The Eighth Army at the Gothic Line,August–September 1944: A Study in Staff Compatibility and Coalition Command

Coalitions are a fact of life in twenty-first-century military operations. From Kabul to Kinshasa, force commanders have been dealing with contingents comprising troops from many and disparate countries, often with peculiar caveats as to how and when they can be employed, and usually with rights of referral to their home governments when it comes to what they will or will not do. Most of these operations have been conducted under the auspices of the United Nations or within the framework of an alliance, like the North Atlantic Treaty Organization (NATO). There are a few exceptions in which low-to-mid-intensity military operations or campaigns have been conducted by a single national force – the Russian Army in Chechnya (1999–2009), the Israeli Defence Forces in Gaza (2006 and 2014), or the Colombian Armed Forces against the Fuerzas Armadas Revolucionarias de Colombia (1964–2017), for example. But, for most western democracies, coalition operations have been the norm. This is nothing new. Almost all the major wars of the twentieth century involved military alliances or coalitions that forced senior commanders to incorporate contingents from other countries, even though their control of those contingents was never complete.

Coalitions were certainly the norm for operations conducted by the British Empire and Commonwealth during the first half of the twentieth century. From the Boer War to Korea, senior British commanders assembled military formations that included components from the self-governing dominions and the British Indian Empire. Such combinations of forces fought in Europe, the Middle East, Africa, and Asia – all at a time when the dominions were asserting their autonomy and nationalist movements in India were steadily gaining momentum. Any of the dominions could have refused to join in, as they did during the Chanak Crisis of 1922, or even walked away once hostilities had started, although none ever did. When it came to preparing and maintaining military forces in peace time, the dominions were free to do as much or as little (mostly the latter) as they saw fit, and when they did participate in imperial military efforts, their forces always came with caveats, like the condition that their contingents not be broken up, or components detached, without their consent. And yet, the military forces of the empire somehow worked fairly well together. How did they manage that? This article takes a reverse-engineering approach to gain some insight as to how the armies of the British Commonwealth became so compatible – and, therefore, relatively easy for commanders to use. It begins with an examination of operations conducted by the British Eighth Army in the late summer of 1944 to examine how common staff procedures and methods allowed Lieutenant-General Sir Oliver Leese, General Officer Commanding in Chief (GOC-in-C) Eighth Army, to assemble his multinational forces and break the much-vaunted Gothic Line. It then looks at how staff uniformity had been developed over months, years, and, in some cases, decades. The two principle findings of this inquiry are that staff compatibility helped Eighth Army operations at the Gothic Line and that serious measures to ensure uniformity of imperial military staffs had been in place for nearly 40 years.

The Eighth Army of 1944 makes for an interesting staff-compatibility study because Oliver Leese was not a great commander, which means that the achievements of the Gothic Line battles cannot be attributed solely to any military genius. Certainly, no military historian has ever ascribed that to Leese. Far from it; he has been rated as little more than a ‘guffawing guardsman’ at worst to a good set-piece battle commander at the very best, albeit one that ‘failed to appreciate the problems of exploitation and [who] was far too optimistic about what armour could achieve’. ¹ Some contemporaries and subordinates did like Leese’s personality. The Canadian official historian of the Italian campaign praised his ‘racy informality [and] the clarity of his presentation’ and one Canadian divisional commander described Leese as a ‘much loved and respected army com[man]d[e]r’. ² He was, as the saying goes, a bundle of contradictions: confident yet quick to scapegoat, details-oriented yet inclined to cut corners, personable and approachable with soldiers and subordinates yet given to a champagne-and-caviar lifestyle. Eldest son of a baronet who succeeded to his father’s title in 1937, Leese had been educated at Eton and later commissioned into the Coldstream Guards in 1914. He distinguished himself as a junior officer, being wounded three times during the First World War and admitted to the Distinguished Service Order (DSO) in 1916. During the interwar period, he passed through the Staff College, Camberley (1927–8), commanded the 1st Battalion of his regiment (1936–8), and served as the chief instructor at the Indian Army staff college, Quetta (1938–40). As deputy chief of the general staff for the British Expeditionary Force (BEF) during the battle for France in May–June 1940, Leese was, by most accounts, level-headed and rock solid; ³ and, after a series of divisional commands in the United Kingdom, he climbed the ladder to corps command in the summer of 1942. That was when the Eighth Army commander, General Sir Bernard Law Montgomery, who had instructed Leese at Camberley and who had admired his steadiness during the crisis of 1940, brought Leese to the Western Desert as a replacement for the ‘pedestrian’ William H. Ramsden in command of XXX Corps. ⁴ And Leese did reasonably well as a corps commander under Montgomery – in North Africa, in Sicily, and in Italy – and he succeeded Montgomery as GOC-in-C Eighth Army at the end of December 1943. At that level though, his limitations showed. His first major battle as army commander was Operation Diadem, the Liri Valley offensive of May–June 1944 that broke the Gustav and Hitler Lines and eventually led to the capture of Rome, but which was also beset by a slow exploitation phase, owing to a number of problems, not least of them Leese’s decision to stuff two heavy corps into a 6 km-wide corridor that had only one major road. ⁵ Perhaps Field Marshal Sir Alan Brooke summed Leese up best: ‘He certainly [was] not anything outstanding as a commander.’ ⁶

His second and last major operation as army commander – Operation Olive – was no easy task (see map). Leese had to break the Gothic Line, a formidable defensive network that stretched across the Italian peninsula, from Massa on the West coast to Pesaro on the east coast. In the sector occupied by the German Tenth Army, more or less opposite Leese’s Eighth Army, there were 8,944 metres of anti-tank ditches, 72,517 anti-tank mines, 23,172 anti-personnel mines, and 117,370 metres of wire obstacles protecting a line of defensive belts that included 3,604 dugouts, 2,375 machine-gun posts and 479 anti-tank gun positions. ⁷ Not all the Gothic Line defences were complete by the end of August 1944; but harried German efforts, including the conscription of Italian labour battalions during the summer, had ensured that the line was indeed defensible by the time the Eighth Army reached it. For his part in punching through the line, Leese had a truly polyglot army – two British corps, one Canadian corps, and one Polish corps, with divisions and brigades from seven different countries. Eighth Army, it should be noted, was one of two army formations in General Sir Harold Alexander’s Allied Armies in Italy (AAI), the other being General Mark Clark’s Fifth (US) Army, to which Leese’s Eighth Army attached another multinational formation – XIII (British) Corps, comprising two British divisions, one Indian division, and a Canadian armoured brigade.

OPEN IN VIEWER

Map.

Allied Armies in Italy (AAI): Concept of Operations, August 1944.

OPEN IN VIEWER

Table 1.

Eighth Army Order of Battle – August 1944*.

Eighth Army
X Corps
10th Indian Division
9th Armoured Brigade
V Corps
1st Armoured Division
4th Division
46th Division
56th Division
4th Indian Division
Corps Italiano di Liberazione
I Canadian Corps
5th Canadian Armoured Division
1st Canadian Division
21st Tank Brigade
II Polish Corps
3rd Carpathian Division,
5th Kresowa Division
2nd Polish Armoured Brigade
Army Reserve
2nd New Zealand Division
3rd Greek Mountain Brigade

*

Source: General Sir William Jackson, The Mediterranean and the Middle East, volume VI, Victory in the Mediterranean, part II (Uckfiled, East Sussex, 2004), p. 263.

Leese was no Napoleon, but he did do some things right. On the plus side, he was instrumental in convincing Alexander to change the original plan, which would have had the Eighth Army advancing through the Apennine Mountains from Florence to Bologna, while the US Fifth Army attacked between Florence and Pisa. ⁸ Leese’s legitimate concern was that his army, which was tank and gun heavy, was unsuited to fighting in the close, mountainous terrain of the Apennines, and he argued, therefore, for a shift of his attack axis to the east: ‘It was obviously advisable to fight this major battle in country where we could best exploit our great advantages in tanks, guns and aircraft, and the most suitable country appeared to be along the Adriatic coast.’ ⁹ Leese was right, and Alexander agreed on 4 August to what he called ‘two-handed punch’, first to threaten Rimini and Ravenna with the Eighth Army, and hopefully draw German reserves to the Adriatic front, and second to attack along the line Florence-to-Bologna with the Fifth Army. ¹⁰ To execute his part in the new plan, Leese’s army would be disposed (from west to east): X Corps holding a very broad front in the Apennines, V Corps attacking out of the Apennine foothills towards the coast, I Canadian Corps assaulting towards Rimini, and II Polish Corps advancing along the coast to capture the high ground northwest of Pesaro. The 2nd New Zealand Division and a Greek mountain brigade would be in army reserve.

Using the Eighth Army in the more open ground of the coastal plain made sense, but the manner in which Leese grouped his forces for the operation did not. The heaviest corps, Lieutenant-General Charles Keightley’s V Corps, which was the designated pursuit force for the operation, was not to advance in the open ground of the Adriatic plain but in the cramped and constricted foothills of the Apennines, not much of them conducive to exploitation by the 1st Armoured Division, let alone the five other divisions that Keightley had under command. ¹¹ Yet, in the more open and rolling ground of the centre, I Canadian Corps had only two divisions – one armoured and one infantry – plus the British 21st Tank Brigade. This initial grouping and deployment of forces made little tactical sense and probably owed more to Leese’s lack of faith in the I Canadian Corps commander, Lieutenant-General E.L.M. Burns, than it did with any consideration of ground or enemy dispositions. ¹² Whatever Leese’s rationale for the corps groupings, his decision would bedevil operations once they got underway.

Pre-battle preparations still proceeded efficiently and expeditiously, largely as a result of competent staff work. With the change in the army group plan, there were only three weeks in which to get ready for operations in the Adriatic sector. There was much to be done and many parts to move. XIII Corps had to complete the capture of Florence and transfer to Fifth Army command. It also had to detach the 6th South African Division and send it to IV (US) Corps, transfer the 4th British Division to V Corps, and release the 2nd New Zealand Division to Eighth Army reserve. X Corps had to close out operations that had been initiated at the beginning of August to capture key mountain passes, transfer the 4th Indian Division to V Corps and take over a massive section of the front in the Apennines, all the while concealing the fact that the sector had been significantly thinned out. II Polish Corps had to drive back the enemy in the coastal sector and secure the Operation Olive start line on the south bank of the Metauro River, which the Poles did by 22 August. And I Canadian Corps had to be reassembled (1st Canadian Infantry Division had been attached to XIII Corps in the mountains) and brought undetected into the line, between II Polish Corps and V Corps. These administrative moves do not exactly make for scintillating reading, but they were important, and complicated in execution. The British official historian described them as ‘the immense organisational, engineering and administrative task of moving 8th Army over to the Adriatic coast in absolute secrecy’. ¹³ While Leese conferred with subordinate commanders, refined battle plans, and made speeches, the staffs at all levels got on with the work of moving the army into place for the attack. None of it was simple. It involved reconnoitring routes to ensure that they could accommodate weighty vehicles like tanks and self-propelled guns, conducting engineering works to strengthen routes where necessary, planning and executing myriad road moves such that they would not be detected by the enemy, and providing adequate supplies to all units and formations, especially those that had moved to new locations. That everything was in place for a start date of 26 August was a remarkable accomplishment of staff work and sound operating procedures.

Some of these procedures and practices had been refined since the Liri Valley offensive of May–June 1944. During those operations, traffic jams, inexperienced staffs, and advances that were too deliberate had all combined to stall pursuit operations after the breaking of the Gustav and Hitler Lines by Canadian and British forces. As alluded to earlier, Leese compounded problems when he crammed I Canadian Corps and XIII Corps into the narrow Liri Valley corridor, although he never admitted it. He blamed Burns for that. But he did later take advantage of a lull in the fighting to pull the Canadians out of the line and give them several weeks to train and work on improving staff procedures and methods for traffic control, which they did. The headquarters of I Canadian Corps and the 5th Canadian Armoured Division, both of which had fought their first major action in Liri, made some personnel changes and practised staff procedures in numerous skeleton exercises. ¹⁴ They even developed a new system for traffic control ‘that resembled the block system used by railways’, whereby routes would be divided into sections, with traffic control posts and assigned waiting areas between each section (or block). ¹⁵ This arrangement allowed for low-priority traffic to be shunted into waiting areas, while higher-priority traffic, like armoured reserves, moved forward on unclogged routes. In the run-up to Operation Olive, Leese also emphasized to all his corps that they must be willing to blow past delaying forces and even ‘gate-crash’ the Gothic Line if the defences were found to be ill-prepared or lightly manned. ¹⁶ Not all formations had a chance to practice ‘gate-crashing’ – many, as noted above, were still engaged in operations against the enemy – but the Canadians, who were in training mode most of the summer, conducted several exercises designed to chase down a disorganized enemy and speed up the tempo of their operations. ¹⁷

Indeed, based on Leese’s ‘gate-crashing’ guidance and the need to maintain momentum in the assault, all three of the attacking corps – the Poles along the coast, the Canadians in the centre, and V Corps in the foothills – built some flexibility into their plans. They determined to close the distance between the Arno River Line (Metauro River in the Adriatic sector) as rapidly as possible. The Poles led with the 5th Kresowa Division, the Canadians with 1st Canadian Division, and V Corps with 4th Indian and 46th Divisions. If those lead divisions reached the Gothic Line proper and found the defences to be poorly organized, they would ‘gate-crash’ immediately. If, however, they found the defences to be formidable, they would bring up reserve divisions and prepare for a set-piece attack. And, as it turned out, this is what happened. First impressions of the lead divisions that reached the Foglia River on 28–9 August were that the defences appeared to be ‘very much stronger than we had expected’, as Leese later relayed to the War Office. ¹⁸ Plans were therefore put in place for a deliberate assault against a dug-in enemy, and reserve divisions were brought forward, as planned, to add weight to the assaults – 3rd Carpathian Division in the Polish sector, 5th Canadian Armoured Division in the middle, and 56th Division on the V Corps front. Leese also arranged additional fire support. Desert Air Force squadrons bombarded the Gothic Line on 29 August and I Canadian Corps, which seemed poised to make the first breach, was augmented with the divisional artillery of both the 4th Division and the 2nd New Zealand Division. Everything seemed to be moving in the direction of a set-piece attack until the lead divisions sent patrols across the Foglia River to reconnoitre the enemy defences.

That changed things – in a way that demanded quick and adroit action. On 30 August, one patrol from the 5th Canadian Armoured Division ‘actually moved along the lateral road [in front of Montecchio] in the open, in broad daylight and drew no fire’. ¹⁹ That piqued curiosity in the divisional headquarters, prompting the divisional commander, Major-General Bert Hoffmeister, to move forward and take a look at the defences from a patch of high ground just south of the river. Not much seemed to be happening: ‘We could by careful examination pick out the odd concrete gun emplacement, and we could see the barbed wire, and we saw the minefields; but there was no life around the place at all. I didn’t expect to see German officers swanking up and down but the whole thing looked terribly quiet.’ ²⁰ It was terribly quiet because the Germans were in the process of executing a most disorganized relief-in-place. The 71st Division and the 1st Parachute Division were in the final stages of withdrawing behind the Gothic Line position, but they were doing so before the 26th Panzer Division – still rushing forward from the Bologna sector – had properly taken their place in the line and the 98th Division had relieved the 71st Division. ²¹ An opportunity for gate-crashing had surely presented itself, not because the defences were incomplete – concrete gun emplacements and obstacles were in place – but because the defences were virtually and temporarily unmanned. Hoffmeister ordered his troops across the Foglia and into the Gothic Line defences, and he did so without preliminary artillery bombardment that could have spoiled the surprise. I Canadian Corps Headquarters was stunned: ‘MONTECCHIO [is] clear [and a] battalion is being put through. Is this correct?’ ²² It was, much to the surprise and delight at Eighth Army headquarters. Leese may not have initiated the action, but he did recognize the opportunity that offered and he took action to support what the Canadians were doing: ‘It was obvious that the lines were not properly held, so I asked that the [air] bombing should cease at mid-day so as to enable me to order forward divisions to send down strong detachments of infantry and tanks to seize the high ground north of the Gothic Line.’ ²³ In support of that objective, staffs at army, corps, division, and brigade headquarters coordinated intricate road movements that ensured the fighting echelons – and the artillery that supported them – were poised to break through the barely manned defences. ²⁴ Some gruelling fighting still ensued as Germans scrambled to plug the breach, but I Canadian Corps did manage to crack the crust of the Gothic Line defences within 24 hours. ²⁵

Unfortunately, Leese’s earlier decision to place the pursuit force (V Corps) in the mountains precluded the full exploitation of this initial success. The Gothic Line was deep, so, by the time the Canadians had fought through the first series of defences on 1 September, they were pretty well spent. They had committed all of their fighting units and had very little left with which to pursue and cut off the remnants of the 1st Parachute Division, which were hastily withdrawing along the coast. Burns threw together an ad hoc pursuit force based on the 21st Tank Brigade and the 2nd Canadian Infantry Brigade, which fought very well on short notice, but it lacked the combat power to punch through to the coast and intercept the withdrawing German paratroopers before they made it across the Conca River on 2 September. By that time, incidentally, Leese’s exploitation force, the 1st British Armoured Division, had only just snaked its way through the mountains to the Metauro River, more than 40 km away from the spear point of the attack. ²⁶

This disappointing stage of a battle that had started with such precision and promise demonstrated both the strengths and the limitations of common staff and operating procedures. In the strengths ledger, units and formations – British, Canadian or New Zealand in this case – could be mixed, matched, and moved efficiently because they were all operating and communicating on identical lines. There was nothing new about British tank units supporting Canadian formations and vice versa. The 25th (British) Tank Brigade had supported 1st Canadian Division during the Liri Valley battles in May and the 1st Canadian Armoured Brigade had a long and productive relationship with the 8th Indian Division. ²⁷ Leese even referred to the latter pairing as ‘great buddies’ as a result of their long and successful affiliation. ²⁸ But, in the limitations ledger, common staff procedures, organizations, and armaments could not compensate entirely for bad tactical decisions. They never can. The forces at Leese’s disposal may have been well-tuned instruments, as was the staff nervous system that animated them, but no amount of good training, modern equipment, or efficient staff work could move an armoured division 40-plus km, from the mountains to the coastal plains, in a matter of hours.

Initial errors notwithstanding, it is worth probing more closely what worked well and why. For one thing, recovering from the initial grouping problem was not difficult. When the pursuit battle ground to a halt 25 km beyond the Foglia River, Leese was able to make adjustments, the first and most important of which was to switch the main effort to the corps in the best position to lead the advance and pursuit: I Canadian Corps. That meant attaching several non-Canadian formations to Burns’s corps. Thus, the 4th Division, the 25th Tank Brigade, the 2nd New Zealand Division, and the 3rd Greek Mountain Brigade all plugged into I Canadian Corps from 9 September. This allowed the Canadians, in conjunction with a supporting attack by V Corps, to break the impasse at Coriano Ridge on 13 September. Burns also used the 2nd New Zealand Division and the 4th Division to push beyond Rimini and the Marecchia River ten days later. It was easy to mix and match brigades, divisions, and corps, because all the formations of the Eighth Army had organized their headquarters in the same way and all worked in accordance with standard operating procedures. For the most part, a brigade was a brigade and a division was a division – no matter the national origin. ²⁹

Artillery support was just as easily arranged across national lines. For the Gothic Line operation, each attacking corps was supported by an Army Group Royal Artillery (AGRA), which supplemented the fire of the divisional artillery organizations. ³⁰ The normal order of battle for 1st Canadian AGRA, for example, was one field regiment and three medium regiments. But, for the Gothic Line battle, it was beefed up with two British regiments – 3rd Medium Regiment Royal Artillery (RA) and 32nd Heavy RA – while its field regiment was detached. ³¹ For the assault at the Foglia River, the Canadians also had the supporting fire of the 4th and the 2nd New Zealand divisional artillery organizations. The attachment of British and New Zealand guns caused no undo friction because it was hardly different than adding more Canadian guns to the mix. British and dominion artillery regiments were organized in the same way, equipped with the same guns, and firing the same types of ammunition, so it was really just a matter of finding the road space to move the British and New Zealand guns forward and placing them in reserved ‘gun areas’ from which they could support the attack. Everyone understood each other, so reallocating fire support was simple. In the 12–13 September attack at Coriano Ridge, for example, 1st Canadian Medium Regiment, 3rd British Medium Regiment, and the 56th British Heavy Regiment supported V Corps from 6pm to midnight on 12 September then switched to supporting I Canadian Corps from 12:30 a.m. to 7:00 a.m. on 13 September. ³² Even Polish artillery regiments, attached later in the operation, proved easy to work with because they too were organized and operating on British lines. Making that particular arrangement work demanded the attachment of two English-speaking Polish liaison officers, through whom communications with supporting Polish gun units was passed; but, as the Canadian after-action report noted, it still ‘worked well’. ³³ It would not have worked nearly so well had the Polish liaison officers been required to translate calls for fire, not only into Polish but into a peculiar national terminology as well.

How did the multinational units and formations under Leese’s command come to be so compatible? For the British, dominion, and Indian armies, this outcome had been decades in the making, stretching at least as far back as efforts to fix the British Army in the wake of its disappointing performance during the South African War (1899–1902). ³⁴ Notwithstanding British victory in that conflict, the experience had exposed very serious problems with the army. In 1903, a royal commission under the 9th Earl of Elgin concluded that the British field force that had assembled to defeat 50,000 Boer rebels was not really an army in any modern sense. ³⁵ The field force that peaked at roughly 250,000 had no general staff to guide it. It comprised different national contingents with different orders of battle and different operating procedures. Even the forces of the United Kingdom – the regular army, the militia, the volunteers, and the yeomanry – did not have identical war establishments, operating procedures, or terms of service. Field formations, which did not exist in peace, had to be assembled ad hoc in war. And, probably most disconcerting, the field force, even though it had been drawn together from across the empire, was small; at least in comparison to the standing armies of the continental powers it was small. The British Army, the core of the imperial armies, had to get better organized and it had to be capable of getting bigger, which in turn meant making the armies of India and the dominions compatible so that their formations could be integrated with those of the United Kingdom in time of crisis. As the Elgin commissioners noted, “[T]he true lesson of the [South African] war is that no military system will be satisfactory which does not contain powers of expansion outside the limit of the regular forces of the Crown, whatever that limit may be.” ³⁶

British political and military authorities set themselves to achieving that goal. ³⁷ They reorganized the British Army itself, reforming the auxiliary forces (the volunteers and the yeomanry) into a territorial force that was organized and equipped in the same way as the regular army; and, from the militia, they formed a Special Reserve that would fill out the ranks of regular battalions on mobilization. ³⁸ They reorganized the War Office and created a general staff, which extended to all the commands and military schools in the United Kingdom and, within a few years, to India and the dominions as well, although the Imperial General Staff had only an ‘advisory’ role when it came to the dominions. ³⁹ They established common organizations (such as the standard infantry division) and common procedures (such as the Field Service Regulations), all of which were accepted by India and the dominions. They increased the number of staff-trained officers to fill positions on capital and formation staffs by boosting the output of the staff college at Camberley, arranging a special course in administration at the London School of Economics (LSE), creating an Indian Army Staff College in 1905, and reserving staff-training vacancies for dominion officers. ⁴⁰ They loaned scores of staff-trained officers to the dominions to standardize their nascent armies and perform functions for they did not as yet have anyone qualified. In sum, they created a system that ensured imperial compatibility and established a framework for expansion.

That system paid dividends. As one British officer who had worked in South Africa at the time later commented, ‘the system made a tremendous difference, and helped towards the united Empire Army, trained and organised on the same lines, which by the end of the Great War, was the strongest military force fighting on the side of the Allies’. ⁴¹ Indeed, the military forces of Britain, India, and the dominions increased by factors of five-to-ten between 1914 and 1918, nearly 8.5 million troops being mobilized among them. ⁴² By 1916, for example, there was a 60-division British Expeditionary Force (BEF) in France and Belgium, to which the dominions contributed 10-plus well-equipped and British Army-compatible infantry divisions and a cavalry brigade, while the Indian Army fielded two cavalry divisions. And, while no credible historian of the First World War would deny that there were inevitable growing pains and rough patches of battlefield performance, a burgeoning body of scholarship has made a convincing case that the armies of the empire actually learned lessons evenly and fought together well. ⁴³ The Canadian and Australian Corps, for example, switched easily between British field armies, occasionally with a British division attached to them, and usually supported by someone else’s gun regiments. The Canadian Corps fought the battle for Hill 70 under Sir Henry Horne’s First Army in August 1917, the battle for Passchendaele with Sir Herbert Plumer’s Second Army in October–November 1917, and the battle for Amiens with Sir Henry Rawlinson’s Fourth Army in early August 1918. Compatible staffs contributed to compatible formations.

The real difficulty was producing enough trained officers to man those staffs. This was one of the most crucial lessons of wartime mobilization. Divisions, corps, and field armies needed trained staff officers to function – captains to perform staff duties at what was called a third-grade level, majors for second-grade duties, and lieutenant colonels at the first-grade level. In fact, the BEF in France alone needed 3,310 of these captains, majors, and lieutenant-colonels by 1918. The trouble was that, all together, the armies of the empire had only 1,433 staff-trained officers in 1914, and nearly one-third of them were above the rank of lieutenant-colonel. ⁴⁴ This critical shortage was badly exacerbated by the inexplicable decision to close the staff colleges at Camberley and Quetta in 1914. In fact, not until December 1915 did an expedient staff school open at Hesdin, in France, where selected candidates followed two intensive six-week courses – a senior staff course that trained approximately 20 candidates per serial for lieutenant-colonel-level (first-grade) appointments and a junior staff course that prepared selected officers for staff appointments at the major-level (second-grade). These courses, which ran at Hesdin between December 1915 and April 1917 and later at Clare College, Cambridge between October 1917 and November 1918, accommodated roughly 20 officers per senior serial and 30 to 50 per junior serial. All told though, they never produced more than 1,000 graduates, which was still clearly insufficient for the needs of the BEF, let alone forces in the Middle East. ⁴⁵ The bulk of the staff deficit was actually made up by a ‘staff learner’ system, an on-the-job training scheme by which officers, mostly captains and majors, learned staff duties while doing them under the direct tutelage of Camberley and Quetta graduates of proven ability. ⁴⁶ Regular divisions and early forming New Army, territorial, and dominion divisions tended to have higher numbers of staff college graduates, so some redistribution of staff talent was necessary as the armies expanded, but military authorities managed to make it work. ⁴⁷ The patchwork system of staff courses and staff-learner appointments eventually produced the complete staffs that the armies of the empire needed to function, but it was not an ideal way of doing things, particularly when it came to ensuring uniformity of what was taught. Military planners all over the British Empire understood this point and would bear it in mind when war came again in 1939.

Maintaining the high level of interoperability that had been achieved during the First World War was a challenge during the interwar years. It had to be nurtured because, after the armistice, everyone demobilized most of their armies, slashed defence spending, and hoped they would never have to fight another great war. Even so, no one questioned the requirement for military compatibility, not least the British. As a British general staff paper from February 1919 stated, ‘The military forces of the Empire can no longer be adequately considered departmentally or even territorially.’ ⁴⁸ None of the imperatives that had forced the British to rethink army organization after 1902 had gone away. The population of Great Britain was still small relative to most other great powers and there were even more global commitments than there had been before the First World War, all of which meant that manpower would continue to be an issue, which, in turn, meant that British political and military authorities still had to be prepared and organized to leverage the manpower of the United Kingdom, the dominions, and India. Suffice it to say that, given the high debt levels and dire economic circumstances of the interwar years, measures to ensure imperial compatibility had to be cheap – and they were. The emphasis, as it had been during the decade preceding the First World War, was on the Imperial General Staff and the professional cadres of the various national armies. Britain, India, and the dominions resumed exchanges and loans of officers. Quarterly correspondence between the chief of the imperial general staff (CIGS) and the general staffs of the dominions and India was another important – and inexpensive – measure. ⁴⁹ In their periodical letters, the various chiefs advised each other of local defence plans, the progress of individual and collective training, defence budgets, and the state of their military establishments. They also circulated the findings of various equipment and organizational trials and the CIGS kept everyone apprised of the newest War Office publications and of the latest revisions to old ones, like the Field Service Regulations (FSRs). ⁵⁰

But the main effort of measures to maintain imperial interoperability was in common military education. This was the investment that had the highest payoff, by far. The dominions continued to send selected officers for staff training at Camberley and Quetta. In fact, they sent even more than had been the case during the pre-war years, especially Australia and Canada, and the rate of dominion officers passing through the staff colleges actually accelerated from the mid-1930s, as rearmament programmes got underway. Australia, for example, sent six permanent force officers for staff training before 1914 and 61 between the wars, nearly every one of whom went back to key command and staff appointments in the commonwealth. The situation in Canada was similar. ⁵¹ A steady trickle of permanent force dominion officers and non-commissioned officers (NCOs) also made their way through the Royal School of Artillery at Larkhill and the School of Military Engineering at Chatham, returning to their own dominion schools where they taught the imperial way of doing things to their own armies. The Imperial Defence College (IDC), established in 1927, also helped. ⁵² The tri-service institution at Buckingham Gate schooled 291 British, Indian, and dominion officers, mostly at the level of lieutenant-colonel, on matters of imperial defence and strategic planning. ⁵³ It also offered an opportunity for a few important soldiers to get to know each other. Alan Brooke (future CIGS), Claude Auchinleck (future commander-in-chief India), and Andrew McNaughton (future commander First Canadian Army) attended the same course in 1927. At one point, they were even in the same syndicate. Being course mates at IDC was no guarantee people would get along later in life, however. Brooke, for example, would later be instrumental in removing McNaughton from command of the First Canadian Army. ⁵⁴ But IDC and other inexpensive measures kept interoperability on life support during the interwar period. As one Canadian general later reflected, ‘Our army was indeed British through and through with only minor differences imposed on us by purely local conditions.’ ⁵⁵ Not that the military education and training were perfect; far from it. Historians David French and Edward Smalley have argued that the curricula of the staff colleges were inadequate because they never really reconciled whether the focus was on either command or staff functions, and, as French writes, ‘placed too much emphasis on training officers in the arts of grand strategy and too little on learning operational techniques’. ⁵⁶ This is a valid argument. Nevertheless, the military education that select British, Indian, and dominion officers received at the staff colleges was uniform, and it imparted a common language that allowed them to work together in war.

Common understanding of the organizations, concepts, and technical language for conducting military operations was vitally important to the mobilization of British Empire armies in the Second World War. In many ways, the ‘start state’ for building formations and the staffs that would animate them was worse in 1939 than it had been in 1914. There were fewer staff-trained officers available, for one thing – just 1,419 of them in the British, Indian, and dominion armies combined – and this time less than half of them were below the rank of lieutenant-colonel. ⁵⁷ The shortage of second- and third-grade staff officers (majors and captains) had the potential to handicap severely the expansion of the British Commonwealth armies, especially after June 1940. Even captains and majors who held the psc qualification (passed staff college) ascended rapidly in the void to general rank, leaving a potentially massive gap at the lower staff levels. Of nine psc-qualified divisional commanders at the Gothic Line, for example, three had started the war as lieutenant-colonels, three as majors, and three as captains. ⁵⁸ This sort of progression was even more pronounced in the dominion armies. All three of the brigade majors who deployed to England with the 1st Canadian Division in 1939 were commanding divisions by the summer of 1944, one of them, Christopher Vokes, at the Gothic Line. ⁵⁹ And the divisional General Staff Officer Second Grade (GSO II), Major Guy Simonds, who deployed with them in 1939, was commanding a corps in Northwest Europe. More staff officers had to be made – thousands of them – and quickly to fill second- and third-grade appointments. Understanding how difficult this task had been during the First World War, military authorities wasted no time in getting started. As early as September 1939, the British Army initiated the first of dozens of wartime staff course serials, the directing staff initially drawn, of course, from the 1,419 trained staff officers on hand. Carefully chosen directing staff condensed the two-year curricula of Camberley and Quetta to four-month serials that conveyed only the basic skills that candidates needed to function as staff officers in formation headquarters – tactics, army and air force organizations and weapons, signals, staff duties, the conduct of liaison, intelligence, administration, and training. ⁶⁰ Nice-to-have items like historical studies and staff rides disappeared from syllabi. Bread-and-butter tactics and staff duties were the core: ‘Infantry Brigade HQ’, ‘Organization and Layout of a Division HQ’, ‘Appreciations’, ‘Office Work in the Field’, ‘Operations Instructions and Liaison’, ‘Orders’, ‘Diagrams Sketches and Messages’, ‘Medical Services’, ‘Administration and System of Supply in the Field’, and ‘Road Movement’; all were tested and confirmed during ‘skeleton exercises’ on the attack, the defence, the withdrawal, and so on. ⁶¹

Camberley produced some 4,000 wartime staff graduates over the course of the war, but even that rate of production was not enough to furnish the staff officers that all the armies needed in 1940, 1942, or even 1944. ⁶² Consequently, other staff schools popped up all over the world. Quetta began running wartime staff courses in January 1940 for the Indian Army and, around the same time, another staff school opened at Haifa to furnish staff officers for Middle Eastern Command. ⁶³ The dominions, Australia and Canada in particular, found that they were not allotted sufficient vacancies at any of the three staff colleges to meet the demands of their expanding armies, so they soon set up staff schools of their own, all teaching from the FSRs and other War Office publications. The Australians, who had established their own Command and Staff College at Sydney in 1938, relocated the newly named Staff School (Australia) to Duntroon in October 1940, when it began offering 12-week courses based on the Camberley curriculum. The first Canadian War Staff Course, which ran at Ford Manor in Surrey from January to April 1941, did much the same, as did 11 subsequent Canadian war staff course serials, all of which were conducted at the Royal Military College of Canada (RMC Kingston). ⁶⁴ And the New Zealanders set up their staff school at Palmerston North in 1941. National peculiarities and the nature of operations in particular geographic areas naturally imposed some modifications of curricula. The Australians, for example, focused more on tropical warfare from 1942 onward. But, it was important not to deviate too far from the Camberley standard because, as one Canadian general put it, common curricula for staff training was essential to ‘organizing our forces in the same way, [and to] writing our orders in identical manner’. ⁶⁵ Exchanges of directing staff helped keep things fairly uniform. Of the 73 directing staff who taught on one or more serials of the Canadian war staff course, for example, 13 were from the British Army and one was Australian. ⁶⁶ Camberley and Kingston also exchanged course papers and directing staff notes. ⁶⁷ By 1942, there were also junior staff schools specifically tailored to preparing officers for third-grade staff responsibilities (captain-level) or appointments in static headquarters – one at Oxford (later Sandhurst) the other at Sarafand in Palestine.

Very nearly all of the staff officers working in the headquarters of the Eighth Army’s brigades, divisions, and corps during the summer of 1944 were products of wartime staff training, the notable exceptions being only the most senior commanders and staff officers, nearly all of whom had graduated from Camberely or Quetta between 1919 and 1939. Leese, for example, was a staff college graduate, as was his chief of staff (Major-General G.P. Walsh), who had started the war as major, and his chief of administration (Major-General E.M. Bastyan), who had been a captain in 1939. ⁶⁸ All four of Leese’s Anglo-Canadian corps commanders had qualified psc between the wars as did nine of fourteen (64 per cent) British, Indian, or dominion divisional commanders. ⁶⁹ The percentage of psc-qualified divisional commanders in the Eighth Army had not changed since the Second Battle of Alamein (October 1942), when the percentage was 63.6, ⁷⁰ although the proportion of brigade commanders with a psc qualification dropped slightly, from 33 per cent (10/30) in 1942 to 26 per cent (11/42) in the summer of 1944. Staff officers at the level of lieutenant-colonel and below, however, were overwhelmingly graduates of wartime staff courses. In the Canadian Corps, for example, the divisional chiefs of the staff, the divisional chiefs of administration, and all of the brigade majors had learned their craft at one of the wartime staff schools. ⁷¹ There was no need for on-the-job trained staff learners. The Canadians trained 1,398 officers for staff duties over the course of the war, most of them at Kingston, but 20 per cent (275) had trained at other Commonwealth schools. ⁷²

Formations from outside the British Commonwealth were also staffed with graduates of wartime schools because, typically, a half-dozen or so vacancies on each staff course serial were reserved for exiled allies. The fourth junior staff course at Sandhurst, for example, had two candidates from Poland, three from the Netherlands, one from the Free French Forces, one from Czechoslovakia, one from Belgium, and six from Canada, while the eleventh serial at Sarafand in the autumn of 1944 included two Poles, two Greeks, one Canadian, and one New Zealander. ⁷³ Everyone learned the British staff language and methods. Thus, when the quartermaster-general staff at Eighth Army headquarters issued orders for road movements and the dumping of artillery ammunition to I Canadian Corps in advance of Operation Olive, the staff officers receiving the orders understood what they were being told to do and they quickly relayed that direction to their subordinate formations in the same language. This is what allowed Leese to move his army out of the Apennines and into position for major offensive operations in the Adriatic sector in less than three weeks. It also allowed him to recover from his initial misallocation of forces and regroup his army for the attack at Coriano.

Those relatively easy working relationships look easiest when examined in comparison with the experience of British Commonwealth forces who worked outside the imperial family, so to speak. The 6th South African Armoured Division, for example, was attached to the Fifth (US) Army from August 1944 until the end of the war, but it did not take long for friction to develop between the South Africans and the Americans as a result of misunderstandings. By October 1944, the GOC and his staff were complaining to South African authorities that American ‘staff procedures [were] cumbersome as compared to the British system’. ⁷⁴ Four months later, they were speaking more directly

about the possibility of a return of the div[ision] to the 8th Army … The Div[ision], although relatively happy in the 5th Army, have always felt not only a natural desire to return to the 8th Army, but they would obviously be happier under a British system of control, staff work, and a manner of conducting operations in which they had been trained. ⁷⁵

This was a common complaint of British or dominion officers who worked with American forces. Field-Marshal Lord Carver, as a brigadier in command of an armoured brigade attached to the 7th (US) Armoured division in October 1944, complained at having to decipher orders presented to him in ‘strange verbiage and format, which included a crudely drawn “trace”, which seemed to fit no map that I had’. ⁷⁶ None of these comments and complaints was fair, of course. American staff procedures were no more or less cumbersome than British ones. They were just different. The Americans found British Commonwealth methods equally unhelpful, as one American after-action report on the fighting in New Guinea noted, ‘Most of the difficulty of supply was occasioned by the mixed administration, Australian and American … Our allies are not familiar with our requirements or methods.’ ⁷⁷ The author was commenting on arrangements during the battle of Buna-Gona (November 1942–January 1943), when the 7th Australian Division worked alongside the 32nd US Division in the Territory of Papua. Terrain, strong enemy resistance, and disease caused problems, but so too did different staff procedures, doctrine, and equipment – so much so that, from that time, the two allies largely fought separate campaigns, as the Australian official historian has observed: ‘Each national army was as a rule given a separate area and this had the useful effect of avoiding the friction that may be produced when forces with different organisations, supply systems, tactical doctrines, and technical vocabularies are mingled.’ ⁷⁸ Friction affects both sides of whatever has been forced to rub together.

If the Eighth Army’s operations at the Gothic Line demonstrate anything, it is how commonality of staff procedures and methods can reduce friction within military coalitions. This explains, at least in part, how Oliver Leese was able to task-organize his multinational army, move it across the Italian peninsula for operations against a formidable defensive position in less than three weeks, and break through those defences in less than seven days. Switching divisions between corps, as necessary, was nearly as simple as reallocating artillery fire because the language of the British, Indian, dominion, and allied headquarters in his army was the language of the FSRs. That level of interoperability had been years in the making, decades in the case of the British Commonwealth armies. But as experience of the Polish corps demonstrates, in expedient circumstances, forces from outside the British system could be converted and made compatible in relatively short order. By the time of the Gothic Line battles in the late summer of 1944, the instruments of the Eighth Army were tuned such that even a limited conductor like Leese could orchestrate them reasonably well in battle.