4

A shattered peace

Ambiguous provisions in agreements for ceasefire, the Political Consultative Conference and army nationalisation

On 26 November 1945, Chiang sent letters of appointment to all delegates of the multiparty peace forum of the Political Consultative Conference (PCC) from the five different groups (as discussed in chapter 2), making it an eventful day in Chinese politics. Zhou Enlai was among the seven Chinese Communist appointees.¹ These appointments could not have come at a worse time. It was announced one day after Zhou returned to Yan’an and on the day Hurley resigned as the US ambassador to China. Hurley’s resignation sent a shockwave across the Pacific from the United States to China. In his resignation letter, Hurley fired a broadside at some employees of the US State Department, accusing them of siding with the CCP.² Truman reacted quickly. On 27 November, he appointed George C. Marshall, the retired army chief of staff, as his special envoy to China with ambassadorial rank.³ Truman hinted that he needed to send someone to China on a difficult mission to prevent civil war. Marshall was the best choice for the assignment owing to his personal qualities and unique experiences in high-level international negotiations during World War II.⁴

The instructions Marshall received from the State Department and Truman before the mission clearly demonstrate that both the Washington leadership and Marshall were aware that pushing for a representative democracy in China was a difficult and long-term challenge.⁵ Although Truman still raised the banner of promoting democracy in his China policy statement at the beginning of Marshall’s mission (which could have been construed as indirect intervention in China’s internal affairs),⁶ the two major tasks he asked his special envoy to accomplish in China were more conservative. Politically, Truman requested Marshall to persuade Chiang to call a national conference involving all major political elements. As discussed (see chapter 2), the Double Tenth Agreement provided for the convocation of the PCC. Despite the date of the PCC remaining undecided, representatives of both parties categorically reassured the US Embassy in Chongqing of their commitment to participate the conference.⁷ If the PCC was held successfully, it could settle outstanding issues about the convocation of the National Assembly and ultimately ‘bring about the unification of China’, as Truman wished.⁸

If the two parties reached a compromise and allowed the PCC to go ahead, Marshall would not need to intervene. Hence, mediating a cessation of hostilities for the two belligerent parties was the most urgent task at the onset of Marshall’s mission. As the Truman administration had specifically requested, a ceasefire in North China should be brokered before any other.⁹ John F. Melby was a US diplomat in Chongqing in 1945. He remembered that the US government at the time considered ‘a cessation of hostilities as the only condition in which the Chinese could work out their own problems’.¹⁰ The following sections will show that Marshall was determined to push a truce deal over the line first. This chapter studies the high politics of negotiation at the beginning of Marshall’s mission.

Marshall’s new dam

Marshall was sent to China to mediate the end of the Chinese Civil War, but direct dialogue between the two parties had actually resumed weeks before his arrival.¹¹ The parties had struck a deal to reopen negotiations in early December and had started to exchange views on a range of issues, including their commitment to a ceasefire.¹²

On 5 December in Yan’an, Zhou submitted a report to the party leadership on the progress of the negotiations. In the most revealing part of the report, Zhou admitted that it was anti-war sentiment, rather than the Communist revolution, that had gained most public support. He acknowledged that the CCP’s continued military campaigns were so unpopular that even the CCP’s sympathisers found them difficult to justify. The report also questioned the wisdom of CCP guerrilla attacks that had severed the Tangshan–Qinhuangdao coal train route, stopping coal shipments to Shanghai. The coal shortage had affected Shanghai’s expatriate communities.¹³ Zhou’s report suggests that the CCP shared the same degree of urgency as the GMD in relation to securing a ceasefire, regardless of Marshall’s presence. This conflicts with two popular arguments advanced by previous studies. First, it has been argued that the reopening of negotiations resulted from a false hope for creating peace through Marshall’s mission.¹⁴ Second, it is believed that the CCP’s short-term tactical manoeuvres in handling Marshall’s third-party intervention in the GMD–CCP conflicts were central to the history of this period.¹⁵

Marshall departed for China on 15 December,¹⁶ the day Truman delivered a statement on US policy towards China. While Truman reconfirmed US support for the GMD government, he asserted that US interests would be best served if the two opposing Chinese parties could resolve their differences via negotiation.¹⁷ On the same day, a Nationalist plane departed for Yan’an to collect the Chinese Communist PCC delegates.¹⁸ After returning to Chongqing, Zhou was conciliatory, stating that the CCP would seek a ceasefire deal with the government before the opening of the PCC.¹⁹

Marshall conferred with Chiang on 21 December. He had resolved to carry out Truman’s mandate. This authorised him to be blunt with Chiang and state that the US was likely to reduce its aid to the GMD government if it refused to forgo the pursuit of military solutions to internal strife.²⁰ On the other hand, Chiang attended the meeting with an intelligence report detailing confidential dialogue between Marshall and Truman. This noted that the United States was committed to back his government even if he refused to make reasonable concessions in the peace talks.²¹ Although the record of the conversation also demonstrates Truman’s position on non-involvement in China’s civil strife,²² it has become an often-quoted exposition of Chiang’s tough stance.²³

During the meeting, Marshall pleaded with Chiang to take a conciliatory approach. He maintained that the American people wanted the two Chinese factions to make major concessions and terminate military conflict. However, Chiang told Marshall that the only way to force the CCP to settle the situation via peaceful means was to send the bulk of his forces to secure North China. This would make it possible for him to accomplish China’s unification. The subtext of Chiang’s statement was that Marshall’s calls for peace would ring hollow unless US military assistance to the GMD was increased to a level that enabled Chiang’s forces to quash the Communist rebellion.²⁴

Marshall held his first conference with Zhou on 23 December. Zhou presented a string of demands, which he claimed would ‘guarantee’ the CCP’s support for a truce. These included the CCP’s participation in accepting the Japanese surrender, a ban on the GMD’s use of puppet and Japanese forces in civil strife, and the establishment of a power-sharing coalition government ‘right now’. Although Zhou presented his conditions for peace to Marshall with a heavy dose of euphemism, he hinted that these conditions must be fulfilled to avoid ‘a resumption of the fighting’.²⁵

Chiang and Zhou’s tough bargaining stance certainly constituted an obstacle to a ceasefire, but the negotiations were still moving towards a resumption. On 27 December, the two parties began to negotiate a truce proposal, but the meetings did not go well.²⁶ Marshall started to exert his influence on the negotiations. He held talks with Chiang and pledged to attend the GMD–CCP peace talks in person, rendering the negotiations tripartite.²⁷ Both parties struck a preliminary consensus after Marshall’s intervention. The accord declared a nationwide ceasefire, but essentially assigned the task of negotiation to the representatives of the two parties and ‘General Marshall’.²⁸ As events developed, Marshall went on to chair a special Committee of Three on the ceasefire agreement and its implementation in which the two opposing sides were equally represented. On 31 December, the GMD government officially approved the PCC being held on 10 January 1946.²⁹

Marshall prepared a US version of the ceasefire plan despite the two parties having reached a preliminary agreement.³⁰ The most outstanding feature of the Marshall truce was that it would grant the GMD the right to undertake troop movement in Manchuria. Marshall showed his draft to Zhou and stressed that the US government was committed to the movement of GMD troops into Manchuria.³¹ While Zhou expressed his complete understanding of Marshall’s position, he indicated that he wanted to think it over. In other words, Zhou agreed to let Marshall mediate his party’s conflict against the GMD, notwithstanding the American’s biased mediation position in favour of his Nationalist opponents. Marshall admitted to Zhou that the agreements reached by both parties were all ‘water over the dam’ now. He nevertheless insisted that the Americans had found a better plan. ‘I have a new dam’, Marshall intoned.³²

Marshall’s truce

Marshall embarked on small-scale ‘shuttle diplomacy’ on 5 January. He presented his truce plan to Chiang. Although Chiang had reservations about certain conditions in the proposal, he found the execution of the truce in general, and the terms about troop movement in Manchuria in particular, rather persuasive.³³ From the GMD’s perspective, Marshall had demonstrated commitment to strive for a satisfactory deal for his ally; he was clearly more supportive of sending GMD armies to Manchuria than Wedemeyer.

After presenting a plan with creative new settlement options to Chiang, Marshall conferred with Zhou. The meeting centred on the Nationalists’ freedom of troop movement in Manchuria after the ceasefire deal was in effect. Zhou had received a directive from Yan’an on 2 January that all troop movements after the ceasefire should be undertaken through consultation because the CCP forces needed freedom for military movements.³⁴ Zhou nevertheless knew that he had to wait for the right moment to raise this. He told Marshall the CCP did not reject GMD troop movement in Manchuria but that it was unnecessary to include words to this effect in the ceasefire declaration. However, Marshall was insistent, arguing that if there was a joint declaration, it needed to be ‘businesslike and exact’ and that the announcement must address the movement of troops into Manchuria. ‘I am not the negotiator … I am the demander’, he stated.³⁵

Given that Marshall was working to secure a ceasefire deal before the inaugural session of the PCC commenced on 10 January, the savagely tight deadline seemed to warrant his get-tough approach. It forced the two belligerent parties to accept a given solution to avoid prolonged disputes. Nevertheless, while Marshall assumed control of the negotiations, his involvement inevitably shifted from negotiation and mediation, becoming something closer to arbitration. This meant that the two parties would let Marshall take greater control of the negotiations. The main drawback of abdicating control over the negotiation process to a third party is that it is likely to prompt the disputants to believe that they have also lost control of the negotiation outcome, making them increasingly fearful of losing.³⁶ When disputants are trapped between the hope of winning and the fear of losing, they are more likely to adopt hard-line positions. The unfolding of events around the negotiation shows that while the two disputed parties ostensibly allowed Marshall to control the discussion, pace and flow at the conference table,³⁷ they did not let the American manipulate them into accepting a less ambitious negotiation stance.

Meanwhile, Chiang was preparing to impose more conditions on the agreement and take an unyielding stand on matters. If Chiang had instead decided to compromise on some points and the situation went badly afterwards, he could always lay the blame on Marshall. He sent Zhang Qun to see Marshall on 6 January with a number of new requests. The most controversial demand was that the province of Rehe be considered part of Manchuria in the ceasefire accord and that the Nationalist forces had the right to take over the entire province from the withdrawing Soviet forces. In particular, Marshall was informed that the GMD had reached an agreement with the Soviets for the takeover of the two strategic towns, Chifeng and Duolun, in the provinces of Rehe and Chahar, respectively. Marshall urged Zhang to compromise. He made it abundantly clear that he would not discuss Chiang’s new demands with Zhou because the CCP would definitely oppose it.³⁸

Marshall’s pleas went unheard. Zhang and Zhou held six Committee of Three meetings between 7 and 10 January. Marshall attended all the meetings as a mediator, only to discover that the two parties had turned his mediation sessions, aimed at finalising a ceasefire order, into a political tug of war. Zhou rejected Zhang’s position on Chifeng and Duolun. Zhou supported freedom of movement for the Nationalist armies in Manchuria, but insisted that this should be written in the meeting minutes rather than the official ceasefire announcement. In turn, Zhang proposed a horse trade: if Zhou accepted the GMD’s package, including the occupation of Chifeng and Duolun, he would accept Zhou’s view that certain exceptions should be recorded in the minutes. Zhang, however, added that the meeting minutes on these specific areas should also be published.³⁹

Given that Chiang had repeatedly instructed Zhang that the takeover of Chifeng and Duolun was a high military priority for the GMD,⁴⁰ Zhang was presenting his concession as a small package, tailor-made for the Nationalists. Here, Zhang had employed a packaging concessions tactic: proposing a trade-off settlement via offering concessions on low-priority issues for major gains on issues of higher priority. In general, the key to making this tactic successful is that its users must create win–win options for their opponents. In particular, its users must ensure that the item offered as a quid pro quo is high on the other party’s agenda.⁴¹ In the GMD’s case, a caveat of Zhang’s approach was that if Zhou considered the proposed package in a negative light, a positive joint outcome would not be achievable.

Zhou was not interested in the offer at all. He steadfastly denied Zhang’s demand over Chifeng and Duolun. His argument was supported by two important claims. First, Zhou declared that the two towns had been overtaken by CCP forces. As the sovereignty of these towns had already ceded to Chinese forces, the presence of GMD armies was unnecessary. According to Zhou, the movement of non-Communist troops into these places would only make the ceasefire order difficult to implement. Second, Zhou challenged his GMD opponents to prove the existence of the alleged GMD–Soviet accord that authorised the takeover of these towns.⁴² If the Nationalists wished to prove that their claim was true, Zhou suggested, Soviet representatives should be invited to attend the negotiations.⁴³ To Zhang’s dismay, he proposed that GMD military movements in Manchuria would be allowed only after consultation with the CCP.⁴⁴

Marshall tried almost everything, including suspending meetings and cutting off Zhou’s tirade,⁴⁵ to prevent the negotiations from descending into political slanging matches, but he soon exhausted his repertoire. Negotiations had made little headway when they approached a desperate late stage on 9 January. While Zhang still refused to change his stance over questions concerning the GMD’s takeover of Chifeng and Duolun, Zhou derided Zhang’s demand as ‘in principle unreasonable’.⁴⁶

Marshall was left with no choice but to see Chiang in person. He pleaded with Chiang to prioritise the ceasefire and noted that it would be better to drop the demand regarding Chifeng and Duolun.⁴⁷ Chiang did not have the authority to tough it out as he was addressing the inaugural ceremony of the PCC the following morning. It would have been a political setback for Chiang if he were unable to announce the agreement to cease hostilities in his speech. Chiang therefore grudgingly agreed to drop his demands in relation to the provinces of Rehe and Chahar.⁴⁸ Almost simultaneously, the CCP interlocutors made last-minute appeals to the Soviet ambassador for Moscow’s direct intervention in the negotiations, but their request was snubbed.⁴⁹ As Zhou told Marshall, ‘There are some similarities in theoretical aspects between the Chinese Communist Party and the Soviet Union, but these two are definitely of two nations.’⁵⁰

The trio held an urgent meeting on the morning of 10 January, only a few hours before the PCC started. The concession Chiang had made the night before gave Zhou an advantage in the arduous negotiations. After receiving confirmation about Chiang’s concession, Zhou told Marshall and Zhang that he accepted the Nationalist military movements ‘into or within’ Manchuria, provided that the movements were reported daily to the truce monitoring headquarters. In the case that the GMD forces wished to move troops to Manchuria through Communist-occupied areas, Zhou proposed an understanding from both parties. This did not need to be written into the official agreement, and he noted that such movements would be discussed with the CCP beforehand. As his time was limited, Zhang immediately accepted Zhou’s proposal. With two of the most contentious issues resolved, the trio agreed to release the order for the cessation of hostilities to the press and return for another meeting in the afternoon.⁵¹

During the afternoon meeting, the trio reached an agreement on issues related to the establishment of a small Executive Headquarters in Beiping to implement the ceasefire order under the leadership of the Committee of Three. Three commissioners of the headquarters representing the GMD, the CCP and the United States would be appointed to take charge of the headquarters with authority to vote and negotiate among themselves. However, unanimous agreement by the three commissioners was required for all actions of the headquarters. The headquarters’ personnel comprised an equal number of GMD and CCP members, along with a similar number of American staff. Contingents of the headquarters (known as field teams) would be sent to supervise the ceasefire at conflict hot spots. The team delegates, who were also equally represented and accompanied by interpreters and communication officers, would conduct investigations on the ceasefire status and report to the headquarters; before making reports, the team had to reach a unanimous agreement.⁵²

Later, the headquarters was assigned to supervise matters concerning the reopening of transportation lines and railway reconstruction, as part of the Restoration of Communications Agreement signed by the Committee of Three on 9 February.⁵³ This was a difficult task, as the agreement’s enforceability was limited by several ambiguous terms. The headquarters commissioners needed to contemplate further negotiations to resolve the problems. One outstanding example was that the agreement provided for the removal of structures (e.g. fortifications) that obstructed the transportation lines, but this begged the question of whether certain structures actually provided safe passage over the railways and roads.⁵⁴

As a result of private talks between Zhang and Zhou, both parties agreed to a grace period of three days, allowing the opposing armies to receive the ceasefire order. In other words, all military action would be terminated by 14 January.⁵⁵ As the following discussion reveals, this arrangement provided the two belligerent parties with a reason to continue fighting for a few more days after the ceasefire order had been issued.

In summary, the GMD and the CCP agreed upon an order for the cessation of hostilities in a joint declaration on 10 January. This provided a nationwide ceasefire, but it was also the result of an inescapable trade-off between the two parties in Manchuria and its neighbouring provinces. The declaration granted an exception permitting the GMD to undertake military movements in certain areas, especially Manchuria. In return, the advance of the Nationalist troops in Rehe and Chahar was halted. As on 9 January, the Yan’an leadership had expected that the GMD forces would reach the town of Chifeng within days; this was a success that the CCP could not have achieved by military means alone.⁵⁶

On the Manchurian question, the two parties had reached an understanding before the signing of the truce that the GMD would not send a large number of troops into Manchuria.⁵⁷ The understanding was hardly enough to compensate for the ambiguous provisions of the ceasefire agreement. The newly appointed three commissioners—Ye Jianying, the chief of staff of the CCP 18th Group Army; Zheng Jiemin, a prominent intelligence chief of the GMD; and Walter S. Robertson, the US chargé d´affaires—were forced to admit at a joint press conference on 18 January that they had no jurisdiction over Manchuria. Marshall was furious. He questioned Robertson, asking: ‘Who do the Commissioners assume will act to settle any sporadic or serious fighting or differences between Chinese factions in Manchuria?’⁵⁸ Marshall should perhaps have asked himself this question, because he had brokered a ceasefire deal for the two sides that almost invited them to keep pursuing a military resolution in Manchuria.

Jockeying for final position

The ceasefire was designed to create a peaceful ambience for the PCC’s inauguration on 10 January. In a directive, Mao described the truce period as a ‘new stage of democratic construction’.⁵⁹ In reality, however, both parties pushed their armed forces ahead nationwide before the ceasefire came into effect, leading to fierce fighting. As the newly appointed US director of operations of the Executive Headquarters, Henry A. Byroade, reported, both parties were ‘jockeying for final positions’.⁶⁰

On 10 January, Chiang Kai-shek instructed Du to secure positions within striking distance of Chifeng, stating that the ceasefire would not come into effect until ‘three or five days’ after its official announcement. In other words, Chiang was granting latitude for Du to attack beyond the legal limit of midnight on 13 January.⁶¹ Nevertheless, Du’s forces aborted the assault near Chifeng on 13 January.⁶² In general, however, eyewitness reports have confirmed that the Nationalists were the habitual offenders in violating the ceasefire order,⁶³ although the 13A commander, Shi, insisted that the CCP violated the truce agreement much more than they did.⁶⁴

On 14 January, Zhou filed a complaint to Marshall about alleged Nationalist ceasefire violations in various contested areas, but it was the GMD troop advancement in the province of Rehe that was the focus of Zhou’s angst. He insisted that if the Nationalists took Chifeng, his party would not consider it a minor incident. Zhou then called for an urgent deployment of ceasefire field teams and reconnaissance planes to the troubled province.⁶⁵ The Americans acted as Zhou requested. A US Marine Corps plane was sent to Chifeng between 14 and 15 January. The aircraft landed safely on the town’s airfield, but the pilot was temporarily interned by Soviet troops for an identification check. In subsequent contacts, the Russians gave assurances that the field team’s plane was welcome to land there.⁶⁶

As the incident was a result of Soviet military presence at Chifeng, it debunked Zhou’s claim that the CCP forces had restored sovereignty there, although it was reported that a group of CCP-aligned militia had entered the town in mid-August and remained active.⁶⁷ From a pro-Nationalist perspective, the GMD government had the right to take the town from the Soviets.⁶⁸ Nevertheless, the Sino-Soviet Treaty did not consider Rehe as a Manchurian province,⁶⁹ and therefore the town was covered by the ceasefire agreement. If Du advanced his forces to occupy the town, Zhou would surely have lodged a complaint against the Nationalists for ceasefire violation.⁷⁰

While the controversy lingered in Chongqing, the pro-CCP local militiamen had no qualms about challenging the Nationalists’ authority. They took control of Chifeng after the Soviet withdrawal finally occurred on 23 January and refused to leave.⁷¹ Commissioner Ye of the CCP began to back away from Zhou’s initial proposal of sending a team to Chifeng after Soviet military presence in the town was confirmed. Ye insisted that the team for Chifeng postpone its departure until his own qualified men arrived from Zhangjiakou (formally Kalgan, the CCP’s North China headquarters). When these men arrived, they turned out to be bodyguards and technicians.⁷² Ultimately, Marshall had to plead with Zhou to press Ye to allow field teams to enter conflict areas.⁷³

After existing in a state of uncertainty, the Chifeng field team was among the first to finally receive approval to depart for conflict spots from 17 January.⁷⁴ It managed to bring the volatile military situation under control after some effort,⁷⁵ but this was achieved at the expense of the unanimity principle. As the CCP militiamen were already stationed in the town, the field team reached a unanimous agreement for GMD troops to move into the suburbs.⁷⁶ The CCP militiamen were to withdraw from the security cordon on high ground outside the town.⁷⁷ This was a concession that Zhou would have liked to retract because it would allow Du’s troops to besiege the CCP militia at Chifeng. His attempt to recant the agreement was unorthodox: the CCP member of the Chifeng team began to absent himself and refused responsibility.⁷⁸

On the other hand, Zhou quickly re-established the vertical decision-making hierarchy within his team. In his telegram to Ye, Zhou issued a sobering reminder of the importance of reporting problems in the negotiations to a higher level. Despite this, he acknowledged his men’s efforts to avoid negotiation breakdowns at the lower level.⁷⁹ The CCP representatives soon understood how to use their veto power over team decisions. The right of veto enabled them to delay sending field teams, refuse responsibility and retract agreements simply by replacing team representatives.⁸⁰ When uncooperative and inappropriate behaviour began to spread, the first casualty was the field team morale.⁸¹ The memoirs of the Executive Headquarters veterans recorded cases of the US and CCP field team members’ prostitution scandals and indecent behaviour towards female civilians.⁸²

While ceasefire implementation remained unconvincing at some contested areas in China proper, a ‘powder keg’ exploded in Manchuria during the truce period. After receiving intelligence reports of a possible GMD troop advance in Rehe, Liu instructed Lin to wage an immediate one-week offensive (starting 13 January) to reopen the communication line between Manchuria and Rehe. Liu’s directive was apparently founded in a belief that a civil war in Manchuria was inevitable and that the option Liu gave Lin Biao was not a choice between war and peace but between waiting for the attack and pre-empting it.⁸³

On 4 January, the NEPAA had been renamed the North-East Democratic United Army (NEDUA), showing the party’s determination to achieve military control over Manchuria. On the day the ceasefire agreement was implemented, a strong NEDUA contingent attacked the isolated GMD battalion at the port of Yingkou and eliminated the entire battalion. According to the CCP’s account, the assault did not end until 15 January, beyond the ceasefire date, although the city was captured on the evening of the 13th.⁸⁴ In the wake of this incident, Marshall called for a team to be sent immediately to Yingkou. Chiang baulked at the idea, citing potential conflicts with the Soviets as an excuse.⁸⁵

The CCP seemed to believe that a peace settlement would soon materialise, but the question was how much power it could obtain from a coalition government-in-waiting. A large-scale military offensive in Manchuria was therefore deemed necessary to achieve the CCP’s ambition. Liu knew that waging war while talking peace was against the people’s wishes.⁸⁶ However, he expected that the war in Manchuria would be short-lived, because once fighting broke out, Marshall would immediately plead for a truce. The peace settlement to follow would end any chance for military action. Liu coined his call for war ‘the final battle in this historical new stage’⁸⁷ and stated that this would still be a positive result even if the battle ended in a draw.⁸⁸

Lin faced a significant task in answering Liu’s call for war. Boosted by the arrival of the elite New Sixth Army (N6A) in early February, the Nationalists continued their advance in Manchuria. Most of Lin’s divisions retreated after presenting a weak resistance.⁸⁹ While the GMD armies kept pushing forward without submitting reports of troop movements to the Executive Headquarters (as the ceasefire agreement specifically provided), a vicious spiral ensued.⁹⁰ On 13 February, NEDUA’s fast-moving foot soldiers encircled an isolated regiment of the 13A at a village north of Shenyang. As the Nationalists’ US long-range artillery was ineffective in close-range engagement against the NEDUA, Lin’s men annihilated the entire GMD regiment.⁹¹ The GMD officers learnt this lesson (about NEDUA’s soft siege warfare tactics) from their defeat. On 15 February, a superior NEDUA force ambushed a regiment of the N6A at another village south of Shenyang. The GMD’s lightly armed relief column was able to sneak behind the NEDUA line and launch a surprise attack, inflicting heavy casualties on Lin’s troops.⁹² As both parties received a drubbing in the fierce fighting, this was deemed a draw, as Liu had wished.

The PCC resolutions: A Machiavellian manipulative device?

Thus the multiparty forum of the PCC was convened concurrently with a shaky truce. Chiang was worried about whether the PCC could make satisfactory progress. He requested Marshall’s comments regarding the PCC on 21 January. In his reply, Marshall hinted that he was mainly concerned with military issues but not the political debate at the PCC.⁹³ Nevertheless, Marshall drafted a brief paper to Chiang, outlining his personal opinion concerning the organisation of a multiparty interim Chinese government.⁹⁴ Chiang considered Marshall’s plan outrageous because, if all went according to Marshall’s plan, the GMD government’s authority to promulgate acts in relation to the CCP-controlled territories would be greatly limited.⁹⁵

However, a recent study claims that Marshall should be credited for his contributions to the subsequent political resolutions at the PCC because Chiang repackaged Marshall’s ideas and introduced them for debate at the PCC.⁹⁶ Participant parties’ proposals recorded in the existing PCC documents, and Marshall’s original paper handed to Chiang on 22 January, nonetheless suggest that this claim ignores three important facts.

First, the key components of one of Marshall’s two ideas in his paper—what he coined the ‘Interim National Council’—did not include any major differences from what the participants had already discussed and agreed to at the PCC meetings held well before Marshall submitted his proposal to Chiang. Second, the most innovative idea in Marshall’s paper—limiting the central government’s power to enact laws affecting the administration of local governments—was too radical even for the Chinese Communists, who had been advocating that provincial governments have their own constitution.⁹⁷

Existing PCC meeting documents do not show any proposal from all participant parties regarding the power of local governments that can be related to Marshall’s original idea. It is not surprising then that the final PCC agreement on central–local government relations endorsed by all participant parties was the exact opposite of Marshall’s idea: local government-enacted regulations must not contravene the laws and decrees of the central government.⁹⁸ In addition, a letter from Marshall to Truman shows that Marshall had been ignorant about the PCC debates. This was partly because he was busy getting the Executive Headquarters to work, but mostly it was because he maintained a calculated silence when the PCC took centre stage, in order to avoid accusations of political meddling.⁹⁹

While Marshall considered his next move carefully, the PCC delegates had needed to find resolutions without the mediator’s participation since the opening of the PCC on 10 January. The multiparty conference, which Chiang chaired, was convened in both public and closed sessions. The public sessions discussed five fundamental themes that had emerged from the Chongqing negotiations, including reorganisation of the government and armies, the National Assembly and the Draft Constitution. The views and recommendations presented at the public sessions upon each of the five subjects would be referred to five specialised subcommittees, which met in closed sessions, for the drafting of specific resolutions. Among these five subcommittees, the Government Organisation, Draft Constitution (later taken over by the Committee for the Reviewing of the Draft Constitution, CRDC, or xiancao shenyi weiyuanhui 憲草審議委員會) and Military Affairs Subcommittees are most relevant to the following discussion. A Steering Subcommittee was established to oversee the work of the five subcommittees.¹⁰⁰ These subcommittees played a crucial role during the PCC, and some continued meeting to discuss unresolved problems after the conference had ended.

At its closing session on 31 January, the PCC adopted resolutions as the result of agreement on the topics established by the conference.¹⁰¹ In essence, the PCC resolutions strove to set an agenda for a power-sharing government that would be legitimised by a constitution accepted by all parties. To put the resolutions into historical context, they were a set of guidelines for post-conflict management vis-à-vis the two civil war rivals: the GMD and the CCP. For the CCP, there could be only one acceptable interpretation of these guidelines, which stipulated that the resolutions must be implemented in two discernible stages. In the first stage, the one-party-ruled GMD regime must be reorganised into a power-sharing government and the CRDC would reach a consensus on the Draft Constitution. The new government would then summon a session of the National Assembly to adopt the constitution.¹⁰² As these two phases were not negotiable from the CCP’s perspective, any attempt to defy the two-phase procedure constituted a violation of the resolutions.

Nevertheless, the PCC resolutions, like many other peace treaties in history, were written ambiguously, emerging from a series of compromises over the conflicting interests of the disputed parties. The ambiguities lay in one open-ended provision concerning the all-important State Council (Guomin zhengfu weiyuanhui 國民政府委員會), the deadline for the completion of the final draft of the constitution and the date for the official opening of the National Assembly. These three elements constituted a ‘cross-textual ambiguity’ that could be open to incompatible interpretations of the resolutions.¹⁰³

The PCC resolutions enabled the establishment of the State Council as the highest level of government, with both legislative and administrative power preparatory to the inauguration of constitutional rule. To place the State Council’s authority in perspective, it would be a multiparty organisation that potentially had the power to settle matters that remained unresolved in the Double Tenth Agreement between the two major parties and make important decisions to reform the existing government.¹⁰⁴ Conversely, it would be more difficult for the two major parties to justify their use of force to settle disputes once the State Council was fully functioning. In short, the State Council was a significant first step towards government reorganisation.

The relevant provision of the resolutions stated that half of the forty state councillors would be Nationalists. Members of other parties and non-partisan candidates would comprise the other half. However, the provision did not stipulate how non-Nationalist state councillor positions (i.e. twenty seats) would be allocated, particularly the exact number of seats the CCP could obtain. Rather, the provision was written with an open-ended structure, whereby the twenty undecided seats would be ‘the subject of separate discussion’.¹⁰⁵ While the provision concerning the State Council specified that the seat allocation was still unsettled, the resolutions also stated that the CRDC would have only two months to revise the Draft Constitution—initially promulgated by the Nationalist government in 1936—so it could be passed in the National Assembly, which the resolutions had scheduled to open officially on 5 May 1946.¹⁰⁶

The ambiguities increased the likelihood of the worst-case scenario: revision of the Draft Constitution would be completed before 5 May, but the governmental reorganisation would be delayed because the undecided State Council seats remained unsettled. The Chinese Communists could then insist on their two-stage theory, arguing that the reorganisation of the existing government must precede the passing of the constitution. Therefore the opening of the National Assembly would be postponed until disputes over the State Council were resolved.¹⁰⁷

The Nationalists, who preferred to maintain the status quo, could argue otherwise. This was because the resolutions had expressly stipulated the time frame for completing the revision of the constitution and the exact date of the National Assembly inauguration to approve the constitution. From the GMD’s perspective, these two provisions were to be considered retroactively specified, meaning that all parties involved must agree on the representation of the State Council before 5 May. Even if this deadline was extended owing to unforeseen circumstances, the terms and spirit of the resolutions had ruled out the possibility of the matter remaining unresolved indefinitely. If the parties involved failed to strike a deal over the membership of councillors, the GMD government could argue for calling the National Assembly ahead of government reorganisation in order to break the deadlock; in this way, all unresolved disputes could be settled via the constitutional assembly.¹⁰⁸

If the GMD acted unilaterally, the CCP and its ally, the Democratic League, would be likely to protest and refuse to attend the National Assembly. The ramifications would then go far beyond the issue of treaty compliance. The frightening reality was that the Nationalists could still have the constitution passed in the absence of a significant number of delegates in the National Assembly. The resolutions stipulated that the Assembly could convene with at least more than half of its delegates (i.e. 1,026 out of 2,050) and could make decisions with three-quarters of those present. Given that the resolutions provided for the GMD to retain 1,200 old delegates, the Nationalists had the power to convene the Assembly.¹⁰⁹ The adoption of the constitution without the CCP’s support would inevitably be developed into a national split.

Notwithstanding the ambiguous provisions, the PCC resolutions were not an agreement that Dražen Pehar would have considered a ‘Machiavellian manipulative device’—a treaty that brought nothing but false hope, generating misunderstandings and worsening the already bad relationship between disputants.¹¹⁰ The parties who adopted the resolutions seemed to find that managing the ambiguities at their core was key to understanding one of the provisions about forcing bills involving changes in ‘administrative policy’ (i.e. important policy changes) through the State Council, which was more conflict prone than others. The provision in question required a two-thirds supermajority vote by the councillors present to be passed.¹¹¹ This created potential points of contention. The CCP and the Democratic League insisted upon having fourteen or more votes (i.e. more than one-third of seats) in the council to secure veto power. The GMD, on the other hand, would try to prevent other parties from reaching that threshold number.

However, most PCC delegates preferred to leave the distribution of the State Council seats open to future negotiations, citing constraints over the discussion deadlines dictated by the adjournment of the PCC.¹¹² Instead, they devised a few provisions, which were written into the resolutions, to make the potential dispute over the State Councillor seats a non-issue. One of these specified whether a given act involving changes in ‘administrative policy’ was to be determined by a simple majority vote of the state councillors present.¹¹³ Given that the Nationalists had already obtained twenty votes, they were likely to win over at least one more vote from the anti-Communist Youth Party and override vetoes consistently. Put simply, a symbolic veto power was all the CCP could achieve, even if it could secure fourteen votes.¹¹⁴ Conversely, if the Nationalists intended to pass a bill that even their political ally deemed unacceptable, all of the twenty non-GMD councillors could combine to vote it down. In other words, the veto power would not be the prerogative of the CCP–Democratic League bloc,¹¹⁵ but the GMD would still hold a dominant position even if it agreed to let the CCP enjoy a nominal veto power. Thus the provision would have substantially reduced the competitive dynamics of the distribution of seats to the council, unless more than one signatory party intended to exploit the ambiguities of the resolutions for partisan purposes.

On the other hand, the provisions stipulated that the president of the National government had the power to nominate any non-partisan politicians as state councillors, but if the nomination was opposed by one-third of the other nominees, the president must make a different nomination.¹¹⁶ Given that the Nationalists later accepted the formula of eight (CCP), four (Democratic League), four (the Youth Party), four (independent) regarding the twenty non-GMD state councillors,¹¹⁷ the stipulation gave the CCP–Democratic League coalition enough votes (i.e. twelve out of thirty-six) to ensure that the four non-partisan councillors would be likely to vote with them rather than the Nationalists. This meant that even if the CCP–Democratic League bloc was two seats short of the required fourteen needed for veto-wielding power, the bloc could still easily gain support from at least two non-partisan councillors and obtain the votes it demanded.¹¹⁸ The ‘inconvenient truth’ here would be that the CCP would have the added complication of needing to negotiate not only with the Democratic League but also with the four independent politicians who might have a wide variety of policy platforms.

In summary, the resolutions concocted by the PCC showed signs of a compromise deal written as a rather rudimentary outline. Simultaneously, the resolutions also included terms and conditions that could minimise the competitive dynamics over certain ambiguous provisions. The only caveat was that the treaty parties must continue to expand interparty cooperation in the post-treaty stage, as the resolutions would not have any genuine value otherwise.

The resolutions adopted by the PCC were good enough to fuel hopes for a permanent peace in China. On 1 February, the Yan’an leadership instructed its forces to terminate military action, declaring a sharp turn from armed insurgency to non-violent popular and parliamentary-based resistance.¹¹⁹ This was perhaps the first time since the signing of the truce that negotiators and mediators from all parties started to believe they were close to achieving their objectives.

A deal is better than no deal: The army reorganisation agreement

By the end of February, the two belligerent parties struck an agreement to reorganise all Chinese armies and integrate the CCP forces into the National Army.¹²⁰ This was an agreement that both parties were obliged to negotiate as the PCC resolutions stipulated. For the CCP, entering into a negotiation with the GMD on equal terms in relation to army reorganisation as a legitimate Chinese armed force might be considered an initial success.¹²¹ The increasing appearance of ‘legitimacy’ could be used to intimidate opponents in negotiations.¹²² The Yan’an leadership saw the army integration negotiation as a worthy experiment, as long as its armed forces were still firmly under the party’s control and the reorganisation timetable was consonant with the CCP’s interests.¹²³

The Nationalists entered the negotiation with an entirely different state of mind. Although the PCC resolutions provided for the GMD army to conduct its existing reorganisation program without scrutiny,¹²⁴ the GMD interlocutors were forced to admit that a joint reorganisation program for all Chinese armed forces was a superior plan.¹²⁵ The Nationalists needed to transform themselves into genuine reformists in the eyes of the Americans, ensuring their appeal in relation to receiving US military aid. More importantly, the GMD urgently needed an army nationalisation deal with the Chinese Communists if they were to have any chance to integrate the opposing armies into a single military system.¹²⁶ Wedemeyer’s planning staff, who had been involved in the army reorganisation negotiations, had pointed out nonetheless that the army integration program required good faith and cooperation from both parties. If either side withheld armed forces to perpetuate its own interests, the plan was doomed to fail.¹²⁷

The United States could not avoid becoming involved in the negotiations for China’s army nationalisation, given the strong military ties between the two nations. However, Marshall did not sit on the PCC Subcommittee of Military Affairs. In mid-January, the GMD and the CCP reached a consensus, inviting Marshall to act as an ‘adviser’ in an exclusive tripartite, three-member Military Subcommittee to develop a plan for army nationalisation with Zhou, Zhang Qun or Zhang Zhizhong.¹²⁸

Marshall entered the negotiation with a new plan to motivate the CCP. He offered an American-sponsored training program for selected CCP army officers from those forces to be integrated.¹²⁹ This transitional program would start with basic training with the prospect of establishing special schools for infantry, artillery and armour in Zhangjiakou. The Americans would provide instructors and equipment enough to train ten CCP army divisions.¹³⁰ The CCP quickly accepted the offer.¹³¹ The military assistance to the CCP gave Marshall the opportunity to gain the leverage to influence the Chinese Communists in the negotiation.

From late January to early February, Marshall talked to the two parties separately on several occasions. Zhou and the two Zhangs also held meetings for an exchange of ideas without Marshall.¹³² These get-togethers, and the multiparty talks in the Military Affairs Subcommittee of the PCC,¹³³ helped the two major parties understand each other’s stance. Marshall arranged seven formal, three-party meetings in quick succession from 14 February.¹³⁴

On 25 February, the negotiations concluded with an agreement. The Executive Headquarters acted as an agency to implement the agreement. The essence of the agreement (titled ‘Basis for Military Reorganisation and for the Integration of the Communist Forces into the National Army’) shows that it was consistent with the CCP’s interests in a number of critical areas. First, the agreement overwhelmingly allowed the CCP to retain control of its combat units. A CCP army consisting of three infantry divisions would be under the full control of a CCP army commander in the first twelve months after promulgation of the agreement. The only major difference was that this army unit would be assigned into an ‘army group’ together with a GMD army.¹³⁵

Although further integration of the two opposing forces would be carried out in the six months after the end of the first year, all CCP infantry divisions, except those being demobilised, would still be separated from their GMD counterparts at the end of the program. The only notable change was that a CCP division and its GMD equivalent would find themselves in the same chain of command under a designated army unit commander of either a CCP or a GMD officer. In other words, the CCP combat units maintained a quasi-independent status up to the divisional level at the end of the integration program. This was despite the fact that it would appear they had merged with the GMD forces in the organisational chart.¹³⁶ A Nationalists’ proposal to mix both parties’ troops in the same infantry division, termed ‘fusion’, received no support and was quickly shelved before the agreement was finalised.¹³⁷ The agreement stipulated that the CCP could retain ten all-Communist divisions alongside fifty all-Nationalist divisions (each consisting of no more than 14,000 men) in the eighteen months after promulgation of the reorganisation. No provision was made for further integration.¹³⁸ This meant that the CCP would possess a strong, streamlined and concentrated combat force capable of rejecting peace treaties unilaterally and fighting again, even if the agreement was fully implemented.

The time frame of the army integration program also coincided with the interests of the CCP. The initial stage of the integration would not commence until the seventh month after the accord was signed. It was agreed that the overall force-to-force ratio of the GMD armies and CCP forces would be maintained at a ratio of 5:1 nationally. In North China, where the majority of CCP troops were based, the troop ratio between the opposing forces was much closer. The agreement allocated a seven-division quota to the CCP against eleven GMD divisions in that strategically important region eighteen months after the signing of the agreement.¹³⁹

CCP leaders in Yan’an concluded that the US-proposed army reorganisation plan would damage the structure of powerful military factions within the GMD, making the agreement more valuable to the CCP.¹⁴⁰ Perhaps the only drawback for the CCP in accepting the agreement was that it agreed to kick-start the reorganisation program by providing vital information about its forces. This included, among other things, a full list of all combat units currently under its control, including detailed information (e.g. strength, unit locations and designations), within three weeks of the agreement being signed.¹⁴¹ Clearly, this commitment could have led the CCP leaders to undertake actions they would rather not. They decided, even before the agreement was signed, to re-designate half of their regular forces as local militia units and concealed the whereabouts of one-third of the military cadre personnel in order to avoid the army reorganisation program.¹⁴² As Zhou later admitted, the CCP was confronting its ultimate fear of losing military power.¹⁴³

Previous studies of the 1946 military reorganisation negotiation consider the GMD an outright loser. According to one study, Chiang was compelled to accept the agreement mainly because the CCP had decided to endorse it. He was quoted as stating, ‘[This deal was] the biggest loss of our government.’¹⁴⁴ However, Chiang also noted that obtaining a deal was preferable to the status quo.¹⁴⁵ In other words, Chiang considered that the army reorganisation agreement did not go beyond his resistance point, even though the deal was not optimal. The problem was whether the agreement was effective enough for the CCP to partially uphold its end of the bargain. Chiang, meanwhile, received the US aid he wanted. Marshall began to arrange a loan worth US$50 million to the GMD government after the signing of the army reorganisation deal.¹⁴⁶

How much is too much?

The situation took another odd turn in mid-March. The PCC met as a consultative body, but it had no legal authority. Although all the parties represented were morally obliged to accept the PCC resolutions, the legal status of the resolutions was subject to recognition from the respective leaderships of the signatory parties.¹⁴⁷ Right-wing elements within the GMD resorted to activism as means of expressing resentment against the PCC.¹⁴⁸ The Second Plenary Session of the GMD’s Sixth Central Executive Committee (CEC) met at Chongqing from 1 to 17 March to endorse the PCC resolutions. The congress was convened in an extremely anti-Communist atmosphere.¹⁴⁹ Strident condemnations pointed directly to some of the political reforms adopted in the PCC resolutions, such as limiting the power of the GMD-controlled National Assembly, enhancing executive responsibility of the legislature, and increasing the autonomy of provincial governments.¹⁵⁰

The PCC members involved were left to handle the dispute by themselves. Their most recognisable mediator, Marshall, returned to Washington on 11 March to report to Truman.¹⁵¹ The PCC Steering Subcommittee and the CRDC convened a joint conference to counteract the CEC’s virulent attack on the PCC resolutions. Meetings held on 14–15 March resulted in a three-point compromise agreement initially proposed by Zhou on some of the resolution’s most controversial components. The compromise deal included allowing the National Assembly to exist as a substantive entity. This abandoned the provision that ensured the accountability of the executive government to the legislature and limited provincial autonomy, as proposed in the original resolution.¹⁵²

It is always difficult for negotiators to decide how much of a concession is too much. In the Chinese negotiations, even the minor party leaders believed that Zhou was giving away too much too early.¹⁵³ However, the PCC resolutions had generally made it difficult to perpetuate the Nationalist’s monopoly of government. ‘The door towards democracy is now pushed open, regardless of how narrow the opening still is’, an elated Zhou told Marshall after the PCC.¹⁵⁴ If the Nationalists reneged on the resolutions, that small window of opportunity would not be open for long. Hence, Zhou needed to address—at least nominally—the interests of the Nationalists’ right-wing groups so that the GMD’s reformists could maintain their roles and pass PCC resolutions at the CEC. Zhou therefore agreed to drop the provision for provinces to draw up provincial constitutions, changing this to a provision that gave provinces the right to self-government.¹⁵⁵ This change would not compromise his party’s autonomy in the Liberated Areas. Neither was Zhou’s compromise on the National Assembly a bad idea. A strong legislature could impose constraints on executive power. It was true that the Nationalists would hold the majority in the assembly, but not all of them supported a Chiang–Song family empire.

Although Zhou was working hard to achieve integrated results at the negotiating table, the CCP leadership in Yan’an decided to pursue its political goals by other means in the face of the Nationalists’ anti-Communist activism. It therefore intensified the war in Manchuria. Liu believed that the GMD could not spare more than about 150,000 troops for Manchuria. If Lin’s troops could eliminate some of the existing GMD Manchurian forces, the Nationalists would soon find themselves fighting a losing battle. While the CCP was toughening its stance, the GMD–Soviet relationship turned from bad to worse. Both governments were deadlocked in negotiations over the proposed economic cooperation in Manchuria (a topic that receives further attention in chapter 5). From mid-March, most cities and prefectures situated to Shenyang’s north, including the transportation hub of Sipingjie, fell into CCP hands as the result of Soviet non-cooperation.¹⁵⁶ On 18 March, Liu telegraphed Zhou, telling him to renege on all concessions that had recently been made to the GMD.¹⁵⁷

The Nationalists’ CEC ratified the PCC resolutions in total on 16 March,¹⁵⁸ but a five-point revision was proposed for the PCC resolutions vis-à-vis the Draft Constitution. Two main areas for revision not only went beyond the scope of Zhou’s three-point compromise agreement of 15 March but also stirred up most of the controversies. The first recommended revision in question was a demand that ‘in drawing up the constitution, the Programme of National Reconstruction shall be taken as the fundamental basis’.¹⁵⁹ The Program of National Reconstruction (Jianguo dagang 建國大綱) refers to a twenty-five-point set of concise guiding principles for the National government originally drafted by its founding father Sun Yat-sen in 1924. Notably, the document only provided general rules for governing and was not custom-made for one-party dictatorship. If the GMD intended to use it to nullify the PCC constitutional principles, the CCP could use the Sun dogma stated in the same document, such as the right of direct suffrage and self-governing local authorities, to argue for the legitimacy of the PCC constitutional agreement.¹⁶⁰

The second proposed revision in contention was that ‘the Control Yuan shall not have the right to ask for dissolution of the Legislative Yuan’.¹⁶¹ The Control Yuan (Jiancha Yuan 監察院) and the Legislative Yuan (Lifa Yuan 立法院) were two of the five branches of the GMD government, of which the former functioned as an agency to monitor the government. The proposed amendments were clearly an attempt to push beyond the boundaries of Zhou’s three-point compromise and create a political environment of weak checks and balances in which the president reigned supreme. For the Nationalists who proposed the amendments, however, the revisions might create more problems than they solved.

Moreover, the CEC’s five-point revision could only be considered a summary of opinions because the Nationalists did not have the authority to revise a multiparty resolution unilaterally. PRC writer Xiao Jiansheng has argued that the CCP leaders should not have used the revision as an excuse for ‘breaking up’ and ‘civil war’. Xiao’s book has been banned in China since 2007.¹⁶² To date, mainstream studies in China contend that the GMD blatantly tore up the PCC resolutions during the CEC and that the CCP was forced to resist it.¹⁶³

The controversy over the GMD’s ‘proposed revision’ against the PCC resolutions was that it allowed ‘irreconcilable elements’ within the party to show their belligerent stance towards the PCC.¹⁶⁴ Even if the Yan’an leadership did not directly mimic such a radical stance, it was unlikely to respond in a cooperative manner. In a hard-line bargaining situation, negotiators tend to match the other party’s tactics with their own.¹⁶⁵ As it turned out, the CCP leaders chose to meet belligerence with belligerence. Using the GMD-proposed revision as a pretext, Zhou triggered the falling-out with the Nationalists at a press conference in Chongqing on 18 March. The CCP thereafter stopped using the expression ‘new stage of democratic construction’ and dropped policies couched in those terms.¹⁶⁶

A shattered peace

In an address at the Bureau of Personnel Administration in 1925, Mary Parker Follett, an early proponent of win–win dispute resolution, elaborated her visionary fear of resolving conflict by merely making compromises. ‘If we get only compromise, the conflict will come up again and again in some other form’, she warned.¹⁶⁷ Follett’s premise was progressive for the time. Although not many took her ideas seriously in the 1920s, the history of the Chinese Civil War does not seem to contradict her. As the discussion in this chapter shows, both the Nationalists and Chinese Communists continued to find reasons for compromise—on the basis of the Double Tenth Agreement—over conflict-prone issues in relation to ceasefire, restoration of transportation, the PCC resolutions and army nationalisation—and they did this both with or without their American mediator. There was no shortage of compromise along the way.

Zhou’s report to the Yan’an leadership shows that the CCP’s ambitious military strategies were unpopular with the Chinese public and the international community, if not politically absurd. A peace deal was needed for the CCP as much (as demonstrated in chapter 2) as for the GMD. Hence, it was the necessity to secure an immediate ceasefire, and not just Marshall’s personal clout, that prompted both parties to engage in a series of compromises and let the multiparty PCC forum proceed. When the US-brokered ceasefire negotiation pushed ahead rapidly, however, both parties bargained hard for an agreement in which they could take more and give less.

The bargaining game played out in an extraordinary way. Marshall took over the negotiation process for ceasefire and tried to exert influence through the US government’s inherently biased stance. Unlike his predecessor Hurley, Marshall intervened at an early stage of the ceasefire negotiation instead of waiting until talks teetered. He enforced an ambiguous agreement exception for the GMD’s troop movement in Manchuria and used it to entice Chiang to accept the truce. Given the new negotiation space that the ambiguous provision about Manchuria opened up for both parties, this might have been a worthwhile endeavour for Marshall. In this sense, Marshall could have considered his mediation a success because he accomplished the task as Truman had requested: brokering a ceasefire in China proper, including North China, in a short time frame.

Importantly, Marshall received a grateful letter from Mao, thanking him particularly for his ‘fair attitude and approach in handling the ceasefire’.¹⁶⁸ Mao was in no way suggesting that Marshall was an unbiased mediator here. Rather, the letter can be understood as an acknowledgement of Marshall’s apt use of peaceful mediation techniques in the context of the United States’ biased stance towards Chiang’s government.

Yet, while Marshall enforced an ambiguous ceasefire agreement, he must have accepted the risk. The Chinese Communists acceded to Marshall’s demands regarding the GMD troop movement in Manchuria, but they were determined to benefit from fighting to counter the losses experienced in negotiations. A previous study has considered Liu’s short-war expectation and the notion of fighting a ‘final battle’ during the ceasefire period as exemplifying the CCP policy-makers’ high-risk decision-making in the Chinese Civil War.¹⁶⁹ As discussed, both Chiang and Zhou reacted to Marshall’s pressure with more stringent demands. Marshall’s plan to trade Manchuria for a truce in China proper also made compatibility between the ceasefire agreement and its execution a major concern. The Executive Headquarters was initially ‘a short term, shoestring organisation’,¹⁷⁰ but it was tasked with a number of highly challenging assignments of ceasefire implementation, disarmament and the repatriation of Japanese soldiers, restoration of transportation, and army reorganisation.¹⁷¹ Chapter 6 shows that the headquarters was soon overwhelmed by the large volume of difficult tasks to the extent that the entire organisation’s effectiveness was jeopardised.

Nobody could have expected Marshall to get everything right the first time in such a high-stakes negotiation, but the following chapters show that his handling of the Manchurian and Rehe issues soon returned to haunt him. Most importantly, Marshall would have been dealing with the side effects of pushing the two parties too hard in the ceasefire negotiations. All too often, unnecessary intervention would only make the situation worse. Marshall’s shift from high-pressure tactics to moderation after the commencement of the PCC was nothing but common sense.

Admittedly, the reaction from both parties to Marshall’s intervention in the ceasefire negotiation almost pushed it towards collapse. Zhang Qun had been hailed as the wise man of the GMD, but the package deal he offered to Zhou was unattractive at best. The difficulty of proposing a package in negotiations is that the initiating party must ensure that the deal offered is a low priority for itself and simultaneously greatly beneficial to the other party. Recent research has shown that trade-off deals are often accomplished by trial and error and a meaningful exchange of information between disputants.¹⁷² In other words, Zhang needed to find a mutually acceptable solution with Zhou. However, as long as Chiang drew the line between winning and losing the negotiation on the occupation of Duolun and Chifeng, Zhang’s bargaining range was perhaps too restricted to make any meaningful offer.

In her recollections of Marshall’s mission, Zhang Zhizhong’s daughter Suchu suggests that the Chinese were less prone to compromise.¹⁷³ However, the Nationalists and the Chinese Communists seemed quite susceptible to give and take when negotiating the agreements discussed in this chapter. Perhaps Zhang Suchu has identified the risk that could have potentially aggravated relations between the disputed parties in accepting ambiguous agreements but not the culture of making compromise itself. The experiences of Wedemeyer’s analytically minded planning staff may have provided the basis for an observation with which Follett would have agreed: in a high-stakes negotiation, acting in good faith towards the other party to achieve integrative solutions is more important than simply accepting compromises.

Nevertheless, relations founded on good faith, cross-party cooperation and unanimity were hard to come by, if not totally out of reach, in the GMD–CCP post-ceasefire relations. Hence, the participating parties of the PCC added ambiguous provisions to the resolutions, using them to bridge the gap between the disputed parties. The PCC case shows that the ambiguous provisions paradoxically made the post-treaty conflict predictable. The participant parties of the PCC established different terms and conditions in an attempt to reduce the risks of post-treaty conflicts caused by ambiguities within the resolutions. In this sense, the ambiguous and somewhat confusing provisions in the PCC resolutions support the view that ambiguity was the reality in treaty negotiations; there was no need to negotiate an all-embracing agreement to achieve the negotiation objectives. An ambiguously worded settlement, however, requires all parties involved to face a bitter truth: a bumpy road exists on the way towards long-lasting peace, as significant challenges remain. The post-treaty relationship between the erstwhile enemies, as Pehar has observed, involves a long process of conflict transformation, which can only take root slowly.¹⁷⁴

The regrettable turn of events that unfolded after Zhou’s 18 March press conference, however, indicates that both the GMD and the CCP found it difficult to cope with their differences. On 21 April, the CRDC convened its final meeting. Some of the attendees believed that a compromise agreement had been struck and the assistant clerk was instructed to read out the revised Draft Constitution. The CCP representative, Li Weihan, unexpectedly filed an oral declaration at the very last minute: the CCP had substantial reservations about vital articles in the revised Draft Constitution. Li further asserted that the constitution draft prepared by the committee could only be considered as meeting minutes because the committee did not have settlement authority. He later explained that the reservations were made because the discussion draft was outrageously doctored by GMD representatives. A GMD representative, Wu Tiecheng, replied that the work completed by the committee could be considered meeting minutes until it passed the National Assembly.¹⁷⁵

The exchanges between Li and Wu created another serious dispute between the two parties. The GMD insisted that the PCC resolutions had set a two-month deadline for the review of the Draft Constitution. If the Chinese Communists were unhappy about the review after the deadline was due, they would have to take their case to the National Assembly.¹⁷⁶ The CCP, on the other hand, asserted that the reviewing process had not been completed. The GMD’s request for the CCP to produce the list of delegates to the National Assembly before completion of the constitutional review was therefore not in accordance with its ‘two-stage’ interpretation of the resolutions.¹⁷⁷

In a statement also issued on 21 April, the CCP declared its refusal to tender the list of nominees for the National Assembly, after accusing the GMD government of failing to deliver on its promised democratic reforms. The statement was particularly critical of the GMD’s disagreement with the CCP’s demand for ten Communist members to sit on the State Council, hinting at a Nationalist plot to deprive the CCP of veto power at the highest decision-making level of the multiparty interim government.¹⁷⁸ The statement added to the increasingly heated rhetoric over the CCP’s claim regarding violations of the PCC resolutions regarding the GMD. Notably, the CCP had already announced (on 30 March) that it would not submit the list of nominees for state councillors under the present circumstances.¹⁷⁹ A CCP Central Committee meeting originally scheduled for 31 March in Yan’an to pass PCC resolutions was also postponed indefinitely.¹⁸⁰

Meanwhile, the CCP failed to meet the deadlines for mandatory submission of the lists of its military units in accordance with the army reorganisation agreement, although the GMD had submitted the relevant lists in March and April 1946, respectively.¹⁸¹ The CCP had also stopped the reconstruction of main railways in the provinces of Hebei and Shandong since late March, using unresolved disputes over the removal of fortifications on and along the railways and the GMD’s large troop movements to justify its unilateral decision.¹⁸²

The GMD then opposed Marshall’s plan to supply equipment for the proposed CCP military training school at Zhangjiakou before the CCP force’s completion of the army reorganisation.¹⁸³ Ultimately, Chiang’s opposition forced Marshall to shelve his plan to supply and train the CCP armies. Marshall’s unsuccessful attempt to provide military assistance to the CCP gave Zhou an additional justification to accuse the Americans of helping the Nationalists to wage war against them.¹⁸⁴

On 24 April, Chiang decided that the National Assembly scheduled for 5 May should be rescheduled after consulting leaders of his own party and the PCC members of other parties.¹⁸⁵ The much-anticipated 1946 legislative session was therefore postponed, and no party could propose a new date for it.

In replying to the CCP’s accusations against them, the Nationalists contended that these were not true, and affirmed their disinterest in refuting the charges.¹⁸⁶ Rather, they preferred to answer all allegations in a lengthy memorandum to Marshall. This declared that the GMD’s proposed amendments to the constitutional agreement adopted by the PCC were merely a display of opinions. The revision was proposed in accordance with the understanding reached in the PCC that all involved parties were obliged to propose more effective approaches pertaining to the constitution for consultation and agreement. Regarding the dispute over the quota of state councillors, the memorandum states that the GMD had already proposed a new plan—that only twelve votes would suffice to block a vital decision—while the CCP would hold eight seats and the Democratic League would have four members on the State Council. As the memorandum concludes, ‘But even this the Communists would not accept.’¹⁸⁷

The Nationalists could argue that the CCP had already accepted the eight-seat quota back in early February, when it had prepared an internal list of eight nominees for state councillors.¹⁸⁸ Arguably, the proponents of such an argument overlooked that the polarised interparty relations had worsened since early March. On 17 March, Mao had decided that all plans—including joining the government and the participation of National Assembly—had to return to the drawing board if Chiang conspired to renege on the PCC resolutions. Therefore, Zhou was under orders not to submit any list of names to the GMD.¹⁸⁹ Hence, if the Nationalists’ new plan on exercising a veto was as attractive as they described, they should perhaps have clarified their position to the CCP: they had no intention of exploiting ambiguities in the resolutions.

The actions and reactions of both parties created more uncertainty. On 23 April, Zhou ridiculed the GMD’s passive response to criticism.¹⁹⁰ Disputes after the Nationalist CEC, the subsequent reaction of the CCP, and the further CCP–Soviet alignment in Manchuria together buried the short-lived but celebrated period of negotiated peace. The unfolding conflagration in Manchuria would make the death of the peace efforts apparent to all.