Warlords, regionalism, and national unity Regional devolution

Wu Pei-fu, the 'Jade Marshal' or 'Scholar Warlord', was the dominant commander in the Chihli Clique throughout the early and mid-1920s. Wu was probably the best field general during the Warlord Period, but politically he was relatively naive, and was betrayed on several occasions.

After the death of Yuan Shikai in 1916 unified military rule from Peking gave way to disseminated military rule. Within a few months the country was divided into a great number of what were known then as satraps, none of them stable or lasting, all based on regional ties, all dominated by warlords. China had become, as Sun Yat-sen had predicted it would, a sheet of shifting sand. Though there continued to be national governments in Peking they wielded very little power, and came and went with bewildering frequency.

China is a vast and diverse country. The regional diversity is expressed in dialects, often mutually unintelligible, in cuisine, in traditions and customs – and in identity. Before there was an empire there were many independent states, whose names survive in the alternative names of provinces (QiLu/Shandong, ShuBa/Sichuan, Yue/Guangdong).

In the many periods of disunity since the founding of the first state in the third century BC, regional power holding always emerged to fill the void left by a collapse at the center. The process of devolution and fragmentation was one that China knew well. The most famous period of disunity came after the end of the Han dynasty, when China was divided into three. The Romance of the Three Kingdoms (Sanguo yanyi), an immensely popular novel written more than 1,000 years after the events it described (and almost certainly apocryphal), told of the anguish of division and civil war through a string of stories of courage, treachery, and intrigue. The stories were known to every Chinese, whether educated or not; they appeared as opera plots, as oral stories, and in cartoons. Disunity was as inevitable as unity, said the Three Kingdoms stories. Some people behaved badly in times of troubles, others came into their own – but the evil men often won; the most evil of all, Cao Cao, triumphed over the greatest strategist, Zhuge Liang, a man of brilliance and humanity.

It may seem a stretch to use a novel as a guide to understanding reactions to disunity and uncertainty – but the mentality portrayed in The Romance of the Three Kingdoms had a formative influence on young men of the early Republic, men such as Mao Zedong, who had all read the novel as boys. Theirs was a Three Kingdoms reaction to disunity: think things through carefully, devise stratagems, and know that the solutions will require force as well as intelligence. The answer was to combine Zhuge Liang’s brilliance with Cao Cao’s ruthlessness.

Warlords and their armies

The rise of regionalism and regional identities had been encouraged by the disappearance of universal examinations in 1905, and by the loss of the law of avoidance. After 1916 the center’s ability to make appointments at the provincial level disappeared, and regional rulers came to power, often soldiers, who called themselves military administrators (dujun); other people called them warlords.

These men saw disunity as opportunity for their particular regions. The negative reactions to warlordism in the civilian world reflected the fear of chaos, of the country falling apart – the fear that had haunted China’s rulers since the beginning of the empire. This fear lived in the metropolitan world of the emperors and the bureaucrats. It was not shared by warlords, men who focused on one region only, nor by many of the people whose lives they controlled, whose horizons did not extend beyond a region and its culture.

In the civilian elite’s stereotype, a warlord was a deceitful, devious, illiterate man, sunk in backward patterns of behaviour, uncouth and filthy. Zhang Zongchang, the “Dog-Meat General,” who ruled Shandong for many years, fitted the stereotype. He was uneducated, a bandit by origin, loud-mouthed, cruel. His proudest “possession” was his large harem, in which there were women from China, Japan, Russia, and western Europe. He lived by violence, he lost his power by violence, and he died violently (after he had lost power), shot at the station of his former capital, Jinan.

Few warlords were as awful as Zhang. Some were progressive figures, complex men who blended self-interest with a genuine interest in the future of China. The most famous of this type was Feng Yuxiang, a mass of contradictions, blunt and devious, a personal power seeker and devout nationalist.

Other warlords were local strongmen who looked after their own regions, and in some cases gave them the most secure and stable periods of rule they were to know in many decades. In Shanxi, Yan Xishan, who ruled the province for more than three decades, is remembered as a model ruler; in Guangdong, Chen Jitang, who controlled the province for most of the 1930s, is considered a local hero; in Guangxi, the rulers of the province from 1925 to 1949, the “Guangxi Clique,” are revered for their martial spirit, which gave the province the name of “China’s Sparta.”

Tuzi buchi wobian cao. “The rabbit doesn’t eat the grass beside its nest.” Source: traditional

The better warlords understood the old proverb about a rabbit not eating the grass beside its own burrow, and they tended to show concern for the people of the region they controlled. They provided stable government, which, even though it came with tax swindling and rampant corruption, was preferable to chaos or anarchy. Tax income stayed in the region – except for the amounts that warlords salted away in Tianjin, Shanghai, or Hong Kong (cities under foreign control) – for the time when their rule came to an end.

The men referred to as “petty warlords” did the most damage to Chinese society. They really were bandits, uncouth and crude. They exploited and vandalized the regions they controlled. Their rule was often short. When they were overthrown by other warlords they went back to banditry or joined local militias.

The number of men under arms expanded dramatically in the early Republic. By the early 1920s there were at least 1.5 million soldiers, and an equally large number of armed men not serving in formal military units – irregulars, militiamen, bodyguards, and bandits. There was a twoway traffic between the organized and the informal armed worlds.

Warlordism had a strongly inhibiting effect on one aspect of Chinese society where there might otherwise have been change. The emancipation of women, which had just begun in China’s cities, was impossible in areas under indifferent or bad warlord control. Girls had to be protected by their families from the unthinkable – rape – and so many of them lived cloistered lives at home.

The warlord system provided immense numbers of jobs – either directly, as soldiers, or indirectly, in the manufacturing and service industries that catered to the military. The continuing growth of China’s population facilitated the expansion of the military. As the population grew, employment opportunities did not. Most of the jobs in the new factories were for young women. There were more and more young men in the rural areas for whom there was no work. A few could emigrate – to Manchuria, Southeast Asia or North America – but the closed nature of migration flows limited this solution to a few regions of China, all of them coastal.

Young men from regions with no established migration chains had only a few opportunities for off-farm employment – peddling, moving to the city, or going into the military.

Warlord finances

The foreign banks, like the concessions, contribute largely to the amenity of Chinese civil war and political strife. Once loot is turned into money and deposited with them by the looter, it is sacred and beyond public recovery. Cases have been known in which generals, far from expecting interest on their deposits, have been eager to pay the banks a small percentage for the privilege of being allowed thus to cache their gain. At a town up the Yangtze [Yangzi], a Chinese military commander visited the American- Oriental Bank and said that he wished to deposit with them, instead of in his own headquarters, what he politely called his records, and left thirty large trunks with the bank. He was presently defeated, and the bank manager was a little disturbed as to what he should do if the incoming conqueror were to demand that these records be handed over. But the in-coming conqueror felt equally insecure, and was more concerned to get his own records safely in to the bank than to obtain those of his enemy. Another huge batch of trunks was brought in, and the bank manager, much relieved, had both sets of trunks piled one beside the other.

Arthur Ransome, The Chinese Puzzle (London: Allen and Unwin, 1927), pp. 123–4.

The Generalissimo and Operation Ichigo I

Dai Li reported that the Japanese High Command was deploying 150,000 troops south from Manchuria and North China and rebuilding a bridge over the Yellow River. The troops were presumably moving to join a major offensive in central and east China, probably sometime in the coming months. This operation, called Ichigo, would eventually involve half a million well-armed frontline Japanese soldiers, the largest number of troops ever used in a campaign in Japanese history.

The reason for Ichigo was the destruction of the Japanese merchant fleet down to 77 percent of its prewar level, which was seriously restricting Japan’s importation of raw materials from Southeast Asia. Strategists in Tokyo hoped to solve the crisis by creating a continental corridor linking Japanese- occupied territories from Korea to Manchuria, through North, Central, and South China down to Indochina, and then through Thailand and Malaya to Singapore. Such a corridor would require uninterrupted control of the railways between Hanoi and Peking and to Dalian in Manchuria. The plan would in effect expand Japanese-occupied China west and southwest, thereby securing the empire’s rear and allowing raw materials (such as oil, minerals, and possibly food) as well as troops to be transported from China and Southeast Asia to the port of Pusan in Korea and then across the narrow Tsushima Straits to Japan. The sizable territorial gains that the plan required would also result in the destruction of Chennault’s new airfields. On April 6, Chiang received intelligence reporting that Japan’s goal was not just to destroy the easternmost airfields but to open up “the Big [Dalian to Hanoi] Asian Railway.”  

Consequently, while the Germans were retreating on every front in Europe and the Japanese were falling back in the Pacific, the Imperial Army was planning its biggest offensive of the war in China, bigger than anything MacArthur had faced. In Chungking as in the other Allied capitals there was increasing talk about the possible collapse of Hitler’s regime sometime during 1944. Chiang assumed that at Tehran Stalin had agreed to enter the war against Japan after Germany’s surrender, although the Allies had not informed him of this. Once in control of Manchuria and possibly Peking and Tianjin, Chiang thought, the Soviets would “talk tough and ask for a regional [CCP-dominated] government and regional autonomy.” But there was nothing at the moment he could do about it. In Yan’an, arguments for serious CCP cooperation with the Kuomintang lost what little force they had retained among the Chinese Communists. In Politburo meetings in December, Zhou again admitted his “capitulationist” (proKMT) views.

Chennault warned Chiang that the expected Japanese offensive in central China was imminent and would eventually extend into Hunan and Guangxi. The Y Force, he thought, would be needed to resist the enormous Japanese attack. Stilwell’s recurrently inaccurate intelligence office in Chungking issued an “emphatic dissent” to the prediction of a massive Japanese offensive. Stilwell informed Chennault that “the current crisis in [Imphal] India” had priority and he instructed the air officer not to send the Generalissimo “a gloomy estimate of the military situation” in China. Only when Chennault’s pilots spotted 239 Japanese troop trains steaming south and west did Stilwell’s Chungking headquarters finally conclude that a major offensive was indeed imminent. On April 17, the Japanese 37th Division, with hundreds of tanks and armored personnel carriers, crossed the Yellow River on repaired bridges and rolled across the flat Henan wheat fields. Their goal was to sweep away the Chinese forces between the Yellow and Yangtze rivers, clearing the rail lines between Peking and Wuhan. The motley forces of General Tang Enbo’s First War Zone fell back in disarray, but the 28th and 31st Army Groups held on at the key city of Luoyang. For good or bad, the Generalissimo was on the phone personally directing the defense. He also appealed to Stilwell, his chief of staff, for five hundred tons of gasoline from the B-29 stores to fuel Chennault’s fighters supporting the Luoyang defenders, but Stilwell, believing that the Generalissimo had brought the situation on himself, declined.

The Chinese divisions in Henan fought as usual with no armor or motor pool, a few old cannon, and for every three soldiers two old, mostly Chinese-made rifles. The only advantage the Chinese fighters inside China enjoyed was tactical air support from the ninety or so operational planes of Chennault’s 14th Air Force that had been assigned to Chinese army operations inside China. The Luoyang defenders resisted for fifteen days, losing 21,000 Chinese soldiers and officers before Chiang’s withdrawal order came.

During the night of May 11, as the collapse in Henan proceeded, the 72,000-man Y Force waded across the Salween River in western Yunnan province to link up with Stilwell. According to the American liaison officers, the Chinese expedition, with its up-to-date equipment and extensive tactical air support, fought well and bravely. Scattered downpours soon merged into the near constant torrential rain of the monsoon. But U.S. planes dropped supplies and ammunition and the Chinese slogged on. At the same time, Stilwell’s Chinese divisions and Merrill’s Marauders, all battling mud and floods, reached the outskirts of Myitkyina.

Despite the rains, Stilwell was hopeful that his five Chinese divisions could soon link up with General Wei’s twelve “Y” divisions to the east. But in mid-June, Wei’s effort to seize Long Ling, the center of the Japanese line on the Salween front, collapsed when a counterattack by only 1,500 Japanese troops drove back 10,000 Chinese fighters. Chiang was extremely angry and demanded that Wei “spare no effort” to renew the attack and capture Long Ling. Chiang ordered two more Chinese armies from North China to join Wei’s force in Burma. These armies were desperately needed in the Ninth and Fourth war zones, which were awaiting the next phase of the grand Japanese offensive. By diverting them south, Chiang again demonstrated his commitment to an Allied victory in Burma.

American pilots continued to report that elite Japanese units from Manchuria were still pouring down the newly occupied rail line into Wuhan. Chennault again appealed to Stilwell to use his emergency authority to divert supplies, transport capacity, and combat elements of the B-29 command to meet the emergency in East China. Stilwell responded that “until the emergency is unmistakable, the decision will have to wait.” Two days later, the Japanese Sixth Army surged out of the Wuhan bulge into Hunan while smaller forces advanced from Canton and Vietnam. This second stage of Ichigo threatened to seize most of Hunan, which was the main rice bowl for Free China, as well as the bases of the 14th Air Force in Hunan, Jiangxi, and Guangxi. The pilots of Chennault’s 23rd Group flew three to four missions a day through and under the weather. They wreaked havoc among the Japanese, but their losses were horrendous—equaling or surpassing that of U.S. bomber groups in Europe. Nearly half the pilots of this group of three squadrons were killed or taken prisoner that summer, among them three squadron commanders.

The Generalissimo and Operation Ichigo II

According to the Chinese War Ministry, since the fall of Rangoon, from May 1942 through September 1944, 98 percent of U.S. military aid over the Hump had gone to the 14th Air Force—and, the ministry could have added, to the B-29 operation and the upkeep of the large and increasing number of U.S. military personnel in China. The United States had provided the two million or so men who were in the Chinese army but not in the X and Y forces a total of 351 machine guns, 96 mountain cannon, 618 anti-tank rifles, 28 anti-tank guns, and 50 million rounds of rifle ammunition. Of these items, only 60 cannon, 50 anti-tank rifles, and 30 million rounds of ammunition were provided before the June 1944 battle of Changsha; the rest afterward. Further, the War Department had decreed that the new Z Force of thirty Chinese divisions to be retrained and armed by the United States (as the second group of the ninety total divisions that Roosevelt had promised to support) would receive only 10 percent of the total Lend-Lease allotment for China. The American team assigned to this project calculated that if divided between thirty new divisions, these supplies of arms and ammunition would in each case amount to “practically zero.

In early June, Stilwell was back in Chungking and discussed the military crisis with Chiang and later in Kunming with Chennault. Both repeated their earlier requests for a B-29 raid on Japanese depots at Wuhan. Stilwell promised to forward them to Washington, but when the department replied in the negative, he sent back a brief acknowledgment to Washington: “Instructions understood and exactly what I had hoped for. Pressure from the Generalissimo compelled me to send the request.” After his meeting with Chennault, Stilwell also stuffed into his pocket and then promptly forgot Chennault’s request that the two hundred fighters of the 14th Air Force assigned to protect the B-29s be deployed against the looming offensive in East China.

According to Stilwell, on June 5 in Chungking Chiang told him “the situation in East China” (the expected offensive into Hunan) was to be solved by air attacks and asked him to “suspend shipment of arms and ammunition over the Hump” in order to concentrate on shipment of fuel, parts, and ordnance to the 14th Air Force. Stilwell promised Chiang that he would assure that the 14th Air Force received 10,000 tons a month in supplies but he also seemed to interpret Chiang’s instructions as an order not to use U.S. transport to deliver from any location arms and ammunition to any of Chiang’s armies resisting the Japanese in East China. It is doubtful that Chiang was as categorical as Stilwell recounts in suggesting air power alone would “solve the situation,” and Stilwell himself does not quote him as saying no American or Chinese arms and ammunition should be sent to Chinese forces in East China. In his diary, Chiang simply reports that he and Stilwell “discussed fuel supply and weapons distribution” and Stilwell “politely promised to do as I wished. His attitude was the same.” Actually only a “trickle” of American arms had been going to Chinese forces not related to the Burma campaign, but Chiang believed he had a compelling reason to control and even to stop any future American arms from going to his commanders in East China. At that moment Xue Yue, commander of the Ninth War Zone in Hunan, and General Zhang Fakui, commander of the Fourth War Zone in Guangdong and Guangxi, had come under increased suspicion of disloyalty— reports about which Stilwell’s headquarters believed were true.

By apparently directing at least that shipments over the Hump be entirely devoted to supplies for the 14th Air Force and failing to insist that the Americans send arms and ammunition to the Chinese forces in Hunan—especially the threatened cities of Changsha and Hengyang—Chiang showed that concern over disloyalty among his generals was more important to him than the successful defense of these cities. As vividly expressed in his diaries, he had enormous political and military interest in defeating the Japanese in these looming battles—he knew the outcome would have a great effect on American and domestic support for him. Thus he would take strong measures to try to defeat the Japanese in Hunan, but, fearing supplies would fall into the hands of Xue Yue, he apparently would not ask the 14th Air Force to drop even Chinese-made ammunition and weapons to the defenders, nor did he discuss the air-supply issue with Chennault.

Chennault was outraged with Chiang for not aiding Xue, but he did not raise the matter with him, apparently assuming the effort would be futile. But if Stilwell, who never hesitated to oppose Chiang’s decisions, had had any question about the wisdom of Chiang’s orders as he interpreted them, he could have sought clarification, or urged him to reconsider. This would have been useful at least for the record. But Stilwell, like Chiang, had his own potent political motives in these decisions. As will be seen, over the next two months he clearly preferred that Chiang suffer a serious defeat in Changsha and elsewhere in East China in order to enhance his own prospects of taking over command of the entire Chinese Army. Chiang and Stilwell would both bear some responsibility for the coming fall of Changsha and Hengyang. Perhaps because of this, in his diary Chiang would not later blame Stilwell for these specific defeats.

Xue Yue’s headquarters was some distance away from Changsha and as the Japanese tightened their noose around the city, Chiang concentrated on helping and directing General Zhang Deneng, whose Fourth Army comprised the defenders. Chiang continued to press for increased logistical support of the 14th Air Force’s attacks on the Japanese attackers and lines of supply and he ordered six armies from four war zones to deploy immediately to Hunan. When it became clear that these units would arrive too late and Changsha would fall, General Zhang, on June 26, without orders evacuated the city with 4,000 of his men and trucks reputedly filled with his personal effects. Chiang had him shot, although he was one of his favorite generals. The Japanese occupied what remained of the wasteland of the city that had avoided its fate for so long. Most of the population had long fled, including the American and Chinese staff of the Xiangya Hospital. The loss of Changsha exposed Hengyang and Guilin to the south and to the northwest, Chungking. Stilwell’s headquarters in the provisional capital began planning for evacuation. Roosevelt quickly dispatched Vice President Henry Wallace to China to “calm” Chiang Kai-shek and promote cooperation between the Kuomintang and the Communists.

At this time came the almost simultaneous news of the attempt on the life of Hitler and the resignation of the Tojo cabinet in Japan. Tokyo’s surrender, Chiang thought, would “not be long” in coming. He believed that if he could avoid a break with the Americans by temporarily giving Stilwell command with some controls, the crisis with America would pass and after the war the aggressiveness of the Russian and Chinese Communists would eventually push the United States into opposing them.

But immediately Chiang was to start down the path toward another defeat that would strengthen Stilwell’s hand and perhaps encourage his warlord opponents. Having lost Changsha, victory in Hengyang, a hundred miles due south of Changsha, would, Chiang felt, be “Heaven’s blessings” and would “dissolve the diplomatic crisis and deliver us to safety.” The idea that the Nationalist Army was about to collapse would be eliminated by such a victory. Again, Chiang did not want to support Xue Yue, the war zone commander involved—so instead, as at Changsha, he directly conducted the defense of Hengyang with loyal forces in the city and elsewhere, which added to the chaos. But this time he had a general, Fang Xianjue, who was willing to fight.

In his mission to save the city, Chiang relied heavily on Chennault’s close air support. The tactic worked well at first. For a period in early July, fighters and bombers of the 14th Air Force so disrupted Japanese supply lines that the attack on the walled city ground to a halt. Then for one week, lack of fuel grounded the American pilots. In accordance with Stilwell’s wishes, the War Department still refused to allow Chennault to use aviation fuel from the B-29 depot in Chengdu.

Meanwhile, on the ground, Fang repulsed three waves of Japanese assaults on Hengyang, reportedly killing 7,602 Japanese soldiers. But he lost most of his regulars—19,380 men—and by mid-July mostly auxiliaries and service troops were in the front lines. Resupply became a critical issue. Chiang again did not fly or parachute in supplies, and Chennault, without telling Chiang, again requested that Stilwell authorize a single air drop of ammunition (presumably Chinese manufactured), in this case at Hengyang. Stilwell refused, saying he was concerned this would set a precedent for further demands that could not be met. On July 20, Stilwell’s new chief of staff in Chungking, General Tom Hearn, suggested authorizing Chennault to drop a “token” of two hundred tons of ammunition into the city, but Stilwell in his reply recalled with satisfaction what he said was Chennault’s old promise “to beat the Japs with air alone.” In fact, neither Chennault nor Chiang had ever claimed they could beat the Japanese without the Chinese Army playing a key defensive role. But Stilwell went on to tell Hearn that if Chennault “now realizes he cannot do [this], he should inform the Gissimo, who can then make any proposition he sees fit.” This part of the message was inadvertently not passed to Chennault by Stilwell’s staff. But in a segment of the same message that did make it to Chennault, Stilwell, turning aside a request for supplies for Hengyang from Bai Chongxi and clearly thinking of his own imminent takeover of the Chinese Army, grumbled, “I do not see how we can move until a certain big decision is made.” He added sarcastically: “You can tell the Chinese we are doing our best to carry out the plan the Gissimo insisted on.”

In reply to Stilwell, General Hearn indicated that he was aware of “the pending big decision” (Stilwell’s new appointment), but he recommended that in the meanwhile “drastic action” be taken “immediately” to aid the Chinese in southeast China. Chennault had offered to convert one thousand tons of his Hump quota to arms and ammunition for the besieged forces with or without Chiang Kai-shek’s approval. Stilwell turned down these proposals as well as additional pleas for supplies from Xue Yue and Bai. In yet another clear reference to U.S. pressure on Chiang for his command of the Chinese Army, Stilwell told Hearn: “The time for half-way measures has passed. Any more free gifts such as this will surely delay the major decision and play into the hands of the gang. The cards have been put on the table and the answer has not been given. Until it is given let them stew.” The meaning was apparent. Hearn informed Chennault that Stilwell agreed that in order to restore the situation in the east, a “real operation” was required, but “he is working on a proposition which might give this spot a real face-lifting and is loath to commit himself to any definite line of action right now. Consequently, we must hold off in making any proffers of help to the ground troops until things precipitate a bit more.”

Bai Chongxi once again urged Chiang to pull back from a besieged city, in this case Hengyang, and concentrate on attacking the enemy’s lines of communications. But Chiang still thought it was necessary to show the Chinese people as well as the Americans that the Chinese military was fighting the enemy head on and could win. Still, as at Changsha, the imperative to win at Hengyang was not strong enough to compel Chiang to insist that ammunition and other supplies be flown to the defenders. Chiang told General Fang inside Hengyang to keep fighting and ordered nearby armies to try to relieve the city. Stilwell’s U.S. Army observers reported that the 62nd, 69th, and 37th armies did in fact take heavy casualties trying to break through the Japanese forces surrounding the city, while three other armies also suffered severe losses while attacking Japanese supply lines. Henyang was finally secured by the Chinese on August 1. After the war, the government built a memorial of 5,000 skulls collected from unburied Chinese soldiers. The Chinese armies were not winning, but no one could say they were not fighting.