7

Towards an all-out civil war in China

Short-term solutions, long-term success and the veto players

The Manchurian truce of June 1946 displaced the civil war to China proper, together with the vicious cycle of attack and counter-attack. Marshall therefore needed to deal with the two revenge-seeking parties at the negotiating table. He rejected Yu’s stingy offer of extending the ceasefire for just one day, but Chiang had prepared to prolong the truce anyway. On 21 June, the GMD unilaterally announced an eight-day extension of the Manchurian truce to 30 June. As the extension was not the result of an agreement with the CCP, Chiang’s statement turned the partial truce into a one-way ceasefire. The extension gave Marshall a little more than a week to negotiate three complex agreements with the two warring parties. The subjects at issue were permanent ceasefire in Manchuria, implementation of the army reorganisation agreement and restoration of transportation.¹ This chapter is about Marshall and the members of Chinese minor parties’ last-ditch peace bid.

Short-term success versus long-term win

To exacerbate Marshall’s agony, the two parties almost simultaneously showed strong disinclination to resolve the disputes one by one or stage by stage. Zhou found it totally unacceptable that while they were being pushed to make concessions on military issues, the Nationalists made no commitment at all on political reforms.² Chiang demanded that the three agreements together must be signed at the same time and that there would be no agreement until all three agreements were settled.³ While Zhou’s approach can be interpreted as the tactic of evaluating the problem in its entirety, Chiang’s prerequisite could not be treated as a common holistic approach.⁴ This was because Chiang’s approach not only issued an all-or-nothing ultimatum to the CCP but also reduced the effectiveness of the negotiation: the first two agreements reached would never be done deals if there was no agreement on the third. The immediate impact of Chiang’s harsh proviso was that it made all involved parties lose a sense of purpose to argue their case. ‘I believe there should be no objection … after all, the paper is not being signed right now’, Zhou commented.⁵

Paradoxically, the negotiation went relatively smoothly. Marshall was able to call the first Committee of Three meeting since he returned from the United States, on 22 June. The two warring sides soon reached two vital but unsigned agreements on a complete termination of hostilities in Manchuria and the reopening of transportation trunk lines in China proper. An unsigned settlement was also passed on granting the American representatives in the field teams and Executive Headquarters with final decision power in case of disagreement.⁶

The final hurdle of the negotiation was far more difficult. Marshall needed to find a mutually agreeable amendment for agreement implementation of the army reorganisation deal of February for the two belligerent parties. The disagreement between the two parties finally boiled down to the redisposition of troops. In other words, it was all about who could station its armed forces in which localities to the exclusion of the other party in implementing the agreement.

Chiang drove a hard bargain. He let CCP troops garrison two large cities that were already in CCP hands—Qiqihar in northern Manchuria and Zhangjiakou—but expected a lot in exchange. His demands included, among other things, the CCP’s evacuation from the city of Chengde and all cities in the province of Shandong occupied after 7 June as well as the CCP’s complete withdrawal from all base areas in northern Jiangsu Province.⁷ Chiang’s demands were allegedly out of safety concerns against CCP attack.⁸

Zhou replied with a counter proposal that the CCP could pull its troops out of the main railway in Shandong and reduce its troop numbers in northern Jiangsu on condition that the GMD forces would not enter these areas and take over the CCP-aligned local governments.⁹ Given that the CCP had been using the rice supply from Jiangsu to wage economic warfare against the GMD, northern Jiangsu was not the territory that Zhou would concede.¹⁰ The Nationalists, however, believed that Zhou had already pledged the CCP’s withdrawal from northern Jiangsu last year, and they felt that the CCP now must be getting ready to withdraw.¹¹

Marshall was aware that if they failed to reach an understanding on the amendment of the army reorganisation agreement before 30 June, the other two unsigned agreements would fall victim to the renewal of hostilities. He therefore pinned his hopes on helping the two disputants to reach a settlement on key areas and getting it signed as a preliminary agreement. On 29 June, he successfully won over Zhou’s support for a compromise solution, particularly on a deal that the CCP forces in Jiangsu Province would not be garrisoned beyond the latitudinal south of Huai’an city in central Jiangsu.¹²

When Marshall brought Zhou’s plan to Chiang, he quickly ran into a stone wall. Chiang bristled at the condition that allowed the CCP to continue its occupation of northern Jiangsu. He was sceptical about the relevance of short-term success (e.g. ceasefire) to a long-term win and had no desire to accept any agreement unless it would ensure that there was no more trouble from the CCP.¹³ He insisted on the complete withdrawal of CCP forces from Jiangsu into the north of the Lanzhou–Lianyungang railway near the Jiangsu–Shandong border within a month. Chiang’s demand was tantamount to giving the CCP one month’s notice to vacate the entire Jiangsu–Anhui base, covering an area of approximately 95,000 square kilometres. In his own defence, Chiang told Marshall that he had let his forgiveness of the Chinese Communists exceed what God might think of as fair.¹⁴

On 1 July, Chiang issued a directive on the prolongation of restraint from aggressive action by his armies.¹⁵ The directive did not state a deadline as it was intentionally omitted to suit Chiang’s thirst for revenge. His wishful thinking was that the deliberate omission of a ceasefire deadline would give his army the freedom to attack at any time, as the line between offence and defence is often blurry.¹⁶ Chiang’s ceasefire order made the Committee of Three and the Executive Headquarters a lot less relevant, as it was directly issued by the GMD government. According to Chiang, this would set him free from constraints imposed by the ceasefire supervision organisations.¹⁷ Marshall had little choice but to pass the buck to Zhou and Chiang. Zhou conferred with Chiang and Chiang’s associates in early July but made no headway.¹⁸ The minor party and non-partisan PCC delegates weighed in to remind the two warring factions of the catastrophe ahead if they failed to compromise. Their appeals were ignored.¹⁹

Mao matched Chiang every step of the way. On 1 July, he issued an order prohibiting the use of force except in self-defence.²⁰ Despite having a no-first-use of force directive, Mao had already issued an attacking order instructing his main force to sally out and took the war further south into the GMD’s heartland.²¹ Zhou told Marshall on 26 June that the CCP did not have an aggressive military plan.²² He had perhaps been misinformed.

‘Stripping of the revolution to its military core’

The Nationalists launched a major offensive in late June aimed at pushing the CCP forces away from the trunk rail route in Shandong,²³ but it was battles between the encircled CCP forces and the GMD besiegers at the outskirt of the metropolis of Wuhan in eastern Hubei Province that triggered the full-scale civil war via a massive CCP breakout offensive. The CCP troops in a frantic breakout through the Nationalists’ siege lines were a force, 60,000 strong, of well-equipped troops commanded by Li Xiannian.²⁴ Li’s forces were deployed in one of the eight enclaves that the CCP agreed to concede in the Double Tenth Agreement. Mao decided to renege on the promises in early November 1945, citing concerns over the killing of civilians after the withdrawal of the CCP troops. Li’s units soon received an instruction to hold out against the enemy siege for another year. Since Li’s forces were stationed in places at the core areas of the GMD regime, the Yan’an leadership intended to use their presence to tie down a large number of the GMD troops.²⁵

The Nationalists saw Li’s army as a threat and had been tightening the ring of encirclement for months. Ceasefire field teams had tried desperately to de-escalate the tension and brokered four local truce agreements from January to June 1946, respectively, including the one signed by the Committee of Three in May.²⁶

Without Mao and Chiang’s fundamental change of heart on their respective military strategies, however, no local truce could be held up indefinitely. This was due to the fact that these local truces required both parties’ troops to remain in situ in order to bring an abrupt end to military engagements. The troop movement restrictions virtually compelled the bulk of the two opposing forces to concentrate in a confined area, even though a demilitarised zone had been implemented in the first place. The American field team members observed in early June that the fighting had been caused by the opposing forces being too close to each other.²⁷ Hence, the Committee of Three needed to negotiate further approaches to support the local ceasefires implemented by the field teams.

The two belligerent parties had entered into negotiations at the Committee of Three since March for the possibility of transferring the encircled CCP troops to other places. Zhou proposed that two-thirds of the CCP troops should be moved back to the base areas where they came from, while the remaining 20,000 combatants were to be demobilised on the spot. He nominated the city of Anyang and the county of Wuhe in the provinces of Anhui and Henan, respectively, as relocation options. The two places were at a great distance from where Li Xiannan’s forces were presently deployed, and there were GMD-controlled areas in between. The Nationalists had to let a corps-sized enemy pass their controlled areas by train or on foot.²⁸

Besides, these two places were strategically located in two bitterly contested war zones. The city of Anyang was a frontline headquarters of the GMD forces, and it was only about 60 kilometres south of CCP-held Handan—the capital of the CCP’s Shanxi–Hebei–Shandong–Henan Liberated Area. Wuhe was a CCP-controlled county, but it was within striking distance to the city of Bengbu—one of the GMD’s subheadquarters for its counterinsurgency campaign against the CCP on the Anhui–Jiangsu border. The Nationalists preferred to find a local solution for the problem because Zhou’s plan virtually forced them to allow a large group of enemy troops to trespass on their areas before these enemies crossed into the other side of the battlefield and turned the guns against them.²⁹

Zhou further claimed that if Li’s armies continued to be held up at their present locations, the local population would bear a disproportionate share of burden of providing subsistence for these troops whereas other CCP-controlled areas had ‘stores of food’.³⁰ Nonetheless, Li Xiannian admitted in a meeting with Field Team No. 9 on 22 February 1946 that the CCP did not have sufficient funds to support his 60,000-member army, which consumed around 90,000 catties (i.e. 45,000 kg) of rice every day.³¹

From a ceasefire negotiator’s point of view, Zhou’s proposal might not have been a bad suggestion.³² It offered total withdrawal of the CCP forces from the conflict zone as a next step for the in situ ceasefires, but the Nationalists rejected it on practical grounds. Zhou, in his turn, issued warnings that if the GMD launched a general offensive against Li’s forces, the CCP would deem it to be the start of a full-blown civil war.³³ The negotiations at the Committee of Three dragged on for two months, but were able to reach a truce on 10 May only with a special provision provided for the immediate evacuation of a thousand sick CCP personnel with family members, administrative and medical staff to the city of Anyang.³⁴

Zhou paid a visit to Li’s besieged army on a fact-finding tour of the Committee of Three just before the signing of the truce. During his visit, Zhou enforced a mandatory order for Li’s troops to plan a breakout. At the negotiating table, however, the truce was signed and executed, as the evacuation of the sick CCP personnel began on 15 May.³⁵ The evacuees considered their mass departure a plan to minimise the non-combat personnel and therefore help the main fighting units prepare for the planned breakout.³⁶

On 23 June, Mao ordered Li’s troops to launch the breakout offensive at once.³⁷ According to CCP sources, the GMD armies fired the first shot of the all-out civil war by launching a general offensive on the 26th, which resulted in counter-attacks by Li’s forces.³⁸ The American field team members, who witnessed the breakout, reported that both parties were at fault. The GMD besiegers had been sending large reinforcements to stifle the mobility of the enemy, but Li’s forces had filtered out a substantial number of men and finally executed a pre-planned breakout.³⁹

While the Nationalists threw approximately 200,000 to 300,000 troops at the enemy, Li’s armies fought tenaciously with a well-planned, three-pronged breakout into the Nationalists’ porous defence. The battle soon devolved into a melee of brutal but disjointed breakouts, pursuits and intercepts that spread over six provinces in China proper for more than three months.⁴⁰ Li’s combat forces were drastically reduced to some 20,000 men and broke into small groups at the end of the breakout campaign. Some of these contingents fled into several remote and hilly enclaves between the borders of a number of major provinces in China proper, while the other groups returned to their home bases in North and East China.⁴¹ The supply problem of Li’s forces was now becoming easier to fix, as troop numbers were greatly reduced after ferocious fighting. When a stricken contingent of Li’s armies arrived at the Hebei–Henan–Anhui base area in October, the troops received good food and new cotton-padded clothes.⁴²

Despite heavy casualties on Li’s troops, veterans and historians alike in China nowadays attribute the CCP’s ultimate victory in the civil war to the breakout offensive for disrupting the Nationalists’ war plan and troop deployment.⁴³ A veteran CCP negotiator confirmed that Li Xiannian had personally instructed his representatives in mid-June to sustain the peace talks as cover to prepare the breakout operations.⁴⁴

In the conflict-prone province of Jiangsu, the two combatant parties had amassed their best armies on the north bank of the Yangzi River since late June. Chiang’s corps of 120,000 men was eager to attack northward with the intention of rooting out the CCP insurgents in the region. The CCP, on the other hand, dispatched a combat-capable field army of 30,000-plus troops down towards the lower reaches of the Yangzi.⁴⁵

The CCP forces initiated a pre-emptive strike against soft spots in the GMD’s defence on the north bank of the Yangzi on 13 July. The Yan’an leadership admonished its generals at the onset of the offensive that the battle would not be a hit-and-run operation but rather was a dramatic prelude to a series of all-out offensives. The CCP’s logistic systems, which relied heavily on manpower, partially contributed to this military strategy. The CCP armies could not accomplish an operational concentration of forces for a major offensive without massive peasant mobilisation. The peasants, who had been doing weeks of hard labour in moving military equipment and supplies to the front lines for the CCP, would have been disappointed if the CCP combatants retreated rapidly after a few skirmishes.⁴⁶

Chiang had to decide whether he should order his armies to go toe to toe with the CCP in an all-out offensive that would take place on his doorstep, as Nanjing and Shanghai were alarmingly close on the other side of the Yangzi or whether he should at least consider scaling back his harsh demands and supporting Marshall’s quest for more short-term successes at the negotiating table. Chiang claimed that he received a tacit revelation from God on 14 July, which gave the go-ahead for military actions against the CCP forces in Jiangsu.⁴⁷ Chiang did not disclose the fullness of the divine intervention he received. It therefore remains a puzzle whether Chiang’s divine sources had reminded him that, during a Committee of Three meeting in May, the deputy chief of staff of his Wuhan Field Headquarters, Wang Tianming, had pledged in all sincerity the GMD’s commitment to settle all disputes exclusively by peaceful means.⁴⁸

The two opposing forces staged full-scale military confrontations near China’s capital for the next two months. Strategically and tactically, the campaign concluded with a CCP victory since, although not giving the CCP forces territorial gains, it inflicted 54,000 Nationalist casualties at the cost of 16,000 men.⁴⁹

According to Mao, his armies’ military success in Jiangsu hinged on the principle of striking the vulnerable points of a numerically superior enemy through efficient and rapid concentration of forces. A series of such battles in quick succession could effectively destroy an enemy’s combat power.⁵⁰ As discussed, Mao later elaborated this principle in a popular treatise, entitled ‘Concentrate a superior force to destroy the enemy forces one by one’.⁵¹ It disseminated the idea that the nationwide civil war could be won by annihilating the GMD forces in a piecemeal fashion. Following Mao’s lead, the CCP propaganda further glorified the battles in Jiangsu in 1946 as the ‘seven battles and seven victories in central Jiangsu’. It remains a mainstream narrative about the history of the battles today.⁵²

Mao’s way of warfare required his forces to accomplish successive combat missions with minimal breaks and achieved victory through enormous sacrifices of human lives.⁵³ When the main body was taking a break, in the case of the CCP army’s ‘seven victories’ in Jiangsu, alternative combat units were brought forward to play a key role in the battle. The Nationalists were deprived of a lull to regroup or withdraw.⁵⁴ The strategy was effective in maximising enemy casualties, but it also contributed to the civil war’s further perpetuation.

Yan’an had a well-thought-out game plan: while the GMD armies suffered massive casualties in Jiangsu, the CCP combatants in other theatres of war would be surging forward in force from their respective bases and spreading the war to the Nationalists’ territory.⁵⁵ Such an approach can be considered a radical version of the strategy of taking the fight to the enemy, and it undoubtedly ensured that the cancerous tumour of civil war engulfed China as quickly as possible. In 1947, the two warring factions adopted a similar strategy and rushed their troops towards each other’s territories. The strategy was later known by its sarcastic appellation of ‘the enemy is coming over to my house, but I am going to the enemy’s house, too’ (Di dao wo jia lai, wo dao di jia qu 敵到我家來, 我到敵家去).⁵⁶ To paraphrase Benton’s description of the unprecedented level of military activism of the CCP during the 1930s, Mao’s strategy was another ‘extreme case of this stripping of the revolution to its military core’.⁵⁷

The repeated operational failures drove Chiang to send more troops to the battlefield. For instance, the Nationalists launched diversionary attacks and infiltrated the CCP’s rear base in northern Jiangsu in mid-September. The CCP main forces on the north bank of the lower Yangzi were therefore forced to disengage. The GMD troops also started another round of large-scale offensives in Shandong almost simultaneously.⁵⁸ These attacks were part of a large military campaign in the second half of 1946 planned by Chiang and executed by his generals in a bid to drive the CCP forces out of the war in China proper.⁵⁹ While the GMD forces suffered heavy casualties, the Nationalist records show that Chiang’s all-out offensive failed to destroy the CCP’s combat power, albeit it resulted in territorial gains.⁶⁰ The two warring parties together dragged China into the whirlpool of a full-blown civil war in the second half of 1946.

Hostages of the war

The most destructive aspect of an all-out war is when it becomes dominant and overrules the political, diplomatic and economic logics of the states or armed actors.⁶¹ In the midst of the full-scale civil war, Chiang, Mao and even Marshall were almost simultaneously hijacked by the war they chose to fight or looked to mediate. In mid-July, overzealous Nationalist secret agents conducted two separate but fatal extrajudicial assassinations of prominent leaders of the Democratic League, Li Gongpu and Wen Yiduo, in the south-western China metropolis of Kunming. The Li and Wen assassinations immediately attracted condemnation of Chiang’s repressive regime nationally and internationally. This topic has been widely discussed by a number of authors without a solid consensus over who was responsible for ordering the death warrants. Some suggest that the murders must have been incited or ordered at the top level of Chiang’s regime, if not by Chiang himself. The assassinations must have been carried out in favour of a more explicit emphasis on the GMD’s repressive policies, not to mention that the city of Kunming had been at the forefront of the student-led anti-GMD movement.⁶² Some contend that Chiang was unlikely to risk international censure by ordering the killing of the two minor party figures.⁶³

For Chiang, the assassinations happened at the most inopportune times, as he needed to play the blame-shifting game against the CCP about who was responsible for the outbreak of the full-scale civil war. The seriousness of the two assassinations was unprecedented. Wen was shot to death in broad daylight on his way home after attending the memorial service for Li, who had been assassinated four days earlier.⁶⁴ When the assassins fired the fatal shot at Wen, Li’s murder had already created a furore. No government in its right mind would order extrajudicial killings under the circumstances.

An internal investigation carried out by the Nationalist government quickly solved the case. Two low-ranking army officers from the Kunming Garrison Command underwent a fast-track trial on 15 August. Both men received the death penalty and were allegedly executed in less than two weeks.⁶⁵ Chiang was angry at Huo Kuizhang, the garrison commander at Kunming and the alleged mastermind of the two assassinations. He nevertheless believed that Huo’s many sins should have been forgiven under the circumstances.⁶⁶ Huo was dismissed from his job but was never prosecuted for his role in the Li–Wen case.⁶⁷

While political dissidents in GMD-held areas were living in fear after the assassinations in Kunming, the American mediators were unexpectedly pulled into a military conflict and became disputants themselves. On 29 July, a US Marine supply convoy clashed with CCP combatants in a deadly skirmish at the town of Anping about 60 kilometres south-east of Beiping. The US investigating officers reported that the marine-escorted trucks were en route to Beiping carrying food and supplies for the Executive Headquarters and the UNRRA when the marines were ambushed. A CCP report obtained by Zhou contradicted the Americans. It charged that the incident was a combined invasion by the US Marines and the Nationalist armies of the CCP-held areas.⁶⁸ At Mao’s suggestion, the CCP insisted upon the Americans firing the first shot.⁶⁹

The Executive Headquarters conducted an investigation into the incident but failed to reach a tripartite consensus over the three parties’ role in the fighting. The disputes therefore remained unresolved. A study based on CCP sources, however, found that the tragic incident was a result of an ambush set up by the CCP forces just outside their territories in eastern Hebei. The Yan’an leadership had no prior knowledge of it. The US Marines in usual patrol formation did not raid any town, and there were no GMD troops on their team. The CCP local commanders and field team members resorted to contamination of vital evidence at the scene and abetting the making of misleading witness statements in the wake of the fighting. The military adventurism of individual CCP commanders was fuelled by a Mao-sponsored mass movement against US military aid to the GMD. As the movement was in full flight in June and coincided with the beginnings of the all-out civil war, it was almost impossible for grass-roots cadres to understand that Mao just needed them to rise up against the United States’ ‘erroneous policies’ not kill US Marines.⁷⁰ This revisionist study contradicts popular notions in China, but it is well in line with what Zhou’s associate Wang Bingnan privately told a US assistant naval attaché afterward.⁷¹

The Stuart Committee

The further acceleration of the military conflict convinced Marshall of the urgent need to bring an American with long experience in China to his mediation effort. At his suggestion, the US government appointed Dr John Leighton Stuart, a Chinese-born American missionary and former president of Yenching University at Beiping, as American ambassador to China on 11 July.⁷² Stuart was knowledgeable about China, but not enough to pass Chiang’s test of cultural competence; he failed to appreciate the imperial Chinese rulers’ two-fold method of quashing revolts both by force and by pacification measures, which justified Chiang’s civil war policy.⁷³ Stuart proposed the creation of an informal five-member committee equally represented by GMD and CCP members with Stuart as chairman.⁷⁴

Marshall, Stuart and Zhou reached an understanding that the initial topics of discussion within the small group would be limited to the dispute over the State Council. The Americans nonetheless advised Zhou not to dictate the outcome of discussions in advance because it was impossible to guarantee anything before the convention of the committee, particularly over the issuance of ceasefire orders once the dispute over the State Council was settled.⁷⁵ The short-term committee would be convened on creating momentum for the peace talks via settling the dispute over the distribution of seats in the State Council. The value of the proposal rested not on providing ready-made permanent solutions but on its potential to create manageable stages on the way to long-term success.

Chiang did not seem to appreciate the value of the Stuart Committee. He expressed his support of Stuart’s idea but set forth a list of prerequisites, which had to be accepted by the CCP beforehand. These requirements were of three sorts. They included the CCP’s agreement to put the January ceasefire order into effect, the restoration of communications and the army reorganisation agreement. Even if the CCP was in full compliance with Chiang’s prerequisites, Chiang would not support the committee unless the CCP accepted five extra conditions. Chiang’s ‘five demands’ were even harsher than those he stipulated in June after the temporary ceasefire in Manchuria. Chiang now required, in addition to his previous request of the withdrawal of the CCP forces from areas situated in Shandong, Rehe and Manchuria, the withdrawal of the CCP troops in northern Jiangsu to the territory located in the northern area of the Lanzhou–Lianyungang railway.⁷⁶ Chiang notified Stuart that he would unilaterally abort the negotiations if his demands were not met.⁷⁷

The plethora of preconditions imposed by Chiang did not coincide with the idea of the Stuart-proposed informal group, which attempted to infuse more flexibility into the negotiation process. In particular, Chiang’s entire proposal suffered from a lack of consistency. His attempted reinstatement of the January ceasefire resolution, which essentially restored the status quo of 13 January, and his demand for the CCP forces to withdraw from northern Jiangsu were utterly incompatible. Zhou, not unexpectedly, rejected Chiang’s terms immediately.⁷⁸

The Americans found no delight in their meetings with Chiang,⁷⁹ but the meetings with Zhou gave them no joy either. Zhou essentially demanded a comprehensive ceasefire and that all unsigned agreements before 30 June should be signed without delay. He requested the reorganisation of the current government so that resolutions could be reached on both the political and military issues simultaneously. If his demands were not met, Zhou maintained, it would have been only too evident that the GMD intended to extend the civil war.⁸⁰ The mandatory nature of Zhou’s demands made them unlikely to become topics worthy of discussion for the Nationalists. In a meeting on 26 July, Marshall could not but ask Zhou, ‘Who is to do the discussing?’⁸¹

The arms embargo

When the Stuart-proposed committee was in tatters before it had even begun, Marshall resorted to radical measures to revive his mediation position. He initiated a US arms embargo on China at the end of July with the support of the Truman administration in the midst of the CCP’s prosecution of US military aid to the GMD. Export licence suspension and the resulting shutdown of shipments of arms and ammunition to the Nationalist government was put into effect from August, marking the beginning of an unprecedented nine-month ban on combat-type military equipment to China.⁸²

In Washington, the chairman of the Nationalists’ Chinese Supply Commission, Wang Shoujing, lobbied the US government to overturn the ban. The Americans deflected Wang’s queries.⁸³ His efforts nonetheless exposed more problems with the GMD’s excessive munitions import dependence and the capacity limits of the United States to maintain a reliable supply of sufficient munitions to Chiang’s gigantic army in a nationwide civil war.

Current Chinese Civil War literature is divided over the issue. Some scholars argue that the embargo was designed for politics instead of policy and that the GMD had large stockpiles of weapons and ammunition.⁸⁴ Others maintain that the negative consequences of the embargo on the GMD forces was not noticeable until six months later.⁸⁵ It was unlikely that the Nationalists did not have war reserve ammunition in storage, given that a full-blown civil war was currently under way in China. It is notable, however, that the GMD military was facing an arms supply crisis even without an embargo. The Nationalists’ capability to replace arms, equipment and ammunition for their US-armed elite forces was almost non-existent, as no appreciable quantity of these items was produced in China.⁸⁶ From the .30-calibre ammunition for their US carbines to small replacement parts, lubricants and coolants for their US military vehicles, the weaponry of Chiang’s crack troops was import dependent. Major GMD arms depots had only a limited range of US military equipment in stock,⁸⁷ in part because the supply of certain items, such as automotive parts, was inadequate to meet the US Army’s own demand.⁸⁸

The .30 carbine cartridges were always in short supply at the rate of consumption of the GMD armies, even though they were the in-stock ammunition in the United States for the US forces. A US War Department document of 1947 has verified that the existing stocks and productive capacity of the .30 calibre ammunition in the United States was inadequate to meet the most basic needs of a country as big as China to wage an all-out war for six months. The sales of ‘war surplus matériel’ from the United States to the GMD regime had been the subject of severe criticism by the CCP.⁸⁹ According to the War Department’s estimation, the total amount of the ‘surplus’ .30 cartridges could only meet 8 per cent of the Nationalists’ needs in a six-month supply of ammunition. The remaining 92 per cent could be fully met only by reducing the stocks currently held for US strategic priorities in the Western Hemisphere and global military training programs by roughly 20 per cent.⁹⁰ As delivering a short-term supply of the .30 cartridges to the Nationalists was a near impossibility for the US government, solutions to a continuing supply would depend on the investment incentives of private US industries. The main obstacle was that almost all ammunition manufacturers in the United States were currently under contract to supply the US forces. These suppliers would be facing both production capacity and contract obligation difficulties in filling the Chinese orders.⁹¹

Unlike their elite counterpart, most of the GMD infantrymen predominantly used locally sourced arms and ammunition, but they did not fare better with the embargo. The most commonly used Chinese-made Zhongzheng Type rifles with the 7.92 x 57mm ammunition (also known as Generalissimo or Mauser ammunition) were not only inadequate in quantity but also in quality, which later became a subject of complaint.⁹² While captured Japanese small arms of varying quality were used by the two warring parties on the civil war battlefield, such weapons also needed captured Japanese ammunition, specifically, 6.5 x 50mm and 7.7 x 58mm bullets. Counterfeit weapons and military equipment—such as customised small arms and substandard lubricants—were also being used on the battlefield. However, the GMD’s Army Headquarters had raised the alarm over the danger of using fake products.⁹³

In mid-August, the US State Department suspended an export licence for the shipment of 130 million rounds of Generalissimo ammunition to China.⁹⁴ This represented a massive reduction in ammunition supply for the GMD, as foregoing estimates carried out by US military intelligence indicated that the rifle ammunition consumption of Nationalist forces was approximately 110 million rounds per annum.⁹⁵ US forces did not use the Generalissimo ammunition, and a constant supply of it was by no means assured in the United States post-war. The 130 million rounds currently on the banned list were specially procured by the United States during World War II for China as military aid.⁹⁶ The ammunition had been in storage in the United States for many years and was due to expire. Even if the ban was lifted, the Nationalists had to send the bulk to a top ammunition manufacturer in Europe or the United States for inspection or recondition.⁹⁷

The Generalissimo ammunition is a special cartridge of German origin, which had been widely used in Europe. Although a number of European countries still regarded Chiang as a friend, most of them could only offer uncertain ammunition supply to the GMD government. The Spanish Army, for example, had just begun to replace its old 7.00mm rifles with 7.92mm rifles and would have been reluctant to reduce the 7.92 x 57mm ammunition stock by sales abroad.⁹⁸ The Nationalists therefore tried desperately to purchase one-hundred million rounds of Generalissimo ammunition from independent manufacturers in Europe. Previous studies emphasise Washington’s diplomatic victory in persuading Canada, the United Kingdom and other European countries to join the arms embargo of China.⁹⁹ Nevertheless, the Belgians and the Danes had expressed genuine interest in exporting arms to China, since the embargo made little economic and political sense in post-war Europe.¹⁰⁰

The United States clamped down on attempts to sell arms and ammunition to the Chinese quite easily in the end. Declassified archival materials confirm that no European ammunition maker had the production capacity to supply such a great quantity of ammunition by itself, which precipitated the diplomatic triumph of the Americans. Leading firearms manufacturer Fabrique Nationale of Belgium was interested in selling one-hundred million rounds of Generalissimo ammunition to its Chinese buyers, but it could fill only half the order and needed to strike a deal with Winchester of the United States for supplying the other half. The deal was off in the end, owing in part to the disapproval of the US government.¹⁰¹ The Nationalists, for their part, apparently encountered difficulties in making payment to the Belgians. The letter of credit that the Nationalists needed for the big purchase was withheld for unknown reasons.¹⁰² This might not be a surprise, considering the GMD’s dwindling war chest. Chiang therefore had only had two options. He could either keep fighting until ammunition and supplies were exhausted, or he could readjust his negotiation position in the hope that the Americans would reverse the embargo.

Perfect falling-out

On 10 August, Marshall and Stuart released a joint press statement in a bid to exert pressure on both parties for a settlement. The statement attributed the negotiation deadlock to both parties’ failure to reach an agreement, particularly on issues related to the redisposition of troops and the CCP-held local governments.¹⁰³ Truman simultaneously sent a stern personal message to Chiang essentially telling him to find a peaceful solution to end the civil war or face loss of support from the United States.¹⁰⁴

Zhou reacted first to the Marshall–Stuart media stunt, although the CCP propaganda published a defiant point of view against the statement.¹⁰⁵ Zhou told Stuart on 12 August that his party was ready to make offers vis-à-vis Chiang’s demands, including a concession of troop withdrawal from positions occupied since 7 June in Shanxi Province.¹⁰⁶ The province had been experiencing the brunt of the full-scale civil war since July 1946 after a strong CCP force laid siege to the GMD-held strategic city of Datong near the northern border of the province. The CCP armies of 80,000 troops launched final assaults against some 20,000 desperate Nationalist defenders of the city in early August after severing the city’s main railway to the GMD-held provincial capital of Taiyuan in the south. If Datong was captured, the CCP forces could combine their armies at Zhangjiakou to attack Beiping.¹⁰⁷

If Zhou wished to maintain his tough negotiation stance, he could not propose a withdrawal from the siege because he regarded all CCP military operations as being initiated to counter or divert the GMD attacks.¹⁰⁸ Therefore Zhou’s new offer expressed his intention to give a constructive response to the Marshall–Stuart joint statement. The problem was that Zhou was vitriolic towards Chiang, and the situation on the battlefield remained fluid.¹⁰⁹ If Chiang did not reciprocate quickly enough, Zhou might change his mind. Much to Stuart’s vexation, Chiang did not shun his business-as-usual attitude. His message, delivered on 14 August commemorating the first anniversary of victory against Japan, was filled with predictable anti-Communist rhetoric.¹¹⁰ Chiang might have argued that he was just trading insults with Mao. In Mao’s latest smear campaign, the Nationalists were painted as ‘paper tigers’ and ‘running dogs’ of the ‘US reactionaries’.¹¹¹

On 26 August, Zhou told the press that the CCP would not enter into any negotiation until the Nationalists made a clear statement that they had relinquished Chiang’s ‘five conditions’.¹¹² Zhou was not wrong for seeking an unambiguous rhetorical shift on Chiang’s part before making any decision, but once he made his position public, he set himself up, as Chiang did, in an inflexible negotiation position. Zhou said repeatedly that he needed the Americans to give him two guarantees as conditions for him to enter the five-member committee: Chiang would drop the five demands, and a ceasefire would go into effect immediately after the meeting. Zhou asserted that the two prerequisites were needed in order to prevent Chiang from negotiating in bad faith and using peace talks as a cover for escalating violence.¹¹³

Marshall was disappointed, replying, ‘If Doctor Stuart and I could guarantee the ceasefire, the fighting would have stopped months ago.’¹¹⁴ Zhou was not convinced, arguing that he had the responsibility to ask for guarantees in order to avoid being forced ‘step by step’ into a concession that he was not ready to commit to. He told Marshall and Stuart that his agreement to consider joining Stuart’s little committee constituted a concession already.¹¹⁵

Marshall paid a visit to Madame Chiang, telling her that the US military equipment had been fuelling Chiang’s ‘sense of false power’.¹¹⁶ He successfully forced Chiang to back down on 9 September. Chiang agreed that a Committee of Three meeting would be held concurrently when the five-member committee met. The Committee of Three would discuss a new truce and all outstanding military issues, but the CCP must designate its representatives for the National Assembly and State Council before the announcement of the truce. Chiang also agreed to let the State Council settle all territorial disputes.¹¹⁷

However, the issue at play was not whether Chiang had retracted his ‘five conditions’ but Marshall’s impression of Chiang’s position during their meeting; Marshall informed Zhou of this in a memorandum dated 10 September. First, Marshall got the feeling that Chiang would not support governmental reforms before the convening of the National Assembly. Second, Marshall believed that Chiang was unlikely to withdraw from the places his armies had recently occupied.¹¹⁸

Zhou was clearly preoccupied when he conferred with Marshall on 11 September. He steadfastly refused to participate in the five-member committee, declaring that Stuart’s informal group was virtually a scam as long as the war continued. Zhou instead demanded that a formal Committee of Three meeting be called immediately in order to re-establish a ceasefire. Although Zhou labelled Chiang’s entire proposal a breach of PCC resolutions, he admitted that Chiang’s idea of settling disputes via the State Council was a topic worthy of discussion by the Stuart-led small group. In other words, Zhou refused to become involved in the Stuart Committee before Chiang halted all military operations, even though he still found common ground with Chiang on the State Council.¹¹⁹

Zhou’s insistence on calling a Committee of Three meeting was at odds with Stuart’s idea of turning the highly publicised but stalemated formal negotiations into a more informal one. From a negotiation view point, informal discussions are essential to increasing the likelihood of concession.¹²⁰ Given that both Chiang and Zhou agreed to take the dispute to the State Council, Stuart’s informal group could have reinvented the entire negotiation in a less competitive environment before both parties gained enough momentum to resume formal negotiations. Instead, as Marshall noted, Chiang effectively botched the Committee of Three meeting from the start by knowingly restricting his negotiators’ ability to make concessions. Zhou, for his part, could not even agree to give Stuart custody of a sealed envelope containing the CCP nominees for the State Council.¹²¹

Zhou was not interested in participating in Stuart’s informal group before getting all the guarantees he requested. The problem was that Chiang had agreed to take the dispute over the CCP local government to the State Council. The offer observed the letter and the spirit of the PCC resolutions and fell within the scope of what Stuart’s small group set out to discuss. Zhou was aware that if he walked away from the negotiation, he would be putting himself in an untenable position. ‘I cannot refuse to consider anything’, Zhou told Marshall.¹²²

While Zhou’s negotiation strategy was in limbo, Chiang provided all the help that Zhou could possibly get. Zhou went to press Stuart on 13 September that the CCP must be given ten seats in the State Council. As mentioned, the disagreement over the State Council seats could have been a non-issue if one of the parties intended to tone down the competitive dynamic in their post-treaty relationship. Although Zhou had been advised not to dictate the outcome of the discussions in advance, he told Stuart that if his request was not accepted beforehand, there was no need for the informal committee to convene. Simultaneously, Chiang’s representatives informed Stuart that the GMD would not discuss the distribution of seats in the State Council in informal meetings, relegating the Stuart Committee useless. The sudden change of heart indicated that Chiang wanted to renege on his previous proposition.¹²³

Chiang’s latest backflip on the peace talks put Zhou in the perfect position to withdraw from the negotiations. Marshall’s office received three strongly worded memoranda from Zhou on 16 September, asking Marshall to convene a Committee of Three meeting ‘immediately’ for a ceasefire. Zhou was unabashed in claiming that all other ‘intricate’ attempts to break the deadlock were ‘non-instrumental’. He suggested that an impromptu Committee of Three meeting was the only hope left to provide ‘a direct and simple’ solution. After doubling down on his criticism on the US-sponsored peace talks, Zhou notified Marshall that he was leaving Nanjing for Shanghai and would not be back until the Committee of Three was called.¹²⁴

Zhou made a public announcement in Shanghai about his withdrawal from the peace talks. He also delivered a bold statement, including a vow to continue fighting the GMD in the civil war.¹²⁵ By the end of September 1946, Zhou began to evacuate his intelligence teams secretly from the GMD-held cities. The aeroplanes of the Executive Headquarters and freighters of the UNRRA were used to ensure the success of the covert operations. Zhou reminded his staffers that the all-out civil war had reached the point of no return.¹²⁶

While the negotiations stalled, the fierce battle for Datong continued to unfold. The CCP attack culminated in early September when the GMD troops were in danger of losing their last line of defence.¹²⁷ The Executive Headquarters and its field teams were powerless to halt the fighting.¹²⁸ The situation therefore continued to deteriorate. The Nationalists announced on 4 September that if the CCP did not lift the siege, they would retaliate by attacking CCP-held Zhangjiakou, even though Chiang had previously agreed to put it under CCP control.¹²⁹ The CCP combatants were forced to end the siege of Datong on 14 September after the Nationalists captured the CCP’s rear base a hundred kilometres to Datong’s north. Chiang’s armies from North China and Manchuria formed a task force to attack Zhangjiakou after the relief of the siege.¹³⁰ The offensive was conducted on a grand scale despite repeated warnings from the CCP that it considered the attack to be a conclusive announcement of a total split.¹³¹

Marshall’s embarrassment was made worse by the fierce military confrontations. He put in writing for the first time his intention to terminate the mediation mission to Chiang on 1 October.¹³² On 6 October, Chiang agreed to halt the offensive against Zhangjiakou by issuing a ten-day and non-renewable truce for both parties to settle all the disputes within the truce period. Chiang made the offer after having some very unpleasant exchanges with Marshall.¹³³ Chiang did not regret it because he was expecting a refusal from the CCP.¹³⁴

Chiang’s predictions were right. The CCP saw his proposal as a threat and turned it down outright.¹³⁵ The rejection justified Chiang’s decision to continue the offensive against Zhangjiakou. Marshall and Stuart called for desperate measures again and issued another joint statement on 8 October. It detailed the unyielding positions of the two warring factions and disclosed the CCP’s rejection of Chiang’s ten-day truce in full.¹³⁶ Marshall then departed quietly to Shanghai in an effort to persuade Zhou to return to the negotiating table.¹³⁷ This was the Americans’ second attempt since August to break the negotiation impasse through the media. A party usually would not take its case to the public unless it wants to show toughness in a position.¹³⁸ When Marshall tried to use the same tactic to foster a reconciliatory atmosphere, he was taking a real risk.

Zhou’s patience was now running out. When he met Marshall on 9 October, Zhou rounded on Marshall about the timing of the joint statements, complaining that they were always released after the CCP’s rejection of demands from the GMD, leading the public to misunderstanding. Zhou went on to excoriate Marshall for his handling of the mediation, including certain expressions within the text of the second joint statement that Marshall deemed trivial.¹³⁹

The Zhou–Marshall conference made no headway. The meeting nonetheless summed up all major differences between the two warring parties. On the question of a ceasefire, the GMD preferred a short-term truce and used it as a litmus test for the CCP’s sincerity of resolving all unsettled issues. The CCP considered the temporary truce as military coercion, claiming that it would accept nothing but a permanent ceasefire provided for the restoration of the territorial status quo of 13 January in China proper and a GMD troop withdrawal to the positions held on 7 June in Manchuria.¹⁴⁰

On the political issues, both parties implemented the PCC resolutions on their own terms. On the quota of seats in the State Council, the maximum concession on offer from the Nationalists was one non-partisan seat to be recommended by the CCP; adding that extra seat to the initial eight and four seats allotted to the CCP and the Democratic League, respectively, made a total of thirteen seats. The Chinese Communists ridiculed the GMD’s ‘concession’. They instead reiterated all the old arguments to justify their ‘original right’ of enjoying fourteen votes in the council with the Democratic League.¹⁴¹

On the convocation of the National Assembly, the GMD government had made a unilateral announcement on 3 July that the assembly would be held on 12 November 1946.¹⁴² Chiang felt that he had every right to confront the obstructive tactics of the CCP and play a leading role in China’s constitutional future.¹⁴³ The CCP in turn declared that it would neither submit names of delegates nor participate in the assembly and cited its ‘two-stage’ interpretation of the PCC resolutions to support the claim that the GMD’s unilateral decision had breached its treaty obligations.¹⁴⁴

Veto players and the third-party bloc

The GMD forces captured Zhangjiakou on 11 October after the CCP troops retreated to nearby rural areas.¹⁴⁵ The Nationalists’ victory coincided with their announcement of the resumption of nationwide conscription.¹⁴⁶ The GMD’s chief of general staff, Chen Cheng reinforced the already-prevalent war psychosis, declaring on 17 October that his armies would be able to win the civil war within months.¹⁴⁷ The Americans were willing to exhaust all means before withdrawing from the negotiations. Under the auspices of Stuart, leaders of the Democratic League, the Youth Party and non-partisan politicians formed a third-party bloc and intervened in the stalemated peace talks.¹⁴⁸ Zhou met delegates of the third-party alliance on 15 October, but he declined to comment on their request for returning to Nanjing.¹⁴⁹

While Zhou was radiating his silent power, all parties deadlocked in the negotiations were trying to make sense of Stuart’s move to deal the third-party bloc into the main game. The third-party bloc’s meddling in the negotiations sent a strong reminder to the two warring sides and the Americans that the PCC resolutions were a deal done by four parties and independent political leaders. The Democratic League, the Youth Party and the non-partisan politicians had the right to interpret the resolutions in their respective ways. Most importantly, the third-party bloc could surpass the military antagonism of the two belligerent parties as China’s democratic third force.¹⁵⁰ This was what the American mediators were unable to do. Nevertheless, the third-party coalition could achieve this goal only in a multiparty bargaining environment.

The minor parties were being kept in the background after the PCC because the only two armed Chinese factions had vetoed agreements and returned to the battlefield. In the state of a civil war, only players who have armies powerful enough ‘to avoid being soundly defeated by a united army in the aftermath of an agreement’ have the ability to veto an agreement unilaterally. As David Cunningham identifies, these armed factions can become the ‘veto players’.¹⁵¹ The military might of the GMD and the CCP therefore made them the only two ‘veto players’ in the Chinese Civil War. As the general secretary of the Democratic League, Liang Shuming remembered, ‘The overall situation has been dominated by the two big forces; others can do nothing against them.’¹⁵²

On the contrary, if the third-party alliance succeeded in negotiating the two warring factions into peace agreements, the civil liberties endorsed by the PCC would have been set free. Once a multiparty bargaining environment was established, life for the two erstwhile belligerents would have been less straightforward. The GMD and the CCP would have had to woo minor parties and non-partisan members, who together would have potentially enjoyed twelve votes in the State Council, either for getting proposals passed or simply because of institutional constraints. Meanwhile, the third-party alliance had come into play. It made the CCP, the GMD and even the Americans simultaneously hold ambivalent attitudes towards the group. The Americans considered the group an alternative for resolving the impasse, but they trod carefully to avoid giving the wrong impression to the CCP that they were trying to win over the Democratic League from it.¹⁵³ Marshall was confronted with the brutal fact that the group was formed by independent yet ambitious politicians with somewhat unique personalities. Democratic League member Luo Rongji was a leader of the group, but his flamboyant personality did not earn more credibility for him. Marshall was troubled by Luo’s habit of smearing things in his random talks with the press.¹⁵⁴

Zhou was in less of a mood to reopen the peace talks. He rather used meetings with the third-party bloc to persuade the CCP’s closest ally, the Democratic League, to boycott the National Assembly.¹⁵⁵ In general, Zhou believed that the third-party bloc needed to be educated because of their illusion of peace and ignorance of Chiang’s conspiracy. The difficult part was that the Democratic League was a minor party alliance. It was not a disciplined Marxist–Leninist party like the CCP, even though the deputy general secretary of the league, Zhou Xinmin, was concurrently a senior CCP cadre.¹⁵⁶ The chairman of the league, Zhang Junmai, was a constitutionalist who played an important role in the CRDC.¹⁵⁷ The CCP quickly perceived that Zhang wavered over boycotting a constitutional assembly.¹⁵⁸

League member Huang Yanpei was a respectable front man of the third-party bloc, a key CCP sympathiser and a member of the People’s Political Council (Guomin canzheng hui 國民參政會)—a GMD-controlled advisory body—who received salary payments from the government. Although the GMD military police raided his home in January 1946, Huang did not stop receiving cash payments as gifts from the Chinese crime boss and Chiang’s staunch ally, Du Yuesheng, who had been trying to lure Huang over to Chiang. Du paid Huang 4 million Fabi when Huang was actively involved in mediation efforts from October 1946 to January 1947. The relationship between the two was not limited to cold cash. The pair stayed in touch via correspondence even when Du was away. Huang might have needed to maintain a good relationship with Chiang’s number one hired killer for financial reasons and safety concerns. However, it did not exonerate him from mismanaging conflicts of interest.¹⁵⁹

Generally speaking, the Democratic League had certain incompatible views on the constitution and army nationalisation with the CCP, not to mention that the inability of many minor-party bloc members to resist corruption had been well recorded.¹⁶⁰ Given that oscillating political stances and personalities would be magnified in a multiparty bargaining environment, it posted a challenge for the CCP in dealing with the third-party alliance in the State Council if a peace settlement could be materialised. The unpredictable interparty politics of a power-sharing government lent veracity to the old assumption that the CCP could conquer China faster by force than by political means.¹⁶¹

The Nationalists did not want the third-party bloc’s direct involvement in the peace talks either. They did not have many friends in the group, as even their political ally the Youth Party had been critical of the GMD. Although Chiang had promised the Youth Party a democratic and constitutional government,¹⁶² it still gave him a hard time during the PCC by advocating a quasi-cabinet system to rein in presidential power.¹⁶³

Hence Chiang needed to put his proposal on the table first instead of allowing the third-party coalition to exert influence on the negotiations. He did it on 16 October in a public statement. It started with Chiang’s usual anti-CCP rant, which the CCP considered bullying. The rest of the proposal was inversely related to its offensive rhetoric. It supported the Zhou-proposed formal meeting of the Committee of Three concurrently with a meeting of Stuart’s informal committee, with eight conditions attached. Some of the stipulations in Chiang’s proposal were either totally identical with or partially different from the CCP’s demands. For example, Chiang’s agreement to calling meetings of the CRDC was answering the CCP’s call that a consensus on the constitution should be reached before formally passing it in the National Assembly. The suggestion for armies of the two warring factions to stay at their present localities pending further decisions was in essence supporting Zhou’s call in August for an immediate nationwide ceasefire. Regrettably, Chiang threatened at the end of the proposal that he would not implement a ceasefire until the CCP accepted his eight conditions and announced its lists of delegates of the National Assembly.¹⁶⁴

The CCP intended to reject Chiang’s plan outright as it failed to address its ‘two conditions’, namely the restoration of positions occupied by the two opposing armies on 13 January and the implementation of PCC resolutions on the CCP’s terms.¹⁶⁵ Another problem was that Chiang’s proposal came only after the Nationalists invaded the CCP base in Jiangsu and captured its regional headquarters at Huaiyin city on 19 September. The CCP construed Chiang’s proposal as a mockery of their defeat under the circumstances.¹⁶⁶

Although leaders of the third-party alliance might tell the press that they were disappointed in Chiang’s proposal,¹⁶⁷ they did not completely oppose it. Luo Rongji and Zhang Junmai found shortcomings in Chiang’s proposal, particularly on the demand for the submission of the list of delegates of the National Assembly before the reorganisation of the government. Zhang nonetheless did not believe that these defects were as important as the ceasefire, because these ‘outstanding points’ were negotiable after a truce deal had been struck. Luo asserted that the CCP ‘would be glad to have and had proposed’ many of the ideas stated in Chiang’s proposal. The problem was Chiang’s attitude. According to Luo, Chiang’s approach was like saying, ‘I give you these terms; agree or not.’ ‘The Communists cannot accept that attitude … There is a matter of face there’, Luo added.¹⁶⁸

Given that the view of the third-party coalition on Chiang’s ‘eight points’ was different from that of the CCP, it was likely to roll out an independent proposal in the event of the CCP refusal of Chiang’s plan.¹⁶⁹ A third-party proposal could be something close to a fifty-fifty split—an approach that modern mediators have been advised to avoid—and unacceptable to both major parties.¹⁷⁰ However, the ‘original right’ of the group to put forward its own proposal would make it difficult for the two warring parties to dismiss it.

The CCP leadership quickly identified the problems. It advised Zhou to actively advocate the CCP’s terms rather than insisting on getting ‘guarantees’ before attending meetings.¹⁷¹ The instruction was similar to the advice from Marshall that Zhou refused to accept when they conferred in Shanghai.¹⁷² However, the Yan’an leadership issued a public statement on 18 October. It contradicted Chiang’s ‘eight points’ with its ‘two conditions’ after a fiery tirade accusing the GMD of waging civil war and treaty violations with the American accomplices.¹⁷³

The third-party group were eager to get Zhou back to Nanjing. They drafted a basic proposal outline at Zhou’s residence on 19 October 1946. The draft was written with inclusive rhetoric, but the article about the two opposing forces holding their current position in a nationwide ceasefire was borrowed from Chiang’s ‘eight points’. The group took the draft and brought Zhou with them to visit the Nationalists’ PCC representative, Wu Tiecheng. The visit set up an informal meeting between representatives of the five signatories of the PCC resolutions. During the meeting, Wu promised to increase the openness of the negotiations, maintaining that Chiang’s ‘eight points’ merely served as a basis for negotiations. The meeting ended with an agreement approved by acclamation that Zhou would return to the peace talks in Nanjing.¹⁷⁴

The third-party bloc and Zhou met Chiang after arriving at Nanjing on 21 October. Chiang, who believed that Zhou and the minor party politicians had no sense of decency, did not throw support behind the group. The entire meeting lasted only about eight minutes, and nothing of any substance was achieved. Chiang left in a hurry for another inspection tour to Taiwan.¹⁷⁵ Chiang admonished his interlocutors not to circumvent his principle: there would be no ceasefire without the CCP accepting his terms. Also, he did not stop ordering his armies to intensify the offensive.¹⁷⁶

Members of the third-party group were shattered when Chiang’s representatives failed to live up to their side of the bargain by engaging in an open-minded negotiation. Some even began to realise that Zhou’s return might be the political manoeuvre that preceded the CCP’s pulling out from the peace talks.¹⁷⁷ The two warring sides simply reiterated their respective non-negotiable terms again and used them to attack each other. Disagreement over the allocation of state councillors’ seats topped the list of the most contentious political issues between the two warring parties—a dispute over the deficit of one state councillor seat, to be precise.¹⁷⁸

As discussed, the State Council had the authority to settle almost all the disputes between the two belligerent parties. Both parties could have adhered to the letter and the spirit of the PCC resolutions much better if they settled the conflict over the quota of the state councillors cooperatively with the third-party coalition and went to the inaugural session of the State Council to dispute all other outstanding issues. Although the third-party coalition members did not reach a consensus over the real causes of the two-party civil war, they agreed that the row over the single state councillor seat was a façade for the GMD and the CCP to hide their true objectives.¹⁷⁹ If Chiang ignored the CCP’s boycott and went ahead with the assembly on 12 November 1946, it would put an end to their mediation efforts.

The Nationalists captured the CCP-held Manchurian seaport of Andong on 25 October. Zhou was furious about it and threatened to withdraw from the negotiation for the last time. Zhou requested the members of the third-party group, who prompted pleas for Zhou to stay, not to advance any proposal without his consent.¹⁸⁰ The demand meant that the group must take concerted actions with the CCP.

The third-party coalition was committed to a quick way to end the conflict. It defied Zhou’s request and forwarded an independent proposal simultaneously to Chiang, the Americans and Zhou on 28 October. The terms of the ceasefire in the proposal were essentially what Zhou had been informed of in Shanghai, but it was the articles about the troop deployment in Manchuria that caused Zhou’s fear. While the proposal recommended that the CCP forces would be redeployed to three cities in remote northern Manchuria, it virtually let the GMD have free access to most of the areas along the Chinese Changchun Railway.¹⁸¹

The proposal showed a clear intention to defuse the ticking Manchurian time bomb by separating the two opposing armies in the most obvious way. It was not at odds with the idea of having a provincial demarcation of garrison lines for the two warring sides that Zhou initially proposed back in June.¹⁸² The unfolding events proved that it might not be a bad idea, as the two opposing forces had committed to launch massive winter offensives in Manchuria.

The third-party bloc nevertheless seemed not to be taking the aggressive negotiating style of the two warring factions into account. The two heavily armed veto players in the Chinese Civil War would prefer the fighting to be sustained before moving into quid pro quo exchanges at the time of their choosing rather than accepting a truce brokered by a minor-party alliance.¹⁸³ Zhou met the third-party group representatives who brought the proposal to him. He threw tantrums to express his disappointment at not being consulted for an unfair deal. It caused huge embarrassment to everyone present, particularly to the group’s leader, Liang Shuming.¹⁸⁴

Chiang, freshly returned from Taiwan, capitalised on the misfortune of the third-party bloc. He made a couple of quick steps in the lead-up to the National Assembly that would end the Americans’ and the third-party bloc’s mediation attempts and precipitate the final split with CCP. Chiang met members of the third-party group right after their clash with Zhou. He asked them not to propose truce deals, stating that he would issue a ceasefire pending the minor parties’ submission of the names of delegates for the National Assembly regardless of how the CCP reacted. The third-party coalition began facing the prospect of breaking up after the meeting.¹⁸⁵

Chiang then rapidly informed the Americans that he was going to make further concessions, including a pledge that his force would stop expanding its territory along the Chinese Changchun Railway before the State Council was reorganised. As the concessions were a modification of the third-party group’s proposal that Zhou deemed unfair, he could be sure that the latter would be too proud to accept them. Chiang noted privately that his concessions merely served the purpose of winning the minor parties over.¹⁸⁶

On 8 November, Chiang issued an unconditional ceasefire order and a statement reiterated the opening of the National Assembly on the 12th.¹⁸⁷ Chiang’s double-sided announcements incited the CCP to release a counterstatement scoffing at the unilateral ceasefire, particularly at the article about conducting self-defence. The CCP considered it the pretext for more military actions.¹⁸⁸ Meanwhile, the Nationalists conveyed messages to the Democratic League about a new formula of eight (CCP), five (Democratic League), four (the Youth Party) and three (independent) regarding the twenty non-GMD state councillors, and proposed once again that twelve votes were all it needed to veto decisions. However, the CCP had shifted its focus to delegitimise a Chiang-controlled National Assembly. Zhou admitted to the third-party bloc leaders that the quota of the state councillors was not an issue any more because the real veto power came from the CCP–minor-party bloc cooperation. He asserted: ‘It’s all the same to me now whether we will be allotted one more seat or one less.’¹⁸⁹

On 11 November, the GMD government postponed the National Assembly for three uselessly short days.¹⁹⁰ The third-party bloc members were devastated. They believed that the dispute over the distribution of seats in the State Council had been completely resolved as the CCP–Democratic League bloc now had thirteen votes to veto decisions that required only twelve. Chiang could have reached out to the CCP and started restructuring the State Council instead of rescheduling the assembly to the 15th.¹⁹¹ Chiang, however, had other ideas. The three-day postponement was his plan to isolate the CCP and split the Democratic League in order to make the National Assembly a GMD rubber stamp.¹⁹²

The temporary alliance of the third-party group disintegrated when the National Assembly convened on 15 November 1946 amid a CCP–Democratic League boycott. The Youth Party and Zhang Junmai’s associates, attended the assembly with some non-partisans.¹⁹³ They tried to make an impression, but the GMD was in full control in the end. The assembly (later known as the Constituent National Assembly 制憲國大) adopted the new constitution of China on 25 December in basically the same form as initially presented, including the much-disputed articles about executive responsibility to legislature and the autonomy of provincial governments.¹⁹⁴ Zhou and key members of the CCP delegation departed for Yan’an on 19 November for the last time, after declaring in a public statement that the National Assembly was illegal.¹⁹⁵ A representative of the CCP’s liaison office in Nanjing declared on 6 December that the CCP forces had launched pre-emptive strikes from all fronts in the name of the PCC resolutions.¹⁹⁶

Marshall was recalled and returned to the United States on 8 January 1947 as chances to reopen negotiations were all but gone. The US government announced its complete withdrawal from the tripartite peace talks after Marshall’s departure.¹⁹⁷ Marshall told Truman before his recall that the CCP definitely considered him to be persona non grata.¹⁹⁸ Marshall had guessed right. China was bracing for nationalist protests against the US military presence in January 1947, following the rape of a Chinese female student by US soldiers in Beiping.¹⁹⁹ A phased withdrawal of US forces soon followed, and by October 1947, US military personnel in China were reduced to 6,532 men, including 1,591 marines.²⁰⁰

The Nationalist authorities issued notices to enforce the withdrawal of all CCP representatives on 28 February 1947.²⁰¹ The door for negotiations was formally shut on 9 March after the last contingent of CCP delegates had left. In preparation for invasion by their ground forces, Nationalist aircraft bombarded Yan’an on 11 March, soon after the withdrawal of the US military observers team.²⁰² The Nationalists now faced a benign State Council as the CCP and Democratic League were out of the political picture completely. They successfully rammed a resolution on general mobilisation of Chinese society for a civil war against the CCP through the council in July 1947 before it was enacted into law in a GMD-controlled National Assembly on 18 April 1948. The resolution gave the Nationalists not only unlimited power to control all forms of labour, services and materials but also the authority to impose restrictions on freedom of speech and expression.²⁰³ The CCP, on the other hand, had no need to enact legislation. It believed that it held the mandate of ‘the people’ to overthrow an oppressive government. The combat-ready armies of the two veto players of the Chinese Civil War were finally able to fight freely along the lighted path of an all-out war.

Marshall: ‘The state of peace appears at times more difficult than the state of war’

In his meeting with Marshall, Hu Lin, one of the founders of the prestigious Chinese newspaper Dagong Bao (大公報, Ta Kung Pao, Impartial Daily), suggested that the difference between the GMD and the CCP was less stark than one would have imagined: both were revolutionary parties motivated by destructive impulses, and both adopted similar tactics against each other.²⁰⁴ The Nationalists’ chief of general staff, Chen Cheng, seemed to agree with Hu. He admitted privately that he learnt revolutionary ideas from the Soviets.²⁰⁵ As discussed, the two veto players in the Chinese Civil War were equally capable of destroying each other on the battlefield. Chiang’s two-fold strategy that prioritised quashing revolts by force was matched by Mao’s all-out offensive approach. Both men virtually collaborated to make the civil war interminable, and they led their respective parties into a state of moral attenuation. As the civil war raged on, the two combatant parties became just what they were fighting.²⁰⁶

Thus, the two opposing forces became locked in a vicious cycle of attack and counter-attack. The Nationalists’ offensive culminated in territorial gains, but the all-out attacks of the Chinese Communists succeeded in inflicting overwhelming Nationalist casualties. While triumph and disaster are two imposters, both sides’ use of truce and peace talks as cover to prepare fresh attacks was astonishing. Chiang purposely omitted the deadline from his unilateral ceasefire order in furtherance of his offensive strategies. Chiang’s tactic was countered by the CCP leaders’ approaches to manipulate the ceasefire monitoring mechanism. The CCP’s evacuation of non-combat personnel from its combat units in preparation for renewed fighting was a stunning example of using the agreed terms of ceasefire for prohibited purposes. Given that the evacuation of injured and sick soldiers can hardly be considered a hostile act, this unorthodox approach, even today, allows belligerent parties to circumvent truce agreements with a carefully crafted list of prohibited acts.²⁰⁷

The total war footing increasingly blurred the line between bellicose rhetoric and practical polices. Chiang might not have intended to make targeted killing the mainstay of his government, but his secret police apparently went astray in the midst of the war. In this sense, Chiang’s decision to pardon the mastermind of the Li–Wen assassinations might have been justified: bad environments often make normal people commit atrocities. There was no evidence that leaders in Yan’an gave orders to attack the US Marines. Mao nonetheless deemed it necessary to shift the blame entirely to the Americans after the deadly firefight. Mao must have known that the barbarity of the full-blown civil war knew no boundaries, and no one had the power to contain the warlike populism of which he was the biggest patron.

The two belligerent parties were sceptical about the long-term value of agreements in the midst of escalated military confrontations. Both Chiang and Zhou stipulated harsh conditions and looked for guarantees of consensus even to enter negotiations. Chiang’s and Zhou’s concerns over the irrelevance of short-term success for a long-term win were not unwarranted. Scholars of today have confirmed that agreements have little long-term value if the relationship between disputants is built on mutual mistrust.²⁰⁸ This perhaps explains why the present historical view in China of Stuart’s five-member committee is basically unchanged since Zhou considered it a scam in 1946.²⁰⁹ It was not until the 1990s that conflict resolution theorists started thinking about long-term success at several intervals after mediation.²¹⁰ Stuart was therefore ahead of his time with his innovative ideas in peace negotiations. This raises the question of whether future studies of Stuart’s contributions to the Marshall mission need to go beyond prior studies that focus on Stuart’s pro-Nationalist stance and distrust of the CCP.²¹¹ The successful implementation of the Stuart-proposed informal yet short-term committee relied on the disputants’ appreciation of the potential of short-term agreements in creating manageable intervals for long-term success. This was exactly what contemporary scholars have set out to explore. Stuart’s plan was snubbed by the two warring factions at the time, and his posthumously published embassy report is metaphorically titled The Forgotten Ambassador.²¹² Given that modern negotiation studies have only begun to scratch the surface of the subject that Stuart raised more than half a century ago, perhaps one could hope that the forgotten ambassador might not be forgotten again.

Stuart’s creative informal committee was the American mediators’ last trump card.²¹³ When the plan failed even before it began, the Americans resorted to desperate measures. The arms embargo was perhaps the most extreme one. Despite early signs of success, the embargo failed to dissuade Chiang from using force to settle problems. Chiang literally ignored the arms embargo despite the fact that his armies were hit hard by it. The quantity of US ammunition held by the GMD Army had reduced to a dangerous level in April 1947, but Chiang kept his field commanders ignorant of this fact.²¹⁴ The Americans started to relax the arms embargo from April 1947 until it was unofficially ended by December of the same year after it was proved to be penalising only the GMD but not the CCP.²¹⁵ The arms embargo ended with the bitter triumph of Chiang’s persistence and the Americans’ failure to erase the stigma of their support of the Nationalists in the civil war.

Marshall and Stuart paid the price for their desperate move of publicising the inflexible stances of the two belligerent parties to the public. Perhaps Marshall’s remark about the unyielding positions of both parties also applied to himself: ‘[The mistake was] a very human reaction, though a very short-sighted one.’²¹⁶ When the tide of war turned against the Nationalists in mid-1947, Marshall was singled out for censure, particularly for the GMD’s defeat in Manchuria. Chiang’s exercise in serial blame-shifting began with talks via diplomatic channels²¹⁷ denouncing Marshall to the GMD military leaders²¹⁸ and ended with the vilification of Marshall in his diary. Chiang’s comments were well recorded, and they ensured that Marshall’s name was tied to the Nationalists’ debacle in the Chinese Civil War.²¹⁹

Stuart claimed that he was aware of Chiang’s psychology.²²⁰ If so, he also must have been aware that both major parties considered the third-party bloc’s intervention to be a problem. The question for the two warring sides was not whether the third-party group was able to broker a fair deal for them but whether they preferred to handle the politics in a more institutionalised, post-settlement environment. Tsou once hypothesised that the CCP would capture control of the government very quickly if the PCC resolutions were fully implemented.²²¹ However, this chapter shows that both the GMD and the CCP preferred a two-party civil war to a multiparty bargaining environment.

As Marshall presciently observed, ‘The state of peace appears at times more difficult than the state of war.’²²² The vigorous political manoeuvring of Chiang and Zhou virtually killed off the third-party group’s last-ditch mediation efforts in the end. Nevertheless, the third-party bloc’s quixotic but futile efforts for peace contained a scathing indictment of the two warring factions. Liang Shuming had given Marshall a little insight into what was going on: ‘Outside of the two principal contesting parties, all other parties and all the people of China unanimously wanted peace.’²²³

Previous studies have justified Zhou’s outbursts against the third-party coalition’s lack of concern over the CCP’s interests by rolling out an independent proposal.²²⁴ Admittedly, the third-party group’s failure to follow Zhou’s script was unacceptable to the CCP. To put things into context, however, the third-party bloc, like Marshall, was seeking ‘some quick way[s] to stop the fighting with a minimum of discussion’.²²⁵ For them, everything else, such as whose army should be in which position on what date or who would get one more seat in the State Council, had become less important. At a meeting held after Zhou’s outbursts, third-party alliance member Huang burst into tears when he talked about the suffering of the Chinese people in the civil war.²²⁶ Everyone cries for different reasons.

When the entire nation was galloping blindfolded down the path of an all-out civil war, the third-party group’s last-minute decision to join the peace talks raised the question of whether the two belligerent parties’ tirades against each other’s putative treaty violations were self-serving attempts to heighten the militancy of the negotiations. The increased competitive dynamic in the negotiations could only give the two veto players more excuses to intensify the war. The third-party bloc’s query about the nature of the row over the one-seat deficit in the State Council shed some light on the convenient rhetorical façade designed for China to stay the course of a two-party civil war. To date, the third-party coalition’s view on the polemic stands as an alternative to the long-running debate, with some pointing to the GMD’s hypersensitivity about maintaining absolute power and others laying blame on the CCP’s obstructive tactics.²²⁷ Hence Zhou’s confession to the minor party politicians about his obliviousness to the quota of the State Council is admirable, as he virtually admitted that the CCP used the disagreement as a tool.