Despite a number of vehicles having been recovered from graves and sacrificial pits, all aspects of the chariot’s employment in the ancient period pose vexing questions, particularly whether they were deployed by themselves as discrete operational units or were accompanied by either loosely or closely integrated infantry. Because even the oracular inscriptions for King Wu Ting’s well-documented reign provide few clues, and the tomb paintings recently discovered that date to the Warring States and thereafter mainly depict hunting scenes and parades, far more is known about the chariot’s physical structure than its utilization. The chariot’s essence has always been mobility, but prestige and displays of conspicuous authority rather than battlefield exploitation may have been defining factors in the Shang.

Some traditionally oriented scholars continue to assert that chariots played a significant role in Shang warfare; others deny that they were ever employed as a combat element. The Shang’s reputed employment of chariots, whether nine or seventy, to vanquish the Hsia is highly improbable given the complete absence of late seventeenth-century BCE or Erh-li-kang artifacts that might support such claims. However, Warring States writers idealistically ascribed differences in conception and operational characteristics to the Three Dynasties: “The war chariots of the Hsia rulers were called “hooked chariots,” because they put uprightness first; those of the Shang were called “chariots of the new moon,” because they put speed first; and those of the Chou were called “the source of weapons,” because they put excellence first.”

The few figures preserved in Shang dynasty oracular inscriptions, Chou bronze inscriptions, and other comparatively reliable written vestiges indicate that chariots were sparsely employed in Shang and Western Chou martial efforts. The chariot’s first recorded participation in Chinese warfare actually occurs about seven to eight hundred years after their initial utilization in the West, ironically just before the Near Eastern states would abandon them as their primary fighting component due to infantry challenges. King Wu Ting’s use of a hundred vehicle regiments for expeditionary actions, already discussed, seems to have initiated their operational deployment, though the only concrete reference to Shang chariots (ch’e) appears in the quasi-military context of the hunt.

Chariots must have been extensively employed in the late Jen-fang campaigns, but no numbers have been preserved. Thus the next semireliable figure is the universally acknowledged 300 chariots that were employed by King Wu of the Chou to penetrate the Shang’s massive troop deployment at the Battle of Mu-yeh, precipitating their collapse. Some accounts suggest that the Chou had another 50 chariots in reserve, while the number fielded by the Shang, strangely unspecified in the traditional histories, could hardly have been less than several hundred. King Wu reportedly had a thousand at his ascension, some no doubt captured from the Shang, though others may have belonged to his allies and merely been numbered among those present for the ceremony. Several hundred were also captured from the Shang’s allies in postconquest campaigns, as well as in suppressing the subsequent revolt.

Nevertheless, chariots seem to have been minimal in early Western Chou operational forces. Scattered evidence suggests that field contingents never exceeded several hundred, with as few as a hundred chariots participating in expeditionary campaigns. Although one of their efforts against the Hsien-yün resulted in the capture of 127 chariots from a supposedly “barbarian” or steppe power, King Li’s campaign against the Marquis of E seems to have been typical. Despite total enemy casualties being nearly 18,000, inscriptions on the bronze vessel known as the Hsiao-yü Ting indicate that a mere 30 chariots were captured in one clash, though a second force of 100 is also mentioned. Somewhat larger numbers were deployed slightly later in campaigns against the Wei-fang, but the maximum figure ever reported for the Western Chou, the 3,000 supposedly dispatched southward against the rising power of Ching/Ch’u in King Hsüan’s reign (827-782), is certainly exaggerated despite the king’s reputation for having revitalized Chou military affairs, as well as unreliable because it is based solely on an ode known as “Gathering Millet.”

The chariot’s effectiveness in the Shang, early Chou, and perhaps even beyond must be questioned in the face of the constraints discussed below, the difficulties that will be examined in the next section, and the lessons that can be learned from contemporary experiments with replica vehicles. However, it should be remembered that although numerous reasons can be adduced why chariots could not have functioned as generally imagined, voluminous historical literature, both Western and Asian, energetically speaks about their employment in battle. Ruling groups were still expending vast sums to build, maintain, and employ chariot forces in the Warring States period, and the Han continued to field enormous numbers against steppe enemies, incontrovertible evidence that rather than being historical chimeras or simply artifacts of military conservatism, they continued to be regarded as crucial weapons systems.

Although all the Warring States military writings contain a few brief observations on chariot operations, only two, the Wu-tzu and Liu-t’ao, preserve significant passages. Primarily important for understanding the nature of the era’s conflict, they still furnish vital clues to the chariot’s modes of employment and identify a number of inherent limitations that would have inescapably plagued the Shang and Western Chou, long before chariots would explosively multiply to become the operational focus for field forces.

Chariots were considered one of the army’s core elements: “Horses, oxen, chariots, weapons, relaxation, and an adequate diet are the army’s strength. Fast chariots, fleet infantrymen, bows and arrows, and a strong defense are what is meant by ‘augmenting the army.’” Several passages indicate that chariots were viewed as capable of “penetrating enemy formations and defeating strong enemies.” Those used in conjunction with large numbers of attached infantry and long weapons were said not only to be able to “penetrate solid formations” but also to “defeat infantry and cavalry.” “When the horses and chariots are sturdy and the armor and weapons advantageous, even a light force can penetrate deeply.” “Chariots are the feathers and wings of the army, the means to penetrate solid formations, press strong enemies, and cut off their flight.” Before the advent of cavalry, they also acted as “fleet observers, the means to pursue defeated armies, sever supply lines, and strike roving forces.”

Passages in Sun Pin’s Military Methods and other works indicate that somewhat specialized chariots evolved in the Warring States, the basic distinction being between faster (or lighter) models and heavier chariots protected by leather armor and designed for assaults. A few of even greater size and dedicated function were thought capable of accomplishing even more: “If the advance of the Three Armies is stopped, then there are the ‘Martial Assault Great Fu-hsü Chariots.’” “Great Fu-hsü Attack Chariots that carry Praying Mantis Martial warriors can attack both horizontal and vertical formations.” Variants with a smaller turning ratio, known as “Short-axle, Quick turning Spear and Halberd Fu-hsü Chariots,” might be successfully employed “to defeat both infantry and cavalry” and “urgently press the attack against invaders and intercept their flight.”

Chariots were deemed astonishingly powerful: “Chariots and cavalry are the army’s martial weapons. Ten chariots can defeat a thousand men, a hundred chariots can defeat ten thousand men.” The Liu-t’ao’s authors even ventured detailed estimates of the relative effectiveness of chariots and infantry: “After the masses of the Three Armies have been arrayed opposite the enemy, when fighting on easy terrain one chariot is equivalent to eighty infantrymen and eighty infantrymen are equivalent to one chariot. On difficult terrain one chariot is equivalent to forty infantrymen and forty infantrymen are equivalent to one chariot.”

These are startling numbers, all the more so for having been penned late in the Warring States period when states still numbered their chariots by the thousands. Even allowing for exaggeration, given that the Liu-t’ao generally reflects well-pondered experience and is a veritable compendium of Warring States military science, the era’s commanders must have had great confidence in the chariot’s capabilities. Nevertheless, it might be noted that the great T’ang dynasty commander Li Ching, upon examining these materials in the light of his own experience at a remove of a thousand years, concluded that the infantry / chariot equivalence should only be three to one.

Chariots were also employed to ensure a measured advance in the Spring and Autumn, Warring States, and later periods when they no longer functioned as the decisive means for penetration. Li Ching’s comments about his historically well-known expeditionary campaign against the Turks indicate that even in the T’ang and early Sung they were still considered the means to constrain large force movements: “When I conducted the punitive campaign against the T’u-ch’üeh we traveled westward several thousand li. Narrow chariots and deer-horn chariots are essential to the army. They allow controlling the expenditure of energy, provide a defense to the fore, and constrain the regiments and squads of five.”

Although certainly not applicable to the Shang, chariots could also be cobbled together to provide a temporary defense, particularly the larger versions equipped with protective roofs. The authors of the great Sung dynasty military compendium, the Wu-ching Tsung-yao, after (somewhat surprisingly) commenting that “the essentials of employing chariots are all found in the ancient military methods,” concluded that “the methods for chariot warfare can trample fervency, create strong formations, and thwart mobile attacks. When in motion vehicles can transport provisions and armaments, when halted can be circled to create encampment defenses.”

Numerous examples of employing chariots as obstacles or for exigent defense are seen as early as the Spring and Autumn period. The later military writings cite several Han dynasty exploitations of “circled wagons” being employed as temporary bastions, including three incidents in which beleaguered commanders expeditiously deployed their chariots much as Jan Ziska would in the West to successfully withstand significantly superior forces. Sometimes the wheels were removed, but generally the chariots were simply maneuvered into a condensed array.


Based on burial patterns and traditional texts, it has been strongly, but perhaps erroneously, argued that three warriors manned the Chinese chariot from inception: an archer who normally stood on the left, the driver who controlled the horses from the middle of the compartment, and a warrior on the right who wielded some sort of shock weapon. Nevertheless, graves with only one or two warriors interred with a chariot are common in the Shang, the Tso Chuan occasionally refers to a fourth rider as if his presence was unusual but not exceptional, and a few Eastern Chou incidents note only two occupants.

Sun Pin’s comment that “those who excel at archery should act as the left, those who excel at driving should act as drivers, and those who lack both skills should act as the right” indicates that driving a chariot was also considered a specialized skill. Archery and charioteering would continue to comprise two of the six essential components of the chün-tzu or gentleman’s education known as the liu yi through the end of the Spring and Autumn period and into the early Warring States, even though Confucius personally disdained charioteering and has been generally perceived as disparaging warfare.

Contrary to possible impression that chariot service imposed lesser physical demands on the occupants than on the average foot soldier, outstanding conditioning and ability seem to have been required: “The rule for selecting chariot warriors is to pick men under forty years of age, seven feet five inches or taller, with the ability to pursue a galloping horse, catch it, mount it, and ride it forward and back, left and right, up and down, all around. They should be able to quickly furl up the flags and pennants, and have the strength to fully draw an eight picul crossbow. They should practice shooting front and back, left and right, until thoroughly skilled.” Incidents in the Tso Chuan confirm that great strength and courage were necessary for the chariot’s occupants to survive the rigors of battle and that constant training was needed to ride in a chariot and simultaneously shoot an arrow.

Apart from acting as an integral component capable of mounting penetrating attacks and engaging other chariots, the mobility provided by chariots had the potential to radically affect the course of battle. However, their actual mode of employment in the Shang remains uncertain despite deeply held traditional explanations. Nevertheless, only three possibilities exist: serving as a command platform for the officers at various levels; providing an elevated platform for observation and archery; and simply transporting the occupants to points on the battlefield, where they fought dismounted as in Western antiquity.

Even though oracular inscriptions have been interpreted as indicating that war chariots were sometimes called up in contingents of a hundred, because of their minimal numbers and novelty most Shang chariots must have been reserved for members of the ruling clan, high-ranking officials, and important officers dispersed across the battlefield. Initially they would have served as transport throughout the several days typically required to reach the battleground and then as mobile command platforms during the engagement.

When the effects of the axle, mounting bar, and box structure are included, the fighters standing in the chariot were elevated at least two and a half feet above the ground. This facilitated observing the battlefield and allowed ancient China’s highly skilled archers to shoot over the heads of any accompanying infantry at the enemy. However, their increased visibility also exposed the officers to attack and identified them as prime targets for enemy archers and spearmen.

Whether even the most skilled archers could maintain their vaunted performance levels despite the bouncing, jostling, and instability experienced while standing in a racing chariot seems problematic. Nevertheless, a few Spring and Autumn episodes recount instances of surpassing skill not just in shooting at ground forces, but also in targeting other warriors in rapidly moving vehicles. For example, even though he was fleeing during the Battle of An in 589 BCE, Duke Ch’ing of Ch’i refused to allow his archer to shoot Han Ch’üeh, who was pursuing him in another chariot, because he was a chün-tzu (gentleman). The archer therefore shot and killed the two occupants standing on either side of Han Ch’üeh.

Even if these incidents are not highly embellished or outright fabrications, they were probably recorded simply because of their exceptional or infrequent nature, thereby implying warriors on rapidly moving chariots rarely achieved such mastery. Questions remain, and the evidence is simply insufficient for a definitive conclusion, but insofar as the infantry dominated Shang and early Chou battlefields, chariot-mounted bowmen need not have been particularly accurate to strike down enemy soldiers with a flurry of arrows aimed in their general direction.

Considerable textual and archaeological evidence indicates that the chariot’s occupants had specialized functions, but the placing of complete weapons sets—a bow and dagger-axe—with both the archer and the warrior on the right side in one tomb at Hsiao-t’un suggests that both occupants may have functioned as archers in the Shang, when the shortness of their piercing weapons would have precluded direct chariot-to-chariot combat. On the other hand, the shields occasionally found in chariot graves, even though probably employed by the warrior on the right in conjunction with his dagger-axe, may have been used to protect the archer during battle, a mode known in the West as well.

After employing missile fire at a distance, closing with the enemy, and becoming caught in the melee, the chariot’s occupants would have had to dismount to engage in close combat because their shock weapons only averaged three feet in length, far too short to strike anyone while standing two and a half feet above the ground in a chariot compartment inset from the wheels and isolated by the bulk of the horses. Whether commanders continued to remain aloof or, far more likely, every Shang clansman was a fighter remains unknown. However, one warrior’s famous refusal to fight dismounted in the Spring and Autumn suggests that the prestige of being a chariot warrior carried considerable emotional weight and that later eras did not view chariots simply as battle taxis, whatever their function in Shang warfare.




Despite extensive speculation, how the chariot and any accompanying forces may have been coordinated remains murky and confused. Operationally the crucial question is whether the foot soldiers, if any, were nominally attached or were integrated in some sort of close spatial configuration that would enable the commander to direct them in the execution of basic tactics. Vestiges in the oracular records, bronze inscriptions, and traditional historical works note soldier-to-chariot ratios ranging from 10:1 to 300:1, with those preserved from the early Chou commonly being 10:1, a reasonable contingent to have been commanded by a chariot officer. However, significant variation is seen with the passage of time, and standing ratios for states in the late Spring and Autumn and Warring States periods lack consistency. Operational and historical discussions in the military writings describe attached contingents as numbering anywhere from 10 or 25 through 72 (plus 3 officers), even as many as 150, further complicating any assessment of prebattle formations and tactical deployments, both of which varied over time and were subject to local organizational differences, as in Ch’u.

Battle accounts and memorial inscriptions perhaps offer a more realistic picture. When chariots were few in comparison with total troop strength, as in the Shang, the overall battlefield ratio was doubtless much higher. According to Shang oracle inscriptions, 100 chariots were occasionally assigned to a fighting force of 3,000, with the chariots apparently acting as an operational group rather than being dispersed and accompanied by a designated number of infantry. Traditional accounts also indicate that the Chou vanguard that penetrated the Shang lines at the battle of Mu-yeh consisted of 3,000 tiger warriors accompanied by 300 chariots. Although this is a highly realistic 10:1 ratio, given the assault’s ferocity the 10 may have simply been running behind their assigned vehicles.

Based on 100 chariots being accompanied by 1,000 men, the famous Yü Ting, dating to the reign of King K’ang of the Chou (1005-978 BCE), similarly indicates a ratio of 10:1. The 100 chariots that had to be dispatched as reinforcements in the conflict with the marquis of E were also basically accompanied by 1,000 men, even though supplemented by another 200 in support. Despite being a Warring States work, the Kuan-tzu indicates that feudal lords were enfeoffed with 100 chariots and 1,000 men and that major lords were expected to supply 200 chariots and 2,000 men for military efforts, minor ones (hsiao hou) only 100 chariots and 1,000 men. However, several other ratios are also recorded, including 300 chariots accompanied by 5,000 men and 800 chariots with 30,000 men; a taxation system specifies seven armored soldiers and five guards for each of the chariot’s four horses, yielding an odd number of 48 men per chariot; and the state of Ch’i supposedly had 100,000 armored soldiers but only 5,000 chariots, still an astonishing number, for a 20:1 ratio.

With the inception of specialized vehicles for dedicated purposes, the ratios were apparently revised and troops attached only to certain vehicles such as the attack chariots. Although the configurations described in texts such as the Liu-t’ao may never have been deployed, they incontrovertibly conjoin ground troops with chariots, further supporting the link’s historicity. However, even if one or another of these ratios characterized Shang armies, the problem of how the accompanying infantry actually functioned in any era remains unresolved, because the early military writings fail to provide any useful information.

The only evidence to date for the existence of prescribed numbers of men being attached to Shang chariots is an intriguing aggregation of graves at Yin-hsü that has rather controversially been interpreted as an array intended to deliberately depict the composition of a chariot company and therefore prompted claims that the late Shang fully integrated ground troops with chariots in an organized manner. The excavators have categorized the graves into eleven types based on the occupants and accompanying artifacts, ten of which are seen as having direct martial implications. Mainly slain by decapitation, all the occupants appear to have been sacrificed on site rather than brought for interment after dying in battle or other violent circumstances.

Five chariot graves laid out in a roughly T-shaped pattern comprise the core of the contingent. Two are offset at the ends of the T; the other three are arrayed in an essentially vertical line that commences somewhat below, giving the group a south-facing orientation. On the assumption that each grave contains one chariot and a crew of three warriors armed in the traditional manner, the site apparently preserves the first concrete evidence that the five-chariot company, long pronounced as a virtual matter of faith, actually existed.

A horizontal line of five graves, each of which contains five warriors with red-pigmented bones, runs across in front of the chariot group, apparently the contingent’s vanguard. Although analysts immediately concluded that they represent the so-called provocateurs employed in later warfare to enrage the enemy and raise the fighting spirit of one’s own troops, the sole reference found in the oracle bones to dispatching a small forward contingent shows a unit of thirty horses being employed as an advance element. Therefore, if the practice of deliberate provocation in early Chinese warfare ever existed other than in the imagination of historical writers, its inception should be traced to the Spring and Autumn, not projected back into the Shang, with small units subsequently being employed in Warring States warfare to deliberately probe the enemy.

More or less to the right of the chariot group lie three groups of graves containing a total of 125 young, strong fighters, some complete but others just skulls, variously distinguished by the presence of ritual objects, red pigment on their bones or skulls, and some sort of headband, all deemed indicative of rank. Without becoming entangled by the numerous details engendered by these finds and the several controversies prompted by their imaginative interpretation, it appears that these burials constitute a contingent assigned to the chariot company.

This force apparently was based on the squad of five, the standard number that would essentially underpin Chinese military structure for the next three millennia. Each squad had an officer, but rather than twenty-five squads there were only twenty, giving a base of 100, a number that well coheres with the Shang practices and penchant for units of 100 and 1,000. Because four higher-level commanders are distinguishable, the twenty squads were apparently grouped into units of five. With the addition of the contingent’s commander, the total reaches the subsequently sacrosanct number of 125 and would essentially cohere with the Chou Li articulation of 100 men to a tsu, but only on the assumption that the officers were not encompassed by the century.36 However, if the officers are excluded and the twenty-five men in the vanguard seen as an integral part of the infantry contingent, the number would again reach the magic 125.

A third aggregation of graves marked by a wide variety of utensils and weapons individually interred with single bodies has been interpreted as representative of the full range of support personnel required by the chariot company. Apart from two officials who seem to have had responsibility for overseeing the food and beverages, they appear to have been divided into groups of five and seven, with the former entrusted with responsibility for the weapons, the latter various other vessels, including the portable stoves and associated utensils. Further confirmation is seen in a grave that contains ten sheep, apparently the contingent’s mobile food supply.

It has been concluded that these graves show that Shang contingents were systematically organized around the chariots and that the additional ground forces were merely supplementary or auxiliary. However, even though the burials certainly seem to represent a deliberate array intended to honor the king, a few objections have been raised and a number of questions remain. Most prominent among the latter is whether, as evidenced by a certain arbitrariness in selecting the graves for inclusion, the military organization discussed in Warring States writings is not being projected back onto the Shang.38 In addition, issues of chronology have apparently been ignored in several instances, because a couple of the graves overlap or intrude upon others, evidence of either sloppiness (which is unlikely given the precision of Shang palace and tomb construction) or ignorance of previous burials, synonymous with the absence of any intent to execute a grand design.

Somewhat broader questions might also be raised, including whether this array is merely an idealization, a deployment appropriate solely for the march rather than any operational utilization just as units paraded in review throughout history, or even a form of organization that only characterized the king’s personal force. This would explain the apparent overstaffing with support personnel, an extremely inefficient and unrealistic approach for a field army, even though later traditions suggest a small contingent entrusted with such responsibilities was attached to chariot squads.

How this contingent would operate on the battlefield also remains unknown. Presumably the vanguard of twenty-five protected the chariots during the archery exchange, but if they raced ahead on foot during the move to melee, the speed advantage of the chariots would be lost. If the chariots dispersed during the engagement, these elite ground fighters could have been assigned as accompanying infantry, one squad to a chariot, for protection. But similar questions also plague any understanding of the function of the 125 men deployed on the right, whether they operated in aggregate as close support, as dispersed units of 25, or shed any connection with the chariots in the chaos of battle, the most likely possibility.

Despite these significant issues, with allowance for possible chronological problems and considerable arbitrariness in selecting the graves, it appears that the aggregate preserves the elements of a late Shang military formation. Nevertheless, the foot soldiers constitute the true power, and as the era of chariot-to-chariot combat had not yet dawned, they no doubt bore the brunt of the fighting subsequent to the initial archery exchanges. Rather than being a key fighting element, the dagger-axe warrior on the chariot probably acted as a bodyguard for the archer, while the driver merely controlled the chariot, presumably under the direction of the archer.

Formations for advancing, prebattle methods of deployment, and alignments for combat (such as seen in the Liu-t’ao) were eventually developed for the chariots, infantry, and chariots intermixed with infantry, possibly beginning in the early Chou. However, dismissing assertions that King T’ang employed nine chariots in a goose formation, the only pre-Warring States confirmation that chariot formations had begun to evolve is a debate found embedded in the Tso Chuan over whether it or the crane formation should be employed for the chariots. Moreover, assertions that the platoons were deployed to the left and right only increase the degree of puzzlement, because in the reality of fervent combat the chariots would have had to shed them to act as a rapid penetrating force, enjoy the requisite freedom of maneuver, and pursue fleeing enemies.

Unless the chariots were broadly dispersed, conjoining more than a few men to them would simply have clogged the immediate area. Mobile combat becoming impossible, the chariot would have been reduced to functioning as an archery platform and localized command point. Presumably this is the reason the Liu-t’ao’s authors subsequently devised a rather dispersed deployment: “For battle on easy terrain five chariots comprise one line. The lines are forty paces apart and the chariots ten paces apart from left to right, with detachments being sixty paces apart. On difficult terrain the chariots must follow the roads, with ten comprising a company, and twenty a regiment. Front to rear spacing should be twenty paces, left to right six paces, with detachments being thirty-six paces apart.” This dispersed formation would allow enemy chariots to pass through while providing the necessary operational space for maneuver.

All the military writers considered open terrain to be ideal ground for chariot operations. The great middle Warring States strategist Sun Pin therefore continued to advocate exploiting it with chariots (and cavalry) even when confronted by superior infantry strength:

Suppose our army encounters the enemy and both establish encampments. Our chariots and cavalry are numerous but our men and weapons few. If the enemy’s men are ten times ours, how should we attack them?

To attack them carefully avoid ravines and narrows. Break out and lead them, coercing them toward easy terrain. Even though the enemy is ten times [more numerous], [easy terrain] will be conducive to our chariots and cavalry and our Three Armies will be able to attack.

Tso Chuan depictions of Spring and Autumn military clashes frequently reduce the chariot’s role to facilitating the exchange of provocative taunts and initiating individualized combat between chivalrous warriors—highly romanticized, unrealistic portraits of bravado that had little impact on the battle’s outcome—yet instances of massed chariot charges undertaken across relatively open terrain are also recorded. The ultimate chariot clash occurred in the spring of 632 BCE at the epochal Battle of Ch’eng-p’u, wherein a coalition of older, northern states vanquished imperious Ch’u’s initial thrust out of the south into the Chinese heartland. Although the rarity of Shang and steppe chariots precludes such clashes from having arisen in antiquity, it can still be pondered as an example of the potential effectiveness of employing chariot contingents en masse. Chin’s commanders secured victory through several unorthodox measures, including targeting the enemy’s weakest components:

When they had finished cutting down the trees, Chin’s forces deployed north of Hsin. Hsü Ch’en, in his role as assistant commander for Chin’s Lower Army, deployed opposite the forces from Ch’en and Ts’ai [allied with Ch’u].

When Tzu Yü, accompanied by six companies of Juo-ao clan troops, assumed command of Ch’u’s Central Army, he said: “Today will certainly see the end of Chin!” Tzu Hsi was in command of Ch’u’s Army of the Left and Tzu Shang their Army of the Right.

Hsü Ch’en covered their horses in tiger skins and initiated the engagement by assaulting the troops from Ch’en and Ts’ai. Their troops fled and Ch’u’s right wing crumbled.

Hu Mao [of Chin] set out two pennons and withdrew his forces. Meanwhile Luan Chih [of Chin] had his chariots feign a withdrawal, dragging faggots behind them. When Ch’u’s forces raced after them, Yüan Chen and Hsi Chen cut across the battlefield to suddenly strike them with the Duke’s own clan forces. Hu Mao and Hu Yen [of Chin] then mounted a pincer attack on Tzu Hsi’s army that resulted in Ch’u’s left wing being shattered and Ch’u’s forces decisively defeated. Tzu Yü gathered his clan forces and desisted from further action, avoiding personal defeat.

Two actions shaped the battle’s course and determined its outcome. First, Hsü Ch’en’s elite warriors initiated contact with an unexpected, concentrated thrust while the preliminary posturing was probably still under way. Its surprising fervency shattered Ch’u’s insecure allies deployed on the right wing, giving Chin’s forces unconstricted mobility and exposing the field to flanking attacks. Second, coordinated feigned retreats by Chin’s right wing and elements of the left wing that had not participated in the initial thrust, masked by clouds of dust that were deliberately created by the branches that had been cut down and attached to the rear of the chariots, easily drew the overconfident Ch’u armies forward in a disordered attack. Presumably Ch’u’s central forces, under Tzu Yü’s personal command, also moved forward to exploit Chin’s retreat and engage any remaining center forces; otherwise, the Tso Chuan account would not have noted that Tzu Yü “gathered his personal forces and desisted.”

In the swirl of battle all the actions were obviously performed by chariot forces exploiting their mobility, any infantry forces that were present having been left behind, perhaps as ensconced forces defending fallback positions. Therefore chariot clashes of the type envisioned by many traditional historians, reputedly common in the second millennium BCE in the West, not only could, but did, occur. The Art of War, which presumably reflects late Spring and Autumn and early Warring States warfare, specifically speaks of “chariot encounters,” implying that these were separate actions that didn’t entail infantry participation: “In chariot encounters, when ten or more chariots are captured, reward the first to get one. Change their flags and pennants to ours, intermix and employ them with our own chariots.”

Somewhat later, while admonishing his 50,000 troops and 500 chariot commanders prior to engaging Ch’in in a major clash, Wu Ch’i proclaimed: “All the officers must confront, follow, and capture the enemy’s chariots, cavalry, and infantry. If the chariots do not make prisoners of the enemy’s chariots, the cavalry not make prisoners of the enemy’s cavalry, and the infantry not take the enemy’s infantry, even if we forge an overwhelming victory no one will be credited with any achievements.” His pronouncement would seem to imply that the three component forces targeted their counterparts rather than engaging in a general melee or that the infantry was tied to the chariots.

As their numbers increased in later eras, rather than just being employed as a single mass, chariots were segmented into operational units ranging from a basic three—center, left, and right—to more tactically specific contingents controlled by drums. However, the low number of just 100 per army in the Shang probably precluded any subdividing, though some analysts have suggested groups of 30 may have been employed. The Shang and early Chou probably mark the transitional period when the chariot’s role had not yet been defined and individual chariots had not yet been integrated with accompanying infantry, a stage that would require rethinking, apportionment, and the development of training and operating procedures to avoid the battlefield chaos that would inevitably result if the chariots were merely added to the mix of combatants.


WWI [blue] and WWII [green] U-boats Compared.

When World War I commenced in early August 1914, the German Imperial Navy had not completed its big-ship buildup. The High Seas Fleet was therefore not strong enough to sail out and confront Britain’s powerful Grand Fleet in a single, decisive battle. Nor was the Royal Navy capable of mounting a decisive attack on the Imperial Navy in its home waters. Hence a big-ship standoff ensued, during which the opposing admirals schemed ways to entrap the other’s fleet in the confined waters of the North Sea by guile and deception. The naval war between these two great maritime powers thus proceeded in a curious, cautious, and unforeseen manner. There was only a single major surface-ship battle—Jutland—and it was brief and inconclusive.

Early in the war both Germany and Great Britain deployed submarines on offensive missions. The initial forays were remarkable. German U-boats sank three British heavy cruisers (Aboukir, Hague, and Cressy) and two light cruisers (Pathfinder, Hawke) with the loss of over 2,000 men. British submarines sank the German light cruiser Hela. Both navies were thus compelled to view the submarine as a grave new threat and they reacted accordingly. The British Grand Fleet withdrew temporarily from its North Sea base in Scapa Flow to safer waters in north Ireland. The German High Seas Fleet sharply curtailed operations in its home waters, the Helgoland Bight.

The British imposed a naval blockade against Germany with the aim of shutting off the flow of war matériel. The British did not strictly observe the prize laws; even neutral ships loaded merely with food were harassed, blocked, or turned back. In retaliation, the German Naval Staff authorized German U-boats to harass Allied merchant shipping. On October 20, 1914, a U-boat, observing the prize laws, stopped, searched, and scuttled the 866-ton British freighter Glitra off Norway. A week later another U-boat, operating in the English Channel, torpedoed without warning a French steamer, Admiral Ganteaume, which was believed to be laden with troops and therefore fair game under the prize laws. In fact the ship was jammed with 2,400 Belgian refugees, including many women and children. Fortunately, it did not sink.

These two U-boat attacks on unarmed merchant ships carried profound implications for the island nation of Great Britain, entirely dependent upon her vast mercantile fleet for survival. An organized U-boat guerre de course might be ruinous. Accordingly, the British government denounced the attacks as illegal, treacherous, piratical, and immoral. Ship owners, merchants, and insurance carriers the world over joined the chorus of denunciation.

The Central Powers, composed of Germany and the Austro-Hungarian Empire, had planned to defeat France in a quick campaign, then turn about and crush czarist Russia. But the plan went awry. The armies in France bogged down in bloody trench warfare; Russia attacked from the east, creating a two-front war. Not having anticipated a long war, the Central Powers had not stockpiled large supplies of war matériel. As a result of the British blockade, by early 1915 the Central Powers were running out of iron ore and oil and other war essentials as well as food.

To this point U-boats, strictly observing the prize rules, had sunk ten British merchant ships for about 20,000 tons. Owing to the shortage of torpedoes—they were still virtually handmade—most of these sinkings had been achieved by gunfire or forced scuttling. The surprising ease of these successes had led the senior German admirals to conclude that if the prize rules were relaxed, even the small number of U-boats available for distant operations could impose an effective counterblockade on the British Isles. The mere appearance of a single U-boat, manned by only two dozen men, whether successful in the attack or not, caused great psychological alarm, compelling the enemy to devote a hugely disproportionate share of his manpower and resources to neutralize the threat. All this would severely impair Britain’s ability to carry on the war, the advocates postulated, and might result in a tit-for-tat deal in which Britain agreed to lift its blockade of Germany.

Neither the Kaiser nor his Chancellor was keen on the proposal. Germany had already incurred heavy criticism from many quarters for sinking merely ten merchant ships. A relaxation of the prize rules would doubtless draw even harsher criticism, especially from neutral nations such as the United States, which had a substantial financial interest in sea commerce and might retaliate by entering the war. Moreover, the number of U-boats available for blockading the British Isles seemed too slight. To announce a blockade and fail abjectly would be worse than no attempt at all.

And yet the proposal would not die. Its advocates argued, not without justification, that the moral arguments were no longer relevant. In its ruthless blockade of Germany, they insisted, Britain had repeatedly violated the prize rules and other traditions protecting sea commerce, most notably in refusing the passage of neutral ships carrying only food. This line of reasoning, and other arguments, finally persuaded the Kaiser and his Chancellor to authorize a U-boat blockade of Great Britain.

The stage was carefully set. The Kaiser publicly declared that from February 18, 1915, onward, the waters around the British Isles were to be considered a “war zone.” Prize rules would no longer be strictly observed. British and French merchant vessels would be sunk without warning or exceptional measures to provide for the safety of the crews. Care would be taken to spare neutrals not carrying contraband, but all neutrals would sail the waters at their own peril. U-boat skippers, the Kaiser further declared, would not be held responsible if “mistakes should be made.”

So was launched history’s first systematized submarine guerre de course. The initial results were less than impressive. In the month of February 1915, the twenty-nine U-boats of the German submarine force sank 60,000 tons of merchant shipping; in March, 80,000 tons. The weakness of the blockade lay in the small number of U-boats available. Owing to the time spent going to and from German bases and in refit, after the initial deployment it was difficult to establish organized U-boat patrol cycles that kept more than six or seven U-boats in British waters at any given time. Notwithstanding the fear and confusion and diversion of resources it precipitated, the first U-boat blockade did not achieve its main goal. First Lord Churchill declared the blockade a failure; British imports in 1915 exceeded those of 1913. The British government refused to entertain any suggestion of lifting the blockade of Germany.

With each merchant ship sinking, the cries of moral indignation intensified. Three sinkings in particular outraged the Americans: the 32,500-ton Cunard liner Lusitania on May 7, with the loss of 1,198 passengers (128 Americans) and crew; the 16,000-ton White Star liner Arabic on August 19, with the loss of 40 passengers (3 Americans); and the liner Hesperian on September 9. So violent was the reaction in the United States (U-boat crews make war “like savages drunk with blood” declared The New York Times), that in early September 1915 the Kaiser called off the blockade of Great Britain and sent many more U-boats to the Mediterranean Sea, where the hunting was less controversial and no less lucrative and there were few Americans.

With victory no closer for the Central Powers, at the beginning of 1916 the chief of the German naval staff, Admiral Henning von Holzendorff, and his Army counterpart urged the Kaiser to authorize a renewal of the British blockade. The Navy now had almost twice as many U-boats in commission (fifty-four versus twenty-nine in 1915) and ever more U-boats were coming off the slipways. The Kaiser was tempted, but the Chancellor and Foreign Minister objected, fearful of another Lusitania, which would almost certainly bring America into the war. After days of vacillation, the Kaiser sided with the Navy, but he imposed complicated restrictions. No passenger liners of any nationality were to be attacked anywhere. No cargo ships or tankers except those unmistakably armed could be attacked outside the war zone.

The renewed blockade commenced in February 1916. Notwithstanding the restrictions and complexity of the rules, all went well for the U-boats for two months: 117,000 tons sunk in February, 167,000 tons in March. Then came another costly error. On March 24 a U-boat mistook the 1,350-ton English Channel passenger ferry Sussex for a troopship and torpedoed it. The Sussex did not sink, but about eighty people were killed in the explosion, including twenty-five Americans. In response to the renewed cries of indignation and a blistering note from Washington threatening to sever diplomatic relations, the Kaiser backed down once more and, on April 24, ordered U-boats in waters of the British Isles again to adhere strictly to the prize rules. As a result, merchant ship tonnage sunk by U-boats in British waters fell sharply for the next four months.

The German submarine force had grown to substantial size by September 1916: a total of 120 boats of all types, many with larger 105mm (4.1”) deck guns. Again the military staffs urged the Kaiser to exploit this force to the fullest. Again the Kaiser vacillated, and finally yielded, but with yet a new set of rules. Skippers were to conduct only restricted submarine warfare (by prize rules) in waters of the British Isles, where there were numerous American and other neutral ships, but they were permitted to wage unrestricted submarine warfare in the Mediterranean. This third and most intense phase of the restricted U-boat war, October 6, 1916, to February 1, 1917, was highly productive for the Germans. The U-boats sank about 500 British merchant vessels for about 1.1 million tons, raising the total bag for 1916 to about 2.3 million tons, most of that of British registry.

By early 1917 the ground war had become a brutal and fruitless bloodletting for the Central Powers and there was deep and widespread unrest at home. The German military staffs urged the Kaiser to authorize unrestricted submarine warfare in all oceans and seas. Using the results achieved in the fall of 1916, the larger number of U-boats available, plus nearly ninety new boats that were to be commissioned in 1917, the naval staff calculated that with an unrestricted U-boat campaign, nearly half of Britain’s still large merchant fleet could be wiped out within five or six months, rendering her not only incapable of prosecuting the war on the continent but also leaving her population in a condition of starvation and rebellion. America be damned, the naval staff said. If she came into the war, Germany would have enough U-boats (about seventy ready for operations in the British Isles alone) to sink all her troop and supply ships before they reached Europe. By that time, too, there was no shortage of German submarine torpedoes; U-boat skippers did not have to rely so heavily on deck guns.

Turning aside peace feelers from President Wilson and others, the Kaiser approved this proposal. He announced to the world that commencing February 1, 1917, U-boats would sink on sight every merchant ship found in British territorial waters. At the same time, he assured the German military staffs that there would be no more pussyfooting or backing down, and he promulgated a radical role reversal for the surface ships of the Imperial Navy: Henceforth they were to support U-boats, rather than the other way around. “To us,” he said, “every U-boat is of such importance that it is worth using the whole available fleet to afford it assistance and support.”

Germany launched this all-out submarine guerre de course in the British Isles with multiple attacks conducted simultaneously with “utmost energy” by about sixty U-boats. To minimize detection by Allied aircraft and submarines, and counterfire from merchant ships, and to take advantage of higher speed for escape, U-boat skippers attacked at night while on the surface. The results were spectacular: 540,000 tons sunk in February, 594,000 tons in March, and an appalling 881,000 tons in April. During April alone—the grimmest month of the U-boat war—the Germans sank 423 merchant ships, of which 350 were British.* Moreover, as anticipated, the campaign scared off most of the many neutral ships trading with Great Britain.

Reflecting the growing anger and outrage in America, President Wilson reacted firmly and militantly to this all-out U-boat campaign. On the third day, February 3, 1917, he broke off diplomatic relations with Germany. At his request, on April 6 the Congress declared war on the Central Powers.

At the beginning of the war the Royal Navy possessed no special countermeasures to fight submarines. Naval tacticians wrongly assumed that since submarines would of necessity spend most of the time on the surface, they would be easy prey for gunfire and ramming. This wrong view was reinforced when the British cruiser Birmingham rammed and sank U-15, the first U-boat to be lost. But in the five months of warfare in 1914, the Royal Navy positively sank only one other U-boat, U-18. Three other U-boats were lost in 1914 (for a total of five) to unknown causes, probably mines.




Sinking of the Linda Blanche out of Liverpool by Willy Stöwer

Beginning in 1915, when shipping losses to U-boats began to climb significantly, the Admiralty diverted a substantial portion of its existing resources to antisubmarine warfare (A/S in Britain, ASW in America) and asked scientists, engineers, academics, and others to help develop ways to destroy U-boats. In the belief that the best defense was a strong offense, the chief ASW weapons to emerge in World War I were these:

• SURFACE HUNTERS. The Admiralty sent scores, then hundreds, then thousands of surface ships out offensively scouring the oceans for U-boats. These vessels included destroyers, frigates, sloops, trawlers, yachts, and heavily armed raiders (Q-ships) disguised as tramp steamers. Some vessels were fitted with crude hydrophones—passive underwater listening devices—which could detect the engine noise of a surfaced U-boat, but only if the hunting vessels were not moving.

In 1916 many of these offensive ASW ships were armed with a new weapon called the depth charge. The best of these underwater bombs, derived from mines, contained 300 pounds of TNT or Amatol and were fitted with hydrostatic fuses which could be set to detonate the charges at 40 and 80 feet, and later 50 to 200 feet. Since early depth charges were rolled from stern tracks (or racks) and exploded at shallow depth, the attacking vessel had to put on maximum speed or risk severe damage to its stern. Therefore, slower vessels could not use the 300-pound depth charges until fuses with deeper settings had been developed. In all of 1916, British naval forces sank only two U-boats by depth charge. In 1917 and 1918, when depth charges had been improved and were much more plentiful, the kill rate by this weapon increased significantly.

• AIRBORNE HUNTERS. When the war commenced, the aviation age was merely a dozen years old. The Royal Navy had acquired about fifty seaplanes and seven nonrigid airships, called “blimps,” to scout for enemy naval forces. Some of these aircraft were diverted to U-boat hunting but, owing to the unreliability of engines, slow speed, limited fuel capacity, tiny bomb loads, and other factors, they were useless against U-boats. It became apparent, however, that when an aircraft appeared near a U-boat, it dived and became essentially immobile. Hence air patrols were useful for forcing U-boats under, thus enabling ships to skirt the danger area and avoid attack. In 1915 the Royal Navy acquired much improved seaplanes (the American-designed Curtiss American) and blimps in greater numbers. These were armed with impact-fused 100- or 520-pound bombs or 230-pound ASW bombs with delayed-action fuses that exploded at a water depth of seventy feet, but the U-boat kill rate by aircraft remained essentially zero.

• SUBMERGED HUNTERS. On the theory that it was wise to “send a thief to catch a thief,” the Royal Navy saturated German home waters with submarines equipped with hydrophones. The early patrols produced no confirmed kills, but the presence of British submarines in German waters, including the Baltic Sea, where German submariners trained, caused great anxiety and disrupted routines. Beginning in 1915, British submarines began to torpedo U-boats in significant numbers. The Admiralty designed and produced a small submarine (R class) specifically for U-boat hunting but it came too late. Had British torpedoes been more reliable, the submarines doubtless would have sunk many more U-boats.

• MINES. From the first days of the war both sides employed moored contact mines, planted in shallow water, usually defensively but often offensively. Defensive minefields were sown to prevent enemy forces from penetrating one’s coastal waters for shore bombardment, interdiction of shipping, or invasion. Such minefields were charted and planted with great care, leaving secret safe lanes for friendly shipping and naval forces. In order to attack British shipping, U-boats often had to negotiate the periphery or heart of defensive minefields, a hazardous undertaking. Many U-boats strayed into British minefields or hit live mines that had drifted their moorings or had broken loose. Offensive mining was more complicated and often hit-or-miss. Surface vessels, operating under cover of darkness in great haste, planted mines in likely spots such as sea-lanes or sometimes even in the safe lanes of the defensive minefields, to catch opposing naval vessels or merchant ships by surprise. Later in the war, both sides employed submarines for minelaying, combining two much-feared naval weapons.

To prevent U-boats from reaching the Atlantic via the English Channel, the British sowed lines of mines across it from Dover, England, to Cape Gris-Nez, France. However, in 1915 and 1916, British contact mines were defective, and not until the Admiralty copied and mass-produced the standard German contact mine could the Dover “field” be depended upon to block the passage of U-boats. When the Dover field was finally effective, it forced U-boats destined for the Atlantic to go northabout Scotland, adding about 1,400 miles (and about seven days) to the voyage.

After the United States entered the war and offered the Royal Navy a secret mine with a magnetic fuse, the Allies put in motion a grandiose scheme to plant 200,000 such mines across the top of the North Sea from the Orkney Islands to Norway. Although American and British forces planted about 80,000 mines in this so-called Northern Barrage, most of these mines were also defective and, other than frayed nerves, caused the Germans small harm. Even so, Allied mines in all areas ranked high as U-boat killers.

• RADIO INTELLIGENCE. When the war began, radio transmission or wireless telegraphy (W/T) was a new military technology at which the British excelled. Taking advantage of a lucky capture of German naval codebooks, as well as an appalling lack of sophistication in German radio procedures and security, the British thoroughly penetrated German naval communications. The British first perfected Radio Direction Finding (RDF) to pinpoint and identify German shore- and sea-based transmitters. Utilizing the captured codebooks, they “read” on a current basis most German naval transmissions. This priceless intelligence enabled the Admiralty’s secret signals-intelligence branch (known as Room 40) to track U-boat operations to a remarkable extent. A British historian wrote that by “early 1915, Room 40 knew the total strength of the U-boat fleet, the rate at which it was growing … the composition of each flotilla … the number of U-boats at sea or in port, and when and if it put to sea … losses, as evidenced by the failure of a U-boat to return, and in most cases, the size of the [U-boat] threat in any particular area.”

Still, these many and varied ASW measures were absurdly inadequate. In all of 1915 the Germans lost merely nineteen U-boats while adding fifty-two boats to the force. In 1916 the Germans lost twenty-two boats while adding 108 boats. Notwithstanding a massive British antisubmarine effort, during the first four months of 1917, the Germans lost only eleven U-boats. To then, the average monthly U-boat loss rate had been only 1.7, a continuing losing battle for Britain because the Germans were producing seven or eight new boats per month.

In the wake of the spectacular shipping losses in April 1917, Britain’s new Prime Minister, David Lloyd George, urged the Admiralty to organize British shipping into convoys, escorted by destroyers, frigates, sloops, and other ASW craft. This was hardly a new idea; defense of sea commerce by convoy was as old as the sail and, as the British naval historian John Winton put it, “as natural and as obvious a tactic as, say, gaining and keeping the weather gauge.”

The Royal Navy had opposed the formation of convoys for numerous reasons. The principal reason, Winton wrote, was that Royal Navy officers had forgotten their history—that the main purpose of the Royal Navy was to protect Britain’s sea trade. Imbued with the aggressive doctrines of the American naval theorist Alfred Thayer Mahan (and kindred souls), who postulated that control of the seas could most effectively be insured by husbanding naval assets for a single, decisive, offensive naval battle with the enemy, they opposed the diversion of naval resources to convoying, which they viewed as mundane and defensive and which, if adopted, would be an admission that Britain had, in effect, lost control of the seas to an inferior naval power.

There were other reasons. First, notwithstanding huge losses of merchant ships on their very doorstep, the Royal Navy continued to grossly underestimate the overall effectiveness of the U-boat campaign on British maritime assets. Second, the admirals insisted convoys were enormously inefficient, compelling faster ships to reduce speeds to those of slower ships, overwhelming seaport facilities during loading and unloading periods, and posing difficult organizational problems in distant, neutral ports. Third, the Admiralty doubted the ability or desire of merchant-ship captains to accept or to follow orders or to station-keep in the required tight zigzagging formations at night or in inclement weather. Fourth, the admirals held, the concentration of merchant ships into a single large body presented U-boat skippers with richer targets, which they were not likely to miss, even with poorly aimed or errant torpedoes.

With the assistance of American naval power, the Admiralty finally—and reluctantly—agreed to a test of inbound convoying in the Atlantic. The first convoy, consisting of sixteen ships, sailed from Gibraltar to the British Isles on May 10, 1917; the second of twelve ships from Norfolk, Virginia,* on May 24. The Gibraltar convoy arrived in good time without the loss of a ship. The Norfolk convoy, escorted by the British cruiser Roxburgh and six American destroyers, ran into minor difficulties. Two of the dozen ships could not maintain the convoy’s 9-knot average speed and fell out. One of these was torpedoed going into Halifax, Nova Scotia. However, the other ten ships crossed the Atlantic in foggy weather, maintaining tight formation, zigzagging all the way, and arrived safely in the British Isles.

With the results of these tests and other data in hand, in August 1917—the beginning of the fourth year of the war—the Admiralty finally adopted the convoy system. It was a smashing success. By October over 1,500 merchant ships in about 100 convoys had reached the British Isles. Only ten ships were lost to U-boats while sailing in these convoys: one ship out of 150. By comparison, the loss rate for ships sailing independently (inbound and otherwise) was one in ten. By the end of 1917, almost all of the blue-water traffic was convoyed. These convoys had been instituted in the nick of time; U-boats sank nearly 3,000 ships for 6.2 million tons in 1917, most of them sailing independently. The historian Winton wrote: “Convoying did not win the war in 1917. But it did prevent the war from being lost in 1917.”

A U-boat skipper remembered the impact of convoying on the German submarine force. Convoying, he wrote, “robbed it of its opportunity to become a decisive factor.” He continued: “The oceans at once became bare and empty; for long periods at a time the U-boats, operating individually, would see nothing at all; and then suddenly up would loom a huge concourse of ships, thirty or fifty or more of them, surrounded by a strong escort of warships of all types.” The solitary U-boat, he went on, which “had most probably sighted the convoy purely by chance,” would attempt to attack again and again, “if the commander had strong nerves” and stamina. “The lone U-boat might sink one or two of the ships,” he concluded, “or even several; but that was a poor percentage of the whole. The convoy would steam on.”

During the final twelve months of the war, convoying became the rule rather than the exception. The British and American navies established large organizations to administer convoys and provided surface and, where feasible (close to land), aircraft escorts, armed with new and improved aerial bombs. In many instances, intelligence from Room 40, accurately identifying U-boat positions, enabled the authorities to divert convoys away from U-boats. After the full convoy system was in place (outbound from the British Isles as well as inbound) in 1918, total shipping losses fell by two-thirds from 1917: 1,133 sunk. Of these, 999 sailed independently. In the ten months of naval war in 1918, only 134 ships were lost in convoy.

Mid-Western American Civil War I

Shiloh was an unexpected battle in an unforeseen place, to which Grant’s army was drawn down the Tennessee River by his victories at Forts Henry and Donelson. Its effect was to open up a new front in the centre of nineteenth-century America, in Tennessee, a crucial state for both Union and Confederacy, since it borders Alabama, Mississippi, and Georgia, and, across the Mississippi, Arkansas and Missouri. To the north it gives on to Illinois, Indiana, and Ohio, all solid and important Union territory, which was to be raided by Morgan’s cavalry in July 1862; eastward it also offered a route into South Carolina. Eastern Tennessee itself, covered by the tail of the Appalachians, was solidly pro-Union, the largest pocket of Unionist loyalty inside the Confederacy; being mountainous and relatively infertile, it was a region of subsistence farming almost without slaves.

At the beginning of the war, Tennessee was spared an outbreak of fighting because the state government, while not seceding, concluded an alliance with the Confederacy. This transparently evasive measure did not stick. Washington continued to regard Tennessee as a state of the Union and its elected representatives continued to sit in both houses of Congress. While the Confederacy also deemed Tennessee a member state, its political leaders formed at best a government in exile. The eastern counties had voted strongly against secession when a convention was held. Richmond was determined to fight to keep Tennessee out of the Union camp, but at first there were almost no opposing forces inside the state until Grant and Halleck appeared to organise the Army of the Tennessee, which was eventually confronted by Bragg’s Army of Tennessee. Thus a new front was opened, or a “line” as fronts were called in the Civil War. The term “front” did not come into use until the First World War, when it was borrowed from the vocabulary of meteorology, on the analogy with weather fronts of low and high pressure. There was an obvious front in Virginia in the region of high pressure between Washington and Richmond. Not so in the West, where troop density was low and there were few cities of importance. Yet gradually central Tennessee would become what a later generation would recognise as a distinctive front, whose crucial features were rivers and railroads. The key to organising the war in the region was to concentrate the scattered forces of the two sides and form campaigning armies. The main components were with Halleck at St. Louis and Beauregard’s Confederate survivors of Shiloh. Other Confederate troops were reaching Tennessee from the Atlantic coast and also from Arkansas. During April 1862, Halleck succeeded, by summoning Pope from the Mississippi front at New Madrid and Island No. 10 and Grant from near Shiloh, in forming an army of 100,000 men. Its generals included many of the Union’s future leaders, including not only Grant, but also Sherman and Sheridan, Don Carlos Buell, Rosecrans, and George Thomas, the “Rock of Chickamauga.” The Union army in the West was organised by Halleck; its three armies were named after the region’s major rivers, the Army of the Tennessee under Grant, the Army of the Ohio under Buell, and the Army of the Mississippi under Pope. Laymen may misunderstand the use of the term “Army.” It was entirely organisational and hierarchic. Battalions made regiments (two), regiments made brigades (three), brigades made divisions (three or more), divisions made corps (two or more), corps made armies (two or more). On the Union side armies were called after the river near which they operated (for example, the Potomac). In the Confederacy, armies were called after the region in which they operated (e.g., Northern Virginia). Armies also tended to be regional in composition, so that the Armies of the Tennessee and the Ohio, having been raised in the Midwest, were largely recruited from midwesterners.

Halleck opened his campaign against Beauregard by advancing on Corinth, a small railroad town in northern Mississippi which the Confederates had fortified. Intimidated by news of Halleck’s approach, Beauregard abandoned Corinth in late May and retreated southward. His army was much depleted by sickness and desertion. He nevertheless initiated a threat to middle Tennessee and Kentucky, and Halleck, rather than engage him, devoted his energies to fortifying Corinth further, thereby converting it into one of the strongest places anywhere in the zone of war. Halleck apparently expected the Southern troops to offer him an advantage by attacking his fortifications, but they did no such thing, instead attacking the Union railroads and threatening advances towards the lower Southern states. Halleck distributed his forces widely in an attempt to safeguard his new area of responsibility, choosing only an advance on Chattanooga as an active measure. In Washington, Lincoln was tiring of Halleck’s lethargy. He respected, however, his powers as an organiser and on July 11 summoned him to the capital to assume the post of general in chief, in McClellan’s place. As Lincoln would soon find, however, and Grant had already painfully learnt, Halleck was as temperamentally averse to offensive action as the young Napoleon. He was also equally dedicated to detail and to faultfinding with subordinates. Command in Tennessee passed to Grant, but the opportunity to strike during the interregnum was lost by the Confederates, since Beauregard, displeasing Jefferson Davis by taking sick leave at that inconvenient time, was also relieved of command, to be replaced by Braxton Bragg. Bragg, though a fighter, was also bad-tempered and alienated most of his subordinates by insulting them. Unlike both Halleck and McClellan, he had an offensive outlook and did not adhere to the Jominian idea that the purpose of a campaign was to manoeuvre an opponent out of position without actually fighting him. As soon as he succeeded Beauregard, Bragg set about confronting Grant in his headquarters at Corinth. His first plan was to march directly against him. He then reconsidered and decided on a roundabout approach through central Mississippi from the west. Grant, conscious of his threat, responded by getting up the forces, but Halleck, with his obsession about defending everywhere, had second thoughts about Tennessee and Kentucky.

While he was absent deploying his troops, Bragg left large detachments in northern Mississippi under Generals Price and Van Dorn, while he transferred to Kentucky, from which he appeared to threaten Louisville and Cincinnati. In early September 1862, he summoned Price and his 16,000 men. Grant, feeling understandable alarm, concluded correctly that the most likely place Price would strike was Iuka, a railroad village near Corinth which was a depot for a large supply of food and warlike stores. He selected a Wisconsin brigade, known after its mascot as the Eagle Brigade, to defend Iuka. Rosecrans led the advance while Grant, with General Edward Ord under command, waited in reserve. Rosecrans advanced to combat behind a cloud of skirmishers, with an accompanying battery. A tremendous firefight ensued. The ground was covered with dense thickets of scrub between which Blue and Gray dodged as the fight raged. By the end of the afternoon two Northern and two Southern brigades had suffered 790 and 525 casualties, respectively, out of strengths of 3,100 and 2,800. Despite the disparity, the Union got the better of the battle, forcing the Confederates to withdraw.

Grant was waiting in reserve only a few miles from the battlefield, but because of the direction of the wind and other factors was prevented by “acoustic shadow” from hearing any sound of the firing at all. He learnt that the battle had taken place from a despatch from Rosecrans only when it was over. He at once joined Rosecrans in a pursuit of Price and the defeated Confederates, but to Grant’s intense displeasure, Rosecrans abandoned the pursuit after Grant left him and Price made good his escape. He and Van Dorn then joined forces. Together they numbered about 22,000 men, whom Price led into southern Tennessee to threaten Corinth, Grant’s railroad base and centre of supply, the linchpin for his outposts at Jackson, Memphis, and Bolivar. By early October Grant detected that the rebel army, now commanded by Van Dorn, had repositioned itself to attack Corinth from the north. By October 3 the rebels were ready to attack. The Union troops, under the command of Rosecrans, were less well prepared, Rosecrans having been dilatory in concentrating his men. They were stationed in the old Confederate earthworks defending Corinth, behind which was a second and better position on College Hill. All day during October 3, the Confederates pressed hard against Rosecrans’s line, losing heavily but refusing to disengage. Instead they pressed onward, mounting one attack after another, pushing the Union troops back into the streets of Corinth. One formation that withdrew was the so-called Union Brigade, composed of regiments disorganised at Shiloh. Once among the houses of the town, however, they rallied and after meeting other units, resumed resistance and held the attackers at bay. General Rosecrans rode about what was left of his lines at this stage of the battle, shouting at his men to hold fast. Aided by Union artillery fire they did so, repelling one attack after another. Eventually the fighting concentrated round a Union earthwork, Battery Robinet, where the Union inflicted heavy casualties. Fifty-four Confederate dead were later found in the battery ditch, among them the colonel of the 2nd Texas, who had been hit thirteen times. At the culmination of the struggle for the battery, the Confederates turned in retreat. They had suffered 4,000 casualties, the Union 2,500. Moreover, the Confederates’ line of retreat was blocked by the Hatchie River, across which Van Dorn sought a crossing. Bridges were hard to find but Rosecrans did not press his pursuit. He was another example of a Union general who lacked the will and insight to exploit a victory when won. Rosecrans halted his army’s advance to the Hatchie for two successive nights, making its pace snaillike. His soldiers were frustrated and many pushed ahead without orders. When the flat bottom land of the Hatchie was reached, the Union troops found several Confederate batteries in place to defend the crossing places, and a murderous fight broke out, reinforced from both sides. Eventually it relapsed into stalemate, as Grant was able to recognise even from a distance. He sent orders to Rosecrans to back off, but as Van Dorn made good his escape, Rosecrans all too typically insisted that he was on the brink of a great victory and that Grant was cheating him of a golden opportunity. Van Dorn found sanctuary behind strong defences at Holly Springs in northern Mississippi, a position too strong to attack without lengthy preparation, as Grant also recognised. Rosecrans was to continue to complain of missed opportunity, but Grant knew better. He was already determined to close down the campaign in central Tennessee and to transfer his effort to a direct thrust on Vicksburg.

The campaign in central Tennessee had not, however, been without benefit to the Union. At its end, western Tennessee was largely swept clear of regular Confederate troops, though not of guerrillas, and northern Mississippi was in Union hands; loyal eastern Tennessee had not been liberated but was under threat of Union invasion. The great Union advantage in the region was that it lay adjacent to the Middle West, where troops could be raised in larger numbers.

The summer of 1862 was otherwise a time of troubles for the Union. The abandonment of the Peninsula Campaign and the humiliation of the withdrawal from Richmond was followed by the South’s assumption of the offensive in the East and its advances into northern Virginia again and then into Maryland. Defeat in the second battle of Bull Run was swiftly succeeded by the costly stalemate of Antietam. And it was not only in the eastern theatre that the war seemed to be going badly for the Union. In the West, Grant was failing to make progress in his campaign around Vicksburg to open up the valley of the Mississippi to Union traffic. There had been large-scale cavalry raids into the dubiously secure Union territories of Tennessee and the liberation of Arkansas underwent setbacks. Worst of all, in July, Braxton Bragg, the Confederate commander in Mississippi, embarked on a full-scale invasion of Kentucky. Kentucky was probably the most borderline of all the border states, counted by both sides as part of their governed territory and with regiments and large numbers of young men in both their orders of battle. The real danger for the Union in Kentucky, however, was not political but geographical. Its northern border was formed by the Ohio River, just across which lay the great city of Cincinnati, still more important than Chicago as an industrial and railroad centre, with a strong Union population acutely sensitive to the danger of military advances by the Confederacy. The way to Cincinnati, moreover, lay across easily marchable territory. If the Confederacy could drive a corridor across its central section, the territory of the Union would be bisected, in exactly the same way as the developing Union campaign in the valley of the Mississippi threatened to bisect the South. It was vital, therefore, that Bragg’s invasion should be defeated.

The difficulty was to organise a counter-offensive. The two Confederate cavalry leaders who had ridden so cavalierly through the region, Nathan Bedford Forrest and John H. Morgan, were still active, while a subsidiary army to Bragg’s, commanded by Edmund Kirby Smith, was advancing from Knoxville towards the Cumberland Gap, historic gateway into trans-Appalachia, from which he rapidly arrived at Richmond, Kentucky, only seventy-five miles south of Cincinnati. There he was confronted by a Union division, but all its troops were newly enlisted and it was swiftly dispersed at heavy loss of killed, wounded, and captured. Braxton Bragg had little enthusiasm for offensive war-making, but he was, at this stage and this place, a better bet than his opponent, Don Carlos Buell.

From Washington, however, Halleck so harassed Buell with directions to advance, to put pressure on Bragg, and to fight that eventually Buell had no alternative. He could not plead lack of strength, since he had by mid-September received the reinforcement of two divisions from Grant, while in Louisville and Cincinnati 60,000 recruits, raised locally, were being put through their training. During September, while Buell prudently retired towards Louisville, Bragg attempted to set the stage for a major battle to settle the balance of power in Kentucky. From his position near Louisville, he sent a request to Kirby Smith, who was then in the area of Lexington and Frankfort, the state capital, to meet him with his 20,000 men at Bardstown, south of Louisville. With their combined strength, Bragg believed he could defeat Buell and so settle matters in the borderlands. He also felt that fighting a major battle would pull the Kentuckians off the fence and bring them conclusively to the Confederacy.

Mid-Western American Civil War II

Buell was at last conforming to Washington’s wishes and during early October appeared in the vicinity of Bragg’s army at Bardstown. He concentrated 60,000 men, to the Confederates’ 40,000. They were now, in Bragg’s temporary absence, under the orders of Bishop Leonidas Polk, who led his men to the small town of Perryville, south of Louisville. What drew him there was a need for water, the Southern summer having dried up the streams. A prolonged drought had left the Chaplin River a string of stagnant pools. As that was the only water available, both sides wanted it. Polk got to it first but was soon attacked by the advance guard of Buell’s army, commanded by the up-and-coming Philip Sheridan. Sheridan was aggressive and directed his division’s efforts to such effect that it defeated Polk’s army and advanced into the streets of Perryville, driving its remnants before them. At this stage, Buell should have completed what was turning into the victory of Perryville and destroyed, with reinforcements, what remained of Bragg’s army. By the meteorological accident of acoustic shadow, however, no sound of the battle raging in Perryville reached the ears of anyone else under Buell’s command. He therefore failed to march to Sheridan’s assistance, though as darkness fell the Confederate line was defended by only a single brigade which would have dispersed if attacked aggressively. Next morning, when Buell positioned his army for a general advance, the ground was empty. Bragg had during the night decided he was beaten and had led his army away.

Perryville was an all-too-typical Civil War battle in its lack of decision, despite high casualties on both sides. The indecisiveness of battles is one of the great mysteries of the war. In the East, particularly from 1864 onwards, it was largely explained by the recourse to digging, which produced earthworks from which it was almost impossible to expel the enemy. In the West, by contrast, particularly in the earlier years, earthworks were less commonly constructed. The explanation therefore seems to lie in two unconnected factors: the lack of a military means, such as large cavalry forces or mobile horse artillery, that could deliver a pulverising blow, and the remarkable ability of infantry on both sides to accept casualties. Casualties at Perryville—4,200 Union and 3,400 Confederate—were certainly high, but neither side seemed shaken. An eyewitness, Major J. Montgomery Wright of Buell’s army, describes the strange phenomenon of the acoustic shadow. Riding as a staff officer on a detached mission, he “suddenly turned into a road and therefore before me, within a few hundred yards, the battle of Perryville burst into view, and the roar of the artillery and the continuous rattle of the musketry first broke upon my ear…. It was wholly unexpected, and it fixed me with astonishment. It was like tearing away a curtain from the front of a great picture…. At one bound my horse carried me from stillness into the uproar of battle. One turn from a lonely bridlepath through the woods brought me face to face with the bloody struggle of thousands of men.” Major Wright witnessed the effect of the struggle on one group, which suggests that the battle was having a decisive effect upon them: “I saw young Forman with the remnant of his company of the 15th Kentucky regiment, withdrawn to make way for the reinforcements, and as they silently passed me they seemed to stagger and reel like men who had been battling against a great storm. Forman had the colours in his hand, and he and several of his little group of men had their hands upon their chests and their lips apart as though they had difficulty in breathing. They filed into a field and without thought of shot or shell they lay down on the ground apparently in a state of exhaustion.”1 Yet despite such efforts the Union line did not break, nor did the equally punished Confederate. Bragg, who rightly recognised he was outnumbered, swiftly decided to withdraw during the night of October 8 and fell back to Knoxville and Chattanooga, abandoning his invasion of Kentucky altogether. The Southern press, and several of his generals, seethed with dissatisfaction; Bragg was called to Richmond to account for his failure, but he had a friend in Jefferson Davis, who accepted his explanations and allowed him to continue in command.

Bragg’s abandonment of the attempt on Kentucky completed a general Confederate failure on the central front in the West. Just before Perryville, Generals Price and Van Dorn had been defeated by the Union general Rosecrans at Corinth in Mississippi. It followed another Confederate defeat at nearby Iuka. Grant, who was engaged in the campaign at a distance, had hoped to trap the Confederates either at Corinth or Iuka and was disappointed not to do so. He blamed Rosecrans, for a movement of his troops he thought dilatory, though the recurrence of acoustic shadow may have played a part. For whatever reason, however, the Confederates had failed in their efforts to reverse the balance of power both in Kentucky and Tennessee, in what proved to be the last unforced Confederate offensive west of the Appalachians. As the fighting died down, Grant gathered his forces to renew his campaign against Vicksburg. The citizens of Cincinnati and Louisville relapsed into calm, after what had been some disturbing weeks. Though it was not realised in Richmond, the failure in the West was a grave blow to the Confederacy, reducing their range of strategic options to the well-worn pattern of keeping alive Union fears of an advance against Washington or feints at Pennsylvania and Maryland, theatres where the North enjoyed permanent advantages. The drive into Kentucky and threats against Tennessee were the only imaginative moves made by the Confederacy throughout the war; their failure and the failure to repeat them confirmed to objective observers that the South could now only await defeat. It might be long in coming, but after the end of 1862 it was foreordained and inevitable.

There were objective observers. Two were Karl Marx and Frederick Engels, then in exile in England, where in March 1862 they composed an analysis of the progress of the Civil War of quite remarkable prescience. Marx and Engels’s interest in the Civil War was not political. As revolutionaries they hoped for nothing from the United States. It was simply that as men with a professional interest in warfare and the management of armies they could not prevent themselves from studying military events, and prognosticating based on their lessons. Marx concluded that, following the capture of Fort Donelson, Grant, for whom he had formed an admiration, had achieved a major success against Secessia, as he called the Confederacy. His reason for so thinking was that he identified Tennessee and Kentucky as vital ground for the Confederacy. If they were lost, the cohesion of the rebel states would be destroyed. To demonstrate his point, he asked, “Does there exist a military centre of gravity whose capture would break the backbone of the Confederacy resistance, or are they, as Russia still was in 1812 [at the time of Napoleon’s invasion], unconquerable without, in a word, occupying every village and every patch of ground along the whole periphery.”

His answer was that Georgia was the centre of gravity. “Georgia,” he wrote, “is the key to Secessia.” “With the loss of Georgia, the Confederacy would be cut into two sections which would have lost all connection with each other.” It would not be necessary to conquer the whole of Georgia to achieve that result, but only the railroads through the state.

Marx had foreseen, with uncanny insight, exactly how the decisive stage of the Civil War would be fought. He was scathingly dismissive of the Anaconda Plan, and he also minimised the importance of capturing Richmond. To that extent, his foresight was defective. The blockade, a major element of the Anaconda strategy, was crucial to the defeat of the Confederacy, and it was indeed the capture of Richmond that brought the war to an end. In almost all other respects, however, Marx’s analysis was eerily accurate, testimony to his grisly interest in the use of violence for political ends. The analysis was published in German, in Vienna, in the review Die Presse. It may not have been noticed in the United States.

Marx, who had the keenest eye for strategic geography, did not discuss the importance of Tennessee and Kentucky as a weak spot in the defences of the Union. Materialist as he was, he had already assured himself that the vastly preponderant industrial and financial power of the North guaranteed its victory. He made insufficient allowances, however, for the necessity of fighting for that outcome and for how relentless the struggle would be.

Cannae – A Legacy

The modern military image of Cannae as what one modern critic termed “a Platonic ideal of victory” appears to have originated with the obsession of one man, Count Alfred von Schlieffen, chief of the Prussian general staff from 1891 to 1905. Schlieffen’s preoccupation with the battle turned on several factors. One was Germany’s strategic position, caught between two potential adversaries, Russia and France, which made a fast decisive victory over one or the other highly desirable. The second factor was the example provided by his predecessor on the general staff, Helmuth von Moltke, who in 1870 at Sedan surrounded the French and captured Emperor Napoléon III through a double envelopment, making a German victory in the Franco-Prussian War nearly inevitable. Finally, all of this was informed by Schlieffen’s reading of ancient history, specifically the first volume of Hans Delbrück’s History of the Art of War, with its extensive description of Cannae. Metaphorically at least, a bulb switched on above Schlieffen’s head, and suddenly Cannae seemed to beam light on everything.

While this revelation could have taken place as early as 1901, it seems more likely that it came as late as 1909, well after the general’s retirement. This is important because until recently it has been assumed that Hannibal’s victory was the inspiration behind the so-called Schlieffen plan, which was supposedly the strategic basis for Germany’s attack through Belgium and into France during the early stages of World War I. Not only has the very existence of the Schlieffen plan as a comprehensive scheme of battle been questioned, but it is apparent that the chief of staff’s recommendations for a war against France from 1901 to 1905 were based on a concentration of force against one flank of the enemy, an approach that simply did not fit Cannae’s footprint, despite subsequent efforts to shoe-horn the plan into this context.

This is not to argue that Schlieffen was not Cannae-obsessed or that he did not come to measure much of recent military history against what he called this “perfect battle of annihilation;” it is simply a question of timing. Schlieffen’s collected works including his Cannae studies were published in 1913, the year of his death. However, they do not appear to have been very influential until after the end of World War I, when they became a touchstone for those who argued that the German Army had lost the first battle of the Marne (and thus the war) because they had failed to keep faith with Schlieffen’s commandments—in the process conflating Cannae and the general’s actual advice. This set the tone for a subsequent generation of German military thinkers, a key segment of whom followed Schlieffen “like Thurber’s owl” and dreamed of future Cannaes made all the more plausible by the advent of armored vehicles.

Schlieffen’s Cannae studies were also influential in military circles outside of Germany during the interwar years. In 1931, for instance, they were translated into English and published by the U.S. Army Command and General Staff College at Fort Leavenworth, availing a new cadre of American officers the opportunity to consider the possibilities.

Hence, World War II would feature a cavalcade of Cannae-savvy military luminaries, a number of whom, especially among the Germans, were looking to inflict such a fate on their adversaries. Heinz Guderian conceptualized his tanks as motorized equivalents of Hasdrubal’s cavalry sweeping around the enemy to seal their fate. Similarly, in 1941, while in the process of chasing the British army in the direction of Tobruk, Erwin Rommel noted in his diary that “a new Cannae is being prepared.”

Even in the face of disaster, the Germans stuck to the theme; thus Erhard Raus, commander of the 6th Panzer Division near Stalingrad, less than three months before his army’s surrender, dubbed a successful day’s fighting around an obscure village “the Cannae of Pakhlebin.” Seven months later, in July 1943, the German assault on the Kursk salient was intended to achieve a gigantic Cannae—only to result in a disastrous attrition of their attacking armored vehicles on both flanks by the well-prepared defensive-minded Russians. The German Army having been bled white by the dream of Schlieffen’s “perfect battle of annihilation,” Kursk marked the last time the Wehrmacht was able to mount offensive operations on the Eastern Front. Defeat beckoned.

Yet the Germans were not alone in responding to Cannae’s siren song. While the British were more cautious, American commanders were offensive-minded and therefore open to Hannibalic feats of arms. In particular, Dwight Eisenhower had nurtured a lifelong dream of emulating the Carthaginian general’s envelopment of the Romans. And for the initial Allied invasion of Germany, Eisenhower envisioned a huge Cannae-like maneuver, employing a double envelopment of the Ruhr. George Patton, the most audacious of the major American commanders in Europe, was also a student of military history and very much aware of Hannibal’s feat. Yet his outlook was hardly predictable, and reflects the contingent nature of even so decisive a victory. Writing in 1939, Patton notes: “There is an old saw to the effect that: ‘To have a Cannae you must have a Varro’ … in order to win a great victory you must have a dumb enemy commander. From what we know at the moment, the Poles qualified …”

World War II ended and was replaced by inconclusive combat in Korea, the standoff of the cold war, and amorphous insurgency in Vietnam; yet the American dream of Cannae lived on, apparently nurtured primarily by forays into ancient military history at U.S. service academies. So it was that after the 1991 Gulf War, victorious General Norman Schwarzkopf was able to proclaim that he had “learned many things from the Battle of Cannae which I applied to Desert Storm.” In fact, the general’s famous “left hook in the desert” more resembled the opening German moves in World War I, but no public-relations-conscious war hero was likely to announce that he had prepared another Schlieffen plan, so Cannae it was.

The future remains ambiguous. Given the predominance of anti-terrorism and anti-insurgency, the possibility of maneuver warfare seems, for the moment, distant. Also, given the sophistication of modern intelligence, surveillance, and reconnaissance assets, the kind of deception necessary to produce truly Hannibalic results may prove very difficult to achieve. But as long as men dream of killing other groups of men in very large numbers, we can rest assured, Cannae will not be forgotten.