As Robert Doughty demonstrates throughout Pyrrhic Victory, French strategy was considerably better thought out than is usually allowed and the evidence considered herein previously suggests that French tactical practice and thought was likewise more competent than is usually accepted. Both the Marne and the Champagne operations of 1918, illustrated the essential problem of offensive operations on the Western Front. The close planning needed for a successful offensive appears to have been only possible in the initial period of an operation, usually indeed only in the first part of the first day of such an operation. After that, whatever measures were taken prior to the attack to continue the struggle, the offensive power of the attacking forces would slowly weaken, due to an increasing lack of cohesion and slowing of its operational tempo, as the defensive forces grew correspondingly more powerful. The weakness in Allied offensive strategy before Foch took charge had been the expectation that one large operation could be used to break the German defences and achieve a breakthrough on the Western Front. Only by mounting a series of co-ordinated offensives could the Germans be successfully pushed back across the Western Front, such as those that the Battle of Soissons and the Champagne operations were part of. Was the difficulty of maintaining the tactical success of operations after the initial advance an inherent and insuperable problem of fighting on the Western Front or a failure of the armies involved to find ways to avoid this? The evidence presented in this book strongly suggests that it was the former. It is difficult to see how with the means available operations could have been conducted more successfully. Without better mobility on the battlefield and communications equipment, such as arrived for the next world war, it was extremely difficult to regain cohesion in units that had been fighting for a day or more, which in turn prevented a high tempo of operations being maintained. Various ways of dealing with these problems were tried; at Champagne the front-line divisions with 21 CA were retired for a day to refresh themselves, a tactic that seems to have worked if we compare the results with those divisions that were kept fighting. It certainly helped maintain better cohesion among the attacking troops.
The major problems for the Artillerie spéciale [AS] were that of its material and its effective integration with the infantry and artillery of the French army. In relation to the latter issue, there was simply not enough time to fully develop combined-arms warfare, although it seems likely that 1919 would have seen an even better developed French combined-arms approach. By then, the French infantry (at every level) should have been more au fait with tank tactics, as there would have been more time to disseminate effectively the established doctrine, particularly in relation to the infantry commanders. Foch had identified that tanks would be a crucial component of the Allied armies in 1919, when he expected the German army to be defeated. He said that ‘the aviation and the tanks should receive the greatest development possible’ for the campaign in 1919. The expected campaign in 1919, of course, never occurred but if it had there would have been a considerable number of tanks involved in the French army operations, including the new Char 2C heavy tank. Pétain wrote a memo to Foch on 8 September 1918 setting out his plans for 1919. He told Foch that ‘the battle of 1919 will be the battle of aviation and the tanks’. He was planning to use 360 heavy tanks and 3,360 light tanks per 30 kilometres of front for the next year’s offensives.
The evidence presented does not show that the performance of the AS was flawless; mistakes were clearly made at various junctures but in the midst of war this is only to be expected. An examination of the performance of the AS during the war shows it to be a military organisation that was adaptable and remarkably effective, particularly considering the primitive material it had to work with. This in itself suggests that the traditional view of the Great War French army, that it was steadfast but rather dim, is at considerable variance with the facts. In reality, it had became very effective by 1918 at fighting both in semi-open and positional warfare, although there is a clear need for more detailed research on this.
In relation to the tanks themselves, the issues revolve around two distinct areas; the technical problems and the organisation questions concerned with manufacture. Concerning the technical issues, most of these were probably inevitable as tracked vehicles were a new and undeveloped technology. However, certain problems really should have been avoided, such as the inadequate width of the St Chamonds’ tracks or the forward-mounted petrol tank of the Schneiders. The technical problems of the medium tanks were eventually for the most part solved by the end of the war, which is some achievement considering that they were in effect tested on the battlefield. The Renault was a very successful design, marred only by technical problems that were inherent in automobile technology of the time. Thus, of the three designs of tank used during the war, the light tank was excellent and the two medium tanks were serviceable. If the Schneider was simply an armoured box on tracks, primarily the result of wanting to get a tank into service as quickly as possible, the St Chamond was the victim of the over-ambition of its designers. However, it can be argued that the St Chamond was wrong for all the right reasons; if the track and engine problems had been cured it would have been a formidable presence on the First World War battlefield, particularly with its 36-calibre long 75mm high-velocity gun, a more powerful gun than initially carried on the Second World War German Pz. Kpfw. IV, for example. Both medium tanks were more sensibly armed than those of the British which, although heavier, carried only 57mm guns, in practice no more effective than the 37mm carried on the Renault. However, there was simply not enough time, resources and experience to resolve all the problems that arose during the war. In addition, 30 St Chamonds were still in combat as late as September 1918, so the effort of their manufacture was not entirely wasted. Where all the French tanks were poor compared with the British designs was in their trench crossing abilities; the Schneider and Renault could only cross a 1.8 metre-wide trench while the St Chamond was only slightly better (2.5 metres) but substantially less than the British Mark V (4.5 metres). Although Steven Zaloga characterises the Schneider as a ‘bitter disappointment’ to the French, this is far too harsh a judgement. The success of the French tanks on the first day at Soissons was as much due to the Schneiders (and to a lesser extent the St Chamonds) as to the Renaults, the latter only coming into action at the end of the day’s fighting. These designs were undertaken in wartime utilising undeveloped technology and brought from specification to use within 13 months, less in the case of the Renault. This is all the more remarkable when it is compared with later tank designs; for example, the German Tiger I took 15 months to enter service and the US M3 took 21 months during the Second World War.
The tanks of the Great War were pushing to the very limits of contemporary automotive technology. The mechanical reliability of the French tanks looks unimpressive but it is worth comparing experiences in the next world war, after a further 20 years of automobile development. For example, the French medium tanks’ propensity to catch fire was surprising to contemporaries but experience since then has demonstrated that this is an habitual problem for tanks. By comparison, fire was a considerable problem for the early Panthers; in a five-day period in July 1943, a quarter of the Panthers in an operation simply caught fire. In an attack on 11 July 1944, one German company lost 10 Tigers; only two were to enemy fire, the other eight caught fire. Eighty-two per cent of the Allied tanks hit during the Normandy campaign then caught fire. By comparison, Steven Zaloga believes the Renault’s performance and endurance was ‘miraculous by Great War standards’, which probably accounts for it still being in service, albeit modified, 20 years after the Great War. Estienne’s complaints in 1917 about the quality of drivers were still a concern for tank commanders in the next world war. For example, there were great difficulties with training tank drivers in Nazi Germany; many conscripts had not even driven before and this was in a society where motor vehicles were much more prevalent than during the Great War. The sorry story of the French heavy tank designs should also be compared with the difficulties in developing a heavy tank in the next world war and after, which are littered with unsuccessful designs such as the US T32 and M103 heavy tanks or the British A39.
While comparing the material of the two world wars, it is also instructive to consider some aspects of armoured operations of the Second World War. Many of these were not well-conducted, with less excuse as they took place after 20 years of armour developments. One good example would be Operation Goodwood during the Normandy campaign. Three British armoured divisions, with over 1,000 tanks, attacked well-organised German positions south of Caen. The tanks were given inadequate infantry support, which was also very badly co-ordinated during the operation. The plan required the tanks to advance on a 2,000-metre front across a 4,000-metre deep plain, with little cover and few hull-down positions available. Over 400 tanks were lost over three days, along with 6,000 casualties across the three tank divisions. In relation to tactics, as late as 1944, the Japanese Army was using tanks for infantry support in a manner that would have been recognisable to Great War armour practitioners; the tanks were used in direct frontal attacks and ‘after the infantry reached their objective the tanks were withdrawn without attempting to exploit the limited gains’. Tank-infantry training was also a problem in the Second World War; American Colonel Wright wrote in May 1943 of the difficulties that the US Army was having with this. The US infantry commanders were all agreed that this training was necessary but always had excuses as to why this had not happened; there was no time available, the infantry training was incomplete. He reported that the US 751 Tank Battalion was almost wiped out in one action because: ‘in four successive attacks, the infantry refused to follow him [the tank commander]. Four times he took the objective and each time he had to pull back, trying to pull the infantry forward, the Germans in the meantime re-obtaining the position.’ This situation would have been very familiar to AS veterans.
While considering the Second World War, it is worth noting that the French army’s performance in 1940 has cast a shadow over the French army’s performance in general during the twentieth century. The evidence considered in this volume strongly supports the idea that the French army developed an effective offensive methodology (often referred to as the bataille conduite or methodical battle) in the First World War, of which tanks were an integral part. That this methodology may have been outmoded from the late 1920s onwards is another argument but it is from the inadequacy of the French army’s inter-war doctrine and dismal performance in 1940 that the bataille conduite and French offensive methodology in 1918 have often been judged so severely. Specifically in relation to the French tanks, for example, André Loez argues that the failure of the French tank force in 1940 had its roots in the disastrous debut of the French tanks during the Nivelle Offensive, primarily because the experience in 1917 cemented the idea of the tank as an infantry-support arm in the collective mind of the French army. This argument fails to understand that the French tanks in the Great War were used as what would be known later as infantry-tanks because that was the only role that the existing technology could support. The AS was quickly amalgamated into the infantry after the war, something that Estienne opposed as he recognised that it could be developed as a highly mobile exploitation arm and needed to be separately organised to do this effectively. However, during the war he accepted that the tanks could do no more than act as direct infantry support, as all his wartime writing demonstrates. Thus, whatever failings can be traced in the inter-war handling by the French of their tank programmes, the French tanks during the war were used in the most sensible way and it simply unhistorical to argue otherwise.
The essential point to consider when evaluating the performance of the AS is that there was no prior experience of tank operations on which to base judgements about how these might transpire. Andrew Lambert has said about the submarine pioneers of the American Civil War that they had to ‘invent submarine operations’ and the AS officers and men had to likewise invent tank operations from scratch. Leader of the Great War British tank corps, Clough Williams-Ellis wrote that the difficulty of training the British tank crews was that no-one had any experience of tank warfare and therefore it was ‘upon the spirit of prophecy alone that they must rely in their preparations’. In evaluating the AS, the important factors are not what was got wrong, as mistakes were inevitable, but how effective the responses were to mistakes. The prime example of such a response was in the reaction to the terrible events of 16 April and the wholesale re-evaluation the AS made to its approach over the two weeks following this event. As has been described, there was a very rapid and effective re-evaluation of tactics made after the Juvincourt engagement, leading to a significantly better showing in the 5–6 May action at Laffaux. Indeed, such was the effectiveness of the AS and its officers that the French army had a sound tank doctrine in place by the end of 1917, after just three engagements, which only required small modifications to accommodate the differences between the medium and light tanks. Despite this, the tank regulations were subject to continuous discussions within the AS right up to the declaration of the Armistice. One example would be the discussion about the ratio of machine-gun to 37mm-gun Renaults, which continued into 1919. The final tank Instruction of 14 July 1918 was adopted in 1919 as the French army’s tank doctrine, with the most minor of changes from that of December 1917, mainly those needed to deal with the retirement of the medium tanks and their replacement by the British Mark V*.
The battles of Soissons and Champagne both illustrate that the French infantry commanders, from divisional level to below, frequently either misunderstood or decided to ignore the tank regulations, the latter being likely in the majority of cases. The French army cannot be accused of failing to promulgate information to its infantry and artillery commanders as numerous notes on tank use were sent out on a regular basis. Of course, in a conflict where infantry losses could be so high, it is easy to see why infantry commanders took every opportunity to mitigate their own losses, even if this was at the expense of the tank units. This led to many arguments between AS and infantry officers; most often when the former tried to insist on sticking to the prescriptions of the tank regulations. For example, the commander of 9 BCL had a ferocious row with General Massent (commander, 7 CA) about using the army reserve tank company on 21 July. This unit, AS325, was in the process of reforming; both personnel and tanks were in no condition for combat and 9 BCL’s commander refused to allow it to fight until it was properly ready. Massent asked Gouraud (commander, IV Army) to dismiss the tank battalion commander and he then issued orders for AS325 to go into combat the next day. The AS commander of IV Army, Chef de bataillon Michel, then countermanded Massent and returned AS325 to the army reserve. It can be observed that it took considerable courage for an AS major to countermand the orders of a General, even though the former eventually got Gouraud’s approval for his actions.
In addition, despite strenuous efforts to train as many infantry units with tanks, there was simply not enough time during the fighting in 1918 to train the whole army in combined-arms tactics. While we can sympathise with the frustrations of the AS officers who believed that the infantry commanders were not sufficiently aware of the tank regulations, this has to be balanced out with the difficulties the infantry commanders were already grappling with; their command was of vastly greater complexity than had been the case in 1914 and infantry losses appeared to be almost impossible to keep low. However, we should not lose sight of the fact that the tank regulations gave very good results initially within each engagement, which then rapidly diminished but this was usually nonetheless sufficient to defeat the German defences, as has been discussed.
However, the French army and the AS certainly appear to have been let down by the ministère de l’Armement in relation to the production of the tanks. The ministers, Thomas and Loucheur, interfered on numerous occasions well beyond both their responsibilities and competence, undoubtedly reducing the effectiveness of the AS. Although most of the French writers after the war attribute little difference between the ministries of Thomas and Loucheur as far as the tanks are concerned, statistics show that production of the Renaults increased substantially after Loucheur became minister. In mitigation, as discussed, both ministers were hard pressed to overcome both the organisational difficulties connected with French industry and the lack of resources, particularly steel. What cannot be doubted, however, is their considerable organisational flair; for example, during his tenure as minister, Thomas increased 75mm shell production from 380,000 per month to over 6 million per month. In addition, in the end a prodigious number of Renault tanks were produced by the Armistice; nearly 2,000 by the Renault factory alone.
The attitude of Thomas and Loucheur to the tanks is probably most easily explained as a simple failure to understand the tanks’ worth. This appears to have been a genuine misunderstanding, although the difficulties they subsequently presented to the tank programme are inexcusable, particularly in their apparent determination to produce most of whatever tank the army wanted least. Here we have an example of civilian rather than military opposition to the tank, in contrast with the popular view that it was the civilians that promoted the tank during the Great War. There is also ample evidence that civilian bureaucratic in-fighting did nothing to help the French war effort; for example, petty arguments were constant between Loucheur’s and Thomas’ staffs; in 1916, a request by the former for more telephones was turned down by the latter, as they felt that Loucheur’s staff spent too much time on the telephone.
Although it has been argued that the French tank effort was both effective and intelligently handled, albeit not entirely without flaws, the question remains as to whether the effort was worth pursuing at all. There have been arguments about the effect that tanks had on the Great War battlefield ever since they first appeared. Estienne wrote in 1931 that ‘in my opinion, the intervention of mechanised chariots on the field of battle gives the historian the appearance of an event as important as the invention of gunpowder and cannon’. By contrast, Foch told Repington in September 1918 that ‘it was an idea of amateurs that tanks and aeroplanes could win a war’ and that tanks were ‘accessories’ to the infantry. More recently a number of scholars have gone further and suggested that armoured warfare in the Great War was ‘marginal’, tanks being a ‘specialized luxury’. However, for the most part contemporary opinion was much impressed by the tanks. Weygand contradicted his old chief, by stating in his memoirs that France had won the war thanks to the tanks, an opinion shared by Hindenburg. Perhaps unsurprisingly, Léon Dutil argues that the Germans lost the war largely because they did not have enough tanks. Herbert Sulzbach refers in his diaries to a meeting in August 1918 where he discussed with his regimental commander a secret paper by Ludendorff. There was general agreement that enemy tanks had an ‘unbelievable effect’ in recent battles. It is difficult to determine whether the weight given in many post-war German accounts to the effectiveness of tanks is not coloured by the quest to explain defeat in terms that did not include a straightforward Allied military victory. This seems unlikely with Sulzbach, who is a reliable front-line witness. Heinz Guderian was writing as an advocate of armour but he was not directly involved in any actions with French tanks during the war and it is clear from his narrative that he was largely dependent on Dutil and the French official history for his information on the French tanks. However, he was an insightful military critic, in addition to being a successful military commander, and therefore a useful judge of operational matters.
What is rather easier to quantify than the effect of tanks in combat is their material and human cost, although in relation to the latter, calculating very accurate casualty figures is problematical. A variety of detailed post-war tallies give similar but not identical personnel losses and cross checking the official casualty list with these suggests we can have confidence that the official figures are broadly correct. The final official figures were for troops lost in combat; 302 killed, 2,350 wounded, with 253 missing, presumed dead, just over 13 per cent losses. This compares with infantry casualties over the war which ranged daily in combat from a maximum of approximately 40 per cent to a minimum of 22 per cent.
Although the AS took troops away from the other arms, the officers were initially mainly taken from the cavalry or the artillery depots and there were thus not many useful officers taken away from the infantry. As the war went on, officers coming into the AS were increasingly unfit for any other frontline duties and thus their service with the AS was a bonus for the army rather than a loss for the other arms. For example, when Maurice Constantin-Weyer started his training with the AS in early 1918, he could only walk with the aid of two sticks, two of his fellow officers had artificial limbs and another had lost one eye. Marcel Rime-Bruneau had been hospitalised for a year from his wounds in 1915 and then spent a year on non-frontline duties, while undergoing a long series of surgical operations, before joining the AS. In addition, the AS was comparatively small; on 1 August 1918, it comprised of 1,017 officers and 18,141 men, including the infanterie d’accompagnement with 45 officers and 1,581 men, serving over 3,500 tanks. This is compared with the artillery service which had over a million officers and men serving 12,000 guns and the air service with 95,000 officers and men serving 3,300 aircraft.
Deygas also points out that the lives of infantrymen saved by the tanks should be taken into account, although this is largely unquantifiable. However, there are numerous accounts of the AS saving infantry units from machine-gun positions, many of which have been discussed previously. In addition, the tank crews frequently dismounted when their tanks were out of action and added their machine guns to the infantry. For example, during the fighting on 18 July, a tank commander from AS35, Brigadier Cellier, dismounted his tank’s machine guns when the tank was immobilised and attacked the German strongpoint that had fired upon him, in conjunction with part of his crew and 15 US infantrymen. The German position surrendered shortly after, Cellier and his men capturing 15 officers, including a colonel, and 700 men.
The losses in material are rather more problematic to analyse, if we want to judge the effectiveness of the French tanks. Although the sources do not vary widely in relation to the total number of tanks lost or damaged, there is an interesting question that was first posed by Balland in the 1930s. He pointed out that only seven of the 29 tanks of Groupement XI (with 1 US ID), on 18 July 1918, actually got into the ‘zone of fire’ and, of those, six were destroyed. ‘Officially, the numbers of tanks destroyed in the two groupements was 30%. In reality it was 80%.’ Thus, it is clear that there is a problem revolving around which tanks are counted as having gone into action; for example, should one count all the tanks that left the waiting position (position d’attente) as getting into action, or should it be all that left the departure position? Or should we count only those tanks that reached the battle zone, as Balland suggests?
Jean Perré gives another variant on this. His figures count each tank’s engagement as a ‘tank engagement’, thus one tank could be engaged multiple times in a battle, each one adding a ‘tank engagement’ to the list. This has the unfortunate effect of masking how many tanks were actually in action against the enemy. In terms of judging the effectiveness of the tanks on the battlefield, Balland’s methodology seems to give a better indication of the vulnerability of the tanks to enemy fire than methods that count the total number of tanks engaged. However, there is simply not enough extant information to apply this methodology to all the engagements of the AS, although it strongly suggests that the official figures for tank losses in combat are lower than they should be. In terms of judging the overall effort and contribution made by the tanks to the French Army, these questions may be largely irrelevant, as it could be argued that the tanks’ impact on the battlefield was as much to do with their presence as with their direct combat fire.
The material cost of the French tank programme was carefully examined by Deygas in the 1930s. He calculated that the amount of steel and other metals used to construct the French tanks was around 35,000 tonnes. This should be compared with the 100,000 tonnes of steel per month being used in 1918 solely for artillery munitions. The total cost of the Schneiders’ manufacture was 25 million francs, the St Chamonds cost 40 million francs and the Renaults cost 150 million francs; a total of 215 million francs. Even adding estimated indirect costs, the creation of this new arm cost no more than 450 million francs. By contrast, the 20 August 1917 Verdun attack cost 700 million francs in artillery ammunition and Malmaison cost around 500 million francs. The consumption of petrol was relatively unimportant; in September 1918, the entire AS (i.e. the tanks and all other AS motor transport) consumed 3,000 hectolitres of petrol. During the same month, the French Army used 300,000 hectolitres for its 100,000 vehicles. Likewise, the expenditure of ammunition was negligible compared with the army’s consumption as a whole. For example, the tanks used 437 75mm shells during the Battle of Malmaison. By comparison, between 17 and 22 October, VI Army’s artillery fired just over 1.5 million shells at the German positions in the artillery preparation for this battle. Deygas estimates that the AS used less than 1,000 tonnes of shells throughout the war. He also points out that tank fire was generally direct-fire on an observed target, unlike artillery fire that ‘wasted’ many shells in general bombardments. These figures show that the French tank programme used a relatively small amount of the resources available to the French army.
It thus seems difficult to justify relegating the importance of the French tanks as they were an integral part of many successful French army operations during the war, for a comparatively small cost. Foch was, of course, right to say that tanks were not a war-winning weapon; only within what is now known as a combined-arms team could the tank operate effectively. It took the French a comparatively short amount of time to arrive at a relatively successful system of tank, infantry, artillery and air co-ordination, with all the pressures of unprecedented industrial warfare to contend with at the same time. Considering the resources devoted to the tanks, they repaid the investment in full. As Dutil points out in relation to the Renaults, a single machine gun required the same two-man crew as a tank, without the latter’s mobility or protection.
It is probably accurate to state that the French Republic had a ‘bureaucratic and unimaginative general staff’ at the start of the war but the evidence shown in this volume suggests that it adapted to the rigours of modern warfare, particularly in relation to armour, remarkably effectively as the war went on. As Michel Goya points out, the French army went through the deepest and most profound changes of its history during the war, finishing it as the most modern army in the world. The French army’s achievement is all the more remarkable considering the technological constraints of the time and its initial disastrous performance in 1914. This suggests a revision of GQG’s reputation is in order and it is clearly time for scholars to examine in detail other aspects of the French army’s performance during the Great War. Returning to Douglas Porch’s categorisation of the French army in the Great War, it is very hard to see how its performance could reasonably be considered unintelligent, particularly in reference to the performance of the officers and men of the AS.