CHAPTER 12
THUS far the war had chiefly concerned China and Japan. Japan was aggressive towards China; considerations of how far this affected Japan’s relations with other countries were peripheral. But from this time onwards, Japan’s relations with the Great Powers became the prime concern of its Government. The war between China and Japan became increasingly difficult to limit to a private war; Japan was faced with problems, rising out of this war, each one of which caused it to consider afresh its policy towards other Powers. Sometimes it experimentally remoulded its policies towards them, only to change them again, with all the repercussions which such instability led to. Japan’s policies became very uncertain. No settled principles guided its action.
Actually, since the days of the Anglo-Japanese Alliance, Japan had pursued a wavering foreign policy, spreading everywhere a diffuse suspicion. It had no sure base in a firm agreement with a greater Power. But until a late period, it seemed that its special, inexorable opponent was Russia. Suspicion of and hostility towards Russia governed its designs. One product of this attitude of mind had been the signing by Japan of the Anti-Comintern Pact with Germany in 1936. This was an alliance which somewhat nebulously pledged the partners to resist the infiltrations of Communism, and, in a secret clause, bound them to withhold aid from Russia should either party be involved in hostilities with the USSR.
As we have seen, there was special meaning in Japan regarding Russia as its hereditary enemy. Russia, as the perpetual threat, had penetrated into the folklore of the people even before Commodore Perry’s expedition in 1853 signalled the opening of Japan’s modern period of history. Ever since the end of the nineteenth century, Japan had regarded Russia as the Great Power against which it was destined to fight for survival. In 1904–5 it had fought the first round; it was convinced that it would have to fight again. The Siberian Intervention not only reinforced that conviction but also demonstrated the rapidity with which Japan could and often did find itself isolated and abandoned by its erstwhile allies. The serious strains which developed between the Army and the civil Government as a result of that crisis were to recur in the future, but this time the civilian Government, using one of the anti-imperialist swings in public opinion which at the time alternated with moods of aggressive nationalism, had prevailed in restraining the military. The Army felt that it had been ordered to drop its legitimate prey when it had been certain of it, and the sense that the civilian Government could not be trusted continued to weigh with it.
After Manchukuo came into being in 1932, Japan stationed there a large part of the Japanese Army. This did not disguise the fact that the eyes of the Japanese Army were focused upon Russia. Much of its training and manoeuvres were made with Russia as the ultimate adversary. Preparations were made to meet the danger of a full-scale Russian invasion; even more elaborate contingency plans were devised for a repeat performance of the Siberian Intervention, this time to be a wholly Japanese offensive against the Soviet Union. Russian belligerency was no less marked, and the exchanges between the two countries became increasingly explosive.
By 1938 2,800 separate armed clashes between the Russians and Japanese had been recorded and their successive repercussions provided the General Staffs of both countries with a perennial headache from which they had no relief in the years leading up to the wider conflict which engulfed the whole of East Asia and the Pacific. Since the rivalry between Russia and Japan is in many ways as important as the struggle between China and Japan in the tragic sequence of events that ended with the defeat of Japan in 1945, we must not overlook the significance of this interminable tension and succession of military clashes. Each was a cautionary tale. If details of most of them are unmemorable, their continuity was not. Three of the incidents had particular significance: the Amur River Incident, the Changkufeng Incident and the Nomonhan Incident. They foreshadowed special features about Total War which affected the perceptions and expectations of all the Powers in the Second World War.
During late June to early July 1937, on the eve of the great Sino-Japanese war that we call the China Incident, Japan and Russia fought the 185th in their series of inconclusive Manchurian border clashes at a site seventy miles up the Amur River from Khabarovsk. This is roughly the same area over which the Russians and the People’s Republic of China fought similar actions decades afterwards. The Amur River Incident with which we are concerned, however, exposed serious Russian weaknesses in the wake of the first Red Army purges to affect East Asia. The Incident began when Russian troops landed on several disputed islets lying in the middle of the Amur River and took off a few Manchurian gold miners whom they accused of trespassing. Soviet gunboats then exchanged fire with nearby elements of the Manchukuoan Army. The Kwantung Army, with the approval of the Army General Staff in Tokyo, ordered the First (Tokyo) Division to respond. This division had been exiled in this remote hinterland for taking a leading role in the 26 February Incident of 1936. Before anything could come of this, more Soviet gunboats steamed to the scene and a number of Red Army regular divisions were put on a war footing. At this point the Army and Naval General Staffs in Tokyo lost their nerve and ordered the Kwantung Army and Japanese naval air and river patrol units in the vicinity to restrain themselves pending a peaceful resolution of the crisis. Elements within the First Division turned a blind eye to these instructions, opened fire on the Soviet gunboats and sank one of them. The Soviets were outraged by this development but with some alacrity agreed to a mutual withdrawal from the disputed points. After the Red Army retired, the Japanese re-occupied the territory.
Western observers, ever keen to extrapolate from such engagements some estimate of the attitude and military efficiency of the Soviet Union in the event of a European war, were unimpressed by the performance of either party in this incident. On the whole it was regarded as ‘somewhat puzzling’. A senior member of the Far Eastern Department in the British Foreign Office put his finger on the problem: ‘This sort of thing would be very serious indeed in almost any other part of the world, but such queer things happen almost every day on the Manchu-Siberian frontier that the sequel to the incident may be no more than some mudslinging by the press of Russia and Japan and an exchange of a few rude notes.’ It was generally concluded that neither country was prepared to adopt extreme measures, nor was it likely that either would gain the upper hand. Thus while the scale of the fighting was considerable, the world was not alarmed. Such indifference seems the more remarkable now given the juxtaposition of the Amur River Incident and the China Incident.
Less significant border clashes flared up between Japan and the Soviet Union during the remainder of 1937 with no sign that either Power was disposed to settle these affairs once and for all. Diplomatically, the Soviets made faces at Japan, concluding a non-aggression pact with Chiang Kai-shek’s Kuomintang in August 1937. Pravda advertised that agreement as ‘collective security’, but the plain fact was that Russia had to contend with the animosity of Germany, too, and showed only token defiance of Japan. Indeed, the British Foreign Office advised Prime Minister Neville Chamberlain that ‘one small ray of comfort to be found is the extreme and expressed reluctance of the USSR to allow themselves to get mixed up in the [Sino-Japanese] dispute. This allows us to hope that the dispute, if it develops, will be confined to China and Japan and that other Powers will not be dragged in.’∗
As Japan sent more and more troops deep into China, the Soviet Union read the handwriting on the wall and began to reinforce its military forces in the East. The West privately counted this a blessing. It ought to make the Japanese more circumspect. There would be no need to encourage this development: Russia would strengthen her defence lines in her own interests, not because it might gain Moscow new friends. British air Intelligence experts – who in retrospect proved more perceptive (except in relation to Japan) than critics have credited – put Soviet front-line air strength at 4,000 aircraft, but 75 per cent of those aircraft reportedly were obsolescent. Soviet aircrews were regarded as badly trained, inefficient and very poorly led. No evidence could be found that Russia then believed that air bombardment of Japanese cities was practicable, although it was generally observed that ‘the threat, slight as it is in reality, has altered the whole of Japan’s strategical outlook and has created a most powerful deterrent to a Japanese attack on any part of Soviet Russia’.† On the eve of a European weekend crisis during mid-May 1938, French Intelligence suggested that Soviet pilots sent into action at Hankow and elsewhere in Central China were winning a string of victories against the Japanese. The British, however, discounted such reports. The difference in viewpoint between the two Powers in East Asia mirrored their respective (dis)inclination to welcome offers of Soviet air support for Czechoslovakia. In terms of what the Soviet Union could do to help preserve British, French and American interests in East Asia, London, at least, took the attitude that there were no significant developments up to the first anniversary of the birth of the China Incident. The French, who had less at stake anyway, remained slightly more hopeful if not optimistic. Both Powers, like the United States, recognized that for the time being the Japanese were in a position to do more or less as they liked in China although not, perhaps, further a field.
Then, late in July and throughout most of August 1938, a major battle erupted at Changkufeng, on the banks of the Tumen River west of Lake Khasan at a spot where the frontiers of Korea, the USS R and Manchuria meet some seventy miles south-west of Vladivostok. Changkufeng was one of the highest in a range of ancient volcanic peaks. It was possible to peer from its summit across intervening marshlands to Posyet Bay, fifteen miles distant, where the Soviet Union had begun constructing new submarine pens and air bases at the port of Novokievsk (Kraskino). Southwest of Changkufeng, about eleven miles away, lay the Japanese port of Rashin (Rajin), where transport vessels unloaded troops and supplies bound for the northern frontiers of the Japanese Empire. The railways, and the port of Rashin as well, were within sight of the ridge of the Changkufeng Hills. Observers there would encounter no difficulty in monitoring the bulk of Japanese railway traffic to the north. There were other vantage points from which the two sides could scrutinize each other. Changkufeng was not unique. However, possession of these crests affected the security of adjacent promontories which the two sides occupied along the same frontier.
Historians and political scientists have made much of this border incident. They have linked it to the great Soviet purges and the efficiency of the Red Banner Front armies. They have drawn conflicting conclusions about how it fits in with what are surmised to have been Soviet and Japanese intentions before, during and after it took place. They have argued over the manner in which the campaign was conducted and the scale of forces employed on both sides. They have inferred contradictory conclusions about the peace negotiations which followed in its wake. They have suggested that it demonstrated the ability or the inability of Japanese authorities in Tokyo and in the field to control the activities of their forces. They have even used it to advance various theories about wider conspiracies to wage aggressive war.
Almost everything about the Changkufeng Incident has been shrouded in mystery. The incident has long been appreciated as an event ranking in importance with, say, the far better known incidents of the German march into the Rhineland, the Italian conquest of Ethiopia, the American occupation of Iceland, or the Japanese advance into Indo-China with the acquiescence of the Vichy régime. Like each of those affairs, the Changkufeng Incident gave rise to accusations that one side or the other was guilty of systematic aggression.
The Soviet purges had decimated the Russian officer corps by the summer of 1938. Russian commanders everywhere were jittery. The initial Soviet occupation of the Changkufeng heights may have been prompted by the defection to Japan of General G. S. Lyushkov, the Commissar of the NKVD (Stalin’s dread secret police) responsible for the whole of Soviet East Asia. In mid-June, Lyushkov’s dramatic escape across the Manchurian frontier in full-dress military uniform, replete with medals, forced the Soviets to effect a thorough reorganization of their forces and a tightening of their border security arrangements. Within weeks the Kwantung Army’s ever-vigilant signals Intelligence monitors were to intercept and decode Russian military communications that provided the Japanese with an inkling of the trouble ahead. Shortly thereafter, the Soviets began to construct fixed defensive lines and reconnaissance positions on Changkufeng.
As the area formed a sector of the Manchukuoan frontier, it was only natural that the Kwantung Army should take a keen interest in its defence. However, local features of the terrain, better lines of communication and the fact that four out of every five inhabitants were Korean (and thus regarded as Japanese rather than Manchukuoan subjects) dictated that defence of the area should fall within the jurisdiction of Japan’s Korean Army, whose major responsibility was the preservation of civil order in Korea. At first the Korean Army was inclined to adopt a watchful but undemonstrative attitude towards the Russian moves. Subjected to ridicule and accused of timidity by the Kwantung Army, the Korean Army Staff wired Tokyo that while the Russian action was clearly illegal, it had little immediate bearing upon offensive operations which Japan was preparing against Hankow and elsewhere. It would not be prudent to conduct a precipitate counter-attack while the China Incident was in full flood: ‘This Army is thinking of reasoning with the Soviets and requesting it to pull back, directly on the spot… In case the Russians do not accede in the long run, we intend to drive the Soviet soldiers out of the area east of Khasan firmly by use of force.’
Imperial General Headquarters in Tokyo considered the matter calmly. It agreed with the Korean Army that the border should be kept under close surveillance and that no hasty action should be taken. If attacked, however, the Korean Army was to press forward at once, seize control of the disputed territory and on no account cross beyond what Japan recognized as the legal frontier. There was no policy of subterfuge or aggressive intent in this: the purposes of the Japanese high command were clearly unprovocative and as moderate as any self-respecting nation could permit itself. Through spontaneous impatience or unendurable provocation, however, Japanese officers at the scene sent out an unauthorized ‘reconnaissance force’ which resulted in a pitched battle for possession of the heights. The Japanese prevailed. In the weeks that followed, the Russians vainly attempted to re-establish themselves on the hills. The struggle intensified, but each successive Russian onslaught failed.
The emergency had arisen while the Korean Army was unusually weak. It normally comprised only two infantry divisions, but prior to the Changkufeng Incident one of these, the Twentieth Division, had gone to take part in the China operation (leaving only a skeletal force at its depot in Korea). Fortunately, Lieutenant-General Suetaka Kamezō, commander of the remaining Nineteenth Division, giving vent to his mortification at long being held in reserve, had imposed an exceptionally arduous training regimen upon his troops that already had earned his division a reputation as one of the crack forces of the Japanese Empire. On the other hand, there was virtually no armour in Korea and only two air regiments (fewer than a hundred aircraft) in the entire country. Japan’s experience in Manchuria and China had shown that any attempt to bring up reinforcements during a conflict was likely to lead to a mutual escalation of the fighting and to a vitiation of efforts to produce a negotiated settlement. The outlook was bad.
Once fighting developed, the world was confused by the evident scarcity of trustworthy reports from the battle zone. There was open disbelief about the veracity of information dispensed by Moscow and Tokyo. Western sentiments were summed up in a note which circulated in the British Foreign Office while the struggle was continuing: ‘both Russians and Japanese are such accomplished liars and so addicted to all forms of bluff and bluster that one can be sure of nothing in this strange affair.’ One curious difference was repeatedly remarked upon. The Soviet press was increasingly hysterical. The Japanese, in great contrast, took pains to underplay the strategical importance of the dispute. Each side, however, claimed to have suffered early setbacks followed by victory against almost overwhelming odds. British and American editors simply printed opposing accounts side by side under sensational headlines. The New York Times commented:
It isn’t easy to guess what is going on. It is still more difficult to guess why it is going on… In all the mystery and madness only one thing is clear from this distance. What takes place in the Manchurian frontier is not a mere border incident. It fits somehow into the great pattern of irrational forces moving behind events that involve Europe and Asia in an ‘indivisible’ struggle.
The little war of Changkufeng was no skirmish. Cramped within a radius of little more than one mile, an elite division of Japanese infantry, backed by a growing volume of artillery support but lacking any air or armour protection, was committed against a force three times as large and comprising no fewer than twenty-seven battalions of Soviet troops, 120–150 aircraft, 100–150 batteries of mostly heavy artillery, and massed tank assaults, often employing waves of 50–60 tanks, from Soviet armoured forces variously estimated at 200–500 tanks. Japanese military planners had long anticipated that in any war against the Soviet Union they would have to overcome odds of three to one. They expected to win. Here was a test case. Suetaka’s under-strength division of barely 7,000 men faced the cream of the Soviet Far East Army, reputed to be a cut above the standard of the Red Army’s European forces. By the time the guns were stilled, more than 40,000 men had been deployed by the two sides either directly in the fighting or acting in support. The intensity of the struggle can be gauged by measuring these numbers against the 70,000 troops which the Japanese would commit to the entire Malayan campaign and the roughly 43,000 utilized in the capture of the Philippines in 1941–2. Nevertheless, even experienced observers generally had no clear idea as to who started or even won the battle.
The truth was that both sides were badly mauled. The number of tanks, guns and aircraft which the Russians committed to the battle awed the Japanese but revealed a lack of imagination and common sense in terms of tactics and results. The performance of the Soviet infantry showed up badly against the initiative, efficiency and tenacity displayed by the Japanese. The effectiveness of the Red Army’s artillery, although more impressive, by no means matched the standards set by the Japanese. Only 3 per cent of Japanese casualties were victims of aerial bombing compared to 37 per cent hit by shellfire. A neutral war correspondent at the scene who had fought on the Western Front in the First World War remarked that ‘the Russian Army do not appear to have learnt anything of the art of war’.
The Japanese, who won the ground and held it, also had reason to be frightened. Although jubilant at their military victory, they knew how narrowly defeat had been averted. They were appalled not only by the rapidity with which the incident had developed but also by its sacrificial cost. The Japanese Army air force was not yet the superbly balanced instrument that it was to become in future. It was in no position to conduct air operations simultaneously in China and Manchuria. The Korean Army was desperately short of aircraft and first-class fliers, and Japan could not afford to be drawn into an air war that would prejudice the struggle underway in China. Military prudence as well as the political will to avoid any general escalation of the fighting compelled the Japanese to keep their own aircraft out of the skies and at airfields far from the scene. The Japanese gunners at Changkufeng, although able to engage the enemy at up to point-blank range, were seriously outnumbered and came perilously close to complete exhaustion of their ammunition. In these circumstances their technical superiority to the much-vaunted Russian artillery was scant comfort. The infantrymen fared little better: their officers suffered exceptionally high losses and the exertions of the Nineteenth Division left it a spent force. The Japanese logistical system had repeatedly broken down, unequal to the demands placed upon it. This was serious. The main railway line between the port of Rashin and Hsinking (the imperial capital of Manchukuo and northern terminus of the South Manchurian Railway) ran within two or three miles of the battle line. In normal times troops embarked at Tsuruga on the west coast of Japan could cross the Sea of Japan and arrive at Hsinking via that railway within seventy-two hours. The Changkufeng Incident therefore exposed fundamental weaknesses in Japan’s strategical position against the Soviet Union.
Fighting in such a confined space as Changkufeng and in such large numbers produced heavy casualties on both sides. The Japanese lost more than 500 dead and nearly 1,000 wounded. Although the Russians failed to overrun the Japanese positions on the eastern banks of the Tumen River and the absolute physical annihilation of the Nineteenth Division was narrowly averted, the human energies and material resources available to the Korean Army were all but consumed by the conflict. Estimates of Soviet casualties vary wildly, but the best guess is that the Russians suffered perhaps as many as 1,200 dead and more than 4,000 wounded. The Japanese set a pattern for the future by taking no prisoners. The casualties which each side suffered, while deeply disturbing to both, were minuscule compared to what Russia and Japan would lose when next they met on the battlefields of Asia.
The Changkufeng dispute was not settled by force of arms alone. Emperor Hirohito himself had already taken steps to express his unmistakable desire to see the Incident brought to a close and had insisted from the beginning that the forces committed to the battle must not advance into Russian territory. He had administered a strong rebuke to his War Minister, Lieutenant-General Itagaki Seishirō (the same man who had been one of the architects of the Manchurian Incident), and the Army Chief of Staff, the venerable old Prince Kan’in Kotohito, for seeking imperial authority to permit the Nineteenth Division and several divisions of the Kwantung Army to commence offensive operations against the Soviet Union if that perchance should seem appropriate to the Army General Staff. This was more than the Emperor was prepared to tolerate, and knowing that the Army’s demands likewise were regarded as absurdly dangerous by Foreign Minister Ugaki Kazushige (himself a retired full general) and by the Navy Ministry and Naval General Staff, the Emperor forcefully expressed his abomination of the Army’s propensity for embroiling itself in such incidents. From the Army’s point of view, the approach to the throne was therefore not only ill-conceived and ill-timed but heaped shame upon Japan’s Imperial General Headquarters. Kan’in and Itagaki left the imperial presence completely shaken, convinced that their careers were in tatters and reflecting that they might have to atone for their misconduct by forfeiting their lives. In typically Japanese fashion, however, the Emperor sent word after them to say that he was content to retain them in the positions that they had so abused: divine benevolence could extend to those who saw the error of their ways.
Under the circumstances, the only way out was for the Japanese to abandon the field to the Russians or face the possibility of repeated counter-attacks by ever larger Russian forces. The Japanese Ambassador in Moscow, Shigemitsu Mamoru, was instructed to seek an immediate end to the conflict. The Russians seemed remarkably indifferent to his pleas, adding to the concern felt by the Japanese side. Any faintheartedness felt by the Russians during the Amur River Incident was dispelled at Changkufeng by battlefield Intelligence coupled with high-level information acquired by Richard Sorge from a member of his spy ring who was a confidant of Prime Minister Konoye Fumimaro and in a position to observe the deliberations of the Cabinet close at hand. The certainty of the Russians that the Japanese authorities in the field and at Tokyo were determined to reach a rapid settlement of the incident at virtually any price laid the groundwork for the transformation of a military defeat into a Russian political triumph.
In the end, Ambassador Shigemitsu, with the full support of Foreign Minister Ugaki, agreed to surrender all of Japan’s tactical gains. That, and the ability of the Japanese central authorities to control their troops on this occasion, impressed military planners, diplomats and their political masters abroad, who evaluated the slender information which had filtered through to them from the Changkufeng Incident. Many drew the ill-founded conclusion that Japan’s wayward field commanders were only given their head when it suited the military authorities in Tokyo. In fact, the truth was almost precisely the opposite.
Odd though it may seem today, one theory which gained currency in Britain and elsewhere at the time was that the Japanese had bowed to German pressure in acquiescing to the only terms that Russia was prepared to accept: it was believed that the German Government had been so intent on resolving the Sudeten problem in Czechoslovakia that they had exerted all their influence upon Japan’s military leaders to persuade them that it would be most unwise of Japan to start another major war prematurely. There was very little substance in this: Germany had no such influence over Japan. Rather more accurately, Western doubts about the Soviet Union’s military capabilities were abundantly confirmed by the events at Changkufeng. The tactical victory of the Korean Army showed that Japanese soldiers were just as able to stop a European Power (at least in a strictly limited war) as they had been in 1904–5. However the strategical lessons learnt from Russia’s massive build-up during the fighting were that Japan could ill-afford to embark upon such adventures lightly and that Japan must liquidate the China Incident with all possible speed. Nearly everyone was convinced that Japan would scarcely be so foolhardy as to attempt any major offensive against the Soviet Union as matters stood. That being so, Russia might safely take a greater interest in European power politics for as long as Japan remained preoccupied in China. In each of these ways, then, the Changkufeng Incident provided a classic example of how events in East Asia were seen by the West to have a profound influence upon the stability of Europe.
Meanwhile, the Soviet Union gnawed its wounds. The celebrated and experienced Marshal Vasilii Konstantinovich Blyukher, who had risen from being a common soldier in the First World War to the command of the Red Army forces that had come to prominence during the Russian Civil War, and who afterwards had seen the Japanese out of Vladivostok in 1922, was obliged for the second time to bathe in the poisoned political waters of military defeat. In the mid-1920s he had been lent as a military adviser to Sun Yat-sen’s young Chinese Republic. Under his tutelage, Chaing Kai-shek had embarked upon the northern campaign that had taken the Nationalist Armies to the gates of Manchuria. But when Chiang had turned against the communists, the man whom the Chinese had known by the name of ‘General Galen’ had been forced to flee. Afterwards, he had become virtually count palatine of Eastern Siberia. Now, having taken personal charge of the Red Army’s operations at Changkufeng, Marshal Blyukher’s second disaster proved fatal. Removed from command, he was executed by 9 November 1938. He had been living on borrowed time. General Lyushkov had given his Japanese and German interrogators information which linked Blyukher with anti-Stalinist circles within the Soviet Union. This Intelligence was fed back to Moscow by Richard Sorge in late October or early November 1938. Whether Blyukher would have survived the purges anyway is doubtful. More reliable and even more distinguished soldiers than he had perished in that lunatic bloodletting.
After the defeat of Japan in the Pacific War, fresh evidence concerning the operations at Changkufeng was put forward at the Tokyo War Trial by Soviet prosecutors. A parade of eye-witnesses, some lavishly decorated as ‘Heroes of the Soviet Union’ for their part in the action, tendered what purported to be incontrovertible proof that the Japanese first began the incident, were repulsed with heavy losses, thereafter continued to attack Soviet defensive positions along the disputed frontier, and treacherously took command of the heights only after a ceasefire had come into effect. The stories told by these Soviet affidavits and witnesses dovetailed together. They were packed with lies, but the Defence failed to rise to the challenge of producing evidence which would refute the Soviet claims. Due to lack of initiative and possibly lack of information such as a larger defence team might have amassed, the Tokyo Trial proceedings bogged down over the issue of conflicting Chinese and Imperial Russian versions of maps which neither side convincingly authenticated. In fact, the area had always been a no man’s land, dotted with a few scattered Korean hamlets whose inhabitants cared nothing for frontiers and who had conducted seasonal religious sacrifices atop Changkufeng for centuries. The wooden stakes marking the original border had mostly rotted away long since, or been scavenged for firewood or been re-sited any number of times by over-zealous border guards. By 1938 there reportedly remained an average of one undisturbed marker for every twenty-five miles of frontier along a meandering line roughly 400 miles long. There were no markers at all within miles of Changkufeng. That was a suitably anonymous pretext for what is now seen as one of the classic limited wars of modern times. It was an archetypal antithesis, perhaps, to Total War, and yet nations stood eager to apply its lessons, imperfectly, on an unimaginably larger scale.
The following year, there was yet another of these incidents. The young turks of the Japanese Army believed that Japan was progressing. The Kwantung Army, recalling the imagined loss of face which it had endured when forced by Tokyo to restrain its subordinate units at the time of the Amur River Incident, distanced itself even further than before from the cautious policies promoted by the Japanese Government and the Imperial General Headquarters. Moreover, after analysing the Changkufeng Incident, Kwantung Army Headquarters was convinced that the Korean Army had bungled the affair and had been insufficiently aggressive from the start. So far as the bright sparks of the Kwantung Army were concerned, the only proper way to respond to Russian incursions was by the application of overwhelming force. Inter-restingly enough, the Russians had come to much the same conclusions about the Japanese.
The Soviet Union laid a trap for the Japanese, a trap which seems to have been intended from the outset to teach the Japanese a punishing lesson. This incident occurred 700 miles away on the fringes of Outer Mongolia, where in the spring of 1939 a series of border incursions by cavalry units of the puppet Mongolian People’s Republic culminated in an attack on 11 May 1939 by a force of seventy or eighty troops who crossed the Halha River and clashed with a Manchurian Army garrison a few miles to the north-east, at the Manchurian town of Nomonhan. The Kwantung Army, chafing at restraints imposed upon it by the General Staff in Tokyo ever since the Amur River Incident, was restless. In April 1939 the Commander-in-Chief of the Kwantung Army, General Ueda Kenkichi, had assembled his corps commanders to inform them that henceforward ‘where the border lines are indistinct, the defence commander shall determine a boundary on his own’. As far as he was concerned, they could feel at liberty ‘to invade Soviet territory temporarily, or to decoy Soviet soldiers and get them into Manchukuoan territory’.∗ It is unclear whether the trespass on 11 May 1939 stemmed from these instructions (which directly contradicted standing orders of Imperial General Headquarters). In the event, the local divisional commander ordered a Kwantung Army cavalry regiment to support the local Manchukuoan forces. When the Mongolian horsemen withdrew across the Halha, the Kwantung Army ordered its own cavalry to pursue and destroy them. A thousand infantrymen from the Kwantung Army’s Sixty-fourth Infantry Regiment were dispatched to back up the cavalry. The trap claimed its first victims. The Japanese cavalry regiment was all but destroyed and the Japanese infantrymen narrowly avoided the same fate. Enveloped by Mongolian tanks and artillery before they had moved away from the river crossing, the Sixty-fourth Infantry Regiment were routed. This defeat was regarded by the Japanese as humiliating in the extreme, but no immediate steps were taken to avenge their defeat. For their part, the Russians issued an official warning that, by virtue of its defence treaty with Mongolia, Moscow would treat any further incidents on the Mongolian frontier as direct aggression against the Soviet Union.
By 18 June reports were received which indicated that Soviet forces had advanced once more across the frontier, driven back the Manchukuoan garrison forces (which may indeed have provoked them), and had bombed three major defensive positions to the rear. Within the Kwantung Army, a debate ensued during which it emerged that some staff officers were inclined to take no action pending the outcome of Anglo-Japanese diplomatic conversations that were attempting to resolve the Tientsin Crisis and other outstanding sources of friction between Britain and Japan. Other voices, however, were raised in favour of inflicting an immediate blow against the Soviet Union sufficient to teach the Kremlin to avoid any further trouble with the Kwantung Army. The hotheads prevailed.
The Russians were far from blind to what the Japanese wanted to achieve at Nomonhan. Master spy Richard Sorge’s top Japanese agent remained privy to the Japanese Cabinet’s fervent desire to avoid war against the Soviet Union at virtually any price. The same agent also acquired detailed information from the South Manchurian Railway concerning the size and order of battle of the forces available to the Kwantung Army. Another Japanese in Sorge’s employment toured the battlefront as a press correspondent, interviewed top-ranking Japanese military commanders at the scene, outlined the limited objectives of the Japanese forces and reported upon the numerical strength and types of Japanese aircraft and armoured fighting vehicles which had reached Nomonhan. Still other agents in Sorge’s network passed on information about Japanese troop movements, the mobilization and state of readiness of Japanese air and armoured formations, and the strength of every unit which either had been committed to the battle or which might be in a position to take part in any full-scale offensive. Sorge himself learned from his contact with German military attachés and others that the Japanese had no intention of using the Incident as a pretext for a general invasion of the Soviet Union. The Kremlin may simply have chosen to disbelieve or to discount Sorge’s Intelligence on this occasion. Moscow may have feared that the Nomonhan dispute was bound to escalate into a wider conflict unless drastic counter-measures were undertaken. On the other hand, the dispute provided a tempting opportunity to pay off a few old scores.
Kwantung Army Headquarters resolved to administer a crushing defeat upon the Russian and Mongolian forces by utilizing virtually all the rapid deployment forces of the Kwantung Army: the Twenty-third Infantry Division, nearly the whole of the Second Air Group, two regiments of light and medium tanks, a regiment of mechanized artillery, and an infantry regiment borrowed from the Seventh Division. This produced a force of some 15,000 men comprising thirteen infantry battalions, 120 anti-tank guns, 70 tanks, 400 vehicles and 180 aircraft, leaving elsewhere only sufficient strength to safeguard the remainder of Manchuria. At this point the Soviet and Mongolian forces in the vicinity of Nomonhan were only less numerous than the Japanese, disposing of 12,500 men, 23 anti-tank weapons, 186 tanks and 266 armoured cars. Recalling recent experience and the Kwantung Army’s formidable reputation as Japan’s most battle-hungry military command, the odds appeared to be stacked heavily in favour of the Japanese. The War Ministry in Tokyo did hesitate to sanction an adventure which could contribute nothing to the China Incident (which was continuing to sap the nation’s military and economic strength), but the Army General Staff felt justified in giving General Ueda its full support for the operation. War Minister Itagaki, who had retained a fondness for the ‘forward policies’ of the Kwantung Army ever since he had joined the conspiracy to assassinate Chang Tso-lin in 1928 and went on to engineer the Manchurian Incident in 1931, now decided to override the objections of his ministerial staff and gave his approval to the plans submitted by the Commander-in-Chief of the Kwantung Army. Significantly, however, General Ueda felt obliged to conceal his intention to open the campaign with an air offensive against the Russians: he was not the kind of man to shrink from the fact that a major air offensive, even if justified as a pre-emptive strike, was bound to provoke the Russians beyond measure. His worry was that revelation of his intentions would heighten the concern of those in Tokyo who wanted to avoid any serious escalation of the conflict: premature disclosure might prompt even the hawkish Army General Staff to abort the whole affair. When news of Ueda’s secret intentions did come to the notice of Tokyo, his worst fears were abundantly fulfilled. A telegram was sent to Ueda ordering him to stay his hand pending the arrival of an emissary from the Army General Staff.
After considering the matter, the Kwantung Army staff responded by resorting to a trick which had served them well in early stages of the Manchurian and China Incidents: they advanced the date of their operations and pounded the Russian air base of Tamsagbulag before Tokyo could stop them. The returning pilots jubilantly claimed to have destroyed ninety-nine enemy aircraft in aerial combat and a further twenty-five on the ground. Tokyo was faced with yet another fait accompli.
Imperial General Headquarters in Tokyo was outraged by this latest example of the Kwantung Army’s irresponsible and insubordinate attitude. The Emperor, too, was infuriated and demanded that General Ueda, if no one else, should be brought to book. Kwantung Army Headquarters, unrepentant, politely invited Tokyo to leave the fighting to the men at the scene. The Army General Staff and political authorities in Tokyo could do little else. They did force the Kwantung Army to suspend attacks against the Russian air bases, but air combat missions continued over the war zone. At the beginning of July, the Kwantung Army’s Twenty-third Infantry Division crossed into Outer Mongolia. They were met by several hundred Russian tanks. At first the Japanese seemed destined to repeat their performance at Changkufeng. More than a hundred enemy tanks were set ablaze. But the Russians held, forcing the Japanese to withdraw across the river after less than two days of combat. Meanwhile, further south, a Japanese tank offensive was launched across the Halha River. After losing forty tanks, the momentum of the attack faltered and the Japanese again had no option but to withdraw. By this time more than half of the tanks deployed by each side had been put out of action after only a few days of combat. The Japanese were in a far worse position than the Russians to sustain such crippling losses. Moreover, the Japanese were alarmed to find that their own medium tanks were easily penetrated by Russian anti-tank guns and that Russian armour was proof against cannonfire from Japanese tanks. Most of the Russian tank losses to date had been due to grenade attacks pressed home by Japanese infantrymen who lacked effective artillery support. In an effort to remedy this imbalance, the Kwantung Army scoured its depots across the length and breadth of Manchuria and managed to gather together nearly a hundred heavy guns to bear upon the Russians. Their gunners laid down a barrage of some 15,000 shells a day and another Japanese offensive began.
The Russians replied with even heavier counter-fire. The Kwantung Army once again had to break off the engagement. The Army General Staff in Tokyo was ready to concede the victory to the Russians then and there, but the Kwantung Army was determined to continue. Then, unexpectedly, the Russians sent their bombers 200 miles deep into Manchuria. The Kwantung Army redoubled their efforts to raise the ban imposed by Tokyo on air strikes upon the Russian air bases. Imperial General Headquarters, however, flatly refused and intimated that the Kwantung Army was reaping a harvest which it richly deserved.
The Chief of Staff of the Kwantung Army, Lieutenant-General Isogai Rensuke, was summoned to Tokyo and on 20 July was told in no uncertain terms that the Kwantung Army must do everything within its power to bring the Incident to an end. If it should prove impossible to resolve the dispute by diplomatic means, then Tokyo would have to order the Kwantung Army to withdraw beyond the boundary now claimed by the Soviet Union. General Isogai was incensed by these instructions, flatly refusing to consent to them. In the end, however, he grimly agreed to convey the wishes of Imperial General Headquarters back to his Army Headquarters in Manchuria (where they were ignored).
By 21 July Japanese Intelligence reports were accurately predicting that the Russians would open their own offensive in a month’s time. The Japanese anticipated, however, that the Kwantung Army could absorb whatever punishment that the Russian and Mongolian side might inflict. The military and political authorities in Tokyo remained as disturbed as formerly at the capricious behaviour of the Kwantung Army, and this certainly contributed towards their unwillingness to reinforce the units already engaged in the campaign. One trainload after another ferried fresh Soviet troops and up-to-date weaponry along the Trans-Siberian Railway. They disembarked and moved up to their assembly points beyond the Halha. In the early part of August, Tokyo reluctantly consented to a lifting of the ban upon air operations over Outer Mongolia: it was a purely precautionary measure, not an escalation of the conflict. News of this soon reached Marshal Zhukov. Then bad weather obscured Japanese aerial reconnaissance over the battlezone for a fortnight. The Russian commander seized his opportunity. The trap was sprung.
Rank upon rank of armoured vehicles backed by massed infantry and coupled with close air support (now far more effective than at Changkufeng) churned through the Japanese and Manchurian lines. Only when the weather began to clear did the Japanese discover to their horror that they were caught in a pincer movement by three infantry divisions and five tank brigades while Zhukov held two further divisions in reserve. This was four or five times the numerical strength of the faltering Japanese and Manchurian forces who were powerless to resist the Russian and Mongolian onslaught. The Japanese lines simply crumpled beneath the weight of the enemy attack. Within little more than the span of a week the remnants of the Japanese and Manchurian forces fled back across the Halha and beyond until they reached what was indisputably Manchurian soil. The Russo-Mongolian forces did not yield to any temptation to follow in hot pursuit. They settled down along the line of what they asserted to be the proper frontiers of the People’s Republic of Mongolia and began to fortify it. This was Changkufeng turned upside down.
Deluded by the thought that Japanese divisional and company commanders at the scene had shown insufficient initiative, that courageous individual actions by common soldiers had somehow parried the Russo-Mongolian forces which (by this reckoning) must have lost any stomach for further fighting, Kwantung Army Headquarters decided to risk everything on an immediate, all-out riposte. Four more divisions as well as all the heavy artillery and quick-firing weaponry left in Manchuria were to be thrust into a counter-attack which was planned to open with a week of relentless night attacks from 10 September. The purpose of this whole operation was not merely to erase the Soviet gains by driving back the Russian and Mongolian armies. That was merely a preliminary objective. It was to be expected that military operations would have to break off for a season of suspended animation in the winter months, but Kwantung Army Headquarters planned its autumn offensive to herald nothing less than the opening of a general war against the Soviet Union in the following spring. Understandably enough, Imperial General Headquarters in Tokyo blenched. It was a perfectly ludicrous scheme. The Kwantung Army was told to abandon it at once.
Many considerations affected the outcome. Events were moving swiftly in international affairs and the chronic internal wrangling of Army politics grew if anything more important. War in Europe was imminent. Hitler’s preparations for the attack on Poland, completed since June, had been suspended while the Russo-Japanese conflict at Nomonhan developed. Then, just as the Soviets launched their offensive against the Kwantung and Manchurian Armies, news came of the Nazi-Soviet Pact. The groundwork for that had been well prepared months before, but the Japanese had been duped by the Germans into discounting evidence of what was afoot. The threat of a Russo-German War now abruptly receded, leaving the Japanese completely isolated and exposed. No Japanese Cabinet could have survived such a shock. The Hiranuma Cabinet duly fell on 29 August 1939.
War Minister Itagaki was replaced by the less mercurial General Hata Shunroku, who had last fought the Russians during the Russo-Japanese War of 1904–5. Hata knew what kind of difficulties would confront Japan in its efforts to seek a negotiated settlement of the dispute: he had taken an active role in negotiating the Russo-Japanese Convention of 1925 by which Japan and the Soviet Union first established normal diplomatic relations in the aftermath of the Siberian Intervention. Furthermore Hata, who had been Commander-in-Chief of the Japanese Expeditionary Forces in Central China in the wake of General Matsui’s disgrace after the Rape of Nanking, was acutely conscious of the quagmire into which the Japanese had slipped in China, and as if that were not enough to make him more cautious than Itagaki as War Minister, Hata had been engaged as Chief ADC to the Emperor in May 1939 on the eve of the Nomonhan Incident and thus knew the earnestness with which Hirohito yearned for peace.
However well suited Hata may have been for the onerous tasks which lay before him, neither his appointment as War Minister nor General Abe Nobuyuki’s selection as Prime Minister could be made without the endorsement of the Army General Staff. That approval was willingly granted, thus setting the seal upon the realignment of national politics that now took shape. The Army General Staff appeared to have taken to heart the painful lessons administered by Zhukov’s battalions, and Japan emerged from the Nomonhan Incident with a consensus between Imperial General Headquarters and the Government which had been conspicuously absent in the recent past. General Abe and his Cabinet were committed to ending the war as soon as possible, and on 2 September, as early news of the German invasion of Poland reached Japan, the Japanese Army General Staff decided to wait no longer but to break off the Nomonhan Incident by effecting a unilateral withdrawal.
The peace negotiations for which neither Japan nor the Soviet Union had shown enthusiasm between mid-July and late August, were rekindled by the Abe Government in early September. On 15 September an armistice was signed, sparing the Japanese Army General Staff the embarrassment of having to implement its decision of 2 September. All concerned recognized that the ceasefire confirmed the comprehensiveness of the Kwantung Army’s military defeat. Detailed discussions held over the months which followed led step by step to further concessions by the Japanese, who finally yielded to the Russians virtually every bone of contention. The end result was marked by a formal accord signed on 18 July 1940: the Japanese diplomats, as usual, were obliged to accept the obloquy of a defeat which quite properly should have been reserved for the military alone.
The military, meanwhile, found other scapegoats, too. Matters were patched up so as to disturb the system as little as possible. Nothing like the great debate over the Washington Treaty System was to be allowed to fragment the Japanese military and political establishment. General Ueda Kenkichi, the ill-fated Commander of the Kwantung Army, and his Chief of Staff, General Isogai Rensuke, were simply replaced by less volatile men. Ueda was never again to be re-employed by the Army. He was so well buried that when he re-surfaced to give evidence at the Tokyo War Trial after the war, nobody even remembered to ask him about what he might recollect of the Nomonhan Incident. General Isogai, after years of disgrace, had the dubious honour to be resurrected as Governor-General of Hong Kong from January 1942 until the war was all but over. He, too, excited no curiosity from either side at the Tokyo Trial: the Nomonhan Incident was a chapter which both sides wanted to forget, except in abstract formulations, and yet it had been one of the most significant armed clashes in the modern history of East Asia.
In geographical terms, the Nomonhan Incident had covered a far wider front than the Amur River and Changkufeng Incidents: it had spread out over an area roughly forty miles wide and eighteen or twenty miles deep. That scarcely accounts for its true significance. It is conventional to judge the relative importance of armed conflicts in terms of their human and material costs. Altogether the Kwantung Army and its protégé, the Manchurian Army, sent 56,000 troops into battle at Nomonhan. It cost Japan and Manchuria 8,440 dead and 8,766 wounded, a devastating 32 per cent casualty rate. The veteran Twenty-third Infantry Division, which had borne the brunt of the fighting, suffered 11,000 casualties, three quarters of its original strength. One can only guess at the strength of the Soviet-Mongolian side but certainly after the opening rounds it was both qualitatively and numerically superior to the forces disposed by the Japanese. It is said that the Russians and Mongolians sustained about 9,000 casulaties. Moscow never disclosed the true figures. The real significance of the Nomonhan Incident, however, extends beyond its human and material balance-sheet. To a great extent it was to determine the shape of the Total War that was yet to come.
From this point onwards the Kwantung Army, although previously regarded as the toughest and most audacious military force in East Asia, simply ceased to exert any decisive influence upon history. To all intents and purposes, it suffered a paralysing stroke at Nomonhan. It withered away and, by 1945, survived only in a twilight of senility which left it incapable of resistance when the Russians came to bludgeon their way across Manchuria. In a larger sense, the Nomonhan Incident provides us with a salutary demonstration of how limited military operations can affect the overall strategical balance between rival nations. It forced the Imperial Japanese Army to abandon any further serious design to mount a northern offensive to secure control over Eastern Siberia. The more bold and radical elements within the Army began to consider the merits of a southern advance instead. That was a strategy dear to certain middle-ranking cliques in the Imperial Navy with whom they were in contact. It was regarded by the Navy’s more level-headed upper echelons as pie-in-the-sky, a mere nostrum that admittedly had served to justify ever-greater naval expansion over the years and in the past had helped to curb the hotheads who up to now had wanted to swarm across the Russian frontier. However cynically the Navy’s most senior commanders may have paid lip-service to the southern advance idea in the past, it had been merely as a theoretical framework rather than as a blueprint for aggression. Now they and their civilian counterparts found themselves unwillingly drawn along by impatient junior officers who demanded deeds and despised masterly inactivity.
The lull in the north lengthened. The Kwantung Army and the Russians continued to probe each other, but the frontier gradually fell quiet. The small-scale incidents, the constant shooting and skirmishes, the espionage and incitements, which had from the start seemed natural to the relationship between these two Powers, dwindled. That is not to say that either side knew peace. At the Tokyo War Trial, Soviet Prosecutors would accuse the Japanese of having violated the frontiers of the Soviet Union on 49 occasions in 1940, 136 in 1941, 229 in 1942 and 414 in 1943. The Soviets would also charge the Japanese with having flown over Soviet territory on 56 occasions in 1940, 61 in 1941, 82 in 1942 and 119 in 1943. In reply, the Japanese would claim that the Soviets had violated the frontiers of Manchuria on 151 occasions in 1940, 98 occasions in 1942 and so on. Whatever the accuracy of these figures (and they cannot be said to be reliable), we must not doubt that these borders remained exceedingly sensitive. There was plenty of rumour, and the outbreak of war on the frontier was still regarded as a very natural possibility by the Japanese. The psychology had not changed: Japan remained malevolent and insecure towards Russia. Russia was still regarded, with deadly cold hostility, as a national enemy, in a way in which China, even at the height of the war between the two Governments, was not. But it became clear that in one way Japan had changed its behaviour. Unless attacked in Manchuria, it was content to do no attacking. More importantly, a powerful current began to pull Japan towards some kind of detente with the Soviet Union. Some Army circles even spoke up in favour of an alliance between Germany, Japan and the Soviet Union. It was an absurdly unrealistic idea. A neutrality pact between Japan and the Soviet Union, however, was a practical proposition. It came to pass in April 1941, by which time Japan was well on her way towards the Pacific War.
Although one cannot say that peace subsisted between Japan and the Soviet Union after Nomonhan, an unnatural silence descended upon their common frontiers while, in other parts of the world, fighting broke out with great savageness.