The French strategic cavalry was composed of ten cavalry divisions. This strategic cavalry would be reinforced by infantry battalions and artillery. Each French corps had a light cavalry regiment assigned (six squadrons). There was no divisional cavalry. French reconnaissance patrols were to avoid combat. In reconnaissance and security the French relied on combined-arms teams to confuse the enemy concerning the location of the main body and force him to deploy. The French thereby separated reconnaissance, which was conducted far forward by the strategic cavalry, from security, which was the responsibility of the corps cavalry and at the infantry division, by local foot patrols.
In August 1914 the French cavalry failed to perform both the reconnaissance and security roles. The French cavalry divisions manoeuvred almost aimlessly. The French corps cavalry remained so close to the infantry that tactical security was non-existent. As a result, the French higher commanders were poorly informed concerning German operational movements and the French infantry was repeatedly surprised.
French Enemy Estimate
Observing the density of the German rail net behind Metz, the Deuxième Bureau, the French General staff intelligence section, concluded that the Germans would concentrate up to 11 corps behind the Metz-Diedenhofen fortress complex and in Luxembourg as a mass of manoeuvre and then shift those forces into Lorraine or Belgium. The French did not obtain any solid intelligence on the location of the German assembly areas during the German rail deployment, and therefore retained the pre-war assumption that the Germans would mass behind Metz. On 9 August the French thought that 17 German acive-army corps opposed them, while four corps opposed the Russians. Since the French had 21 active army corps, the French thought they had numerical superiority. They estimated that there were five or six German corps in Belgium, five to eight corps located at Metz-Diedenhofen-Luxembourg, with more on the way, one to three corps in Lorraine, a corps plus in Alsace. Five corps were unaccounted for.
In fact, the German armies were evenly deployed from Alsace to the north of Aachen. The German 4th and 5th Armies were behind Metz and in Luxembourg, but did not have the decisive role that the French ascribed to them. The French intelligence analysts had been trained according to the theories of Bonnal, who doctrinally employed a large mass of manoeuvre, and were mirror imaging – writing the German plan as a French officer would have written it.
The pre-war calculation of the Deuxième Bureau was that the Germans could attack as of the 13th day of mobilisation. Expecting to find the Germans in the northern Ardennes, Sordet’s Cavalry Corps of three divisions was sent into Belgium on 6 August and reached the area west of Liège on 8 August. On 9 August he found nothing at Marche. Neither he nor French aerial reconnaissance could find any German forces as far east as the Ourthe River because there were no German forces there, nor would there be any there until around 18 August. Sordet’s cavalry had moved ten days too soon. Nor did the Belgians provide much useful information. By 12 August Sordet had moved to Neufchâteau but still made no contact; he then pulled back to the west bank of the Meuse on 15 August and was attached to 5th Army. Sordet reported that it was impossible to supply the cavalry in the Ardennes and that air recon was unreliable in the dense woods. His cavalry corps had conducted an eight-day march without obtaining any information concerning the German forces. In order to find the German 3rd, 4th and 5th Armies, the French cavalry would have had to advance across the Belgian Ardennes to the border with Germany and Luxembourg; it was unable to do so. The German deployment was not completed until 17 August and the German 5th and 4th Armies did not begin their advance until 18 August. The French had great difficulty understanding why the Germans were not as far to the west as they expected them to be.
By 10 August, the French saw indications that the Germans were digging in on the Ourthe between Liège and Houffalize. The French intelligence summary on 13 August reported that in the Ardennes there were only two German corps (VIII AK at Luxembourg and XVIII at Aumetz – the latter was actually XVI AK) and two cavalry divisions. The French were beginning to get the impression that there were no German troops in the Ardennes. This was not an illogical conclusion. It is more than 100km from the sparse German railheads in the Eifel, in the German Ardennes, to the Franco-Belgian border. The Ardennes is thinly populated and heavily forested, with few and poor roads. Crossing it would pose significant problems in supply and traffic control. At the end of the approach march lay the Meuse River, a formidable obstacle. It would seem unlikely that the Germans would commit significant forces from the very start of the campaign into such an out-of-the-way and difficult theatre of war.
In the skirmishes between cavalry and foot patrols during the first week of the war, the French thought that their troops were generally victorious, returning with prisoners, horses and weapons. The chief of staff of VI CA said that ‘this filled them with great joy.’ French pre-war predictions of the natural superiority of the French soldier seemed to be justified.
Between 7 and 10 August the French VII CA had advanced towards Mühlhausen in the upper Alsace and been thrown back into France by the German XIV AK and XV AK. On 14 August the French 1st Army and 2nd Armies attacked into Lorraine. Joffre was fully aware that the German forces to the east of Metz could attack through the fortress to the south into Lorraine: he gave the 3rd Army the mission of attacking any such German sortie in the flank with two corps, while on 15 August he told the 3rd Army to be prepared to invest Metz from the west
By 15 August the French recognised the strength of the German forces in the general vicinity of Liège. Joffre told the commanders of the 4th and 5th Armies that the Germans were going to make their principal effort ‘to the north of Givet’ with a second group marching on Sedan and Montmédy. The 4th Army estimate of the situation on 16 August said that these forces represented the German mass of manoeuvre, and that aerial reconnaissance showed that there were no significant German forces at Arlon or Luxembourg in the southern Ardennes. Joffre based on his plan of attack on the idea that the Germans had left their centre weak in order to strengthen the force north of the Meuse. He therefore decided to break the German centre in the Ardennes. On 15 August GQG ordered 5th Army on the left flank to march north to an area west of Givet. 4th Army was to be prepared to attack towards Neufchâteau. On 16 August the 3rd Army was told to hand over the area between Verdun and Toul to a group of reserve divisions in order to be able attack north of Metz towards Longwy.
The inability of the French cavalry divisions to obtain an accurate picture of the advance of the German 4th and 5th Armies led to serious mistakes in French operational and tactical planning. Due in great part to IR88’s success at Longlier, the French 4th and 9th Cavalry Divisions were pushed out of the way of XVIII AK and were not able to determine what the Germans were doing, nor hinder their movements. The anonymous author of the FAR 25 regimental history said that the French cavalry simply would not fight. From the smallest patrol up to the level of cavalry corps, the French cavalry avoided combat and when it unexpectedly did meet German forces, such as at Longlier, the French cavalry withdrew.77 The German cavalry was able to screen the movements of its own forces, while on 21 and 22 August it provided accurate information concerning the French advance.
3 DIC, Morning, 22 August
The Colonial Corps order, issued at 1800 21 August, directed the corps to march to Neufchâteau on 22 August, with 3 DIC on the right, marching through Rossignol, and 5th Colonial Brigade on the left, marching over Suxy. Because the Corps would transit the Forest of Neufchâteau–Chiny, the Corps cavalry regiment, the 3rd Chasseurs d’Afrique, would follow the advance guard. 2 DIC was held back west of Montmédy as the army reserve. XII CA was on the corps left, marching on Recogne and Libramont, II CA on the right, marching on Leglise. The corps order said that the only enemy forces in the area were those of the German 3 KD and 8 KD, which had been defeated by the French cavalry on 17–18 August.
The 3 DIC order of movement was 1 RIC, 2 RIC, Division Artillery (2 RAC), 3 RIC. 7 RIC followed, guarding the corps artillery (3 RAC); the column was 15km long. The movement order for 2 RIC conveys the prevailing attitude in the division: ‘Today a 33km march. Arrive at Neufchâteau at 1100 and billet. No contact expected.’
The advance guard battalion (I/1 RIC) missed its movement time at 0630 because it was in contact with German cavalry patrols. Then the rest of the regiment, which was to lead the main body, missed its movement time because the staffs did not know where the units were located and orders consequently arrived late. At 0800 the Colonial Corps was informed that II CA on the right was three hours behind 3 DIC, exposing the 3 DIC right flank. This was not an auspicious beginning. Heavy fog hindered movement until it lifted at 0700, revealing a clear, sunny sky.
Meeting Engagement, 3 DIC
A reserve cavalry squadron (6/6th Dragoons) provided security immediately in front of the 3 DIC advance guard. The choice of this reserve squadron, when a regiment of professional cavalry was available (the Chasseurs d’Afrique), can only be explained by the fact that the division did not expect contact. As usual, French cavalry stayed close to the infantry for protection. The Dragoons were engaged about 600m south of Rossignol by dismounted German cavalry, which withdrew. The Dragoons advanced through Rossignol and then 500m into the forest of Neufchâteau where they were again engaged by cavalry. At 0740, 23 August the Dragoons were engaged for a third time 1,500m into the woods, this time by infantry, and stopped cold. The commander of 1 RIC was told that this could not be a large German force because Germans were 35km to the east of Neufchâteau, and that it was important to move quickly through the woods. He therefore committed the advanced guard battalion, II/1 RIC. The forest was deciduous, mixed with pines. The undergrowth was very thick, and only the occasional clearing offered visibility up to 50m. A wall of fire met II/1 RIC. Immediately there were heavy casualties; the commanders of the 5th, 6th and 8th companies were killed, the CO of the 7th Company wounded. A violent standing firefight developed at point-blank range. The fight became hand-to-hand at several points. The rest of 1 RIC was committed; all three 1 RIC battalion commanders were killed while standing on the road, as if on manoeuvre.
The remainder of 3 DIC was strung out on the road. 2 RIC was entering Rossignol; the divisional artillery, 2 RAC, was crossing the bridge at Breuvanne; 3 RIC was entering St. Vincent. Two battalions of 7 RIC had taken a wrong turn and were marching cross-country to regain the correct route. At the rear of the column was the corps artillery, 3 RAC.
At about 0930 it was difficult for the commander of 3 DIC, General Raffenel, to judge the seriousness of the fight; all that he could see were the wounded coming to the rear. Although all of 1 RIC was engaged in the woods, he still refused to believe that he was in contact with a major enemy force. His concern was to bring forward 3 RIC and clear the woods.
By 0800 the lead element of the 3 DIC divisional artillery, I/2 RAC, had advanced until it was at the southern entrance to Rossignol, followed by II/2 RAC, whose last vehicles were at the Breuvanne bridge and III/2 RAC, which was south of the bridge. The firefight in the woods ahead prevented 2 RAC from advancing. As would soon become clear, the ground was too soft to move the guns off the road.
At 1015 I/2 RIC was sent into the thick woods to the right of 1/1 RIC, but became completely disoriented and strayed to the right. II/2 RIC was committed on the left. It took heavy fire from an invisible enemy, probably II/IR 63 on its left flank, lost most of its officers, including the battalion commander, and by 1100 the battalion broke for the rear.
German doctrine emphasised that cavalry needed to be aggressive during the battle in developing opportunities to both participate in the battle as well as to operate against the enemy flank and rear. Doctrine also stated that cavalry was the arm best suited to conduct pursuit.
While the 3 KD and 6 KD had been very effective in the reconnaissance and counter-reconnaissance roles before the battle, during the battle they accomplished nothing. The 3 KD commander decided that the terrain prevented the division from accomplishing anything and resigned himself to inactivity. 6 KD was used to guard the army left flank. Neither division conducted a pursuit, either on 22 or 23 August, although the Colonial Corps would seem to have offered a fine target for 3 KD and the right flank of the French VI CA an even better target for 6 KD.
It appears that the cavalry learned during the approach march that a mounted man presented a fine target and that even small groups of infantry were capable of blocking cavalry movement. By 22 August the senior cavalry commanders were thoroughly intimidated: they avoided serious contact and were unwilling to attempt to move large bodies of cavalry anywhere that they might be subject to small arms or artillery fire. Coupled with the unimaginative operations of the 5th Army headquarters, the timidity of the cavalry leaders cost the cavalry the opportunity to have made a major impact in the battle.
Lessons Not Learned
Upon mature reflection, Charbonneau said that the defeat of the Colonial Corps was due to three factors; the superiority of German training and doctrine not being one of them.
The first was the failure of French reconnaissance. On 20 August the French cavalry reported the Germans moving north of Neufchâteau–Bastogne. On 22 August the Colonial Corps cavalry, ostensibly due to fog and wooded terrain, did not detect the German advance. For these reasons, the Colonial Corps was surprised. Why German operational and tactical cavalry had detected the French advance was not explained. On a tactical level, the 3rd Colonial Division and 33 DI were not destroyed because they were advancing rashly, but because the Germans counter-reconnaissance had blinded the French patrols, and the Germans manoeuvred at a rate of speed that befuddled the French division commanders.
Second was the failure of the French theory of the advance guard, that is, the idea that the advance guard could significantly delay the enemy, giving the main body time to manoeuvre. This theory had nothing to do with Grandmaison, but was the essential element of Bonnal’s doctrine, which had been implemented in the French army in the late 1890s. Charbonneau said that the advance guard concept failed if the enemy attacked at once ‘appearing like a jack-in-the-box’, not only against the front but also against the flanks. Again, French defeat was not a result of superior German doctrine, but deficiencies in French tactics.
Third, Charbonneau said the offensive à outrance failed because it did not incorporate the concept of fire superiority. He did not acknowledge that fire superiority was the foundation of German offensive tactics. He did say that disregard of the effects of fire increased in the French army as the lessons of 1870 slipped further into the past. Indeed, to Charbonneau the offensive à outrance had been taught as French doctrine for most of the period before the First World War, thereby absolving Grandmaison of instituting a radical change in French tactics.
Charbonneau steadfastly maintained that pre-war French tactical doctrine and training recognised only the offensive and that his division was defeated because it attacked recklessly. But neither 3 RIC nor 7 RIC made any attempt to conduct an attack of any kind, much less a reckless offensive à outrance. 3 RIC was pinned down by German fire, which eventually destroyed the regiment. There was no attempt by 3 RIC to ignore the effects of enemy fire charge with the bayonet. As Charbonneau well knew, his own regiment, 7 RIC, was overrun while attempting to hold a defensive position.
Given the choice between drawing conclusions from what he had seen with his own eyes and parroting the party line, Charbonneau came down foursquare on the side of conventional wisdom. Charbonneau’s cognitive dissonance is symptomatic of the subsequent problems in the discussion of the Battle of the Frontiers.