Silent Riddles

Our discussion of the origins of modern martial arts is always haunted by twin specters. The first is the claim of great, almost unimageable, antiquity.  The second is arguments from silence.  We have explored at length all of the cultural, political and psychological reasons why martial artists seek the origins of their current practices in the distant past rather than focusing on more proximate, recent, events. Yet it is the paucity of the historical record that makes this realm such fertile ground for mythmaking. To make matters worse, we are not always skillful in identifying, interpreting and contextualizing the historical sources that do survive.

The further back we go, the more deafening the silence become.  The oldest extent Chinese martial arts manuals date only from the latter part of the Ming dynasty.  And yet it is clear that the area has possessed a rich martial culture (or, more likely, several of them) dating all the way back to the Warring States period.

The paucity of the historical record suggests that we will always have more questions than answers about these early periods.  It might be easier to just ignore them all together and focus on that which is more easily documented. If anyone were to ask my advice about writing Chinese martial arts history, that is certainly what I would recommend.

Yet historians and social scientists are not the only ones who weigh in on these questions. The field of archeology also has something to say.  And when it comes to Han dynasty arms, both civilian and military, archeologists have unearthed literally tons of period weaponry. The number of Han dynasty swords that have been found dwarfs that of all other dynasties except for the basically modern late imperial period.  While one can count the number of confirmed Tang dynasties swords that have been recovered on your fingers, Han dynasty caches of surprisingly well-preserved weapons are so common as to be basically worthless to most of China’s museums and cultural institutions.  They are much more interested in the bronzes, fine lacquerware or bamboo documents that are often recovered along with the various jian, dao and ji that are our primary interest.

Students of China’s ancient martial culture are thus faced with a paradoxical disconnect.  On the one hand we have no written texts documenting how these swords, spears and other weapons were used, aside from the occasional reference in a handful of surviving literary texts.  Yet the archeological record suggests that the display and use of weapons was an important aspect of every stratum of Han society.  

In previous posts I have explored a handful of lucky archeological finds which provided written inventories of Han arsenals, or instructions for its officers on buying good quality blades. What other sources might we turn to in an effort to answer such basic questions as, what constituted a complete curriculum of martial knowledge?  Or to put it somewhat differently, what was the universe of martial technology available to the sorts of civilian households (rather than just the professional military) who would have had the time and economic resources to invest in these things?

Two Han Sha in the the Suzhou Museum. These were previously misidentified as jian with large and ornate hand guards. In fact they are the blades of a Han Dynasty polearm.

Eternity in Stone

The answers to these questions emerge from the archeological, rather than the literary or historical, record. But even there caution is in order. We might attempt to do a survey of the number and type of weapons that have been found and extrapolate from those the sorts of arms (and by extension martial practices) that dominated daily life.  Yet that would open a door to various types selection bias, skewing our understanding of the past.

Simply put, the archeological record does not preserve a random sampling of weapons. Many of the Han arms that have survived are being recovered from the oxygen deprived environment of flooded tombs. Only the wealthy could afford elaborately constructed stone tombs, and they were not going to be taking anything as plebian as a crossbow or a hand ji with them as they ascended to immortality. Unsurprisingly, finely crafted swords seem to be overrepresented in the archeological record.  Yet it is unlikely that most individuals went into battle with elegant jian outfitted with jade handguards and pommels.

Fortunately the information within these tombs is not limited only to the “shinning goods” that they contain.  Often the walls of these structures, and their various offering halls, would be carved with elaborate stone relief scenes.  Given the function of a tomb, many of these scenes depicted an ascent into the afterlife, or the attainment of immortality. Other images retold popular stories of filial piety, speaking to the ideological and political concerns of the time.

One of the main ways in which the families of low and middle ranking officials might seek to advance their employment opportunities within the imperial bureaucracy during the Eastern Han dynasty was by publicly displaying the virtue of filial piety. As such, there may have been social and political reasons why tomb building for a beloved patriarch became a growth industry among the petty official class and their cousins in the countryside which went beyond the universal human need to mourn a loss. Mortuary art served as an ideological statement to the living.

This specific tradition of elaborate tomb construction got its start during the Western Han Dynasty, but it was during the later Eastern Han that it really hit its stride.  The Wu family shrines, discussed in a previous post, are probably the most famous example of this mortuary practice, but they are far from unique. Indeed, such structures are common enough to be economically paradoxical.  The construction of a vaulted stone tomb and offering hall would have cost about 20,000 bronze coins at a time when that was a bit more than the annual salary of a mid-ranking official. That begins to give us some sense of the scale of investment was being made in the creation and decoration of these structures. 

The construction of the tomb itself would have required the hiring of teams of specialized masons who were led by both a master builder and an artist.  These individuals would direct the work that was to be carried out by different groups of apprentices.  Most tombs were somewhat modular affairs where the main stones were shaped and stored off site.  Once a contract was reached the family would be free to choose the details of the construction and decorations.  Elaborate images seem to have been traced onto the finished stone using silk stencil patterns.  As such, we often see the same, or similar, designs on tombs in a given area.  Artists could just flip the stencils over to create mirror symmetry for some designs, or in an attempt to work something onto an oddly placed surface.  The services of these artisans were in demand, but modern scholars have determined that they rarely traveled more than a couple of hundred miles to for even the most lucrative jobs.  Hence several regionally distinct styles of carving emerged.

This quick overview has omitted most of the details that archeologists and arts historians have pieced together in their reconstruction of the funerary industry of the Eastern Han. Still, we now know enough about the origins of these images that we can begin to answer some questions.  In addition to spiritual and ideological imagery, Han tombs also contained images of an idealized afterlife where the deceased would basically carry on their duties as the head of a household, or a beloved son or wife, but in greater luxury and peace.  Thus, their walls are often inscribed with scenes of daily life, not as it was so much as what people hoped it could become.  The afterlife was to be full of banquets with the finest food and troupes of acrobats as entertainment.  

These images of one’s eternal home were often adorned with pastoral scenes from everyday life containing everything that the deceased could need.  It is not uncommon to see scenes of women spinning cloth on looms, brewing alcohol or herding domestic animals.  Not all of the scenes are bucolic in nature. Popular stories or morality tales (such as the “Battle on the Bridge”, previously discussed here,) might also show individuals in combat. Interestingly, because of the nature of these popular folktales, the combatants were often civilians rather than soldiers.

While individuals that modern readers would identify as “martial artists” are also shown to these images, our knowledge of how the Han would have understood that term is limited. We need to be conscious of our limitations in mapping these scenes onto any sort of modern understanding of martial practice.  Yet there is much that we can learn from period artwork.  For instance, while a tomb may contain one or more fine swords, its walls will be liberally adorned with images of individuals in official processions carry all sorts of shields, spears and polearms that are much less likely to appear in the tomb itself.  

The amount of tactical information anyone could derive from these images is limited by their abstract nature and the highly ritualized context for which they were created.  Again, these carvings were not meant to be accurate representations of battle, but type of magical machinery to assist the departed in their ascension to the afterlife. 

Still, they are quite suggestive of the role of martial objects and values in Han society.  We should note, for instance, that while images of weapons are surprisingly common, scenes of historical battles or killings, while occasionally present, are much less so. And many of the images of violence that we do encounter are not meant to be an accurate retelling of instances of real conflict, but a dramatic rehersal of well know stories of filial piety or adventure.  

What is more striking is how often arms are displayed in moments of actual peace.  They are frequently seen in the hands of acrobats and performers, suggesting that the pattern of mixing performative and combative elements within the Chinese martial practice may have very deep roots.  Other times we see retainers standing guard with their ji as two gentlemen meet.  Official processions on the road always seems to have some sort of ceremonial armed escort.

Perhaps the most interesting group of images are stylized scenes showing a fully stocked weapons rack in the context of a civilian home or estate.  These are not the endless ranks of identical arms that would be useful to an army or a well provisioned militia.  Instead, we see a potpourri of different pole arms, swords and bows that would be of more use to a hobbyist on a country estate, or someone much more like a modern martial artist. The fact that these weapons typically appear in pairs suggests that they may have been intended for training or bouting rather than battling bandits.  One suspects that the roads of the afterlife were probably lacking in such terrestrial concerns.

While swords are the most likely weapon to be found by archeologists within these tombs, the arsenals incised on their walls heavily favor polearms. While much of stone imagery is highly abstract, enough care was taken in the carving of some of these images that we can even identify specific classes of weapons.  This allows us to begin to build a more complete map of the martial imagination during the late Han.

Modern students are likely to encounter such images in one of two ways.  In some cases we might be lucky enough to get a high-quality photograph of the carving itself.  But in most cases, we are more likely to see a black and white ink rubbing of the underlying stone.  Ironically, a well done rubbing can sometimes reveal details of an image that are not as readily apparent in an average photograph.

Let’s begin our survey of these images by examining the one that appears at the head of this article.  It was found in a tomb in Shandong province dating to the Eastern Han. At the top of the image we first encounter a crossbow (quite common during the period) and smaller abstract icons representing armor and other equipment.

The only sword that we find in this image is a ring hilted dao, at the very top of the rack.  These had come to dominate both the battlefield and civilian fencing by the Eastern Han, largely (but not entirely) replacing the more expensive and detailed jian that were popular during the first half of the dynasty.  We should note, however, that the artist has not attempted to provide any sort of visual scale to record the sizes of the various weapons. Western rules of perspective were absent from Han art which had its own highly developed set of visual conventions. Instead, all of the main weapons have been sized to fill the horizontal width of the stone that the masons were attempting to decorate. Less important objects (such the shields) are shown as much smaller in size. 

Beneath the dao we see a pair of ji, or halberds heads, that have been fitted so that they can be used as an ax-like, hand held, weapon.  It is not uncommon to encounter images of individuals carrying these weapons in battle scenes. We even see one of the female combatants wielding this weapon on the “Battle of the Bridge.” Further, this image shows that both the vertical and horizontal blades of these weapons have been fit with their own, customized, squared off wooden scabbards, similar to how Han dynasty Jian were equipped. Given their elaborate fittings and presence in this sort of a scene, we may conclude that this was a weapon with a certain degree of cultural recognition.  It was certainly part of the era’s martial imagination.

Another carved image in which we can see the hand ji in use, paired with a dao. The individual behind the mounted figure carries a hand ji and shield.

Beneath this we see two more Ji, both of which have been fit with their customary poles.  The style of the horizontal blade on this pair is different, and the upward curve would have made them useful for intercepting or manipulating other pole arms.  Indeed, this style of Ji head is commonly seen in period art and is distinct from the blade geometry seen on the hand Ji (better suited to a piercing chop).

Beneath this we see two spears with iron lugs, or wings, emerging from the socket.  I am not entirely familiar with the archeological literature on Han dynasty spears, but I have seen almost identical weapons which were produced and used in Northern Korea at about the same time. So one suspect that this was not an uncommon design choice.

The next set of weapons are sometimes misidentified online as jian, perhaps because of their guards.  In fact, they are actually sha, a unique type of polearm in which a sword-like blade with a long tang (rather than a socket like a spear) was fitted to an oval shaft.  These weapons were given an oversize bronze guard which swept upward at both ends. The shaft of these weapons was typically a little less than a person’s height.  With a long internal tang and an oval hilt they overcame many of the problems that make it difficult to cut with other sorts of “hewing spears,” including their flexibility and the difficulty in indexing a round handled for proper blade alignment. The sha, in contrast was deadly with both the cut and the thrust. 

During the Eastern Han the sha was treated as something of a prestige object as they were technically difficult to make compared to the humble socketed spear. Indeed, many were even fit with elaborate bonze scabbard furniture. Yet despite this expense, records indicate that they were popular with the both the military and civilian households. 

The popularity of the sha also remind us that the smiths of the period were still turning out larger numbers of jian-type blades, even if they were not being deployed as side arms with the same frequency as during the Western Han. The continued popularity of the jian blade when mounted on both the sha and the pi (discussed below) may necessitate some reconsideration of the reasons that are often given for its replacement by the dao as the preferred sidearm.  For instance, it is often said that the jian was replaced as it was more expensive to produce and more delicate than the dao which had only a single cutting edge and a thick spine. Yet mounting the supposedly delicate blade on the end of a long lever would only have multiple the sorts of mechanical forces it was subject to. Apparently this was not seen as a problem by period armorers, suggesting that fragility and cost may not have been their primary concern. One must begin to wonder whether the ascendency of the dao had more to do with a change in military doctrine than any inherent deficiency in the jain, which continued to be deployed in larger number, but now in the guise of the of sha and pi. Indeed, the Eastern Han was very much the age of the sword staff.

Like a jian, the sha’s blade was fitted with the sort of squared-off scabbard commonly adorned with bronze furniture (much of which survives in the archeological record).  It is the fact that the Sha in these images have scabbards which can be mistaken for hilts (as well as the lack of scale in these carvings) that leads some individuals to misidentify this weapon as a long jian.  Indeed, the similarities between their guards and other scabbard furniture even led some archeologist to misidentify them as jian with “oversized” guards in cases where the conditions inside the tomb did not allow evidence of the Sha’s wooden shaft to survive.  As such we still occasionally see sha blades misidentified as jian in some museums. This occasional confusion is actually helpful in that it underlines the fundamental similarities between these arms.

The next set of weapons down carry on these themes, though they do present more of an interpretive challenge. I suspect that this is a set of pi, another popular polearm during the Han dynasty.  Like the sha, these were fitted with a long, jian like, double edged blade. Typically these were also fitted to the shaft be means of a tang rather than a socket. But they tended to not be as elaborate as the sha and they lacked its characteristic bronze crossguard.  The real question is how we should interpret the hour glass swelling on the weapons shaft, as that is shared by practically many period carvings of the weapon. 

Finally, at the bottom the rack we see what appears to be a rough-cut pole with the nubs of its trimmed branches still showing.  Finding a pole in this position in not uncommon in this genera of images.  Further, it provides symmetry to the lone sword at the top of the image. Likewise, a single gou-rang (often used in conjunction with the dao) is leaned against the rack, as are symbolic representations of two wooden shield at the bottom. The fact that everything else appears in groups of two again suggests training or sparring rather than raiding or organized warfare.   

Our next image (this time a photograph of the actual stone) elaborates on and repeats many of the basic patterns that we see throughout this genera.  The tip of sword-type weapons can be seen always pointing to the viewers right, whereas polearms face the left.  Weapons are typically displayed in sets, and polearms are present in greater number and variety than the swords.  Further, it is the same mix of polearms that we saw in the previous, less artistically complex, example.

Consider the similarities and difference with the previous image.  It is of much higher quality and includes various animals playing on the weapons (monkeys, birds, ect…) as well as a representation of their owner. Again, this is a pastoral scene rather than one of violence.  Here the crossbow has been exchanged for a recurved model.  The swords are once again placed at the top of the rack.  This time we see both a ringed hilted dao and a jian with its characteristic disc pommel.  These share rack space with another set of hand ji and a gou-rang.

Beneath them we once again see a set of pi (this time with scabbards), and below that an almost identical set of Sha.  These are followed by a set of spears.  The one on the top again looks as though it might have wings protruding from its socket, while the last of the set has a particularly broad blade.  Finally, we see two ji. Rather than being a perfectly matched set they have two different styles of heads, both of which are commonly seen in period art.  Rather than being a slave to symmetry, this artist seems to have been more concerned with exploring some of the finer details of the period’s weapons and animal life.

Our next image was found on a set of tiles rather than a single stone blocks.  As such, it would have been less expensive to produce, but the basic imagery remains unchanged.  The main importance of these scene, however, is not the details of the weapons themselves but the context in which we find them.

Compared to the previous arsenals, this selection is rather modest in its ambition.  It breaks with convention by only displaying a single example of each weapon rather than a set.  At the top we see what appears to be a fully realized trident with some sort of pommel. Beneath that we have ji, a broad headed spear, a ring hilted dao, a shield and both a cross bow and a recurved bow.

However, these arms are not displayed as part of a military scene.  They are presented as a single element of commercial life during the Han dynasty.  These are weapons seen as products. They share equal space with women producing cloth and are visually subordinated to other economic activities such as the brewing of large amounts of alcohol and animal husbandry.

During the Han dynasty weapons were mass produced in government workshops (typically run by the local rather than the imperial administration) and sold through markets, just like other iron/steel product and luxury good. Looking at some of these images, it is hard to shake that the impression that the deceased is flexing their economic muscle, rather than displaying their purely martial values.  Of course, owning land and other economic enterprises necessitates the protection of the same.  So it is probably not a surprise to discover that in China’s long history landlords have often taken an interest in the promotion of local martial arts traditions, particularly when they have needed a militia. Still, even in an afterlife where such mundane concerns have been left behind, the collection and display of weapons continues to serves social, ritual and economic functions.

Our next image continues to carries this same this theme forward.  Photos of this rubbing can be found on LK Chen’s webpage, and they were kind enough to send me a copy for study. Here we see two gentleman meeting.  One is unhitching a horse from a wagon that is laden with the weapons that are later displayed on the rack that takes up most of the image.  The implication that viewers are apparently meant to gain from this composite scene is that a shipment of arms has just been delivered from the factory or shop that produced them.  Again, they are shown as one aspect of the normal flow of commerce. The living are meant to understand that the acquisition of luxurious goods is something that will continue for the deceased.

Following with the previous explored patterned, this image is also dominated by polearms.  It is possible that the top of this stone was lost, or an artist had to modify their pattern to fit on a certain sized slab. But in this case, we cannot see the top of the rack which would likely have included any swords.  Instead we start with a set of sha that are practically identical to that which we have already seen.  We then have what is very clearly a set of winged spears, followed by what I guess to be a pair of pi (this time without scabbards).  The final weapon in the collection again appears to be a roughly cut pole.

The motif of a horse being unhitched from a cart of goods is not uncommon in Han funerary art and appears in other contexts as well.  A tomb in Jiangsu revealed a scene in which a seated audience (making up the far right hand border) watches a performance by professional acrobats, one of whom is literally jumping through a set of hoops.  However, off to the right of the image we can see that a cart and horse are once again present displaying a selection of arms that also make appearances in entertainment scenes. Period literary accounts make numerous mentions of sword dancers and other sorts of martial displays.  In this case the style of art is more abstract and its harder to interpret the weapons. But at minimum we can make out what appears to be a ji, a spear and roughly cut pole.  Further, the individual at the head of the cart seems to be wearing a ring hilted dao, which was the most common type of sword during the Eastern Han.  In this instance its almost possible to overlook the weapons completely in the scene of dance and frivolity.

The final image that we will be reviewing is also a rubbing displayed on LK Chen’s webpage, though I have not been able to locate any other information on when or where it was discovered. Still, it shares so many similarities with the very first image that one is left to wonder if it also came out of a tomb in Shandong.  While well preserved this image is also more abstract than some of the others.  Rather than showing a typically constructed weapons rack, here the arms are suspended from pillars that seem to run from floor to ceiling.

A crossbow again sits at the top of the image, and it is dominated by polearms.  But it puts more emphasis on swords than any of the others that we have reviewed so far.  In this image we see three swords appearing at the bottom of the rack, but the visual conventions of signifying and displaying them are the same.  Again, their tips point towards the viewers right while the polearms direct their blades to the left.  The first is a short ring hilted dao.  In this case the cravers have given us some additional detail.  We can see that the ring is somewhat squared off, and the hilt has been wrapped with cord.  Next, we have a somewhat longer jian, easily discernable by its disk pommel and characteristic scabbard fittings.  Finally, we see a longer ring hilted dao.  The most notable omission from this collection is the lack of a set of hand ji.  Perhaps the additional swords took their place. There does seem to be something that looks vaguely like a mace at the top of the left hand pillar, but without additional perspectives on this image it is difficult to determine exactly what the artist intended.

All of the standard polearms appear in this scene, though this time the artist has alternated them on the rack rather than grouping everything into orderly pairs.  This arrangement leads one to wonder whether the ji and spear/pi, or the ji and sha, were seen as natural combative pairs in training.  Indeed, this may be a subtle reference to the ability of a ji to manipulate other types of arms through its ability to ward, entangle or hook in addition to its more obvious modes of attack.  Or this could simply be a variation in regional artistic convention.  Without more information on this stone, and a larger sample set of images, it may not be possible to say with much certainty.  Sitting directly above the swords, and beneath the polearms, we should also note the presence of two slender rods.  Again, without some additional data it hard to know how to interpret these.  Perhaps they are meant to be iron rods, or maybe a type of light javelin such as the chán which, according to Sima Qian was a favorite weapon of the Xingnu.  But without any discernible features (such as a clearly pronouced spear head) it is impossible to say more.

A set of chán, or short throwing spears, once a favored weapon of the Xiongnu. These weapons are constructed entirely of metal, including the shaft. There is some suggestion that this type of weapon was thrown at opponents.


While jian and dao (some with rich jade fittings) are commonly found within upper class Han tombs, we must look to their walls for a better understanding of complete range of weapons that comprised that era’s martial imagination. It can be difficult to interpret the exact nature of the weapons that are being used in many of the more fluid and abstract scenes of action.  While certain weapons, such as the ji or ring hilted dao have obvious iconographic markers, other sorts of pole arms seem to blend together. It can be very difficult to determine if a given warrior holding a socket headed spear, a sha or a pi in most of these images.

Fortunately, these more abstract action scenes were often complimented with highly detailed still life images of racks of weapons that the deceased would have access to in the afterlife.  Swords and bows are present in all of these collections. Given that these images all date to the Eastern Han it is not surprising that ring hilted dao and crossbows are most common, though the jian might also appear on the period’s weapon racks.  And the iconic gou-rang, or hooked buckler, was almost a mandatory addition for any well-appointed arsenal.   

Yet polearms seem to have dominated the era’s artistic imagination and received the most attention. Among their ranks we see several types of ji, easily discernable by the shape of their horizontal blades and often juxtaposed in the same image. They appear as both mounted polearms and compact hand axes.  Their dominance of these scenes suggests something about the importance of the ji in all of its varieties to not just Han soldiers, but also martial artists.

Next, we see a wide selection of socket headed spears. Like the ji these appear in several varieties, including those with winged sockets and others with notably broad blades. Indeed, spears form the backbone of every heavenly arsenal examined in this post.

Also very common was the sha.  Combining both a double-edged jian blade and elaborate bronze fittings it is no surprise that these versatile weapons were considered to be status symbols in both this life and the next.  They, along with another polearm that I have tentatively identified as the pi, were also an essential aspect of any well stocked arsenal.  

Finally, many of these scenes included a number of miscellaneous items that would have helped to properly outfit any martial artists. These included rough cut poles, and small representations of period armor and various styles of wooden shields. It is important to acknowledge the inclusion of these items, yet it is also critical to note their small size and placement. In all honesty the type of armor worn, as well as the ubiquity of shields on the Han battlefield, had a critical impact on the development of the era’s other weapons. One suspects that the ji was so popular precisely for their ability to hook and encumber the shields that most period infantry relied upon.  Yet in the eyes of the those who constructed these tombs, and those who commissioned them, they played only a supporting role in era’s martial imagination.

There are limits to what one can hope to accomplish in any single blog post, but this short essay has demonstrated that perhaps the weapons of the Han dynasty left a richer documentary record than it at first appeared.  Rather than relying only on the artifacts recovered as tomb goods, or on the hand full of written references that have come down to us, the iconography of the era’s stone tombs bears silent witness to richness of the era’s martial imagination. While fully interpreting these images requires an actual engagement with the fields of art history and archeology, it is sure to enhance our understand Han weapons and individuals who used and collected them.

Recommended Reading

If you are interested in learning more please see the following works:

Wang, Hongzhen. 2011. Chinese Stone Carvings: Treasures from the Han Dynasty 2,000 Years Ago. New Work Press, Beijing.

Sun, Zhixin Jason. 2017. Age of Empires: Art of the Qin and Han Dynasties. Metropolitan Museum of Art and Yale University Press: New Haven and London.

Liu, Cary T., Michael Nylan and Bary Barbieri-Low. 2005. Recarving China’s Past: Art, Archology and Architecture of the “Wu Family Shrines.” Princeton University Art Museum and Yale University Press: New Haven and London.