Measuring Reliability In The Wartime Transport of Provisions: The Case of Mao Yuanyi (1594–1641)

A military strategist and advisor, Mao Yuanyi (1594–1641) was one of the most prolific writers of the late Ming period on military matters and participated in the Ming war against the Jurchen in Liaodong in the early seventeenth century. In his seminal study on the conduct of war, the Records of Military Preparedness (Wubeizhi), he exhaustively discussed the costs and benefits of the transport methods available at the time, including carts, pack animals, and water transport. Among all the transport methods that he considered, Mao clearly favored what he called “human transport” (renyun), which exclusively relied on the labor of human bearers. By drawing on Mao’s writings on the transport of provisions, this study analyzes his forceful argument in favor of employing human labor and illuminates the practices of organizing wartime logistics. It also sheds light on the manner in which the costs and benefits of transport methods were being evaluated and how the notions of efficiency and reliability came into play in calculations concerning transport in the late Ming period.


INTRODUCTION
When we peruse scholarly writings of sixteenth-and seventeenth-century China on military matters, we often encounter extensive discussions of logistical and provisioning challenges. 1 Whenever a military conflict erupted, the task of provisioning troops inescapably confronted those who found themselves-willingly or otherwise-in the position of planning and overseeing military campaigns. While their careful planning was essential to transporting provisions successfully, of even greater importance was to coordinate the work of countless seamen, bearers, carters, and animals, who shouldered the physical burden of the transport efforts. Among them, according to late Ming sources, human bearers in particular played a critical role in military transport. During power incapable of inventing and adopting labour-saving devices." 8 The role that such labor-saving innovations historically played in economic development in China has been closely debated in the so-called "Great Divergence" debate on the relative productivity of the Chinese economy in the early modern and modern periods. 9 For instance, in assessing the structure of the rural economy in the Yantze delta region, Philip Huang emphasized that the "vast farm-labor supply removed incentives for labor-saving capitalization and dictated that change take the direction of labor-intensifying involution." 10 Meanwhile, Kenneth Pomeranz has argued that China and Western Europe were similarly on "a labor-intensive path" until the late eighteenth and nineteenth centuries, when, according to Pomeranz, the use of fossil fuels and access to New World resources in Western Europe "obviated the need to manage land intensively." 11 However, previous studies on relative productivity in China have rarely considered the manner in which human labor was compared with labor-saving devices or technologies in the late Ming period. How was the notion of saving labor discussed and implemented? How were priorities and preferences identified and established? Inquiry into these questions also requires us to examine the availability and cost of human labor in late Ming society. 12 This study's analysis centers on an argument put forward by Mao Yuanyi. But Mao's argument for the labor-intensive method of human transport not only illuminates the practices of organizing logistics during wartime. It also sheds light on the manner in which the costs and benefits of transport methods were being evaluated and how the notions of efficiency and reliability came into play in calculations concerning transport in the late Ming period.

MODES OF TRANSPORT IN THE RECORDS OF MILITARY PREPAREDNESS
As Mao Yuanyi opened his section on military logistics in his Records of Military Preparedness, he almost immediately struck a cautious note and alerted the reader that the realities of war did not always allow a swift transport of supplies and provisions: Since ancient times, we have sought to make sure that when armies march, provisions follow. However, when provisions must be transported a thousand li (576 km), the soldiers [already] have a hungry look. 13 Therefore, if war continues over an extended period, nothing is more beneficial than military colony farming (tuntian 屯田). If we are to open up wasteland, we cannot avoid discussing the matter of farming. Those who discuss farming attach great importance to enhancing the natural advantages of the land with human labor. A single achievement can yield twice as many profits. If military colony farming becomes constant, experienced farmers should be consulted. 14 Focusing on the methods of military colony farming, Mao then examined the historical precedents of military colony farming. Military colony farming was an ancient form of provisioning troops, and soldiers were expected to work the land when there was no military engagement. As one of the earliest references to military colony farming, Mao cited a Warring States text and wrote that "[people] farmed for three seasons and learned military skills for one season." 15 While emphasizing the historical roots of military colony farming, Mao was also keenly aware that much had changed since ancient times, most especially in the area of irrigation techniques.
Mao detailed the irrigation techniques of drawing water from rivers, wells, and rain and snow by citing an illustrated work on irrigation titled Western Methods of Hydraulics (Taixi shuifa 泰西水法), produced at the beginning of the seventeenth century by the Jesuit missionary Sabatino de Ursis (1575-1620) with the collaboration of Xu Guangqi 徐光啓 (1562Guangqi 徐光啓 ( -1633. 16 Mao did not provide much commentary on the text itself. But he noted that the sections and illustrations he includedroughly three quarters of Ursis's work-were selectively chosen as they "can be of assistance to military colony farming." 17 He closed the section on military colony farming and irrigation techniques by quoting passages on water-powered mills from Wang Zhen's 王禎 Agricultural Treatise (Nongshu 農書), a widely consulted fourteenth-century work on agriculture and sericulture. While Wang Zhen's treatise contrasted dryland farming in the north with irrigated farming in the south, 18 such a distinction is not emphasized in this section of the Records of Military Preparedness. But Mao's choice of treatises to be included in this section not only highlights the extent of agricultural knowledge that can be amassed by studying printed materials available in the late Ming period. It also underscores, in Mao's view, the importance of agricultural knowledge in administering military logistics.
Judging from the manner in which Mao introduced the subject of military colony farming, he appears to have believed that the ultimate objective of implementing military colony farming was to prevent hunger among soldiers in distant places and obviate the need to transport provisions over a long distance. At the same time, he acknowledged that success in implementing military colony farming did not completely eliminate the need for transporting provisions. In order to provision large contingents of soldiers and animals on the move, gathered crops still had to be sent to distant encampments, and military colony farming could only be a partial solution. Furthermore, while one had to wait for a certain period before being able to harvest crops from the cultivated land, soldiers and animals could not subsist without food and water for a protracted period. In the absence of logistical support, soldiers and animals were left to live off the land, which often created more problems than it solved. Although Mao did not explicitly explain why he chose to exhaustively discuss transport methods in the Records of Military Preparedness, perhaps his underlying assumption was that the transport of provisions was an inescapable component of the warfare of his time.
Shifting his focus from military colony farming to the challenge of provisioning soldiers on the move, Mao then characterized the deployment of armies, large and small, as follows: If small armies march far, it is inevitable that we support them by transporting provisions. If large armies suddenly congregate, it is also inevitable that we support them by transporting provisions. There are only two types of transport: water and overland transport. Water transport consists of canal and sea transport, while overland transport consists of cart, mounted, and human transport. 19 Mao's breakdown of military transport offers helpful hints about his structural understanding of the transport of provisions. It allows us to understand how Mao, one of the foremost experts on military matters of his time, conceptualized and differentiated various means of transport on waterways, by sea, and across land. MEASURING RELIABILITY IN THE WARTIME TRANSPORT OF PROVISIONS RIVER AND SEA TRANSPORT (HECAO 河漕, HAIYUN 海運) While Mao chose to examine canal and sea transport in the Records of Military Preparedness, his discussions primarily described existing navigation routes without offering extensive commentary. He also drew almost exclusively on earlier works on waterways and sea lanes. As he discussed river and canal transport, for example, Mao entirely relied on the work of Pan Jixun 潘季馴 (1521-1595), "the foremost hydraulic engineer of Ming times," 20 who detailed the course of the Yellow River from its purported source in the mythological Kunlun Mountain to the mouth of the River in today's Zhejiang province. 21 In the section on sea transport, on the other hand, Mao relied on the illustrated geographical survey of the Ming realm by Luo Hongxian 羅洪先 (1504-1564), who in turn had based his work on an earlier geographical study by the Yuan-dynasty scholar Zhu Siben 朱 思本. 22 In the Records of Military Preparedness, Mao introduced the four sea lanes described by Luo (see Figures 1 and 2). The first route takes the reader from Fuzhou 福州 in Fujian province to the coastal area near Jinghai Guard 靖海衛 on the southeastern tip of the Shandong peninsula. Second, one sails from Liujiagang 劉家港 near today's Shanghai to the wharf of Nuizhuang 牛壯 along the Sancha River 三叉河, the downstream of the Liao River 遼河 in today's Liaoning . This map views the Shandong peninsula (in the middle) from the east with the north facing the right of the page. Choson Korea and the Yalu River appear in the lower right corner, and Huai'an prefecture and the Yellow River are depicted in the top left corner. (Mao Yuanyi,Wubeizhi,juan 141, province. 23 Third, we journey south from Zhigu 直沽 near Tianjin to the mouth of the Yangtze River. Finally, we start at the mouth of the Liao River and reach Liujiagang at the mouth of the Yangtze River. 24 Of the four routes introduced by Luo, the second and fourth connected the Yangtze delta region with the Liaodong peninsula and were almost identical except for the direction in which they are described. Considering that the other two routes, one from Fujian to Shandong and the other from Tianjin to the Yangtze delta region, are not paired with ones in the opposite direction, it may be plausible to suggest that the sea lane between Liaodong and the Yangtze delta region held particular importance in sea transport at the time of Luo's writing, and perhaps also in Mao's mind. 25 But, similar to his section on river transport, Mao did not offer extended commentary. We are left to speculate on how Mao gauged the value of Luo's maps and descriptions of the four sea routes in light of the military challenges that the Ming faced in Liaodong at the beginning of the seventeenth century.

OVERLAND TRANSPORT (LUYUN 陸運)
In stark contrast to his reliance on earlier texts in the sections on river and sea transport, Mao Yuanyi asserted himself more forcibly in his assessment of overland transport methods. Whenever he consulted the past literature on military matters in the . This map also views China's southeastern coast from the east, with the north facing the right of the page. It depicts Nanjing and the Yangtze River on the right and Fuzhou on the left of the page. (Mao Yuanyi,Wubeizhi,juan 141, from his own writing. While little commentary is included in the sections on river and sea transport, all of the three sections on overland transport contain Mao's detailed analysis. In addition, Mao entirely omitted descriptions of overland routes across the Ming realm. Had he wished to include them, the wide availability of route books during the late Ming period would certainly have made the task feasible. 26 Similarly, for reasons that are not clear, he chose not to include maps of overland transport routes. Instead, Mao solely focused on carefully analyzing the merits and costs of overland transport methods. Perhaps reflected in this contrast is the likelihood that Mao had acquired greater familiarity with the overland transport of provisions, in comparison to his experience of-or interest in-water transport. Despite their relative brevity, the sections on overland transport in the Records of Military Preparedness meticulously address technical and practical concerns, such as costs, capacity, efficiency, and reliability. Mao's assessment of overland transport methods, therefore, allows us to take a close look at the military transport of provisions during the late Ming period. Cart Transport (cheyun 車 車運 運) Mao's analysis of overland transport consists of three sections, with each corresponding to cart, mounted, and human transport. "Cart transport is an ancient principle (zhi 制)," wrote Mao as he opened his section on cart transport. "Its method ( fa 法) has long been neglected and cannot be examined," he added, "and as for examining its purposes, we can only examine [what goes on] today." 27 This opening passage is immediately followed by his assessment of four types of cart transport that were presumably in use at the time of his writing. Also provided are numerical figures indicating how heavy a load each mode of transport could transport. The first was what he termed the "human-cart" (renche 人車), which was driven and controlled by carters. Two persons pulled and pushed the cart, and it could transport a load of up to four bushels (shi 石, approx. 414 ℓ). Unfortunately, Mao's description of the hand-held cart is not accompanied by any illustrations. But his contemporary Song Yingxing 宋應星 (1587-ca. 1666) in his Exploitation of the Works of Nature (Tiangong kaiwu 天 工開物) provided commentary on the single-wheeled cart with illustrations. 28 Song noted that this so-called "single-wheeled push-cart (dutuiche 獨推車) of the south can be managed by the strength of a single person. It can carry a load of two bushels (207 ℓ). It must stop whenever it encounters a pit on the ground. The farthest it can travel is merely 100 li (58 km)." 29 Judging from the difference in the volume of load, Song's single-wheeled push-barrow was likely smaller than Mao's hand-held cart, which required the labor of two persons. Nevertheless, Song's illustration and description provide vivid pictorial images of late Ming carts, which Mao might have observed in his native Zhejiang province, or in the Liaodong region (see Figure 3).
The second type of cart that Mao discussed in the Records of Military Preparedness was the "ox-cart" (niuche 牛車). The ox-cart was powered by two oxen, and two persons controlled the movement of the oxen and the cart. This type of cart, according to Mao, could carry a load of up to twelve bushels (1.2 kℓ), three times as much as the human-cart driven by two people. 30 Thirdly, Mao wrote that the "mule-cart" (luoche 騾車) was led by a team of ten mules and could carry a load of up to thirty bushels (3.1 kℓ). In relative terms, the mule-cart could carry more than twice as much load as the ox-cart, and over seven times as much as the human-cart. The mule-cart, however, was clearly not Mao's preferred method. From the point of view of cost efficiency, he was clearly of the view that the mule-cart did not offer any advantage over the other types of cart. While Mao did not explain in detail, he emphasized that "its costs are incalculable." 31 The fourth and final type of cart transport, according to Mao's categorization, was the so-called "supply-cart" (zizhongche 錙重車). As Mao explained in the Records of Military Preparedness, the invention of the supply-cart is normally attributed to the Ming military commander Qi Jiguang 戚繼光 (1528-1588), who helped rebuild the Ming defenses in both south and north China. The functional purposes of the supply-cart were twofold. Tactically, its wooden structure could provide cover for soldiers on the battlefield; logistically, it could transport a significant volume of supplies and provisions for armies deployed away from urban centers or sites of agricultural production. According to Qi's original design, a team of FIG. 3. Left, a single-wheeled cart pushed by one person employed in the south; right, a single-wheeled cart pushed by one person and drawn by two donkeys. (Song Yingxing,Tiangong kaiwu, eight mules drew each wagon, and eighty supply carts formed a division (ying 營) (see Figure 4). On top of the cart was placed a large single-sided shield (pianxiangpai 偏廂牌), which "resembled a city wall when seen from a distance." 32 Qi added that his "supply-cart" could transport twelve and a half bushels (1.3 kℓ) of grain, beans, and preservable food, slightly exceeding the capacity of Mao's ox-cart. Thus, one division of supply-carts could transport a combined load of 1,000 bushels (104 kℓ) of food, which was deemed sufficient to feed 10,000 men and horses for three days.
On balance, Mao Yuanyi concluded that the ox-cart was preferable to all the others as far as cart transport was concerned. As he drew such a conclusion, he found an encouraging support in the Classic of Changes (Yijing 易經), one of China's most canonical texts. "The ancient saying goes," wrote Mao referring to a passage in the Classic of Changes, "horses are used to reach a distant place, while oxen are used to transport a heavy load. Thus, oxen are the basis of cart transport." 33 Unfortunately for our analysis, he did not address the question of how widely each cart device was being used for military or other purposes. Nor did he discuss climatic or terrain conditions under which each type of cart could and could not function as designed.
FIG. 4. A mule-led "supply-cart" depicted in Qi Jiguang's Practical Account of Military Drill (Lianbing shiji). The caption reads: "Each cart weighs about 2,000 catties (approx. 1,194 kg)." While the harnesses of the five mules in the front are not connected to the mules in the rear in this illustration, Qi Jiguang wrote in the accompanying passage that that a team of eight mules led a cart loaded with military provisions and supplies. In another section of the text, Qi noted that each cart was led by ten mules, as depicted in this illustration. (Qi Jiguang,Lianbing shiji,juan 1,19:111;Lianbing shiji,Zaji,juan 6, Mounted Transport (qiyun 騎 騎運 運) Turning to the next mode of overland transport, Mao focused his attention on the method involving three members of the horse family: horses, mules, and donkeys. Although Mao did not explicitly explain how those draft animals differed from those employed for cart transport, the very distinction that Mao drew between cart and mounted transport implies that these were pack animals, which carried heavy loads either on their backs or on pack saddles. Of the three sections concerning overland transport in the Records of Military Preparedness, the one on mounted transport is the shortest, merely consisting of three lines. Moreover, Mao's commentary makes it evident that he considered mounted transport to be both costly and inefficient. "There are three types of mounts," noted Mao as he explained how they differed from one another in terms of the volume of load they could carry. "Horses and mules are capable of carrying up to one and a half bushels (155 ℓ), while donkeys are capable of carrying up to one bushel (104 ℓ)." 34 Compared with the four types of cart transport that Mao discussed, namely the human-cart, ox-cart, mule-cart, and supply-cart, the volume of load which pack animals could carry indeed appear relatively small. According to Mao's own estimates, even the smallest of the four-the hand-held cart driven by two carters-could transport a load of up to four bushels (414 ℓ), nearly three times as much as a pack horse.
In addition, in order to be able to reliably travel a long distance with loads, every mount had to be fed with an adequate amount of fodder each day. Sources indicate that during the Ming period cavalry horses were provided with a daily average of three pints (sheng 升) (3.1 ℓ) of beans and one bundle of grass, weighing about fifteen catties ( jin 斤) (9 kg). 35 Without adding further commentary on mounted transport, Mao then almost bluntly stated that "while the costs are considerable, the load that can be transported is small. This is not a fine method. [Only] when there is no alternative should this method be used." 36 Interestingly, Mao closed this section on mounted transport by noting that "camels of the northwestern frontiers are capable of fulfilling the duties of several horses." 37 No mention is made of whether he had personally seen camels transporting heavier loads. According to a memorial submitted by a Vice Minister of Revenue in 1619, however, the then Governor of Liaodong appears to have proposed employing camels to transport provisions, for "a camel can carry three times as much load as a horse or donkey but eats no more than a horse or donkey." 38 Mao was perhaps aware of such a discussion among Liaodong officials and might have considered the use of camels in the overland transport of provisions in the region. But purchasing or hiring a sufficient number of camels perhaps proved difficult. Mao concluded that camels could not become a viable alternative to horses because "they cannot be obtained in large numbers." 39 Human Transport (renyun 人 人運 運) The last of the three modes of overland transport examined by Mao Yuanyi employed neither the principles of mechanical engineering nor the physical stamina of pack animals. It solely relied on human labor. Of all the three methods, Mao clearly favored this method of transport. Well versed in literature on military matters, Mao was keenly aware that there existed in China a historical precedent for transporting military supplies and provisions by coordinating the work of human bearers. He found such an example in the biography of the Yuandynasty general Dong Tuanxiao 董摶霄 (d. 1358) in the Standard History of the Yuan Dynasty (Yuanshi 元史). Included in Dong's biography is the following proposal that he presented to the Yuan court in 1356, when widespread insurgencies were seriously undermining the Yuan emperor Shundi's 順帝 claim to power. The matter in question was the transport of provisions in the strategically vital region of Huai-Hai 淮海, where the Huai River met the Yellow River in central China. Among the key localities in the region was the coastal prefecture of Haining 海寧, located to the east of Xuzhou 徐州: In the area around Haining Prefecture, boats cannot pass, and military provisions can only be transported overland. But the people of the entire Huai-Hai region have suffered from repeated banditry, and we should provide comfort. For the time being, soldiers should be ordered to transport provisions. 40 Dong's primary objective was to transport provisions in the lowlands of today's Jiangsu province. But he assessed that water transport was not an option, and banditry had already depleted the resources of the local communities. In this context, Dong proposed employing the following method of transport, making the most of the readily available labor of his soldiers: If everyone walks ten paces (16 m), then thirty-six people can cover the distance of one li (576 m). 360 people can cover ten li (5.8 km), and 3,600 people can cover 100 li (58 km). Everyone carries four pecks (dou 斗) (41 ℓ) of husked grain (mi 米) in a lined cloth sack, which is sealed and marked with a seal. People do not rest their shoulders, nor does husked grain touch the ground. People are lined up so that they form a straight line. Everyone walks 500 laps and a total of twenty-eight li (16 km) a day. [In other words,] everyone walks fourteen li (8 km) with a load and another fourteen li without, and we can transport a total of 200 bushels (21 kℓ) of husked grain a day. If a pint (1 ℓ) of husked grain is distributed [per person], we can provide for 20,000 people. This is the procedure (shu 術) of transporting provisions over 100 li a day. 41 In the Records of Military Preparedness, Mao only quoted this latter portion of Dong's proposal, which detailed his method of transport relying on human labor. Mao quietly left out the first portion of the passage, where Dong succinctly explained the geographical and socio-political context of his proposal.
As we have seen above, Dong's biography in the Standard History of the Yuan Dynasty illuminates the two central premises of his transport method. First, enlisting the labor of local residents in the Huai-Hai region would further aggravate their suffering caused by banditry and social disorder. Second, in place of local residents, 3,600 soldiers were available and could be assigned the task of carrying provisions as bearers amidst growing threat posed by rebel forces in the region. Mao was likely aware of these premises of Dong's memorial as he wrote elsewhere that soldiers not assigned to combat duty would be an ideal source of labor for the purpose of transporting provisions. 42 In his discussion of human transport in the Records of Military Preparedness, however, Mao did not address the issue of labor availability, at least explicitly. Instead, he focused on the technical aspect of Dong's proposal. Without specifying who were to serve as bearers, he went on to suggest a modification to Dong's design and argued that the bearers should be able to substantially increase their workload: According to [Dong] Tuanxiao's method, thirty-six people were assigned to each li, and everyone was responsible for ten paces. If they walked 500 laps a day, that meant that everyone walked fourteen li with a load and another fourteen li without. Even the least strong should be able to double the distance. According to [Dong] Tuanxiao's method, each sack contained four pecks (41 ℓ) of husked grain. If we consider its weight, it does not exceed fifty catties (30 kg). Even the least strong among the men should be able to double the volume of his load. 43 On the one hand, Mao appears to have acknowledged the overall validity and effectiveness of Dong's approach. On the other hand, he clearly felt that Dong did not optimize the use of men's physical resilience. Thus, he suggested that all bearers should be able to increase their workload fourfold. 44 Still, this was not the sole modification of Dong's method that Mao put forward (see Figure 5).
Perhaps perceiving that the scale of the military threat that the Ming faced in Liaodong was different from Dong's challenges in the Huai-Hai region centuries earlier, Mao further multiplied Dong's figures of bearers by a factor of five. That is to say, FIG. 5. The page explaining Mao's method of human transport in the Records of Military Preparedness. Note emphasis marks in the shape of water drops on the right-hand side of the text. (Mao Yuanyi,Wubeizhi,juan 141, five teams of thirty-six bearers, or a total of 180 bearers, formed five separate straight lines in each li to transport provisions. Mao explained, "if we assign 180 bearers to each li, we can transport 4,000 bushels (414 kℓ) a day." 45 With thirty-six bearers assigned to every li, Dong's original method enabled the transport of 200 bushels (21 kℓ) a day. With 180 bearers assigned to every li, therefore, the total volume of transported grain would increase fivefold to 1,000 bushels (104 kℓ). Then, if all bearers could double both the distance that they walked and the volume of grain that they carried, the total volume would further increase fourfold-at least in theory-to 4,000 bushels a day.
In the Records of Military Preparedness, Mao did not directly discuss the practicability of his method of human transport. As shown in the introduction of this study, however, sources suggest that the method of human transport had indeed been implemented for military transport in the late Ming period. In the case of the Ming official in charge of logistics during the Japanese invasion of Korea at the end of the sixteenth century, for example, he was instructed to position 200 soldiers unfit for combat every ten li-or twenty soldiers every li-over a distance of 500 li (288 km) for the purpose of transporting provisions and supplies. 46 Twenty men assigned to each li were even fewer than thirty-six men assigned to each li under Dong Tuanxiao's method. Perhaps such an adjustment was oftentimes necessitated by the realities of war and a lack of human labor in wartime society. As far as we can judge from Mao's writings, he appears to have firmly believed that he would be able to transport 4,000 bushels of grain a day by assigning 180 men to each li, five times as many as under Dong's method, and nine times as many as under the method indicated in the case above during the Japanese invasion of Korea.
In the passages that followed, Mao further elaborated on his method of human transport and underscored its advantages over the other transport methods. He calculated the costs of implementing his method of transport and provided the following estimates: If we pay each person two pints (2 ℓ) of husked grain and have him eat a half for himself and provide for his family with the rest, the daily expenses for each li is merely three bushels and six pecks (362 ℓ). If we make groups for meals, three cooks can feed [those assigned to] each li. Thus, a transport over 100 li (58 km) only requires the labor of 18,304 people. 47 To our eyes, the task of enlisting the willing cooperation of over 18,000 individuals may seem extremely challenging. Mao, however, undoubtedly considered this number to be relatively small. Inserting the particle "only" (er 耳) twice into the passage, he stressed the relative smallness of the figures both in terms of the total wages and in terms of the number of laborers necessary in order to implement his method of transport. With all things considered, Mao concluded that "this is the most convenient aspect of human transport." 48 A CASE FOR HUMAN TRANSPORT For Mao Yuanyi, who personally participated in the Ming war against the Jurchen armies in Liaodong in the early seventeenth century, the transport of provisions was not merely a matter of intellectual exercise. He presented a forceful argument in favor of human transport at a time when Nurhaci's consolidation of power in Liaodong was gathering momentum at the expense of the Ming hold over the region. Especially in the aftermath of the crushing defeat of the Ming at Sarhu in 1619, 49 deciding on the manner in which supplies and provisions should reach the encampments of Ming troops likely assumed renewed importance. Mao might have felt that any decision on the method of transport could potentially translate into the crucial difference between victory and defeat on the front lines.
In the Records of Military Preparedness, however, Mao presented his case for human transport in a rather general form, perhaps in part because he sought to broaden the appeal of his magnum opus. To take one example, Mao did not link his assessment of various transport methods to any specific geographical regions. Mao's choice to leave out the first portion of Dong Tuanxiao's proposal further adds to the impression that he intentionally removed regional references from his section on human transport in the Records of Military Preparedness. Consequently, Mao's case for human transport leaves the reader searching for further context as to why he considered human transport to be the most efficient means of wartime transport, compared with all the other methods considered in the Records of Military Preparedness.
As if to address such concern on the part of the reader, Mao provided a more thorough discussion of human transport in one of his collected essays titled Daring to Speak (Maoyan 冒言). In his own preface to the collection, Mao explained that he had completed the essays in 1618 in his mid-twenties, well before he participated in the Ming war in the Liaodong region. 50 Although it remains difficult to determine whether these essays preceded or followed his writing of the Records of Military Preparedness, the first two essays-one on military colony farming and the other on human transport-specifically address the military situation and challenges in the Liaodong region. 51 This indicates that the Liaodong region had long held a prominent place in Mao's thinking about military transport and logistics.

DURATION AND WAGES
In Daring to Speak, Mao's essay detailing the method of human transport is simply titled "Human Transport" (renyun 人運). In examining this essay, we quickly realize that Mao's method of human transport was clearly intended for the Ming war against the Jurchen in the Liaodong region. As in the Records of Military Preparedness, Mao also chose to begin his discussion by introducing the method proposed by the Yuan general Dong Tuanxiao. But what clearly distinguishes this essay from the Records of Military Preparedness is that Mao carefully took note of the critical difference between Dong's challenges in the low-lying plains in today's Jiangsu province and those of the Ming across the vast mountainous terrain of Liaodong at the beginning of the seventeenth century. Almost dismissing Dong's method as outdated, Mao in this essay introduced his subject matter in a forthright manner: The method of human transport has been in existence since ancient times. According to historical records, it was started by Dong Tuanxiao, a man of the Yuan period. However, if we wish to implement his method today, it is certainly bound to fail. What Dong Tuanxiao proposed was to transport [provisions] in Haining. Few men were involved, and the distance was short. He used soldiers to transport military [provisions] and simply fulfilled what was needed. 52 In contrast to the Records of Military Preparedness, Mao in this essay immediately highlighted the limitations of Dong's method. With the ongoing conflict in the Liaodong region in mind, Mao indicated that Dong's plan-covering the distance of 100 li with 3,600 bearers-was an appropriate measure under Dong's specific circumstances. But it was not a method that could be applied directly to his challenges at the beginning of the seventeenth century.
Mao responded to his challenges in Liaodong by devising his own method of transport based on the foundation laid by Dong some two and a half centuries earlier: The method [of transporting provisions] in Dongning Circuit [of Liaodong] has yet to be implemented uniformly. Now, if we focus on the two Circuits discussed by the Military Commissioner ( jinglüe 經略), there is a distance of 240 li (138 km) to Liaohai 遼海 Circuit and 210 li (121 km) to Hai-Gai 海盖 Circuit. 53 The Military Commissioner has mentioned that there would be an annual need of 1,120,000 bushels (115,920 kℓ) of husked grain and beans, and that the costs would total 588,750 taels [of silver]. We have yet to encounter the following problems: carts breaking down, oxen collapsing, and sacks being swapped. If these problems were to occur, nothing would be better than proposing human transport. 54 While Mao in this essay does not appear to have directly leveled criticism at the Military Commissioner, his writing makes plainly evident how much the two men disagreed over how best to transport military supplies and provisions in Liaodong. According to a letter that Mao wrote to a friend in 1621, the Military Commissioner had taken a position in favor of employing ox-carts in Liaodong. Mao explained that each ox-cart required a team of two oxen and two carters to operate, perhaps resembling one of the cart vehicles depicted in Song Yingxing's Exploitation of the Works of Nature (see Figure 3, right). Mao, on the other hand, did not share the Military Commissioner's favorable views on the ox-cart. He firmly held that cart transport was both costly and inefficient.
For Mao, the immediate logistical challenge was to transport provisions over a combined distance of 450 li in the area under the jurisdiction of Liaohai and Hai-Gai Circuits in Liaodong. 55 Before detailing his own method, Mao once again briefly explained the basic ideas of Dong's human transport: thirty-six bearers were assigned to each li, and everyone walked twenty-eight li a day to transport a total load of 200 bushels. As we have seen above, Mao argued that bearers could increase their workload fourfold. In other words, while Dong called for a transport of 200 bushels a day, Mao expected each line of bearers to transport a total of 800 bushels (83 kℓ) a day. With five teams of thirty-six bearers assigned to each li, therefore, Mao's method theoretically enabled a daily transport of 4,000 bushels (414 kℓ). In term of the volume of load that could be transported, this was where Mao left the reader in the Records of Military Preparedness without providing additional variables. Between Mao's discussions of human transport in the Records of Military Preparedness and Daring to Speak, one of the notable differences is that he addressed the issue of time and duration in his essay in Daring to Speak. Mao was clearly responding to logistical challenges not merely in the short term, but also in the long term. Over the course of a year, which consisted of roughly 354 days according to the lunar calendar, or about 384 days when there was an intercalary month, Mao expected the transport of provisions in the Liaodong region to continue for 280 days, or nine and a half months. At the rate of 4,000 bushels of grain per day, this meant that the transport of 1,120,000 bushels (115,920 kℓ) of grain was possible annually. Mao did not explain how he arrived at the expected duration of 280 days for the transport of military provisions. Nor did he specify whether he expected the bearers to work continuously for 280 days, or with days of rest in between. Perhaps Mao had in mind the ancient passage about dividing the four seasons for farming and military training, which he cited in his section on military colony farming in the Records of Military Preparedness. The passage indicated that in ancient times people farmed for three seasons in a year and learned military skills for one season. Scholars have long understood this passage to mean that while the three seasons of spring, summer, and autumn were apt for farming, the winter months permitted people to spend time away from the fields and hone their military skills. Mao might have borrowed this framework of dividing a year into two groups and envisioned three seasons, or roughly 280 days, of military transport. He might have also assessed that the transport of provisions was possible only for 280 days due to icy climatic conditions in Liaodong's winter, or that one million bushels of grain transported over 280 days was sufficient to provision the Ming soldiers stationed in the region for the entire year. Mao's text also remains silent as to how he expected the bearers to spend the remainder of the year. One thing that is clear from his writings, however, is that he did not expect the bearers to receive wages when they were not transporting provisions.
As far as wages were concerned, Mao calculated that two pints (2.1 ℓ) of husked grain was needed daily for each bearer. Based on this assumption, the costs of transporting provisions over the distance of 450 li with the labor of 81,000 bearers-or 180 bearers in each li-totaled 1,620 bushels (160 kℓ) of husked grain a day and 453,600 bushels (44,770 kℓ) over 280 days. 56 Mao further argued that his method of transport could also be applied to transporting grass fodder for a shorter duration of 227 days. He noted that one bundle of grass weighed fifteen catties (9 kg) and that each person should be able to take seven bundles at a time. This meant that everyone was expected to walk ten paces with a total load of 105 catties (63 kg). If we assume that one could enlist the labor of the same number of bearers over a distance of 450 li-81,000 in total-as for the transport of grain, this method would have enabled the transport of nearly eight million bundles of grass fodder over 227 days. As in the transport of grain, each bearer was to receive a daily wage of two pints (2.1 ℓ) of husked grain, and Mao expected the total expenses to be 367,540 bushels (36,276 kℓ) of husked grain. 57

SOCIAL RAMIFICATIONS OF HUMAN TRANSPORT
Apart from its geographical focus on Liaodong and consideration of duration, Mao's essay in Daring to Speak also differs from the Records of Military Preparedness in that it considered the resulting effects of human transport on the region's economy. Noting that "the people of the Liao region must eat millet (su 粟) of the Liao region," 58 Mao wrote that if Ming armies transported their own provisions using the method of human transport, they would be able to reduce the likelihood of disrupting the balance of supply and demand in local grain markets: Now, if officials send out 450,000 [bushels of grain], it means that 450,000 bushels will remain among the people. Should there be an additional supply of 450,000 bushels, we will certainly be able to purchase grain at fair prices. We would not need 300,000 taels; there would be no use for carts or oxen. The only inevitable item would be cloth sacks. 59 Rounding down the figure of 453,600 bushels that could be transported over 280 days, Mao argued that his method of human transport would not only enable a reliable transport of provisions, but also help stabilize local grain prices in the Liaodong region. As a positive consequence, according to Mao, stable grain prices would also guarantee that, when necessary, officials would be able to locally procure grain at fair prices.
At the same time, Mao was probably aware that his method of human transport posed inherent risks. By design, human transport required a substantial number of hired hands, and preventing fraud among the bearers was a serious challenge. In his essay, Mao repeatedly underscored the need to securely seal every cloth sack, presumably in an effort to thwart attempts of deception on the part of the bearers. He did not describe the exact ways in which bearers could swap sacks in secret for personal gains. The practice of "swapping sacks" (yidai 易袋), however, is mentioned twice in this relatively brief essay. It may indicate that actual cases of fraud had indeed alerted Mao to the risk of losing grain in transit. Perhaps he felt that dishonesty among the bearers could seriously derail the Ming effort to provision its soldiers in Liaodong, and that preventative measures had to be put in place. Mao likely knew, however, that completely rooting out fraudulent practices was extremely difficult, if not impossible. 81,000 bearers, spaced ten paces apart over the distance of 450 li, could not constantly come under the watchful eyes of officials, some of whom might have also chosen to collude with bearers. Mao could only hope that clearly marking every sack with an official seal would dissuade the bearers from cheating. "If the sacks are marked and sealed securely and reach the granaries where they should be unloaded," noted Mao, "we can prevent the evil practice of swapping sacks." 60 We also know that Mao was paying attention to the Sino-Korean borderland, for he discussed the ground conditions of Kuandian 寬奠 and Aiyang 靉陽 Forts (bao 堡) on the western bank of the Yalu River (see Figure 6). 61 The strategic importance of the two forts was soon to assume added importance as the Jurchen gained control of the area and began to seriously challenge the defenses of both the Ming and the Choson. According to the Choson official Kim Sokchu 金錫胄 (1634-1684, who strongly advocated an anti-Manchu position at the Choson court during the reigns of King Hyonjong 顯宗 (r. 1659-1674) and Sukchong 肅 宗 (r. 1674-1720), Kuandian and Aiyang Forts were located opposite from Ch'angsong 昌城 and Pyoktong 碧潼 on the Choson side of the border and could become bases for a military attack against Korea. Kim Sokchu also noted that the terrain in the areas to the south of Ch'angsong was relatively less precipitous than in the surrounding areas, and that the rivers there flowed slowly. For these reasons, according to Kim, the Choson armies chose to march north through this region to cross the Yalu River in 1618, on the eve of the Battle of Sarhu. It was also through this region that Jurchen armies launched their invasion of Korea in 1627, and again in 1636. 62 From the point of view of Mao Yuanyi, however, such relative ease of movement for large armies did not automatically translate into the ease of movement for provisions. "In the areas surrounding Kuandian and Aiyang Forts," he wrote, "we are unable to use the cart. If we were to use pack horses (matuo 馬馱), the costs would be even greater. If we employ human labor, we will need no horses." 63 By emphasizing that the terrain of the border region along the Yalu River made the use of the cart practically unsuitable, Mao provided-conveniently for his argument-yet another reason to adopt his method of human transport.

THE EFFICIENCY AND RELIABILITY OF HUMAN TRANSPORT
Mao Yuanyi's writings suggest that there existed strong support for the use of carts among Ming officials at the time of the war against the Jurchen. Mao clearly felt that he had to persuade such advocates of cart transport of the advantages of using human labor. In the final section of his essay on human transport in Daring to Speak, he mounted a stout defense of his method of human transport, most likely aimed at the cautious critics of the method: Those who caution against human transport merely question whether there will be a sufficient number of people. What they do not know is that a proposal has been made to hire 30,000 carters [for cart transport], not including public and private petty laborers (zayi 雜役) who cut grass or feed oxen. Now, [under the method of human transport] we will only need 81,000 bearers over the vast expanse of the two Circuits [of Liaohai and Hai-Gai]. If we separate the bearers into groups and provide meals, and if three cooks feed those assigned to each li, in total just over 1,000 petty laborers would be needed. Why should we worry that people will be insufficient in number? 64 Mao maintained that human transport would be far less costly than cart transport. Assuming that each cart required the labor of two oxen and two carters, hiring 30,000 carters essentially entailed hiring 15,000 carts and 30,000 oxen. In Mao's reasoning, if one considered the daily expenses of feeding the oxen, providing for the carters, and performing routine maintenance of the carts, no one could choose cart transport over human transport.
With respect to the challenge of enlisting the labor of a significant number of bearers, Mao's apparent optimism might have derived in part from the expectation that a large portion of the garrisoned soldiers in Liaodong, such as Liaoyang 遼陽 and Guangning 廣寧, could be reassigned to transport duties. 65 At the same time, Mao made an economic argument that his wage structure had greater appeal than the one proposed by the proponents of cart transport. Even in the absence of military labor, Mao would likely have remained confident that local residents in Liaodong would willingly provide their labor for wages. According to Mao, his critics proposed to hire carters by offering three-hundredths of a silver tael (1 g) per person a day. Over the course of a year, the carters were expected to work and receive wages for eight months, slightly shorter than under Mao's method of human transport. Thus, each carter would have earned nine-tenths of a tael (34 g) of silver a month and seven taels and two maces (267 g) by the end of the eighth month. 66 At the time, a peck of grain (10 ℓ) was valued at one mace and seven-hundredths of a tael (6 g). Based on this price, the annual wage of seven taels and two maces would have allowed one to purchase forty-two pecks and three pints (439 ℓ) of grain, or an average of less than two pints a day. Under Mao's method, on the other hand, each person received a daily wage of two pints (2 ℓ) of grain, which amounted to six pecks (62 ℓ) over a month. Since Mao expected the transport of provisions to continue for 280 days, the total annual wage in kind totaled fifty-six pecks (580 ℓ) of grain. When compared to the forty-two pecks of grain that the carters could have purchased with their wages, the economic advantage of his method appeared obvious to Mao. "In my humble opinion," asserted Mao, "the people of the Liao region will willingly come forward." 67 Finally, perhaps reflecting his concern that such an economic argument alone might not be sufficiently compelling, Mao highlighted the adaptability of the human body to a wide range of working and climatic conditions. "Humans differ from carts or oxen," noted Mao, "in that they can work shifts, and that we are able to rotate them. They can work throughout the year without interruption, stopped only by heavy rain or snow. Such days [of heavy rain or snow] do not exceed thirty days a year." 68 For officials in the Ming civil and military administration, such a line of argument likely proved most effective as their performance was subject to constant scrutiny by peers and censorial officials. In Mao's view, not only could humans carry heavy loads and walk long distances at a relatively low cost, they could also provide crucial reliability for military transport under testing conditions. As Mao pointed out earlier in the essay, carts could break down, and pack animals could collapse while transporting provisions. In stark contrast, according to Mao, the human body had proven remarkably resilient. Admittedly, the same humans could also commit fraud by swapping sacks. Mao must have reasoned, however, that the overall majority of the bearers were obedient and dutiful. Curiously left out of his argument is the likelihood that humans could also collapse while transporting provisions, succumbing to illnesses, injuries, or exhaustion. If Mao considered the fragility of the human body, he in the end chose not to address the matter in his writings. As we come to the closing passages of Mao's essay on human transport in Daring to Speak, we cannot help but wonder about the human toll of his method of transporting provisions in wartime Liaodong in the early seventeenth century, and about the intriguing interplay between technology, efficiency, and reliability in the mind of the erudite late Ming scholar.

DISCLOSURE STATEMENT
No potential conflict of interest was reported by the author.