Hoffmann was unwilling to believe that this was anything but a manifestation of the fog of war. Still, it was just possible that the Russians around Soldau had given François more than he had been able to handle. Hoffmann, never a man to minimize his own exploits or abandon a good story, described calling the errant battalion commander to the phone and peremptorily ordering him to turn his men around and advance until he found an enemy to engage. Then he instructed an aide to take a staff car, drive towards Montowo until he found either Russians or Germans, and report the actual situation.26
The explanation was not long in coming. When the 3rd Brigade had deployed that morning its commander kept the battalion in question in brigade reserve. Its commander had been marked before the war as having “bad nerves,” which might mean anything from an unfair efficiency report through a slightly overactive imagination to acute dipsomania. Brigadier-General Mengelbier was neither the first nor the last commander to cope with a weak link by putting him in a position where he could be directly supervised. This time everything went wrong. Mengelbier had proposed to establish his command post with the battalion, but was unable to find the unit in the fog. The anxious major for his part lost touch with his regiment, his brigade, and his division. The few reports he received from stragglers and fugitives were of Russians everywhere. In a kind of homing instinct he marched his men back to Montowo, picking up most of two machine-gun companies and part of a cavalry regiment along the way. By the time he reached the town he had apparently convinced himself of the truth of the story of disaster that was his sole protection against disgrace unthinkable to a German regular officer. Then the telephone caught up with him.27
If confusion ended as farce in the south, it threatened to become tragedy in the north. Below’s and Mackensen’s corps had spent an uneasy night after their victory on the 26th. Instead of merely bivouacking in the field they were ordered to dig in—a task easier outlined on a staff map than executed in pitch darkness. Trenches and rifle pits wound up facing every point of the compass. Guns painfully positioned at midnight were revealed at dawn to be targeted on their own rear areas. Company officers were not particularly sorry when orders came to abandon the area before their superiors had time to inspect the scene. The men were even more pleased to find the Russians running faster than the Germans were chasing them. Dismounted cavalry scattered the few stragglers remaining in Bischofsburg. Airmen reported a corps’s worth of Russians on the road south. Civilians described the enemy’s flight in lurid terms.
In XVII Corps headquarters, the previous day’s misgivings gave way to a euphoria sanctioned by higher authority around 12:30 p.m., when the army order to march south finally reached Mackensen. Below, arguing that his men were too fatigued to get very far, had asked Mackensen to assume full responsibility for the pursuit.28 By their own accounts at least, the regimental officers were overwhelmed with volunteers. The corps’s two cavalry regiments, the 5th Hussars and the 4th Mounted Rifles, began the chase, followed by machine guns, artillery, and infantry riding trucks borrowed from the supply columns or bicycles commandeered from civilians. When the cavalry, its horses blown, bivouacked for the night, the infantry kept moving. At 2:15 a.m. on August 28 an improvised flying column consisting of a battalion of the 175th Infantry supported by the regimental machine gun company, a battery of field guns, and two troops of cavalry reached Passenheim. The village had been ransacked by the retreating Russians, who left behind an entire ammunition column and a war chest containing 200,000 rubles.29
Thus far, so good. But at 9:00 p.m. on August 27, Ludendorff’s evening was disturbed by a phone call from I Reserve Corps. One of its cavalry patrols had just arrived with information. Strong Russian forces, estimated as at least a division, had in fact occupied Allenstein. When the troopers left at 4:00 p.m. to report, Russians were still marching in. Apparently the enemy in that sector was much stronger than anyone had thought. Possibly, moreover, XIII Corps, instead of turning against Scholtz as expected, was trying to extend north and east to join Rennenkampf’s II Corps, which according to the best German information had reached Rastenburg by 11:00 a.m. on the 27th. If this happened the rearward communications of XVII Corps and I Reserve Corps would be well and truly severed. Below therefore proposed that the full strength of both his and Mackensen’s corps be turned against Allenstein the next morning in order to drive the Russians there south onto the guns of Scholtz, Morgen, and Goltz.
Army command approved the change in plans. The I Reserve Corps and elements of XVII Corps were ordered to attack Allenstein the next day. The two corps must strive to mount a joint attack by noon, but I Reserve Corps was definitely to be in action at that time. To ensure that contingency, Below was to move out without waiting for Mackensen. August 28, declared Ludendorff, must be the day of decision.30
Mackensen’s headquarters still had not established a direct telephone link with army headquarters in Löbau. With Allenstein in Russian hands, even a fast car would have to make elaborate detours over questionable roads to maintain contact. The army staff therefore expected Below to pass the revised orders along to Mackensen. He did not—at least not in the form intended. Instead Colonel Posadowsky, chief of staff of I Reserve Corps, drove to Mackensen’s headquarters in person and informed him that the entire XVII Corps, including the detachment at Passenheim, was to move on Allenstein, extending only security detachments south towards Jedwabno and Ortelsburg.
After the war Mackensen wrote that this turn against Allenstein was so much opposed to his view of the situation that at first he hesitated to execute the operation. Army headquarters only noted that Below’s headquarters “misunderstood” the new orders.31 The confusion is not hard to understand. The general army order issued at 10:00 p.m. said that “elements” of XVII Corps would cooperate with I Reserve Corps around Allenstein. “In addition,” the former corps was to drive south “in the direction of Willenburg.”32 However, the specific orders sent to I Reserve Corps stated that on August 28 XVII Corps was to advance on Ortelsburg while attacking Allenstein. The two towns were in opposite directions from the corps’s current position. The missions were not exactly mutually contradictory. But in the face of an enemy whose strength and positions remained uncertain, pursuing them simultaneously might put Mackensen’s corps in the position of the proverbial chameleon on a plaid shirt. Instead of striking a concentrated blow, it risked the fate of being too weak everywhere, and too tired from overmarching, to accomplish anything.
In his study of the Tannenberg campaign, N. N. Golovine argues that the German orders were intended to provide a barrier against Rennenkampf, rather than to expedite the annihilation of Samsonov. According to his interpretation, Ludendorff intended Ostgruppe to remain more or less in position while XX Corps and I Corps pursued the 2nd Army and drive it south. These latter corps would then disploy on the right flank of Ostgruppe, with the Goltz Division and the 3rd Reserve Division in reserve, to face Rennenkampf should he move south. Golovine asserts that the orders to march on Allenstein did not reach Mackensen directly because “some calmer person” at army headquarters intervened to prevent this. Posadowsky was sent to XVII Corps Headquarters by mistake, because Below did not understand either the situation or his orders.33
The thesis, though provocative, does not appear tenable in light of the German records. Fatigue played a significant role in the mix-up. Mackensen was sixty-five, Below fifty-seven. Their principal staff officers were men in late middle age. While they were collectively fit enough none was as yet accustomed to, or indeed fully aware of, the constant stress of campaigning under conditions of modern war. These were not the circumstances of 1916 or 1917, when staffs spent months at a time in the same comfortable surroundings. The corps of Ostgruppe had been on the move since the start of the campaign. Questions of interpretation that seem obvious in a scholar’s study or the academic precincts of a war college can loom much larger late at night in a field headquarters to men whose lifetime routines of eating, sleeping, and moving t
heir bowels have been rudely and systematically shattered.
Other human factors were at work as well. The obvious course for Below, requesting clarification from army headquarters, incorporated two less-obvious risks. The first was looking like a fool in the eyes of a new set of superiors. The second was hearing what one did not wish to hear. While I Reserve Corps had done well in limited actions with a solid block of active troops close at hand, Otto von Below might well have been pardoned for feeling uneasy at the prospect of taking his civilian soldiers into Allenstein against a strong concentration of Russians while XVII Corps marched in the opposite direction. Instead Below and his staff, after consulting with the army liaison officer with I Reserve Corps,34 apparently put together the two orders in their possession and developed an interpretation performing three functions. It fulfilled their understanding of the letter of army headquarters’ directives. It preserved XVII Corps as a concentrated striking force. And it allayed the unspoken, perhaps unconscious, fears of what might happen were I Reserve Corps to be left on its own at Allenstein.
That Below was uneasy about his solution is indicated by Posadowsky’s appearance at Mackensen’s headquarters. Using a corps chief of staff as a liaison officer was hardly common practice in the German army, but its wisdom in this case was plain. According to Mackensen, only the respect he had for Posadowsky’s reliability kept him from disregarding the orders he bore. With many misgivings, Mackensen finally prepared to march west in support of Below.35
Had it been necessary, the 8th Army could probably have taken up the positions Golovine suggests. Perhaps Ludendorff had this in the back of his mind. But the army orders of the night of the 27th, based on the information available at the time and taking into account Scholtz’s constant concern for his northern flank, reflected a firm decision to continue the attack against the 2nd Army as long as possible. If they were flexible enough to be well adapted to a defense against an unexpected onslaught from Rennenkampf, this reflected the professional skill of the staff officers responsible for developing them under such pressure.
The Germans had more than a well-drawn operations order in their favor. Martos had decided on the night of August 26 that he could not safely advance on Osterode as ordered without clearing his left flank of the threat posed by Scholtz’s corps. After checking the 41st Division’s halfhearted attack, at 4:00 p.m. Martos committed XV Corps to the counterattack around Mühlen whose limited results have been described earlier. An hour later he was summoned to the telephone by Samsonov’s chief of staff. Throughout the day Martos had been begging army headquarters to order XIII Corps to cooperate with him. Now he was informed instead that Samsonov wanted XV Corps to advance to Allenstein on August 28 to “cooperate” with Kluyev and Blagoveschensky.
Martos exploded. He shouted that it was impossible to disengage his corps from the fighting in progress, and insane to march it north leaving behind an undefeated enemy. Rather than attempt such follies, Martos declared, he would prefer to be relieved of his command on the spot. Whatever its mixture of conscious bluff and simple bad temper, the challenge worked. Postovsky temporized, informing Martos that he would call again in an hour.
By this time Second Army headquarters was a scene of confusion dominated by impressions. Learning of I Corps’s defeat and retreat was a shock. So was becoming aware of the scope of the disaster that had overtaken the 2nd Division. In the light of these events and Martos’s report, Samsonov seems to have decided by early evening that the main German strength was on his left, facing XV Corps, I Corps, and what remained of the 2nd Division. His mind and his plans changed accordingly. The formal orders for August 28 were that the latter formations hold at all costs. Instead of Martos joining Kluyev, XIII Corps would march south from Allenstein and, under Martos’s tactical command, combine with XV Corps in an “energetic offensive” against the presumed flank and rear of the German position. The VI Corps would move west to the area of Passenheim and cover the 2nd Army’s right.36
It was these orders, intended to concentrate the Russian center corps against the Germans in the southern sector, that set the actual stage for what so many later general accounts described as a Cannae, or at least the possibility of one. Neither Ludendorff’s directives nor the gratuitious coups of the radio interceptions were nearly as significant for the developing Russian disaster as the command decision taken at 2nd Army headquarters during the evening of August 27—a decision made for a most logical set of reasons.
Max Hoffmann suggests that if nothing else the length of the German line, extending as it did from Usdau to Mühlen, should have convinced Samsonov that he was facing a larger force than he thought.37 But Samsonov’s orders reflected more than careless thinking or blind stubbornness. He had received no concrete information from either Rennenkampf or Zhilinski about the whereabouts of the German troops who had fought at Gumbinnen. As the situation on the 2nd Army’s own front grew more complex, Samsonov and his staff had enough to occupy their minds. Unless otherwise informed, it was logical to assume that Rennenkampf had remained in touch with a force retreating westward, or into Königsberg. In any case Samsonov had no immediate fear of an attack from the north. The VI Corps’s fragmentary reports did not indicate the scope of that formation’s collapse. As for XIII Corps, its staff and commander had settled in for the night at Allenstein convinced that there were no Germans anywhere in the vicinity!
The process of reaching this remarkable conclusion illustrates once again the problems of assimilating tactical intelligence information at the operational level. Kluyev, after two days on scraps of reconnaissance, decided to see for himself at least by proxy. He sent out two airplanes, one to check the positions of XV Corps, the other to scout the roads to the east over which VI Corps was supposed to be advancing. The latter observer reported a corps-strength force marching westward. Though he had been unable to determine if the troops were German or Russian, XIII Corps’s staff was firmly convinced that the airman had spotted VI Corps advancing as ordered. The time and place were right. No one at XIII Corps headquarters had received any information encouraging even a suspicion that German troops might be that far south. Kluyev promptly sent the pilot back with a message for Blagoveschensky and orders to land beside one of the columns and see that the dispatch was forwarded.
Around 5:00 p.m. a Russian plane, flying low as ordered, was brought down by a fusillade of small arms fire. Though the pilot said nothing about the origin or purpose of his flight, presumably he was one of the more surprised men in either army—not because he had been shot down, but that his captors were Germans. His failure to return generated no alarm at corps headquarters. Those new-fangled contraptions were always breaking down. And when a cavalry patrol of Kluyev’s 36th Division reported that it had been fired on by a column advancing westward, the division commander assumed his men had simply mistaken a trigger-happy advance guard of VI Corps for Germans! He neither attempted to verify the information nor passed it on to corps and army headquarters. The troops in question in fact belonged to I Reserve Corps, which had halted for the night only ten kilometers northeast of Allenstein.38
Kluyev’s behavior cannot be dismissed as simple incompetence. Defeat is a possibility in every military operation, and can be produced by a broad variety of circumstances. No commander can afford to reflect too deeply on potential failure. Taking counsel of possibilities can bring on fears that paralyze the will to act. For Kluyev, starting at ghosts was less important than imposing his will, and the will of his superiors, on the Germans in the next day’s fighting.
Second Army headquarters was operating from the same matrix. In the context of the information—or lack of it—provided by XIII Corps, Samsonov’s decision to push forward in the center seemed the best way of alleviating what he saw as the major, indeed the only, immediate German threat: the growing pressure on his left flank. Like Prittwitz, Hindenburg, and Ludendorff, Samsonov was a nineteenth-century general trying to cope with a twentieth-century problem: the gap between c
ommunications and mobility. The generals of World War I faced in greater measure than any of their predecessors the test of knowing at least in outline what was going wrong, without a corresponding possibility of adjusting the situation. Railroads might have enhanced strategic mobility, but the tactical and operational pace was still determined by the muscles of men and horses. The wars of the nineteenth century had increasingly shown that an army commander holding a corps or two out of action in the manner of Napoleon was simply depriving himself of that many troops. Modern battlefields were too large to permit the ready shifting of large reserves either to shore up weak spots or to turn stalemate into success. Samsonov was able to do nothing directly to support his left except relieve Artamonov of command and hope his successor could straighten out the mess. But it seemed reasonable to hope, particularly in view of Martos’s reports during the day, that XV and XIII Corps acting together might give the Germans something to worry about on their left flank. As late as 11:30 p.m. on the 27th, Samsonov informed Zhiliski that Mühlen was in Russian hands, and that German troops had been observed retreating southwest. If all went well the next day, Kluyev and Martos might overrun their immediate opponents and decide the battle, or at least one part of it, favorably.
This consideration almost certainly influenced Samsonov’s next decision. At 7:15 a.m. on August 28, he told Zhilinski that he was going forward to XV Corps headquarters in order to take control of the attack in the center. He and his staff left Neidenburg at 8:00 a.m. and drove northeast along the road to Jedwabno.
Critics generally agree that this was the final step in the 2nd Army’s destruction, that Samsonov ceased from that time effectively to command anything. At best the move has been interpreted as a brave but limited man’s desire to influence events slipping out of his grasp, a despairing gesture of personal courage in the face of disaster. It was both less and more. Since crossing the frontier, the exigencies of maintaining communications with its far-flung units had forced Samsonov’s headquarters further and further behind its subordinate formations. The point had been sharply and repeatedly noted by Zhilinski. Samsonov’s pride was wounded by his superior’s observations, but vanity alone did not impel him to leave Neidenburg. He was going to reinforce what seemed like success, a maxim taught in every military school in Europe. At least his presence should remove some of the strain from the commanders on the spot, and ease cooperation between XV and XIII Corps. Samsonov’s detractors tend to overlook the fact that on the other side of the line Hindenburg and Ludendorff spent most of their days not at headquarters, but with subordinate formations believed in need of guidance or encouragement. Samsonov’s announcement that he was temporarily closing down the 2nd Army’s signal station was not a gesture of resignation but an example of bad staff work. Samsonov expected to restore communications once he re-established his headquarters farther up the line.
Tannenberg: Clash of Empires, 1914 (Cornerstones of Military History) Page 41