Note: This was intended to be the "decisive battle" for May, the month of the initial engagement that opened this nearly half-year long campaign. My own sloth and inattention prevented this post from publication in that month, but the actual history of this engagement was such that you could almost pick a handful of dates as the "battle of Khalkhin-Gol". Was it the original Mongolian incursion in early May? The destruction of the Azuma force in late May? The Japanese offensive in early July? The second Japanese attack in late July? Or the final Soviet offensive in late August?
This is the third and final entry in the "The Imperial Japanese Army of WW2 - What Went Wrong?" series. The first was were Bataan, the second Kohima. The only link between the three - which were widely separated geographically, fought against three different enemies, and included a win and two disastrous losses - was the armed force of Imperial Japan.
I believe that all three speak eloquently about that force.
Much of what they say speaks well of the courage and resourcefulness of the Japanese soldier and his officers. But, I believe, they all have some bitterly scathing observations on the fundamental failings of the Japanese military and particularly the Japanese Army as both an organization and for its influence on the policies of the Japanese nation during the early years of the 昭和時代 Shōwa jidai, the reign of Emperor Hirohito.
Forces Engaged: Imperial Japanese Army (大日本帝國陸軍, Dai-Nippon Teikoku Rikugun) The IJA force that did the bulk of the fighting at Khalkhin-Gol was the 23rd Infantry Division (第23師団 Dai-nijūsan Shidan) of so-called Kwantung Army (関東軍, Kantō-gun).
The 23rd, also known as the "Sunrise Division (旭兵団, Kyokuhei-dan) was a relatively young outfit, having been raised in the southern city of Kumamoto only the year before. It was a "triangular" infantry division, formed of three infantry regiments of three battalions each. Divisional troops included the division artillery, a reconnaissance regiment, engineer, and transport regiment.
The notional strength of a typical triangular division was about 12,000 all arms and included 36 x 75mm cannon (in the divisional artillery) as well as 12 x 74mm Type 41 mountain howitzers and 18 x 70mm Type 92 infantry howitzers. The infantry regiments had an antitank-gun company for a total of 18 x 37 or 47mm AT guns.
In this organization you can already begin to see some problems for the Japanese grunt.
First, the artillery strength of an IJA infantry division was weak both in numbers and in weight of metal. The 70mm guns were little better than big mortars and, because they had no fire direction section, practically limited to direct fire. So in practice the direct support artillery for the IJA infantry consisted of something like 48 cannon, none larger than 75mm.
The experience of WW1 had shown that projectiles smaller than 100mm were almost useless against well dug-in infantry or armor.
Add to that the training and equipment deficiencies of the Japanese redlegs, which included antiquated or lacking commo gear, unrealistic training, and a Verdun-era reliance on preplanned fires. In a war where the artillery would prove to be the King of Battle the IJA had a very unimpressive royal house.
Another huge issue with the IJA was supply and transport and it showed in the 23rd's "transport regiment". "Transport" could as easily be horses, mules, or human porters as trucks. This owed much to the barely-mechanized condition of the IJA (and the larger Japanese society supporting it) but also to the primitive conditions on the frontier. "Roads" were often mere tracks and were uniformly unpaved; sandtraps in summer, bogs in winter. Transport in Manchuria depended on railways; typically an IJA staff plan for the Manchurian region considered any major operation beyond one day's drive (250km) from the railheads unsupportable (Drea, 1981).
The other units we know to have been present at Khalkhin-Gol include:
26th Infantry Regiment (7th Infantry Division), roughly 3,000 infantry and support troops
(Note: the authorized strength of IJA infantry units seems bizarrely huge to me, used to the smaller organizations of a modern army. For example, an IJA line infantry company was authorized around 200 troops and a battalion close to 1,000. Actual strengths at the time of Khalkhin-Gol apparently ran about 80%, so a typical battalion would have carried some 800 on strength for a total of 2,400 plus the regimental troops which included the signal, AT, and cannon companies.)
This unit was borrowed from the experienced 7th ID to stiffen the green 23rd ID for the early July attacks into Soviet Mongolia.
8th Border Guards Regiment Probably 3,000 to 3,500 infantrymen (Note: These "border guard" units were from the Quisling government of "Manchukuo", the Japanese puppet state of north China. I can find little about them other than they appear to have been organized similarly to an IJA infantry regiment, although likely without the regimental troops particularly the artillery.) In effect a light infantry outfit composed of Chinese troops with Japanese officers (or, at least, a Japanese commander is listed for the unit)
Participated largely in the static defense in July and August.
3rd (Medium) Tank Regiment (3TR, part of the Yasuoka Detachment) probably about 600-800 troops with 41 vehicles including 26 x Type 89 I-Go medium tanks, 4 x Type 97 Chi-Ha medium tanks. 7 x Type 94 tankettes, and 4 x Type 97 Te-Ke tankettes.
4th (Light) Tank Regiment (4TR, part of the Yasuoka Detachment), 600-800 tankers with 35 x Type 95 Ha-Go light tanks, 8 x Type 89 I-Go medium tanks, and 3 x Type 94 tankettes.
(Note: I'm finding some discrepancies in the strengths of these tank units. The above vehicle numbers are from the Wiki entry that cites Coox (1985) as the authority for these numbers. They conflict slightly with the numbers given for the units in Drea (1981) of 25 medium tanks (type unspecified) for 3TR, 35 light tanks and 7 medium tanks for 4TR. Drea (1981) also includes armored cars as either assigned or attached to both of these units. If I cannot find another authority I will simply have to accept that there are issues with unit strengths)
13th Artillery Regiment and 1st Independent Artillery Regiment (Kwantung Army troops?) Probably about 2,000 artillerymen with an estimated 24 cannon of various types. I cannot find MTO&E for these units.
However, I have encountered orders of battle for other "independent artillery regiments" in the IJA as well as the strengths listed in the 1944 U.S. Army TM-E 30-480 Handbook on Japanese Military Forces. So my best guess is that these units would have included about 1,500 to 2,500 artillerymen and 24 x Type 91 105mm or possibly Type 4 140 mm howitzers. However, some independent artillery units were largely armed with 75mm fieldpieces - typically 16 - and only 4 each 105mm and 150mm.
Air support was provided by the 2nd Army Air Division (2nd Hikoshidan). In early May a Rinji Hikotai (Provisional Air Group) was tasked to Hailar (about 160km NE of Nomonhan). This unit included 19 x Ki-27 ("Nate") fighters 6 x Ki-15 ("Babs") and 6 x Ki-30 light bombers.
By June IJAAF air strength included 4 reconnaissance A/C with the HQ element, 15 reconnaissance Ki-15 A/C directly attached to 23rd ID, and two combat wings with 125 aircraft: 12th Hikodan with 88 Ki-27 fighters and 9th Hikodan with 24 Ki-30 light bombers and 13 Ki-21 ("Sally") medium bombers.
Rising from less then 500 in May to about 20,000 all arms, ~90-100 cannon (and most of those 76mm or smaller), ~130 tanks, and ~250 aircraft by the end of August, the IJA ground forces were largely under the command of LTG Michitarō Komatsubara (小松原 道太郎, Komatsubara Michitarō), the 23rd ID commander.
Soviet Union - the Soviet forces in the Khalkhin-Gol area consisted of Soviet troops from the Trans-Baikal Military District and local Outer Mongolian troopers; the Outer Mongolians were to the Soviets what the Manchukuoans were to the Japanese.
From the original handful of Mongolian cavalry from the 6th Cavalry Division that ambled across the Halha River on or about 11 MAY 1939 the Soviet forces increased over the spring and summer.
The following are listed as engaged in the May actions:
6th Mongolian Cavalry Division Don't get these guys wrong; they weren't the wild riders of Genghis Khan. By 1939 smart cavalrymen realized that the big fat horse was transport, like a truck or a bicycle.
They probably carried a sword or three (they were still Mongols, after all...) but were primarily armed with rifles and dismounted to fight. Typical Soviet cavalry division organization in 1941 is listed as "...four cavalry regiments, a horse artillery battalion (8 x 76mm guns and 8 x 122mm howitzers), a tank regiment (64 x BT-series tanks), an anti-aircraft battalion (eight 76mm AA guns and two batteries of AA machine guns), a signals squadron, a field engineer squadron and other rear echelon support units and sections. The total authorized strength of a cavalry division included 8,968 personnel and 7,625 horses..." (Wiki, 2013)
Bykov Detachment 1 reinforced rifle battalion (with 8 x T-37 tanks and 21 BA-6 and FAI armored cars attached), 1 FA battery; probably 1,000 all arms
149th Motorized Rifle Regiment (MRR) (36th Motorized Rifle Division) less one battalion, and
175th Artillery Battalion
Total Soviet-Mongolian strength in May is listed as 2,300 all arms (including about 1,300 Mongolians), 12 x 76mm guns, 8 x 45mm anti-tank guns, 4 x 122mm howitzers, 4 x SU-1-12 self-propelled artillery vehicles (this thing...
...a sort of giddy harumphrodite with a 76mm cannon on a pedestal mount in back of a GAZ 4x6 truck.
As goofy as it looks it seems to have worked, proving my old drill sergeant's striction that if it's stupid and it works, it's not stupid.), 8 x T-37 tanks, 5 x HT-26/OT-26 flamethrower tanks, 39 assorted armored cars including BA-6 and FAI types.
By August, though, the Soviet First Front Army looked like this:
6th Mongolian Cavalry Division
7th Armored Brigade
1 x medium armored car battalion
1 x reconnaissance battalion (medium and light armored cars)
1 x machine gun battalion
601st Infantry Regiment from 82nd Rifle Division
11th Tank Brigade (-) (2Bns - BT-5)
36th Motorized Rifle Division
5th Machine Gun Brigade
82nd Rifle Division minus the 601st Infantry Regiment.
57th Rifle Division
11th Tank Brigade (-) (2Bns - BT-5)
6th Tank Brigade (-) (3Bns - BT-7)
8th Mongolian Cavalry Division.
9th Armored Brigade
(as 7th Armor)
6th Tank Brigade (-) (1Bn BT-5)
212th Airborne Brigade.
Soviet air support (the Soviet air arm was the Военно-воздушные силы, Voyenno-Vozdushnye Sily) initially consisted of two air groups; 70th IAP (Fighter Aviation Regiment) assigned 24 x Polikarpov I-16 and 14 x Polikarpov I-152 (also known as I-15) fighters and 150th SAP (Composite or "Mixed" Aviation Regiment) assigned 29 x SB medium bombers and 15 x Polikarpov R-5 light bomber/reconnaissance aircraft.
Reinforcing units included 22nd IAP (3 x I-152 and 28 x I-16 Type 10 fighters) and 38th SBAP (Composite Bomber Regiment, 59 x SB bombers).
Approximately 50,000 to 55,000 all arms, ~ 500 cannon (including 100+ over 100mm), ~500 tanks, ~800 aircraft under GEN Georgy Zhukov, GEN Grigoriy Shtern and GEN Yakov Smushkevich
The Sources: Here we the English-speaking and -reading publics have the same problems we had dealing with the Battle of Tsushima; all the primary sources are not in English.
Add to that the inevitable difficulties in dealing with two societies that - at least in 1939 - dealt with information as weapons. The opening of the former Soviet archives since 1989 has helped, obviously. And there has been some good work done on the subject in English, which I'll try and present here. But the initial difficulties persist and are reflected particularly in the sources available on-line.
One of the most worthwhile sources is the 1981 Leavenworth paper by E.J. Drea; Nomonhan: Japanese-Soviet Tactical Combat, 1939.
While this work is largely focused on the battalion-level (it reports and discusses the events recorded in the war diary of the IJA 2nd Battalion, 28th Infantry) it provides a clear, authoritative account of the larger engagement.
On the other hand, LTC Charles Otterstedt's 2000 War College paper The Kwantung Army and the Nomonhan Incident: Its Impact on National Security does a good job of describing the broader geopolitical context of the events with an emphasis on the politico-military finagling of the boys in the Kantō-gun.
The Wikipedia entry is well researched and cites Drea (1981) as well as the other most commonly-cited published source, Alvin Coox's 1985 work Nomonhan: Japan Against Russia, 1939.
One recent source it cites appears well worth pursuing; Stuart Goldman's Nomonhan, 1939; The Red Army's Victory That Shaped World War II, just published last year.
Someone named Ryan Schultz produced a fine work of scholarship for his thesis for Oberlin College in 2011. That paper, The Failure of Japanese Tactics at Changkufeng and Nomonhan and Lessons Left Unlearned is found here.
Among the more useful internet resources is a PDF version of an English translation of a 2002 paper in Russian by one M. Kolomiets entitled Boi u reki Khalkin-Gol from the Russian journal Frontovaya Illyustratsiya ("Frontline Illustrated").
Another is Bellamy and Lahnstein (1990) The New Soviet Defensive Policy: Khalkhin-Gol 1939 As Case Study from the September issue of Parameters. This article at "Warbird Forum" does a decent job of covering the air war from the Japanese side, while something called "Håkans aviation page" has a fairly good summary of both Soviet and IJAAF air operations.
Since the engagement at Nomonhan/Khalkhin-Gol took place over the course of months there is little use in subdividing the events into the "campaign" versus the "engagement" itself, so the course of the fighting will be summarized in a single section. However, since the situation in western Manchukuo/eastern Mongolia is so little known today I consider it worthwhile to briefly discuss the setting for the events of 1939.
The Situation: What really brought the two armies together in the plains of western Manchuria was the Russo-Japanese War of 1905. Before that time it was Russia and Qing China that had tussled over what the Chinese called Northeast China (東北失地) and the Russians called Приаму́рье (Priamurye) or "Outer Manchuria", the light pink corona around the dark red of the Manchurian heartland in the image below.
In particular the northwestern portion of Manchuria adjacent to the somewhat notional state of Mongolia was a very ill-defined piece of real estate. This area may well be among the most desolate and forsaken on earth, characterized by sandhills and grass and very little else. The people who have passed over the region for centuries are largely nomads, Mongol and Manchu, following the herds that must move on or worry the grass from the loose, sandy soil and turn the steppe back into a barren sea of blowing loess and sand dunes.
With no real interest in any particular piece of the steppe the idea of a "border" would have seemed risible to the nomads. Everything was just "the land"; a place to graze horses and sheep, to pitch a yurt on, for horse-lords to ride over and pass by.
But the rise of settled nations inevitably led to the establishment of borders. Where the eastern edge of "Mongolia" and the western edge of the Manchurian part of China met the "border" was still just as sketchy as ever.
Otterstedt (2000) sums up the place nicely:
"Perhaps the most vague border issue was that on the flatlands between western Manchuria and eastern Outer Mongolia. For over two centuries, the vicinity of Nomonhan has been the boundary line separating the pastoral plains of the Halha Mongols of Hulun Buir, and the Kalmuks of Outer Mongolia (but)...(n)o definitive borderline was every drawn or observed."Where Japan marches onto the scene is after its defeat of Russia in 1905. Japan - which was already beginning to catch the imperial fever incubated in the European powers - latched onto Korea and portions of northeast China, specifically what it called Ryojun (or Port Arthur, now Lüshunkou, 旅顺口) and the adjacent bits of the Liaodong Peninsula and something called the "South Manchuria Railway". The latter becomes more important, but the former was the first critical bit of business because the IJA formed a garrison for the Port Arthur/Liaodong region (that the Japanese government designated 関東州, Kantō and in English we call Kwantung.
This outfit was called the 関東都督府陸軍部 or "Kwantung Governor General Prefecture Army Section" and consisted of an infantry division, an artillery battalion (typically described as "heavy siege artillery" but in my opinion more likely gunners for the Ryojun artillery defense batteries), and a half-dozen separate battalions of supposed "railway guards" that were in fact regular Japanese infantry; about 10,000 troops all arms.
After 1919 this organization was renamed the Kantō-gun, the "Kwantung Army". Coox (1985) describes the mindset that informed the Japanese government at the time of the formation of this army.
"Chinese mounted bandits, demobilized soldiers, vagrants, and other lawless elements poised a familiar threat in Manchuria, especially in the rural sectors. But the major concern of the Japanese High Command was the possibility that the Russians would seek revenge for their recent defeat. Consequently, the traditional defensive or reactive emphasis of the Japanese Army was transformed...and a clear-cut offensive strategy designed."The Army retained this aggressive mindset throughout its existence; it was the intellectual center of the "Northern Expansion Doctrine" (北進論, Hokushin-ron) which insisted that the most profitable direction for Imperial aggression was to the north and west into the Siberian hinterlands combined with the conquest of China.
From 1906 to 1931 the Kwantung Army wasn't able to actually do anything much about this; they were separated from the Soviet border by the Chinese province of Manchuria.
Mind you, the Japanese wanted all that Manchurian goodness.
Manchuria, to the Hokushin-ron types in the Kantō-gun meant lots of lovely raw materials, new places to hawk Japanese goods, along with the sort of lebensraum that the fascist governments of the Thirties and Forties seem to have been obsessed with. You'd think with a restive Korea already inside the imperial borders that the Japanese Army would have been happy to leave the wilds of Manchuria alone, but you'd have counted without the Kwantung Army.
In 1931 a group of expansionist officers led by a couple of characters on the Army staff, COL Seishirō Itagaki and LTC Kanji Ishiwara, set off what is today called the "Mukden Incident", a sort of Reichstag-fire false flag plot that was designed to implicate the Chinese in an attack on the South Manchuria Railway and provide Japan with a casus belli.
It went slicker than water off a cat's ass and within six months Japan effectively owned all of Manchuria, which it set about organizing into a supposed-state called "Manchukuo" supposedly run by the deposed Qing Emperor, Puyi. In fact it was a military depot for Japan run by a Kwantung Army that set about trying to make its "Northern Strategy" into reality.
And that's where it ran into the Soviets.
Here's Otterstedt (2000) again:
"With the founding of Manchukuo, the border around Nomonhan became the scene of disputes between the Outer Mongolian "People's Republic" and the Hsinking (Manchukuoan) authorities. Since the strength of the Kwantung Army far exceeded that of the Soviet Far Eastern Army between 1931-1935, no border disputes broke out during that period. (H)owever, after the Soviets gradually built up their Far Eastern military strength and instituted far more stringent control over Outer Mongolia...(b)order disputes erupted with increasing frequency..."The Wiki entry for "Japanese-Soviet border conflicts says that
"Imperial Japanese Army recorded 152 minor incidents on the border of Manchuria between 1932 and 1934. The number of incidents increased to over 150 per year in 1935 and 1936, and the scale of incidents became larger. In January 1935, the first armed battle, Halhamiao Incident (哈爾哈廟事件 Haruhabyō jiken?) occurred on the border between Mongolia and Manchukuo."Clearly there was trouble brewing on the empty places around Manchuria, and both the Japanese government through the Kwantung Army and the Soviet Far Eastern Army were willing to mix it up.
By 1939 the forces on either side of the border were substantial. The Kantō-gun included 8 infantry divisions, two cavalry and tank brigades and something like thirteen "border guard" units supported by three air brigades; a total of some 250,000 troops. The Soviets, having to cover much longer exterior lines, had between 20 and 30 rifle divisions, 5 to 7 cavalry divisions and 6 to 8 mechanized rifle brigades; perhaps 500,000 troops, 2,500 tanks, and 2,500 aircraft (Otterstedt, 2000).
Between 1937 and 1939 the border spats escalated into serious fighting.
In June of 1937 three of some sort of Soviet river gunboats or small craft cruised by troops of the IJA 1st Division fortifying Kanchazu Island (乾岔子, a sort of instream bar of the Amur River the Russians called "Bolshoi Ussuryski Island"(болъшои Уссурыскии Остров) on the northeastern border of Manchuria. The Japanese opened up with 37mm AT artillery and sank one of the small craft and damaged another.
The Soviets, busy purging everyone in sight and other business in both Europe and Asia, chose not to retaliate. Another, more serious "incident" occurred in the summer of 1938. In the Changkufeng Incident (Хасанские бои and 張鼓峰事件, Chōkohō Jiken) the IJA 19th Infantry Division supported by Manchukuo units pushed elements of the Soviet 39th Rifle Corps off high ground near the Soviet-Korean border in July, but Soviet reinforcements returned the favor in August, shoving the IJA units back to or close to their start lines.
The 1938 encounter was not decisive but played a big role in clarifying the Kwantung Army's ideas of what it would do if challenged on the Soviet borderlands; in April 1939 the Kwantung Army drew up The Principles for the Settlement of Soviet-Manchurian Border Disputes which laid out the first two cardinal principles of the way it saw the Manchurian borderlands:
(1) The basic policy is never to invade and never to be invaded.
(2) If an enemy violates the frontiers, he must be wiped out at once.
It's worth noting that this little policy document was never approved - hell, it was barely even known - by the Army chiefs of staff or the Imperial Foreign Office in Tokyo. In effect the Kwantung Army had just re-written Japanese foreign policy for its Soviet borders, without consulting the Japanese government or even notifying their government that had done so.
Where this directly impacts our engagement is in the question of "what IS the frontier?" In the vicinity of Nomonhan the two rivals had very different opinions on the subject. The Japanese through their Manchukuo proxy stated that the valley of the Halha/Khalkhin-Gol River was the border. The Soviets through their Outer Mongolian proxy had, instead, a rather arbitrary line on the steppe some considerable distance east - you'll note from the map below that it ran southeast from a bend in the Halha to Nomonhan then turned south and from there (off the map) southeast to rejoin the river valley again.
Mind you, unlike the northeast where the Soviets and the Kwantung Army were eyeball-to-eyeball, nobody much was farkling about on this portion of the border. A handful of Manchukuoan border guards were station in bustling downtown Nomonhan while the Mongolian cavalry patrolled the heights west of the Halha. Everybody might have been tense, given all the wars and rumors of wars but there were no serious signs of trouble along the Manchukuo-Mongolian border in the spring of 1939. That's where the we stand on the morning when a company from the Mongolian 6th Cavalry Division waded across the salty waters of the Halha River heading east to find some better grazing for their horses.
The Engagements: As noted above; entire books have been written on this business. I will, instead of giving you tactical details, attempt to summarize what happened in each major part of the Nohomhan engagements and discuss what I believe it reveals about the IJA.
May: First Encounters
Alerted to the Mongolians by their attack on the little Nomonhan garrison a battalion of Manchukuoan cavalry mounted up and chased the
But a couple of days later the pesky Mongols returned, this time in larger numbers, and refused to budge when the Manchurian troopers waved their Arisaka rifles at them. At this point the border-crossers seemed more than a mere nuisance; LTG Komatsubara, 23rd ID commander, tasked his division recon (actually about a battalion-size force of mixed horse cavalry and armored cars) backed up by the 1st BN, 64th Infantry Regiment (-) to run off the damn Mongols.
Drea (1981) notes that "Japanese staff officers (at 23rd ID) were having difficulty even locating Nomonhan on their operational maps, so there seemed no reason to believe that additional force might be required..." So the 1,000-man-odd IJA force rucked up and headed off for the border. The Japanese force swung south down the Halha River valley in 14 MAY and, sure enough, the invaders scampered away.
And returned the next week.
By this time Komatsubara was beginning to get seriously peeved. He ordered his task force commander, COL Yamagata (commander of the 2nd/64th), to deal with these pesky interlopers once and for all. So on 28 MAY the Yamagata Force played military hardball; the colonel divided his forces, swinging the recon unit under LTC Azuma out west along the Halha to encircle the Soviets, size the Halha bridge and be the anvil while his infantry swung the hammer and smashed the enemy. Azuma was so confident he left his AT guns behind - why bother? They would just slow everyone down and who needed antitank guns to shoot horses?
Instead of a bagatelle of NKVD border guards and Mongols, however, the IJA ran into a buzzsaw composed of motorized riflemen, three battalions of "machinegun" troops as well as the Mongolian 6th Cav and elements of 11th Tank Brigade. The 149th Motorized Rifle was supported by its regimental artillery, as well.
Yamagata's 2/64th was stopped dead, and worse, Azuma's recon company was attacked from both sides and cut off. Surrounded, hammered by tanks and artillery, swarmed by infantry the Azuma force was destroyed between 28/29 MAY; out of the original some 200 troops more than 100 were killed outright and another 30-some wounded. On the 29th the survivors of Azuma's outfit and the 64th Infantry withdrew to the northeast.
What happened and why; assessing the IJA's performance
Among the IJA's most persistent problems were a combination of underestimating enemy capabilities and overestimating the value of Japanese "spirit". The failure of the May expedition showcased both. The Yamagata plan was a Custer-level clusterfuck, dividing the attack force where extremely poor "intelligence preparation of the battlefield" (IPB) left the attacker with little or no knowledge of the locations or capabilities of the enemy forces. The failure to anticipate the presence of Soviet armor or artillery - or the "enemy's most dangerous course of action" - left the IJA force without the capability either to defeat the enemy they found or to break off contact effectively once engaged. Only through sheer fighting ability (and Soviet weakness in numbers) were the surviving Japanese units able to avoid disaster and consolidate on the east bank of the Halha.
Drea (1981) says that 23rd ID and its higher were inclined to take their licking and move on "...because they believed that the Nomonhan desert was worth no more Japanese blood." But the Soviets were not so inclined. Reports from the border indicated troops showing up on the western side of the Halha and aggressive aerial activity. Continual small attacks on the Manchukuoan 8th Border Guard units continued. Finally in mid-June GEN Komatsubara requested, and received, Army permission to take the offensive and "restore the border".
After the disasters of May Kwantung Army decided to take no chances in June. 23rd ID was reinforced with two tank regiments, two artillery regiments, and four infantry battalions (26th IR plus 2/28th IR) from 7th ID. The date for the attack was set for late June or early July, and with the preponderance of force the Kwantung Army expected a sokusen sokketsu, a short and decisive battle that would end the fighting.
Instead the right wing was stopped by combined arms counterattacks and forced back across the Halha by 5 JUL. The left wing managed to push aside weak Soviet and Mongolian defenses for the first two days of fighting but were counterattacked and stopped cold short of the Halha on 5 JUL.
The IJA would never again manage to advance past the river.
The IJA left wing continued to try and batter past the Soviet defenses for the next five days or so, including a night infantry assault on 10 JUL. It got nowhere.
Logistical nightmares made everything difficult, especially rounds for all the artillery and anti-tank guns. Indeed, Japanese anti-tank weapons and ammunition supply were so deficient that many, perhaps most, of the Soviet tanks disabled and destroyed were by teams of infantrymen attacking with explosives and firebomb "Molitov cocktails".
These insanely heroic groups were fortunate in that the Soviet BT-5 and BT-7 tanks were poorly protected against fire; their gasoline engines tended to "brew up" much more quickly than the T-34 which succeeded them.
For the next two weeks the Soviet and IJA forces engaged along the east bank of the Halha stretching about 2 to 3-miles north of the Holsten River conducted a series of small scale, back-and-forth actions without result.
The troops of the IJA 23rd ID suffered miserably throughout this period. Except for the units along the Holsten River they often had to count on the infrequent rain for water; no water reached the Nomonhan front from the Japanese railheads and the waters of the Halha and its nearby wetlands were too saline to drink. Rations and ammunition came in painfully small quantities. The Soviet logistical train, on the other hand, included over 2500 trucks to haul supplies to the fighting front.
Finally GEN Komatsubara tried again.
On 23 JUL 23rd ID kicked off another offensive, this time a division-sized deliberate attack. With nasty memories of the devastating Soviet steel rain of early July Kwantung Army had sent 23rd ID substantial artillery support, bringing the IJA fire support up to four regiments including the 23rd ID divisional artillery (82 cannon, including 16 x 105mm howitzers, 16 x 150mm howitzers, 6 x 150mm guns, and 12 old Type 38 120mm howitzers). Under this fire the 64th and 72nd Infantry Regiments attacked the Soviet troops defending the Kawatama Bridge; largely the 5th Machinegun Brigade and 149th Rifle Regiment.
IJA artillery fired 15,000 rounds on 23 JUL, a total of 25,000 rounds over the next two days to try and shoot in the infantry; a heavier weight of metal than any fired by IJA artillery since the assaults on Port Arthur in 1905, a barrage that shot up half the Class V (ammunition) stocks in the Manchurian theatre, to try and blast open the Soviet defenses.
The assault made small gains but ground to a halt on 25 JUL. Not only had the "big barrage" failed to smash the Soviet infantry (who showed the facility for digging in deep that so frustrated the Germans two years later) but the Soviet counterbattery and defensive fires were heavier than all but the most intensive IJA concentrations.
Over 5,000 Japanese troops had been killed or wounded around the Halha since late May. When the artillery rounds began to run short GEN Komatsubara called off the attack.
What happened and why; assessing the IJA's performance
Again, the IJA operations staff made a succession of critical errors in both planning and execution. Again, IPB was pathetically faulty. The attackers had no real notion of the defenses they were attempting to breach. Little intelligence collection was done, even less planning was done to reflect that intelligence rather than the preconceptions of the Kwantung Army and 23rd ID highers.
Again the IJA assault of 2 JUL was planned as a double envelopment that depended on everything going perfectly. Instead the plan ran into the worst sort of trouble; a strong defense capable of defeating each wing in detail while the two attacking wings were unable to support each other.
Artillery support, while much improved, continued to suffer from a lack of tactical coordination. We haven't discussed this in much detail, but both combatants on the Halha noted that the biggest single difference between the two sides' artillery was that the Soviet gunners were much more nimble at adjusting onto new targets as well as working together with infantry and armor. Japanese fires, while often deadly accurate, tended to be sparse and limited by what their gunners could see around them. Japanese field guns also tended to have short ranges for their sizes; Soviet artillery was able to shoot counterbattery fires at ranges that IJA cannon of similar calibres could not reply.
One thing I should mention here that had a major impact on the fighting around the Halha and that was the terrain.
One of the first things I was taught as a soldier was; know the ground. And the second thing; know the ground. As simple a matter as a hill here and a valley there can make the difference between living and dying in combat, between winning and losing battles. And for all that there wasn't much in the Khalkhin-Gol region by way of terrain, what there was mattered, and that was that the west bank of the Halha was seriously higher than the right. See?
Possession of the high western bluffs allowed the Soviets to overwatch every Japanese movement on the low ground to the east, and the Japanese learned quickly that if you could be seen you would be shelled. The failure to appreciate the critical terrain and make a more determined attempt to seize it was another IJA planning error.
But of all the other fuck-ups perhaps the single poorest performance was put in by the Japanese tankers.
While the crewmen were individually brave and (in the case of the the Type 97 Chi-ha mediums) the tanks not thoroughly outclassed IJA tank doctrine had barely progressed beyond the waddle-alongside-the-infantry style of WW1. The IJA had just never bothered to develop tank-infantry-artillery combined arms warfare, but even the simpler coordination between infantry and armor was beyond the IJA tank regiments at Nomonhan. Add to that the light armor and low-velocity cannon of the Type 89, the most common IJA medium at Nomonhan, meant that if they could not fight in teams (and typically they could not; Japanese radios were poor even when mounted) they were vulnerable to the higher velocity cannon of the Soviet BT-5 and BT-7 cruiser tanks.
July - August: Stalemate
For about the next month the two sides fought a nasty WW1-style positional battle. Raids, minor assaults, artillery "hates", air-ground attacks. The Soviet artillery continued to pound the IJA positions with 2,000-odd rounds a day, a rate and volume of fire the Japanese could not hope to match or even aspire to. Japanese troops dug in and suffered the relentless fire, Soviet attacks, and probably most of all the misery of an undersupplied army in bad terrain.
"Apathy began to set in..."reports Drea (1981)
"...as all that the (Japanese) soldiers could think of was eating white rice again and washing it down with fresh, clean water. In fact, the battalion (2/28th Infantry) collected rainwater in oil drums or helmets but since their move to Noro Heights (about a week before) there had been only one day of rain. The troops, desperate for water, drank from stagnant, discolored pools and got amoebic dysentery. The two battalion doctors were overwhelmed...thirty or more bloody bowel movements a day might not be considered serious enough to be sent to a rear area field hospital...typhus cases were reported. The continual Soviet shelling left the troops no choice but to defecate in their foxholes...and Japanese officers forbade cooking fires for fear of betraying their positions. At times the soldiers were reduced to eating grass."Amazingly enough the Japanese infantrymen continued to fight hard and, when they could meet their Soviet enemies on equal terms, well. Drea (1981) describes a night engagement on 8 AUG where elements of No. 7 Company, 2/28th Infantry surprised a Soviet infantry platoon and artillery observers attempting to infiltrate 2/28 defensive lines. Machinegun and battalion gun fire hammered the surprised Soviets, whose approach had not been as stealthy as they had assumed. The fire lifted and shifted to allow a platoon of 7/2/28th to close assault the Soviet force with bayonets and swords.
Of the 51 Soviets in the unit only one was taken prisoner.
The grinding stalemate continued through August. Heat, blowing sand, thirst, vicious insects, and the constant drain of deaths and woundings - "normal wastage" in the callous language of WW1's Western Front - continued. Still the Japanese forces tried to fight the sorts of savage close combat fights they had trained for.
For this I recommend to you Drea (1981) pages 66-70; Night Attack II, a well-detailed report of the night raid conducted by No. 6 Company, 2/28th Infantry over 18/19 AUG. The action demonstrates the ability of the IJA infantry to conduct a precise and effective action over difficult terrain, at night, against a well-dug in enemy well-supported with artillery and automatic weapons.
The ironic thing about it was that the entire action was planned and executed because "...the enemy seemed to be preparing for some new action." (Drea, 1981). The raid was a "spoiling attack", designed to disrupt a planned enemy offensive. And for the eighty-some Soviet casualties in the 603rd Rifle Regiment it certainly did just that.
What the Japanese had not figured out is that the Soviet forces had been preparing a surprise for them beyond anything in their worst nightmares.
What happened and why; assessing the IJA's performance
The single biggest IJA shortcoming revealed by the static fighting of late July and early August was logistical. Japanese officers had never been taught to respect supply and transportation, inglorious province of those not "heroic" enough to actually fight. The realities of industrial war, however, took hold of the poor bastards in the Japanese positions on the Halha and literally choked the life out of them. While individual infantry units continued to fight well long after they should have fallen apart the lack of food, medical supplies, water, and ammunition meant that although valor and infantry skills could delay defeat they could not prevent it. Logistics had been the foundation of industrial warfare since the mid-19th Century and because of its sword-swaggering attitudes the IJAs foundation, as it proved at Nomonhan, was built on nothing firmer than the sands of the Halha valley.
The 2/28th was right; the Soviets were "up to something". Kwantung Army intelligence had been intercepting Soviet radio traffic all through the end of July hinting at a general offensive in the works. A more aggressive aerial reconnaissance and patrolling effort would likely have revealed that the Soviets were reinforcing the Khalkhin-Gol front. The IJA intel weenies neither believed the intercepts nor harassed their recon elements.
So when some 200 Soviet aircraft hammered the 23rd ID positions in the early morning of 20 AUG the IJA defenders had no idea what was coming at them.
And what was coming was pretty brutal; the First Front Army had massed two rifle divisions and three cavalry divisions supported by two mech brigades and seven artillery regiments in the first echelon; the second had another rifle division and five more mech or armor brigades. These units now hammered down on both ends of the IJA defenses, manned by little more than a single reinforced infantry division.
A large part of the Soviet surprise was Japanese logistical ignorance. Remember how the Japanese didn't believe you could supply an army in Manchuria more than a day or two's drive away from the railheads? Well, the nearest Soviet logpac assembly areas - the centers for supplying the Nomonhan fighting - were Borzya on the Trans-Siberian rail line and Ondorhaan on the "main road" east from the Mongolian capital of Ulan Bator. Both were some 600 kilometers or more from the Halha:
And yet the Soviet commander, Zhukov, managed to assemble his corps-size assault force and all their food, fuel, ammunition, ash, and trash in a month. The Soviet secret?
First Front Army had over 3,000 trucks by the end of August, including some 1,000 fuel tankers. The IJA literally couldn't imagine such a wealth of vehicles; the idea just didn't fit into the way their brains worked.
The intel pukes weren't the only screwups; weather and bad luck played a part. Drea (1981) notes that
"From August 12 on, there had been only one day of clear weather. On August 19, Japanese pilots did spot a concentration of Soviet vehicles on the Halha's west bank, but their reports were still being evaluated at higher headquarters on August 20. Potentially significant intelligence data gathered by lower echelon units...was not expeditiously reported...most likely because there was no battalion intelligence officer to evaluate or disseminate such information."The IJA infantry battalion had a very lean staff organization which did not include officers for either what a U.S. battalion would have identified as Operations (S-3) or Intelligence (S-2). The IJA infantry regiment had a staff officer identified as responsible for "Code and Intelligence", which sounds like a guy as much in charge of commo as intel. Only at division level did an IJA organization have a purely Intel staff position.
The Soviet attack, rather ironically, was designed as another double envelopment. As you can see from the map, the main force - 57th Rifle Division plus nearly three brigades of motorized infantry and armor with artillery support - would drive north through the 71st Infantry Regiment on the Japanese left.
Fixing attacks would keep the Japanese center pinned along the Halha, while a secondary attack would swing southeast to link up with the main effort around Nomonhan.
Well, it took the Soviets about ten days.
Here's Drea (1981):
"The Soviet armored attack, skilfully directed by General Georgi K. Zhukov of later World War II fame, rapidly turned the southern flank of the IJA's 23rd Division. To the north, progress was slower, but after vicious fighting on Fui Heights, where flame-throwing Soviet tanks finally dislodged and routed the Japanese defenders, the "Red juggernaut" rolled up the Japanese right flank. At the village of Nomonhan, the Soviet armored columns met and sealed off the 23rd Division."The 23rd ID and its attachments were virtually destroyed. The U.S. Army in WW2 considered a unit "destroyed' - that is, completely combat-ineffective, so degraded by deaths, wounds, and terror that it would need to be disbanded or completely rebuilt - if it suffered more than 30% casualties.
The IJA units engaged at Nomonhan suffered 73% loss. Some even more; 2/28 Infantry lost 86% of its original compliment. The 71st Infantry Regiment lost over 90% of the men who had marched out towards Nomonhan in May.
Had Zhukov not halted his forces at the end of August there is no military reason why he could not have brushed aside the shattered remnants of 23rd ID and pushed further into Manchukuo; the Kwantung Army had no combat-effective units within a day's road-march at the closest, and it was purely due to Soviet contentment with its August gains that more of the Mongolian frontier was not lost.
Despite this, the Kwantung Army reinforced the Nomonhan sector. Fighting continued in the region, and the Army was planning another offensive when the Imperial Foreign Office negotiated a peace deal with the Soviets in mid-September. The war in western Europe had broken out and the Soviets had signed a pact with Japan's ally Germany. Suddenly getting valuable soldiers killed in border battles on the Manchurian steppes seemed pretty pointless.
The defeat at Nomonhan - in fact, the entire engagement at Nomonhan - was a nasty wake-up call for the Imperial Army and the Japanese government. The visible failure of the Hiranuma government to both anticipate the Nazi-Soviet Pact or control its Manchurian army forced the resignation of the Cabinet.
The effect of the defeat on the Kwantung Army command and staff was mixed. The IJA General Staff moved against the Kanto-gun; its commander, Ueda and many of his senior staff officers were either replaced or forced into retirement. But the failures of Nomonhan and particularly the poorly-thought-out aggression that caused them, were not really punished, and the issues raised by the insubordination and incompetence of the Kwantung Army at Nomohan were not made public.
Many of the architects of Nomonhan remained influential in the IJA and through the Army on Japanese foreign policy: "It is generally acknowledged by those who held contemporaneous High Command posts that the officers responsible for the Nomonhan debacle became strong advocates for launching the Pacific War" (Coox, 1985).
The Soviet Union held the ground they had taken in 1939 until the final days of August 1945, when the T-34's of the 6th Guards Tank Army ground the bones of dead men into the sandhills above the Halha River on their way to destroy the Kwantung Army and Imperial Japan.
The Outcome: Decisive Soviet grand tactical victory; the defeat at Nomonhan had some degree of impact on Japanese grand strategy in World War II.
The Impact: The most widely held opinion of the larger effects of Nomonhan was the defenestration of the "Northern" faction - the Army officers heavily represented in the Kwantung Army - from Imperial foreign policy and the ascendency of the 南進論, Nanshinron or "Strike South" faction that was largely centered around the Imperial Navy. In fact, Coox summed up this point precisely in a 1992 lecture:
"I feel that if at Nomonhan their affairs would have gone successfully, they would have initiated an offense against us." (Coox here is quoting from a postwar interview with Zhukov) "Their far-reaching plans included the capturing of the eastern part of Mongolia, cutting the Siberian main railway line." The Tokyo war trial jurists agreed and said, in memorable prose that as the door of opportunity closed in the north, the southern gates began to open for Japan." (Otterstedt, 2000)I will be contrary here.
While I agree that the rough handling that the 23rd ID received from the Soviets at Nomonhan surely had the effect of cooling some of the hot heads that dreamed of an Imperial Japan stretching west to Lake Baikal, the economic and logistical realities of 1941 were that the Imperial armed forces needed reliable access to petroleum most of all. At the time no significant reserves were known in the Soviet Far East.
Militarily, politically, and economically the cost and difficulty of expanding the Imperial borders north and west far exceeded any potential gains to be had there. Siberia was then what it is now; a vast emptiness, difficult to traverse and costly to develop.
Add to that the fact that the "Southern" strategy was already part of the official policy of Imperial Japan and had been since 1936.
No question that the defeat at Khalkhin-Gol helped the Southerners. But I'm not really convinced that the "Go North, Young Nipponese!" notion was manifest destiny. There were just too many other factors against it.
I believe that the real military and political impact of Khalkhin-Gol was just what Zhukov said it was; to scare the IJA enough to prevent an opportunist move against the Soviet Union in the summer of 1941.
Simultaneous attacks from west and east would have been a hell of a nightmare for Stalin's STAVKA. At the very least, the counterfactual is that the units that arrived before Moscow in the autumn and early winter of 1941 would have been pinned down in Siberia. Whether that prevents the loss of Moscow...well, that's hard to say. But it sure narrows the margin for the Soviets.
Indeed, perhaps the most significant impact of Nomonhan was not on Japan and the IJA at all, but to train one of the craftsmen of the defeat of Nazi Germany in his craft. Then-GEN Zhukov took west with him the experience of assembling and employing a combined-arms force that he would continue to refine over the next six years. No question but that he was a blunt instrument. The Soviet Army was a hammer, not a rapier. But what worked on the Halha would work on the Don, and the Dnepr, and eventually on the Elbe.
What happened and why; final assessment of the Imperial Japanese Army of WW2
The August battles pretty much sum up my opinion of the Imperial Japanese Army of World War II; an anachronism of a force, a feudal warlord's fighting tail tricked out with a handful or tanks and aircraft but still fundamentally unable to fight a modern war.
Unwilling to change, unable to adapt, this costume-movie army muddled along until confronted with a decently-led, intelligently-organized armed force that knows how to fight industrial war.
At that point the IJA got handed its ass.
This happened over and over again between 1939 and 1945. It happened in the islands, it happened in New Guinea, it happened in India and Burma, and it even happened right here, again, in 1945 when the Soviets came storming back across the steppes.
In my opinion the IJA leaders were charlatans, tricksters who deluded their entire nation into thinking that somehow "spirit" and "warrior virtue" could defeat lead, fire, and steel. As culpable as any warlords in history whose hubris and folly led their peoples to their own destruction, the disaster they created and that destroyed them at least did Japan the favor of striking off its love of military posturing and the sort of chest-thumping "warriors" that have proved so disastrous throughout human history by being exposed for the frauds they were.
Where to start?
The IJA doctrine of light infantry aggression was as fatally flawed in 1939 as the French attaque à outrance was in 1914. While it might have promised a tactical victory or three in the larger picture it meant futile deaths for tens of thousands.
Technically the IJA never learned the lessons of Nomonhan - or any of the other engagements we've discussed. It never developed a fully-integrated air arm. It's armored force never rose above comical. Its artillery - while, like its infantry, well-trained on the tactical level - never mastered the complex fire control and fire support tasks that were routine for all the other combatant armies by 1945.
Above all, it remained logistically shambolic until the end. Japanese soldiers, sailors, and airmen routinely suffered and died lacking supplies that Allied, and even German and Soviet troops took for granted.
The Kwantung Army in 1939 was Japan's overall WW2 problem writ small.
Instead of being an instrument of its nation the Kantō-gun acted as if it were the nation, and made decisions that affected that nation for reasons no better than solving minor military problems. The officers of the Japanese Army in Manchuria - with no better understanding, knowledge, or wisdom about the geopolitical implications of dragging their nation into war with the Soviet Union over a piece of worthless steppe - chose to put their nation on a course it had neither planned nor prepared for.
As individuals the officers and commanders of the IJA of the late Imperial period were no different than any other human. Some were brilliant, some were idiots. Some were sweethearts, some were assholes.
But as a group; as an Army, they were a damn deadly dangerous group of people. They ran their country into a war that they could not win, that they should have known - that the brighter of them did know - they could not win. And in the losing of it they killed uncountable millions, many of them their own people. Seldom in the history of human conflict has a group of ordinary men been so viciously culpable for so many terrible crimes. All with the ideal that they were living the most virtuous of lives; lives full of "honor", "duty", "sacrifice", and "loyalty".
And on the altar of those virtues they sacrificed a world of lives.
There's just nothing much "human interest" about Khalkhin-Gol; its pretty much all blood and misery. Except this - we don't know today exactly where Nomonhan was.
The modern border between Manchuria and Mongolia is a restricted zone; the borderline is drawn close to the 1939 Soviet claims, very near First Front Army stop line in August.
The area is as lightly lived-in today as it was in 1939 or in 939, I suspect. (Stuart Goldman, whose work on this battle I have recommended, has a nice little blog post about visiting the site of the fighting. He mentions the museum in the little town of Khalkgol which seems to be something something of a shithole).
At some point the Soviets (or their Mongolian pals) threw up some one of those usual ginormously hideous concrete-and-steel war memorials that the Soviets seemed to have enjoyed so much over on the west side of the Halha. Pictures really don't seem to do justice to the butt-ugliness of the thing, which appears from a distance like a monstrous wingnut on a stick:
Birds, though? That seems to be another matter; the rare Band-bellied Crake seems to find a happier summer in the valley of the Halha than the troopers of the 23rd Infantry Division. From what I can glean from the Internet the eastern end of Mongolia is nice for birdwatching and grazing and not much more.
In general the plains around the Halha River seem as little worth fighting over as they were seventy-four years ago. Life seems very precarious there, and the little towns always seem to be that close to simply disappearing when the residents just lose interest one day and wander off. The habits of nomadism die hard.
That is what seems to have happened to Nomonhan at some point between 1945 and now.
So far as I can tell there is nothing visible at the 1939 site of Nomonhan today. I can find no pictures or snapshots taken by visitors, even of ruins. No village is shown on the maps. Satellite pictures do show a sort of wrinkle or small irregularity in the otherwise-featureless plain, but of roads, buildings, of any of the works of man; of the small town that once stood by so much fighting and dying...nothing remains now but the grass, the river, the wind, and the sky.