactical War Game (TWG) that was developed in the latter half of the 1950s (TNA WO291/2233, 1959). War games can be classified into three broad categories— training games, planning games, and analytical research games (Jaiswal, 1997). Training games are, literally, designed to train commanders in the effective deployment of military forces in the operations of war. In simulating combat scenarios, they also provide the opportunity to assess the merits of weapons, tactics and doctrines. The purpose of planning games is to evaluate the effectiveness of ‘standard operating procedures’ (SOPS), while analytical research games use controlled experiments to assess concepts and doctrines. From its inception in 1955, the AORG TWG was conceived as a training game. Early versions did not incorporate the use of tactical nuclear weapons. They were, however, introduced in 1957 in the light of gaming experience. Gaming took place on a terrain model divided into two square kilometre consistent with the physical characteristics of the BAOR’s sphere of operations in West Germany. The model was designed to encompass a major land battle up to Corps/Army level
and a game was played by two sides, ‘RED’ and ‘BLUE’, the former based upon the latest intelligence concerning the organisation, equipment and capabilities of the Soviet Army, and the latter based upon all units within the BAOR. In most games the RED side enjoyed superiority in conventional ground forces between 2:1 and 3:1, and it was the RED side that was given the role of aggressor in terms of contact and final assault on BLUE’s main position. The task allotted to the BLUE side was to delay and then hold the enemy on the line of a ‘major obstacle’. Each side was located in separate operations rooms containing identical terrain models and the game was presided over by a neutral controller in a third room equipped with a ‘master model’ encompassing both RED and BLUE forces. Gaming was conducted according to a strict set of rules derived from knowledge of previous operations, weapons trials, contemporary military structures and SOPs. Insofar as early games were played by senior officers (Major Generals and Brigadiers) with considerable battlefield experience of World War 2 and Korea, it has been claimed that a high standard of realism was attained in relation ‘to every sort of activity and foreseeable emergency on the tactical battlefield, although it should be noted that outcomes could be resolved by ‘random number selection’ (Shephard, 1963). Strict adherence to the rules was viewed as essential to the credibility of the game and in this respect the controller fulfilled a crucial role in delivering ‘sound evaluations and comparisons’: the controller was not an umpire exercising personal judgement. The game procedure itself was straightforward. Each commander was given ‘starting conditions’ that included orders of battle, the allocation of weapon, and the task allocated to unit commanders. They also had complete freedom in force deployment and in the conduct of the battle. Play commenced at the discretion of the controller and each game could take up to 6 weeks to complete. Once it had begun, action took place at hourly intervals of ‘combat time’: all moves were made at ‘clock hours’ when both sides received ‘intelligence’ reports. Games were played in sequence by different commanders who were confronted with a variety of starting conditions. Eventually the point would be reached when it was appropriate to engage in comparisons of ‘battle outcomes, weapon effects [and] casualty levels’. It was this final comparative exercise that was used as a guide to future ‘tactical and organisational concepts’, as well as weapons and equipment procurement. During 1957, and continuing into 1959, the AORG modified the gaming process to include nuclear weapons ranging in size from 1 to 500 kilotons (the yield of the weapon dropped on Hiroshima in 1945 was between 15 and 20 kilotons). It was generally assumed that the BLUE side enjoyed superiority in nuclear firepower in a ratio of 7:5. In refining the TWG to incorporate battlefield nuclear weapons, a number of detailed studies were carried out with a particular focus on the needs (advance warning and protection) of friendly troops in the event of a nuclear strike. The titles of these studies are
MW Kirby and MT Godwin—The ‘invisible science’
indicative of the following key factors in nuclear war gaming: (i) Optimum Burst Heights in the Tactical Use of Nuclear Weapons (ii) How Far Can [Nuclear Weapons] Replace Conventional Artillery Support? (iii) Safety Distances for the Tactical Use of Nuclear Weapons (iv) Nuclear Attack Warning and Reporting System (v) Protection of the Soldier on the Nuclear Battlefield (vi) Personnel Targets in the Defensive Use of Nuclear Weapons (TNA WO291/1519, 1957; WO291/2191, 1958; AORG, 1958; WO291/2263, 1959; WO291/2318, 1959) The main conclusion to be drawn from the TWG was that the outcome of a future land battle in Central Europe would be determined mainly by the comparative nuclear firepower of the belligerents. If the supply of nuclear weapons was limited, their role would be to lend fire support to a conventional war both in terms of defence and attack. Assuming that the defending BLUE force enjoyed nuclear superiority, a RED force with a comparative advantage in conventional weaponry and personnel ‘can probably be held and destroyed’. In terms of delivery systems and targets, the AORG identified three classifications: the nuclear gun, the 30–50 mile guided missile, and the ‘longer range system whereby delivery would be via a guided or free-flight missile, or air delivery’. The nuclear gun would be directed at opposing forces in ‘the contact and observation zone’ up to 8 kilometres from the forward defence line. The short range in relation to defending forces dictated the use of a low-yield shell. The 30–50 mile system would necessitate a medium-yield warhead directed at known enemy targets, ‘who are dug-in or otherwise well protected’, gun battery and launcher sites, and large formations of troops in the open. The ‘longer range system’ was based upon the use of a medium or high-yield weapon directed at ‘deep’ targets within the 30–50 mile range, reserve units in rear areas and the enemy’s logistical system. War gaming provided some critical insights into the tactics to be adopted in the main phases of the nuclear battle following the ‘advance to contact’ with the enemy. The key consideration for the Commander of the defending force was that the advent of battlefield nuclear weapons required a radical and fundamental change in traditional military doctrine, bearing in mind that such weapons could destroy well-protected enemy formations and their equipment extremely rapidly over a wide area. In this light, fighting the nuclear land battle would dispense with the need for massed ground forces, but it would also entail the maximum possible dispersion of friendly troops to avoid their decimation by the enemy’s nuclear firepower. For the defending force, . . . the tactics will be to hold the enemy at a distance and destroy him before he can close on the main position in strength. To do
71
this, natural and artificial obstacle lines will be used to check the enemy and force him to concentrate or to expose his troops, and a widespread well-protected observation network will be deployed to give adequate and timely target intelligence for the use of nuclear weapons TNA WO291/2233, 1959).
As for the command and control structures for the nuclear land battle, it was clear to the AORG that command decisions would be far more frequent than in the past. It followed, therefore, that in a battle fought at Corps/Army level the ‘penalties on indecision’ could be disastrous, resulting in unprecedented strain on Commanders at all levels.
Nuclear war gaming: the liberal critique In September 1958, when tactical nuclear war gaming by the AORG was well underway, the War Office, acting under the direction of the Chief of the Imperial General Staff Field Marshall Sir Gerald Templer, circulated a substantial looseleaf document under the heading ‘The Corps Tactical Battle in Nuclear War’. The foreword was unambiguous in its recognition of the implications for warfare of battlefield nuclear weapons: The time has now arrived when the quantity and quality of nuclear weapons becoming available on the battlefield impose a necessity for completely new tactical methods. These are weapons of a new order, and minor adjustments in outlook are no longer sufficient. We shall not find the solutions to the problems of nuclear battle by adapting ideas which were successful in conventional war because we shall not experience those conditions again if nuclear weapons are used. (TNA WO279/279, 1958)
What followed was a description of the conduct of the ‘Corps nuclear battle’ over the divided German frontier in which the BAOR would use tactical nuclear weapons. The document ended with the reassuring observation that although the deployment of nuclear weapons represented ‘a step forward longer than that between bow and bullet’, and notwithstanding the fact that battlefield tactics had altered radically, ‘the principles of war [remain] unchanged though subject to modern interpretation’ (TNA WO279/279, 1958). The BAOR began to be equipped with American-derived tactical nuclear weapons in 1959 with their use subject to a ‘dual key’ arrangement with the US. By 1961, ‘Corporal’ and ‘Honest John’ missiles had been deployed within three regiments. But even as they were being incorporated within the BAOR, doubts and uncertainties about their utility were being expressed so that by 1962 the doctrine set out so clearly in ‘The Corps Tactical Battle in Nuclear War’ was effectually redundant. Coincidentally, there was mounting criticism within the higher reaches of the British defence establishment of the overwhelming nuclear bias of the AORG’s war gaming activities. The leading critic was the distinguished anatomist and expert on the physiology of primates, Sir Solly (later
72
Journal of the Operational Research Society Vol. 61, No. 1
Lord) Zuckerman, who took up his appointment as Chief Scientific Adviser to the Ministry of Defence in 1960. It was significant that Zuckerman had fulfilled important roles in OR in World War 2 when he investigated the impact of bomb blast pressures on the human population and emerged as a firm advocate of the precision bombing of military targets in direct opposition to Bomber Command’s area approach (Peyton, 2001, pp 51–63; Kirby, 2003, pp 126–127, 140–141). Thus, when he was appointed to the Ministry of Defence, Zuckerman ‘could lay claim to a wider experience in the planning of bombing operations than anyone at that time in high command on either side of the Atlantic’ (Peyton, 2001, p 145). Zuckerman’s anti-nuclear stance was rooted in his belief that the British nuclear weapons programme was not so much the product of rational thought but ‘the irrational consequence of a nuclear obsession’ (Maguire, 2007, p 4). In espousing this unpopular view, Zuckerman’s credibility was dependent upon support at the highest level in the military establishment. That was amply supplied by Earl Mountbatten of Burma, wartime Chief of Combined Operations, a Supreme Allied Commander, and currently serving as Chief of the Defence Staff. Mountbatten held Zuckerman in the highest regard, viewing him as ‘the cleverest man in the world’, not least because of his wartime contributions to bombing policy (Peyton, 2001, p 140). Indeed, their friendship and empathy became known in Whitehall defence circles as the ‘Zuckbatten Axis’, a partnership that ‘caused the enemies of either to be enemies of both’ (Peyton, 2001, p 141). It is instructive to note that Zuckerman’s assault on the received wisdom underlying British nuclear policy was registered both at the tactical and strategic levels. In the present context, it was the proposed use of battlefield nuclear weapons by the BAOR that caused him to express profound concerns about their incorporation in AORG war gaming. Even before he joined the Ministry of Defence, Zuckerman had expressed reservations on tactical nuclear weapons in his capacity as chairman of the Air Ministry Strategic Policy Scientific Policy Committee. After 1960, he began to develop a coherent critique utilising his experience of OR in World War 2 and his academic knowledge of the life sciences. An important part of his critique was based upon theoretical analyses of ‘low-yield’ nuclear attacks on British cities, the first based upon a 1 megaton attack on Birmingham and the second, a 20 kiloton attack on Carlisle. The conclusion in both cases was that organised life would cease to exist. A further analysis of nuclear strikes against industrial cities in the north-west USSR produced similar results: casualties would be in their millions and up to 80% of industrial capacity would have been wiped out so that ‘The total results would be disastrous for the whole Soviet Union’ (cited in Maguire, 2007, p 10). As Maguire has observed, Zuckerman’s view of wartime bombing policy—that ‘bigger and ever-more sophisticated [bombs] were not needed’—led him to conclude that for the purposes of deterrence ‘possession of a minimal number of any type of nuclear weapon was sufficient to
prevent attack’. In any event, nuclear weapons ‘could never be used except as an act of national suicide’ (Maguire, 2007, p 11). Commenting on the physical organisation of OR for the Army at West Byfleet in the early 1960s, the veteran OA scientist, Roger Price, has described ‘a couple of mediumsized outbuildings with ‘windows placed so high that looking from the outside was impossible without a ladder. This was a security precaution, as they housed the war-gaming tables on which Army commanders ‘fought’ imaginary battles in Germany’ (Price, 2004, p 235). Unsurprisingly, these ‘imaginary battles’ soon attracted Zuckerman’s critical attention after he joined the Ministry of Defence. His first impression was that there was an unquestioning assumption that so-called tactical nuclear weapons had a practical use on the battlefield. As he recalled in his memoirs, . . . I got the feeling that no-one had thought the matter through. No account had been taken of distant interdiction strikes or of the battles that would be raging on the flanks of the central part of the front held by British troops. It was also assumed that the Russians would not have enough nuclear weapons to reply at a rate more than one to every two or three we used—an assumption which I could not imagine was still valid by the beginning of the 1960s. And I was also worried that the AORG studies had not considered the possibility that the morale of our own troops, as well as that of the Russians, might be shattered in the environment of a nuclear battlefield. (Zuckerman, 1988, pp 272–273)
After further examination of the gaming process, Zuckerman was surprised at the large number of battlefield nuclear weapons that the gamers were prepared to use in order to stem a Soviet-led invasion of West Germany. Taking a NATO Army Corps front based upon the River Weser, Zuckerman calculated that the destructive power of the requisite number of weapons on all forms of life and the amenities of life ‘would destroy the political purpose of the operation’ (Price, 2004, p 170). With the necessity for up to 20 nuclear weapons to be fired in ‘a single afternoon’, was it credible that a meaningful ‘defence’ of West Germany could be organised on the basis of ‘twenty Hiroshimas between lunch and tea?’ (Price, 2004, p 170). So dismayed was Zuckerman at the prospect of nuclear annihilation of military personnel and civilian populations that he used NATO’s annual conference, held in the summer of 1961, to give full vent to his view that for the Alliance to contemplate the use of tactical nuclear weapons was ‘wishful thinking’. Speaking under the title ‘Judgement and Control in Modern Warfare’, Zuckerman addressed the audience with characteristic bluntness: I rubbed in the fact that according to NATO plans, nuclear strikes would be made far behind the battle line—and by both sides, at the same time as short range nuclear weapons were used in the battle zone itself. I reminded the audience of the enormous shortcomings of the command and control apparatus with which the Supreme Commander would have to try to direct a battle
MW Kirby and MT Godwin—The ‘invisible science’
in which nuclear weapons were used. . .. Nuclear weapons, I insisted, cannot be categorised as tactical or strategic. Whatever they were deemed to be, depended on the target they struck. A few Soviet ‘tactical’ nuclear strikes could remove from the map some of the smaller members of NATO. That could hardly be regarded as tactical warfare. After referring to the limitations of abstract operations analysis that left human beings out of account, I summed up by saying that when we spoke about the potential use of nuclear weapons, we had to avoid the conceptual framework derived from the military terminology of pre-nuclear warfare. One could deter with nuclear weapons but, I asked, could one defend? There was silence when I sat down. General Heusinger, shaken by the picture of destruction that I had painted, rose to express his disquiet about NATO nuclear policy. (Zuckerman, 1988, p 289)
It is significant that Adolf Heusinger was Chairman of NATO’s Military Committee and he was joined in his concerns by General Earle Wheeler, Commander-in-Chief of US forces in NATO. As Wheeler asked Zuckerman at the end of the meeting, ‘We have painted ourselves into a corner. How are we to extricate ourselves?’ (Zuckerman, 1988, p 289). Apart from Zuckerman’s concerns about the destructive effects of tactical nuclear weapons, two further points in his speech are worth highlighting—the first his reference to ‘the command and control apparatus’, and the second, ‘the limitations of abstract operations analysis’. It is certainly the case that AORG war gaming assumed that Corps Commanders would exercise firm control of the nuclear battlefield. Yet it was OA scientists attached to the BAOR who raised some fundamental issues concerning the implications for nuclear command and control arising from the structure of NORTHAG. As early as 1958, after referring to the presence of Belgian, Dutch and German Corps within the group, an OA liaison officer commented forcefully that NORTHAG was far from being an integrated fighting force due to differences in language and procedures, some of which were the product of ‘national pride’. Logistical arrangements were complicated to a degree: Not only the requirements of the various corps, but also the methods of demand and response differ. This adds to the difficulties of ‘cross-servicing’ which is quite likely to be needed in certain operational situations. Added to this, the fact that each corps’ [Lines of Communication] may run across the territory of up to three different nations and that each nation jealously guards its sovereignty, although ostensible subscribing to the principle of collective security, makes things no easier. The incentives which exist in peace appear not to be sufficient to break down international barriers and allow reasonable planning and preparatory work on these questions. Although command headquarters get smaller and fewer, logistics headquarters proliferate. (TNA W/15, 1958)
It is unnecessary to possess a military mind to appreciate the immense strains that this structure would be subject to
73
in the event of a nuclear land battle. NORTHAG’s military effectiveness was compromised to the extent that there was ‘no real sense of a ‘nuclear Army’, a deficiency that persisted into the 1980s (McInnes, 1996, p 25). As for ‘leaving human beings out of account’, Zuckerman was joined in his critique by Patrick Blackett, the acknowledged ‘father of OR’ and its foremost practitioner in World War 2 (Gardner, 2003, pp 110–137; Crook, 2003, pp 138–166; Kirby, 2003, pp 86–131; Nye, 2004, pp 75–85). After 1945, Blackett emerged as an inveterate opponent of nuclear weapons. His views had been expressed with eloquence in his polemic, The Political and Military Consequences of Nuclear Energy published in 1948 (Blackett, 1948). This work, drawing on his knowledge of wartime bombing policy, was sufficient to marginalise him in nuclear debate both in Britain and the US for years to come. Yet by the early 1960s, Blackett’s stance on nuclear weapons was not as unorthodox as it had been in the late 1940s. In an address to military officers at the Royal United Services Institution (RUSI) in 1961, he expressed the view that war gaming for nuclear war was ‘wrong and dangerous’. OA scientists in World War 2 had been effective because they worked closely with senior military officers in interpreting operational data. Believing that their advice was reliable, they also took moral responsibility for their work. In the case of nuclear war, however, the absence of operational data meant that OA scientists could not produce ‘a common sense analysis’. In attempting to compensate for this absence, war gaming embodied ‘a nauseating inhumanity’ reflecting ‘a strain of deep social pessimism’ that ‘combined sometimes strangely with an almost neurotic contemplation of destruction’ (Nye, 2004, pp 94–95). Although Blackett’s main targets were located in the United States—the Rand Corporation, the Hudson Institute and Princeton University—it is reasonable to assume that his RUSI audience was able to make the connection to AORG war gaming.
‘Flexible response’ and the Central Front Game (CFG) At the December 1954 meeting of the NATO Council, formal approval was given to American proposals for the deployment of tactical nuclear weapons. Characterised as the ‘New Look’ for American defence planning, this decision was an open recognition that previous commitments to the expansion of conventional forces would not be met, and that nuclear firepower, in compensating for NATO’s numerical deficiencies, would act as an effective deterrent to aggression by the Warsaw Pact. The full implications of the decision can be summed up as follows: The role of conventional forces [in Central Europe] would be to provide a ‘shield’ behind which the nuclear ‘sword’ could be readied. Above all but the lowest levels of armed conflict described in terms of ‘border incidents’ and ‘police actions’, the role of conventional forces in the defence posture of the alliance was widely understood as being to act as a ‘tripwire’ that would initiate a nuclear response to any aggression. The use of
74
Journal of the Operational Research Society Vol. 61, No. 1
conventional forces in this manner would reinforce deterrence by providing an unambiguous threshold beyond which the alliance would use its nuclear weapons . . . . In short, the purpose of conventional forces and nuclear weapons was to deter the widest range of contingencies by ensuring that any aggression would be met with massive retaliation. (Buteux, 1983, pp 2–3)
NATO’s commitment to ‘massive retaliation’ lasted for less than a decade. Although the construction of the Berlin Wall in 1961 was a critical factor in precipitating a fundamental reappraisal of American and ultimately NATO nuclear strategy, of equal significance was the launch of the first Sputnik in 1957, an event that foreshadowed the capability of the Soviet Union to target inter-continental nuclear missiles (ICBMs) on the United States. The launch of a new era in Western nuclear strategy was heralded in a notable speech on strategic policy by the US Secretary of Defence, Robert McNamara, in 1962. Here, he outlined the concept of ‘flexible response’ as a means of reducing the threat of a strategic nuclear exchange precipitated by the Soviet Union. Given the arrival of nuclear parity between East and West and its concomitant—mutually assured destruction—the threat of devastating nuclear attacks on Soviet cities was no longer a credible deterrent. McNamara therefore set out the case for a wide range of Western military responses from ICBMs, submarine-launched ballistic missiles and strategic bombers, to limited non-nuclear counter-force operations (Baylis, 1995). In confronting a Warsaw Pact attack on West Germany, NATO would initially deploy conventional non-nuclear forces, thereby providing the Soviet Union with an incentive not to launch an ICBM attack on the continental United States. Only if NATO forces were being overrun would commanders resort to the use of battlefield nuclear weapons. Unleashing the nuclear ‘sword’ in this way would take the form of ‘controlled escalation’. If the use of the most powerful tactical nuclear weapons failed to halt the Warsaw Pact, then the only degree of escalation left was a strategic attack on Warsaw Pact cities. Lying behind ‘flexible response’ was the doctrine of ‘assured destruction’ in the sense that if the Soviet Union engaged in a strategic attack on American cities and industries, the US would retain a sufficient ‘second strike’ capability to deliver an overwhelming attack on their Sovietcounterparts. Despite British and European concerns that limited nuclear war in Central Europe might well lead the Soviet Union to launch a ‘pre-emptive strategic strike’ if the Red Army was being annihilated, NATO agreed to adopt a ‘Flexible Response Strategy’ in 1967 on the assumption that in presenting Soviet military planners with multiple uncertainties, it was the optimal deterrent to Soviet aggression (Freedman, 1980, pp 20, 29, 104; Buteux, 1983, pp 102–103, 219–220, 228, 242; Baylis, 1995, pp 301–303, 374–376). Within Britain, NATO’s adoption of ‘Flexible Response’ was registered by the inauguration of ‘Exploratory Study 147’ in April 1969 (TNA DEFE10/396, 1969). It remained in existence until at least the late 1970s by which time it had become ‘the umbrella framework for a wide range of . . .
land and air studies ranging from equipment value for money studies, through examinations of the air battle (counter air vs. offensive air support to NATO deployment and studies in support of the early Mutual Balanced Force Reduction Talks’ (G. Beard to authors, 2008). In the present context, the key project within the wider study was the ‘land and air battle’ in the form of the ‘CFG’) (TNA DEFE48/78, 1970). This was undertaken by two of Britain’s leading OA scientists—Nigel Beard and Geoffrey Beare—working with colleagues within the Defence Operational Analysis Establishment (DOAE), formed in 1965 to incorporate OA provision for all three of the armed services. The purpose of the CFG was to determine ‘the present comparative capability of the NATO and Warsaw Pact forces, and to study the steps that might be taken, within resources similar to those now allocated by the various nations, to improve the NATO posture’ (TNA DEFE10/396, 1969). ‘Improvement’ was specifically interpreted to mean ‘an extension of the time during which NATO could sustain conventional operations without resort to nuclear weapons’. The game itself was based upon a major Warsaw Pact invasion of West Germany where NATO’s 245 ‘Standard Battle Groups’ were defending against an invasion force of 38 divisions—equivalent to 456 battle groups producing a force ratio of 1.86:1 in favour of the Warsaw Pact. In contrast to the NORTHAG-based TWG, the CFG encompassed ‘Europe from the Baltic to the Alps and with air and land engagements assessed by computer model’ building on the work of Martin Beale (Beard to authors). Embracing both NORTHAG and the American-dominated Central Army Group (CENTAG), the CFG used a computer model to generate the progress of battle within each corps area of responsibility over a period of 24 h. As Beare recalls, this was displayed to a ‘command team’, comprising military experts and senior scientific staff, who would then make militarily credible decisions on, for example, the disposition of air effort, commitment of reserves etc. for both NATO and [Warsaw Pact] forces. In this broad way strategic command decisions could be represented and their significance examined in a scientifically controlled way. The underlying models were deterministic and any particular point in the battle could be replicated and replayed with different decisions. (G. Beard to authors, 2008)
All of this stood in marked contrast to the TWG that was now viewed as ‘neither scientific, nor objective, nor a means of comparing various strategies as it was cumbersome and slow’. Insofar as the earlier game assumed that a Warsaw Pact invasion would always be directed at NORTHAG, it removed any uncertainty as to the likely trajectories of attack. The CFG, however, allowed for an element of surprise in this respect so that ‘it focusedthought on “the optimum scale and position of [NATO] reserves” and the direction of air attack’ (Beard to authors). More to the point, ‘the most important foundation of the [CFG] was that we could support all the numbers that went into it’ (N. Beard to authors, 2008).
MW Kirby and MT Godwin—The ‘invisible science’
The CFG was based upon a major battle fought in the following stages: (a) Covering force battle. Comparatively small forces deployed as close to the frontier as possible with the task of establishing the strength and direction of the enemy attack and disrupting and delaying him for as long as possible. (b) Aggressive withdrawal. This phase will be fought by the main forces. Their aim is, by means of a mobile battle, to disorganize the enemy forces, inflict on them more casualties than they themselves receive and to halt them wherever possible. (c) Main defensive position. Here those forces remaining to the Army Groups will make a concerted stand with the aim of preventing any further advance of the [Warsaw Pact] forces. It is at this stage that the decision will have to be taken on whether to resort to the selective use of nuclear weapons (TNA DEFE48/13, 1970 [emphasis added]). The ‘Main Defensive Position’ is of particular interest in confirming that the use of tactical nuclear weapons would only be considered at this point in the battle—presumably at the time when Warsaw Pact forces were in a position to overwhelm their NATO counterparts, leading to a decisive victory. A typical CFG could be played on a 6-h or 12-h game cycle between Command Group decisions. With computer assistance (an ICL 1907E), one game cycle could be played in ‘one day of real time’ with the ratio of real time to game time set between 4:1 and 2:1. Typically, a full game lasted for 2 weeks with preparation for a game adding up to a further week. In the belief that this was still too slow despite computer assistance, OA scientists complemented the CFG with the linear-programme-based ‘NATO Deployment Model’ (TNA DEFE48/486, 1972). This was a ‘simplified’ version of the CFG and was used ‘to examine a wide range of deployment options against a variety of threats’ (G. Beard to authors, 2008). Its primary function was to identify ‘the minimum time it took Soviet forces to break through NATO defences where there were several different points of attack at which the Warsaw Pact forces were assumed to be able to concentrate their attack’ (Beard to authors). The final component of the CFG was the deployment of air power by both sides. For the Warsaw Pact, it was assumed that there were ‘first and second echelon air forces’ consisting of 1250 ‘attack’ and 1600 ‘air defence’ aircraft. The bulk of these aircraft were stationed in two groups along the border with West Germany where they were faced by NATO’s Second and Fourth Tactical Air Forces. The latter consisted of 935 attack and 267 air defence aircraft. Of these, 102 Canberra and F111 aircraft were to be held in reserve ‘for nuclear operations’ if NATO ground forces located in the ‘Main Defensive Position’ were being overwhelmed. Constraints of space preclude discussion of the role of air
75
support, but playing of the CFG produced the consistent result that despite their numerical inferiority NATO aircraft were substantially more successful than the Warsaw Pact in reducing the capability of enemy battle groups. The CFG assumed that both sides were heavily dependent on their medium tanks and it was in their role as ‘tank busters’ that NATO aircraft delivered a substantial advantage over their Warsaw Pact counterparts. This was, in part, the result of the inferior armament of the latter and their relative inability to identify suitable targets. The conclusions derived by OA scientists from Project 147 in general and the CFG in particular formed the main agenda item for a meeting of the Chiefs of Staff Committee held in November 1971. Tabled under the auspices of the Defence Policy Staff (DPS), the Chiefs of Staff were invited to note the following conclusions: In detail, Project 147 confirms the now current assessments and views: that the main source of WP strength on the ground lies in their numerical superiority and firepower of their tanks; and that an exchange of tactical nuclear weapons, with about equal numbers being used by both sides, causes roughly the same level of casualties on the two sides and therefore moves the force ratio to NATO’s disadvantage. DOAE in Project 147 tested various land and air campaigns and verified that WP breakthrough was inevitable on one or more Corps fronts within a day or two, even if other Corps were able to hold out for a longer period. The analysis assumed that . . . the conduct of operations would require an aggressive withdrawal from the Demarkation Line (DML) to maintain defensive positions about 100 kms to the west but did not fully take into account the effects of air attack and battle field confusion on WP and NATO movements. In our view, the disruption of movement by blockage of roads with damaged vehicles accompanied by air attacks . . . could influence and extend the duration of the conflict perhaps by some one or two days. This would lead to a conventional phase . . . lasting some two to four days. Project 147 has also shown that extension of the duration of conventional operations, amounting to one or two days might however be achieved if the final NATO defensive position were moved westwards from the present area to the present rear Corps boundaries, a distance of more than 100 kms. (TNA DEFE 4/262/2, DP 25/68, 1971)
After noting that further work was necessary on the deployment of air power ‘in the case of surprise attack and full aggression in the Central Region’, the DPS focused particular attention on the most controversial result of Project 147—the movement westwards of NATO’s ‘final defensive position’ in order to extend the phase of conventional operations. This would provide the incalculable benefits of fighting the battle without any significant increase in conventional forces at the same time as it would gain ‘precious time for consultation and critical decision-making relating to nuclear escalation’. It would also provide NATO with greater flexibility in anticipating enemy movements via the deployment
76
Journal of the Operational Research Society Vol. 61, No. 1
of reserves. There were, however, major disadvantages to a ‘defence in depth’, the first relating to West Germany, whose government had been consistently in favour of a ‘Forward Defence Strategy’ on the grounds both of ‘emotion’ and ‘politics’. Moreover, for NATO to depart from the established principle of forward defence ran the risk of conveying ‘a sign of weakness’ thereby encouraging the Warsaw Pact ‘to pursue their aggression with increasing determination and the acceptance of greater risks’. Above all, such a fundamental issue was about to be subsumed within ‘a searching examination of NATO’s strategy overall’ so that it would be counterproductive for Britishdefence planners to openly advocate ‘the concept of fighting in greater depth as a means of extending the duration of the conventional battle’. That ‘searching examination’ would be the inevitable by-product of impending NATO–Warsaw Pact discussions on ‘Mutual Balanced Force Reduction’ (MBFR) in Europe. Substantive talks commenced in October 1973 and from their inception NATO’s negotiating stance in favour of reductions in conventional and nuclear forces was informed by ‘DOAE data and analytical techniques from Project 147’ (TNA DEFE4/262/2, DP25/68, 1971). This was open recognition by the DPS of the relevance and importance of the work of OA scientists on Project 147 (TNA DEFE 4/262/2, DP25/68, 1971). Their work on the Central Front Game and other associated studies had provided defence planners with ‘a more up to date’ appreciation of the political and military situation both in NATO and the Warsaw Pact. As the DPS concluded, The development by DOAE of a data base against which the force capabilities of NATO and the WP can be assessed, together with the preparation of associated analytical models for applying the data base to specific scenarios has been the most valuable by-product of the [Central Region Study]. The data are updated to take account of changes in equipments and variations in concepts; this and the associated methodology and techniques can be used to assist in solving the variety of operational problems which will require scientific analysis in the future. (TNA DEFE4/262/2, DP25/68, 1971)
In the event, the MBFR negotiations dragged on for years due to intractable problems in measuring the relative size of NATO and Warsaw Pact forces in Europe and in establishing a credible system of verification. Agreement was eventually reached in 1989, one year before the collapse of the Soviet Union and the Warsaw Pact. There remains to be considered a controversial aspect of the movement to ‘Flexible Response’ insofar as it had unlooked for and unpleasant consequences for the DOAE and its OA scientists. A foretaste of things to come was provided by Air Chief Marshall Sir Denis Spotswood, Chief of the Air Staff and in that capacity a member of the Chiefs of Staff Committee. While the Chiefs of Staff voiced its agreement with the conclusions of the Central Region Study and expressed admiration for the high quality of the work carried out under Project 147, Spotswood was determined to fire a
warning shot at the remit of the DOAE. The examination of ‘inter-service problems’ was entirely legitimate, but it was not . . . the task of DOAE to tackle single-Service problems, nor was it equipped to do so, although he [Spotswood] had detected an unwelcome tendency for it to become involved in this area. In this context he considered that the DOAE had not yet discovered a satisfactory means of analysing the air battle. (TNA DEFE 4/262/2, DP4/262/2, DP25/68. 1971)
In examining the postwar structure and remit of OA, a key development was the formation of a unified Ministry of Defence in 1964 and its concomitant—a centralised OA facility in the form of the DOAE in the following year. The key driver in this was Solly Zuckerman with powerful support from Lord Mountbatten in his capacity as Chief of the Defence Staff and Peter Thorneycroft, the Conservative Government’s Defence Secretary in the early 1960s. All three were united in the belief that abolition of the separate service ministries and centralisation of OA would be conducive to the more effective formulation of defence policy as well as greater efficiency in military procurement. It is a fact, however, that the movement to ‘flexible response’ had the effect of intensifying inter-service rivalries thereby frustrating Zuckerman’s vision of a fully coordinated national defence policy. As Beare has commented, notwithstanding the coordinating remit bestowed upon the Ministry of Defence in 1964, ‘The Services remained fiercely jealous of their independence’ so that the defence budget continued to be split three ways (G. Beard to authors, 2008). On the DOAE itself, the recollections of Graham Jordan, recently retired as a senior OA director within the Ministry of Defence, are illuminating. Jordan worked as an OA scientist in the DOAE from 1967 to 1982 and his recollection of that period is that the DOAE’s analytical work on the conventional phases of ‘flexible response’ gave rise to ‘polarisations based on single service interests’. In this context, The services saw DOAE as a threat because it was set up to look at things on a tri-service basis and (by implication) to draw comparisons between the contributions made by the individual services to achieving particular defence objectives. And without the sponsorship and protection of a strong ‘purple’ central staff (as originally intended) DOAE was vulnerable. (Jordan to authors, 2008)
In a recent address to the International Symposium on Military Operational Research, Jordan expanded on this theme by highlighting the consequences of Sir Hermann Bondi’s attempt in the early 1970s to determine the appropriate allocation of resources between the main service branches in the event of conventional military operationsin Central Europe. Bondi was then serving as Chief Scientific Adviser to the Ministry of Defence and his conclusions on the conventional phase of ‘flexible response’ were ‘explosive’. As Jordan recalls,
MW Kirby and MT Godwin—The ‘invisible science’
The Army were told that conventional land warfare was not likely to last more than a week before . . . Soviet forces broke through in Central Europe. The Navy were told that the idea of taking a month to get US reinforcement convoys flowing across the Atlantic to Europe was just too slow to be of any use. And the [RAF], flying aircraft into dense air defences, without defence suppression or stealth to protect them, and dropping only unguided bombs, [was] a very inefficient and expensive way of destroying tanks. The response against this study was massive. (Jordan, notes for speech, 2008)
The effect was to undermine the careers of officers who lent support to these conclusions, while Bondi himself ‘withdrew from direct sponsorship of [OA] studies’. This culture of ‘blazing rows accompanied by blood-curdling threats’ was largely absent during World War 2. The main exception to this was Patrick Blackett’s suggestion that success in the anti-U-boat war in the North Atlantic was dependent upon the expansion of RAF Coastal Command at the expense of Bomber Command’s area campaign against Germany (Kirby, 2003). In general, however, OA scientists enjoyed exceptionally close working relationships with senior officers, at least in relation to the delivery of tactical advice. During the Cold War, however, there were no operations going on, and the real threat to a service was from having its resources cut back, perhaps because of a particular piece of [OA] that said that another service could do something better, or that some type of operation was simply not worth doing, or some piece of equipment not worth buying. (Jordan, notes for speech, 2008)
‘Thinking the unthinkable’ It was the arrival of thermonuclear weapons in the early 1950s that encouraged OA scientists in the Admiralty to embark upon a programme of work designed to model a strategic nuclear exchange under the umbrella of ‘global war studies’. As early as 1950, the Chiefs of Staff had commissioned a joint Admiralty – Air Ministry committee to examine the role of OR in modelling and predicting future nuclear warfare at the strategic level. The resulting report, prepared by the Maritime Air Defence Committee (MADC), focused on the North Atlantic and Mediterranean theatres where the main concern was the defence of convoys. Although much use was made of wartime operational data, the Committee embellished its report with analysis of the likely costs of new equipment, from up-dated aircraft carriers to improved sonar detection of enemy submarines. ‘Slice costing’ was to cast a long shadow forwards in defence planning for nuclear warfare and it is in this respect that the MADC report may be viewed as a pioneering example of ‘big’ OR. In 1954, the Director of Operational Research (DOR) in the Admiralty embarked upon a study of a strategic nuclear exchange in the form of an all-out nuclear war between East and West. ‘Thinking the unthinkable’ in this context was rooted in the assumption that global nuclear conflict would be followed by a period of
77
conventional ‘broken back’ warfare in which maritime forces would fulfil a vital role in determining the ultimate outcome of the conflict. OA studies were therefore applied to the calculation of naval losses in the nuclear phase and the minimum force requirements in the conventional phase. A series of war games was played where ‘the will to conduct conventional operations was assessed by comparison with the effects of wartime bomber raids on centres of population’ (Pratt, 1981). The conclusion drawn was that ‘a military will and capability’ could survive a nuclear war provided the right decisions were taken on defence policy, in particular the need to invest considerably in civil defence in the form of shelters and supplies. In the mid-1950s, this would have required a major departure from current policy: splitting the defence budget four ways would hardly prove popular in defence circles (Moore, 1997, p 88). In spite of this, the Chiefs of Staff were sufficiently impressed to authorise the establishment of a Joint Global War Committee with a dedicated study group—the Joint Global Warfare Working Party. Again, resort was had to war gaming (Moore, 1997, pp 78–79) and once again the issue of civil defence was highlighted as an integral part of defence strategy. The key calculation was that in the event of a strategic nuclear attack on Britain consisting of 58 megaton weapons within 1 nautical mile of accuracy, an unprotected population would suffer 50% casualties. This would fall dramatically to 3.5% if everybody took refuge in the best possible shelters (Moore, 1997, pp 79–80). It was the budgetary implications of national survival that probably accounts for the DOR’s continuing commitment to the study of civil defence. Thus, in 1958, he proposed that a ‘no shelter’ policy should be negotiated with the Soviet Union to be policed by a joint inspection system. This would be a simple procedure, as the construction of shelters would need to be in close proximity to population centres making their construction difficult to conceal. As the DORconcluded, if an effective inspection system was in place, this could pave the way for substantial reductions in strategic nuclear forces on both sides. In the event, it was the Foreign Office that vetoed this novel approach to nuclear defence policy, dismissing it as ‘faintly ridiculous’, ‘peddling’, ‘gimmicks’ and therefore ‘not worth pursuing’ (TNA FO 371/136916: jkt UN1192.171/G, undated). In 1959, defence planners returned to the issue of strategic nuclear war in the form of a new Global Warfare Committee chaired by Solly Zuckerman, soon to be appointed Chief Scientific Adviser to the Ministry of Defence. The Committee was supported by a team of military and civilian personnel known as ‘JIGSAW’ or the Joint Inter-Service Group for the Study of All-out War. From the standpoint of the Ministry of Defence, the wide ranging membership of JIGSAW was to be welcomed in that it provided a central OA capability that had hitherto not existed. It is unsurprising, therefore, that Zuckerman proved to be a strong, though at times critical supporter. In examining JIGSAW’s programme of work, it is evident that its primary concern was to identify the scale of a
78
Journal of the Operational Research Society Vol. 61, No. 1
nuclear attack that would produce ‘a situation . . . . Unacceptable to any rational government’ (TNA DEFE 19/9, 1966). With the Soviet Union in mind, JIGSAW adopted the assumption that ‘breakdown’ of an industrialised nation would occur if 50% of the total population or 75% of the urban population was affected by ‘a widespread or sudden’ nuclear attack. It is in this context that Zuckerman extrapolated from his wartime experience and also from contemporary studies of the impact of nuclear attacks on Birmingham and Carlisle. It was the methodology adopted in the latter that was used to assess the ‘first’ and ‘second-order’ effects of a nuclear attack on the Leningrad area. This included nine cities with populations in excess of 50 000 and accounting for 83% of the total population in the area. On the assumption that each city was attacked with a 1 megaton ground-burst weapon, the study concluded that the Leningrad area would no longer be ‘an asset to the USSR’. Extrapolating from this to the whole of the Soviet Union showed that 180 deliveries would ‘put paid to at least 80–90% of the country’s industry’ (TNA DEFE19/91, 1966). Although JIGSAW was confident about its ‘breakdown’ theory, it is difficult to determine the effect of its work on British nuclear policy at the strategic level. What is certain, however, is that the Leningrad study, together with the earlier analysis of nuclear attacks on Birmingham and Carlisle, fulfilled critical roles in informing Zuckerman’s stance on the absurdity of nuclear war. It seems likely also that Zuckerman’s experience of JIGSAW helped to reinforce his belief in the need for a centralised OA facility in order to reduce or at least contain inter-service rivalry, not least in relation to the British nuclear deterrent. The decision that Britain should possess its own atomic weapons was taken in secret in 1946 by a small group of Cabinet ministers in the Labour Government. Coincidentally, the RAF issued an operational requirement for a new jet bomber that would form the delivery system for the British nuclear deterrent as it was developed after 1950. This marked the birth of the V bomber force that was designed to attack key Soviet targets, including Moscow. The V-force, in the form of Valiants, Victors and Vulcans became an iconic symbol of Britain’s military and technological strength in the early post-war decades. Given their engineering complexity and the task allotted to them, it is unsurprising that OA scientists were called upon to devise effective operational procedures both for the aircraft themselves and a wide range of related equipment and weapons systems. Almost from the inception of their work, OA scientists in the Air Ministry and Bomber Command accepted that the effectiveness of the V-force would be limited in the light of anticipated improvements in Soviet air and ground defences (Godwin and Kirby, 2009). In the mid-1950s, the government‘s response to this was to develop a ground-launched ballistic missile as the future basis of the nuclear deterrent. But given the time scale of some 10 years, it was realised that there could be a gap in Britain’s deterrent posture by the mid-1960s. It is this scenario that led OA scientists to devise operational procedures that would extend
the life of the V-force. Constraints of space preclude detailed analysis of their work, but it is possible to identify the key issues that they sought to resolve. This is especially the case in relation to the Valiant, the first of the V bombers to be deployed operationally. In 1956, OA scientists in Bomber Command predicted that during the next 10 years there would be ‘No . . . developments . . . , either in fighters or guided missiles [that] will make any serious contribution to low-level defence; no country has yet proposed any practicable low-level defence scheme which is worthy of serious consideration’ (TNA AIR14/4202, 1956). Despite this major benefit to low-level flight, there was a considerable disadvantage in the form of severely curtailed range. However, this could be mitigated by carrying extra fuel tanks, in-flight fuelling, combining low and high-level flight, and by using air bases outside the UK. In these respects, OA scientists calculated that by utilising Middle East airbases, and with under-wing tanks, the offensive capability of the Valiant would be greatly enhanced: 50% of all targets could be reached via a low-level flight path that would be virtually invisible to enemy radar. In addition, V bombers could be equipped with stand-off nuclear weapons that could be launched at considerable distance from the enemy’s air defences. In the event, the American (‘Skybolt’) and British (‘Blue Water’) development programmes for such weapons were cancelled after 1960 at the same time as OA scientists were beginning to express concern about the fatigue implications of low-level flight. In calculating the durability of the Valiant, OA scientists were able to make use of ‘fatigue meters’ attached to the aircraft. With builtin assumptions concerning ‘g’-force levels, a Valiant was assumed to have a standard fatigue life of 6100 h with a standard flight programme being set at 316 h at high altitude and 44 h at low altitude per aircraft per year. This result, which did not include training flights, pointed to a major fatigue issue for the Valiant that was likely to emerge in 1968. In fact, it came as early as 1964 when the Valiant fleet was grounded in the wake of fatigue problems in landing. Two further studies undertaken in relation to the V-force by OA scientists are worth highlighting, the first concerning the vulnerability of aircraft to ballistic missile attacks on their airfields and the second the implications for the V-force of the ‘four minute warning’. In the case of missile attacks, OA scientists were acutely aware that their attempts to deal with this problem were very much dependent on estimates of future Soviet capabilities. Their ultimate recommendation favoured dispersal of the V-force far beyond the UK. Thus, if aircraft were stationed in Nairobi, the Maldives, Salisbury (Rhodesia) and Canada, ‘ample target coverage’ was available if in-flight fuelling was used. Even remote bases in Singapore or Bermuda would have some targets in range, and if V bombers could be equipped with a longrange stand-off weapon then an even greater number of targets was available (TNA AIR 14/4210, 1959). But basing aircraft overseas in this way had complications. Not least, the
MW Kirby and MT Godwin—The ‘invisible science’
enormous distances involved would entail a considerable refuelling capability, even in the case of the most up-to-date V bomber available—the Vulcan Mark 2. For example, it was found that operations from Nairobi would require a stop at Cyprus or refuelling on the outbound journey from the UK. The Maldives could be reached only if tankers were available from Cyprus, although with two refuellings, the Maldives could be reached without an intermediate stop. Nonetheless, these were significant restrictions, especially since some refuelling scenarios entailed pilot bale-outs. OA scientists therefore concluded that although overseas dispersal was valuable, it was only with an enhanced version of the Blue Steel standoff bomb that it would become particularly worthwhile. As noted already, however, Blue Steel was cancelled in 1960. The ‘four minute warning’ of an impending Soviet missile attack on the UK became possible in 1960 following governmental agreement with the United States to install specialised tracking equipment at Fylingdales on the North Yorkshire moors. Although official emphasis was placed upon its relevance to civil defence, the resulting Ballistic Missiles Early Warning System (BMEWS) was primarily intended to warn RAF Bomber Command of an impending attack so that the V-force could be scrambled. Four minutes was a very short period especially since Fylingdales itself was a prime target for attack, all the more so since the installation was conceived at the outset as an essential component of Anglo-American defence cooperation. OA scientists in Bomber Command were directly associated with the BMEWS system: it was their calculations of incoming missiles warning times that produced the ‘four minute’ scenario. Critical to their work were assessments of the physical characteristics of missiles, their deployment and firing trajectories. In these respects, it was assumed that missiles with a range between 650 and 1500 miles would be selected for use against the UK, and that they would be fired on their maximum range trajectory, or on a low trajectory with a re-entry angle of 15 degrees. In practice, it proved impossible to supply a uniform warning time for all UK targets since different missile capabilities and different targets would have unique values. It is also unclear as to how improvements in Soviet missile design or enhanced facilities at Fylingdales may have altered warning times over the duration of the Cold War. The one clear fact is that when Fylingdales became operational, the only way to increase warning times would have been to move targets as far to the west as possible (TNA AIR14/305, 1961). It was the Nassau Agreement of December 1962 between Britain and the US that marked the beginning of the end of the V-force as Britain’s strategic nuclear deterrent. From that point onwards, British nuclear strategy would be based upon the American Polaris missile system to be launched from British-built nuclear submarines. The Polaris system became operational in June 1969 so that for the greater part of the 1960s the V-force remained in service as the primary deterrent. It was the decision to transfer the nuclear deterrent to the Navy that helped to politicise Britain’s OA effort in the early
79
1960s. Reference has already been made to Zuckerman’s commitment to the centralisation of military OA. It remains to be said that the role of OA in the Air Ministry and Bomber Command added weight to his views. This is confirmed by contemporary correspondence between Terry Price and Zuckerman, as well as Price’s memoirs. Price was certainly of the opinion that the Air Staff was guilty of ‘talking up’ the efficacy of the V-force to the extent that he doubted the accuracy and objectivity of its calculations. This was especially the case on the issue of the ability of the V-force to penetrate Soviet defences. Following a conversation with an American official who expressed ‘disquiet’ on the capabilities of the V-force, Price commented to Zuckerman that . . . we in the Ministry of Defence have so far always had to rely on estimates by the Air Staff, who have assured us that they have no worries on the score of penetrability. I have never been entirely happy about the situation but until the Ministry of Defence has an operational research group of its own, I do not see how we shall be able to test which of these two views is the correct one. (TNA DEFE19/102, 1962)
Reinforcing this perspective is the so-called ‘manned bomber incident’. Although no documentary sources have come to light to confirm its veracity, it has been highlighted by retired OA scientists as evidence of the Air Staff’s inflated claims for the V-force. The scenario is the early 1960s following Zuckerman’s arrival at the Ministry of Defence. Concerned about ‘penetrability’, he asked the Chief of the Air Staff to provide details of OA work on this issue. The reply has been summarised by Graham Jordan as follows: The Chief of the Air Staff said that it was his job to advise the Secretary of State for Defence on the capabilities of the Airforce (sic), not the Chief Scientific Adviser. And his advice was that the ‘manned bomber would always get through’. And if he chose to take advice from his [OA] scientists in coming to this view, then that was his private business, and no, the Chief Scientific Adviser could not see his report. (Jordan to authors, 2008)
Zuckerman makes no reference to this in his memoirs, but it is reasonable to presume that if the incident is true, he must have experienced a sense of d´ej`a vu, bearing in mind that he had adopted a deeply sceptical view of the objectivity of OA scientists in lending uncritical support to Bomber Command’s area offensive against Germany in World War Two (Kirby, 2003, pp 140–141).
Conclusion This paper has focused on the role of OA during the Cold War, a period that generated degrees of apparent certainty in the formulation of defence policy. In the European theatre, in particular, a combination of fixed geography and wellestablished logistic structures were key factors in determining the appropriate mix of land and air forces and their operational
80
Journal of the Operational Research Society Vol. 61, No. 1
efficiency in the event of aggression by the Warsaw Pact. With the end of the Cold War as a military phenomenon, there was much public comment on the social and economic benefits of reductions in defence expenditure in the form of the so-called ‘peace dividend’. For Britain, however, the dividend failed to materialise not only because of the continuing commitment to the ‘independent’ nuclear deterrent, but also because of new calls on Britain’s armed forces. After 1990, the Cold War ‘standoff’ was followed by a sequence of expeditionary operations in a variety of locations, from peace-keeping duties in the former Yugoslavia and Sierra Leone, to high intensity warfare and anti-insurgency campaigns in Iraq and Afghanistan. In contrast to the ‘symmetry’ of the Cold War era, these operations have been viewed as ‘asymmetric’ in the sense that adversaries have sought to avoid full-scale confrontations with conventional military formations and to achieve victory by methods that maximise their own strengths and exploit the weaknesses of their opponents. For OA scientists, asymmetric operations have presented unprecedented challenges in the delivery of scientific advice to commanders (Forder, 2004). Thus, in the post-1990 world OA scientists have broadened the indicators of ‘effectiveness’ and ‘success’ in operations beyond ‘warfighting’ to embrace a range of civilian and political criteria, including the establishment of democratic political structures and public tolerance to casualties in the field. In terms of methodology, war gaming and simulation have continued after 1990, but in the new asymmetric environment the portfolio of techniques has expanded. This is especially the case in relation to ‘soft’ methods of analysis, including multi-criteria decision analysis, influence diagrams and the modelling of complex socio-technical systems. There has also been renewed interest in the analysis of past conflict in order to comprehend more fully the effects on the modern battlefield of ‘shock, surprise, breakthrough and manoeuvre’. The onward march of computer power has been registered in the improvement of data handling and visualisation, the use of complex algorithms and the non-deterministic formulation of Monte Carlo simulations. Far from the end of the Cold War diminishing the need for OA, post-1990 field operations in challenging socio-political environments have rendered military commanders ever more dependent on the work of OA scientists up to and including the provision of advice and assistance in the front line. Acknowledgements —The authors are grateful to the Defence Special Interest Group of the Operational Research Society for comments and advice on their work. Thanks are also due to Nigel Beard, Geoffrey Beare, Paul Sutcliffe, David Daniel, Graham Jordan, David Faddy, Geoffrey Hawkins and George Rose. The authors are indebted to the Leverhulme Trust and the Ministry of Defence for funding support. Errors of fact and/or interpretation are the sole responsibility of the authors.
References Baylis J (1995). Ambiguity and Deterrence: British Nuclear Strategy 1945–1964. Clarendon Press: Oxford.
Beard N to authors, 3 April 2008. Beare G to authors, 3 April 2008. Blackett P (1948). The Military and Political Consequences of Atomic Energy. Turnstile Press: London. Buteux P (1983). The Politics of Nuclear Consultation in NATO 1961–1980. Cambridge University Press: Cambridge. Crook P (2003). The case against area bombing. In: Hore P (ed). Patrick Blackett: Sailor, Scientist and Socialist. Frank Cass: London, pp 167–186. Forder R (2000). Operational analysis in defence: A retrospect. Paper presented at NATO Partnership for Peace Symposium on The Applications of Operational Analysis to Defence Decision Making, Garmisch Partenkirchen, Germany. Forder R (2004). Operational research in the UK Ministry of Defence: An overview. J Opl Res Soc 55: 319–332. Freedman L (1980). Britain and Nuclear Weapons. Macmillan: London. Gardner J (2003). Blackett and the black arts. In: Hore P (ed). Patrick Blackett: Sailor, Scientist and Socialist. Frank Cass: London, pp 126–137. Godwin M and Kirby M (2009). V is for vulnerable: Operational research and the V bombers. Defence Stud 9: 149–168. Jaiswal N (1997). Military Operations Research: Quantitative Decision Making. Kluwer: London. Jordan G to authors, 22 February 2008. Jordan G (2008). Notes for Speech to ISMOR, 28 August 2008. Kirby M (2003). Operational Research in War and Peace: The British Experience from the 1930s to 1970. Imperial College Press: London. Kirby M and Godwin M (2008). Operational research as counterfactual history: A retrospective analysis of the use of battlefield nuclear weapons in the german invasion of France and Flanders, May–June 1940. J Strategic Stud 31: 633–660. Larnder H (1978). The origins of operational research. In: Haley KB (ed). OR’78. North Holland Publishing Company: Amsterdam, pp 3–12. McInnes C (1996). Hot War Cold War: The British Army’s Way in Warfare 1945-95. Brassey’s: London. Maguire R (2007). Scientific dissent amid the United Kingdom government’s nuclear weapons programme. Hist Workshop 63: 113–135. Moore R (1997). A JIGSAW puzzle for operational researchers: British global war studies, 1954–1962. J Strategic Stud 20: 75–91. Nye M (2004). Blackett: Physics, War, and Politics in the Twentieth Century. Harvard University Press: Cambridge, Mass. Peyton J (2001). Solly Zuckerman: A Scientist out of the Ordinary. John Murray: London. Pratt T (1981). A rose by any other name. J Naval Stud 7: 1–9. Price T (2004). Political Physicist. Book Guild Ltd, Sussex. Shephard R (1963). War gaming as a technique in the study of operational research. Opl Res Q 14: 119–130. The National Archives [TNA] (1944). CAB 90/5. Operational research in the Royal Air Force. S.A.C. (44), 6 April. TNA (1945). ADM 219/230. Future of Naval Operational Research Department, 26 July. TNA (1949). WO291/2459. The Army Operational Research Conference, 21 April – 30 April: Organisation and Role of the [Scientific Adviser’s] Department in the War Office and Army Operational Research Group, Appendix c. TNA (1956). AIR14/4202. The effect of low-level tactics on the operational range of the valiant, March 1956. TNA (1959). WO291/2233. AORG, The army operational research group tactical war game: Report No. 6/59. TNA (1957, 1958, 1959, 1962). WO291/1519, AORG, Optimal burst heights in the tactical use of nuclear weapons: Report No. 2/57, March 1957; WO291/2191, Safety distances for the tactical use of
MW Kirby and MT Godwin—The ‘invisible science’
nuclear weapons, Memorandum No. J.5, May 1958; AORG, nuclear attack warning and reporting system, Memo No. J.8, December 1958; WO291/2263, AORG, Protection of the soldier on the nuclear battlefield, Memo No. K.9, October 1959; WO291/2318, Personnel targets in the defensive use of nuclear weapons, Report No. 9/59, November 1959. TNA (1958). W/15. ORS (BAOR) Liaison Letter No. 14, June 1958. TNA (1958). WO279/279, The corps tactical battle in nuclear war, April 1958. TNA (1959). AIR 14/4210. Target cover from remote dispersal bases overseas, October 1959. TNA (1961). AIR14/305. Calculations of the minimum warning times obtainable from the BMEWS Fylingdales Moor, of ballistic missile attacks on the United Kingdom, April 1961. TNA (1962). DEFE19/102. B.T. Price to S. Zuckerman, 9 June 1962. TNA (1965). DEFE 19/102. C. Moore to S. Zuckerman, 7 January 1965. TNA (1966). DEFE19/91. A resume of JIGSAW studies by I.J. Shaw, 19 April 1966. TNA (1969). DEFE 10/396. Ministry of Defence, Operational Analysis Committee: Military Priorities Sub-Committee, NATO Land/Air Studies. Exploratory Study No. 147, March 1969.
81
TNA (1970). DEFE48/13. First report on NATO land/air studies, March 1970. TNA (1970). DEFE48/78. Defence Operational Analysis Establishment, Description of the Central Front Game, Memorandum 7013, October 1970). TNA (1971). DEFE4/262/2. Ministry of Defence, Chief of Staffs Committee, 40th Meeting 23 November 1971: Confidential ANNEX; Chief of Staffs Committee, Operations in the Central Region of Allied Command Europe, DP 25/68 (final), ANNEX A to DP 25/68, 23 November 1971. TNA (1972). DEFE48/486. Defence Operational Analysis Establishment, Results of a Series of Central Front Games to Check the NATO Deployment Model Ready Reckoner, Working Paper No. DJ/147/7.6/43, Project No. 147, October 1972). TNA (undated). FO371.136916. jkt UNI192. 171/G: Brief on the Paper ‘A Possible Approach to Disarmament’ by the former Director of Operational Research, Admiralty. Zuckerman S (1988). Monkeys Men and Missiles: An Autobiography 1946-88. Collins: London.
Received August 2009; accepted September 2009