Pages Navigation Menu

Swords, Cutting and Military History

The debate on Japanese swordsmanship based on skeletal trauma

The debate on Japanese swordsmanship based on skeletal trauma

photo credit Lorianne DiSabato

Share Button

How did medieval Japanese warriors (bushi) fight and act on the battlefield during the classic 14th century Kamakura period?  What were their weapons of choice?  Sword?  Bow? Spear?  Today, through movies, books and stories, we think of the sword as the primary weapon of Japanese.  Yet histories and battlefield accounts suggest that bow and spear were the primary battlefield weapons.   So, which is it?

While histories, battlefield records and common stories do provide some information, many of the claims made in these are contradictory, obscure, or obviously enhanced to the point of being legend.  Academics have turned to investigating the remains of those killed in battle to prove the primary fighting style of the period warrior.

This kind study is known as battlefield archeology, conflict archeology or — when focusing specifically on forensic analysis of fallen warriors —  bioarcheology. As forensic sciences improve,  is becoming a major area of research.

Extensive digs at the site of the Seige of Kamakura in 1333 (a part of the Genkō War which saw the end of the Kamakura Shogunate), as well as from adjunct burial mounds, have been heavily investigated by forensic anthropologists and archeologists since the 1950’s, with many sets of remains being recovered.

Research on these remains are at the heart of the argument over Japanese battlefield swordmanship during the Kamakura period.  Depending on how the results are interpreted, the warriors involved were either skilled and honourable swordsmen engaged in one-on-one duels, or the remains are those of poorly armed and armoured foot soldiers slaughtered by mounted samurai using naginata or spear.


Samurai bushi (warriors)


A quick look at the history

Today it has generally been agreed by historians that the bow was the primary battlefield weapon of the time.  An excellent distillation of the view of various historians on this subject can be found in Dr. F.A.B. Coutinho’s paper for the Japanese Sword Society of the United States, Use of the sword, part 1 – Some comments on the form and use of the tachi in battle.

In this paper, he quotes a number of references which suggest the sword was far from the primary battlefield weapon of the bushi;  though not all academics disagree.  Instead, investigation into old records make it appear that the bow was the major weapon of war:

“…a samurai’s worth was measured by his prowess with the bow rather than the sword.” Stephen Turnbull, The Samurai Swordsman Master of War, Tuttle Publishing, Tokyo (2008)


“Warriors, of course, also carried swords, but until the mid-sixteenth century, these were not considered principal battlefield armaments. During the Heian period, swords were auxiliary weapons, analogous to the side-arm worn by modern [20th century] military officers. They were sometimes employed in combat, when warriors fought on foot at very close quarters, or when they run out of arrows or were otherwise deprived of their primary weapons…”  Karl Friday, Valorous Butchers: The Art of War during the golden age of the Samurai, Japan Forum 5 pp 1-19


The disagreement between historians is, in part, what has driven the exploration into the bioarcheology.


The research of Myra Shackley

One of the first academics to use forensic evidence to prove  of their theory of battlefield behaviour is Myra Shackley, who, in her 1986 paper Arms and the Men; 14th century Japanese swordsmanship illustrated by skeletons from Zaimokuza, near Kamakura, Japan, analyzed the wounds found on skulls from the Zaimokuza site.

In her paper, she puts forward the concept that warfare of the time was changing due to the attempted Mongol invasions.  She suggests that warfare was changing from a series of independent duels of honour between mounted samurai — Mongols weren’t interested in fighting honourably —  to the more formal use of mass formations of foot-soldiers (ashigaru).  Ashigaru were thought to be lightly armed and armoured rabble, and not to have a high level of military skill;  some academics have suggested there was a general reduction of fighting skills at the time, and that even sword skills fell into decline.

In her research she examined 26 skulls from Zaimokuza which exhibited measurable injuries, 95% of which were sword cuts (25 out of 26).  Of these 25, 21 had cuts across the hairline or on top of the skull;  some were single cuts, others were “skidding cuts”, where a blow hit at a sharp angle, making a first large cut then skidding into a second notch, and so on (see image below).  There were a few cuts on the left top of the skull, but none on the right-top, likely because most warriors were right-handed.


Frontal cut and skidding cut on skulls

In detail she adds:

…11% of skulls showed different combinations of these injuries. Most (78%) of the cuts were relatively shallow (1.5mm deep) and 90% had not penetrated the internal lamella of the brain. Of the remaining 10% half were exceptionally deep cuts through both lamellae into the brain without removing any bone, but in the remaining 5% large pieces of bone had been removed, exposing the brain. Such cuts were sometimes inflicted at a shallow angle, slicing away a piece of the victim’s head. Several injuries were inflicted by the extreme tip (kissaki) of the sword, which had sometimes been twisted while extracting the blade, removing a roundel of bone. Some depressed fractures from blunt instruments were noted although these were rare (8% of skulls); skull no. 451, for example, had an impact injury caused by a small diameter blunt instrument driven into the left eyebrow ridge, possibly a kashira (sword pommel) or the end of a staff

She notes the accuracy of many of these blows, saying that the majority crossed suture lines (where the bones of the skull merge), or occur just under the hairline;  from this she draws it was unlikely the victims were wearing helmets, or that helmet provided little protection.  From the angle of the cuts she determines that the attackers were likely mounted bushi using swords.  The deeper wounds mentioned she attributes to the use of heavier swords in single combat;  from this she draws the conclusion that the victim and their opponent were likely high-ranking bushi, following the samurai code of conduct and conducting an honourable duel.

Finally, drawing on the research of others, she notes that no injuries to the back of the skull were found on the skulls of the victims.  Injuries to the back of the skull have been thought by previous researchers as being inflicted on the victims as they were running away from battle.  From this she draws the conclusion that the warriors were following the old tradition of honourable combat.

In her conclusion, influenced in part by her own study of “ancient sword techniques”, she states that her research showed bushi of the period were practicing honourable combat,  were highly skilled swordsmen, and followed classic samurai protocols:

…the pattern of sword cuts on the Zaimokuza bushi reflects swordsmanship along classical lines, dominated by a light sword, either tachi or katana (the two are differentiated only by their method of carrying, not by blade widths) used from horseback. Most attacks seem to have been made by a mounted warrior using a stroke aimed directly at the centre of an opponent’s head, cutting him across the temple with the boshi of the sword.

and that: “the supposedly ‘bureaucratised bushi’ (Warner and Draeger 1982) of late Kamakura times still fought according to the martial traditions of the classic hero-warriors of the Minamoto bafuku, more than a century previously.”


The work of Antony Karasulas

This theory, that the bushi of the Kamakura period were expert swordsman, with the sword being their primary weapon after the bow, has been challenged by Mr. Karasulas in his paper, Zaimokuza reconsidered: The forensic evidence, and classical Japanese swordsmanship.  He conducted a series of test-cutting exercises on cow pelvises from both standing and horseback levels in an attempt to recreate the damage found on the skulls from Zaimokuza:

Through a multitude of tameshigiri, or test cuts, with a 65 year old katana, a 350 year old tachi, and a modern katana, Karasulas is able to determine that it is highly unlikely that the cuts on their warrior skulls were solely due to a sword. Karasulas then concludes that the weapons likely used were either the naginata or the yari, or a combination of the two. Both can be used on horseback and can inflict cuts much like those found on the skulls, but with the presence of its curved blade, Karasulas seems to believe that the naginata was the chief weapon used. Furthermore, because many of the blows were inflicted on top of the skull, Karasulas deduces that one side was mounted and the other side was on foot.

It should be noted that both the naginata and yari (spear) were used to both cut and thrust; many spear blades of the time were shaped like a sword (Stab Marks Possibly from a Spear (Yari) on a Skull Excavated from the Medieval Zaimokuza Site, Kamakura City, H. Fujita).  Both were also in common use at the time period, and have been historically confirmed as being used by men on horseback.

 In Japan, yaris first appeared in the literature in1334, 1 year after the fall of the Kamakura shogunate (Kondo, 1998). However, yaris were thought to have been used in combat in the late Kamakura period, and art from that period depicts soldiers with yaris.   (H. Fujita, Stab Marks Possibility from a Spear)

Karasulas rejects Shackley’s conclusion that the battle was between high-ranking warriors.  Rather than suppose a complicated bout of ritualized fighting, he draws the simple conclusion that: if one side is on horseback and the other on foot; and if the victims apparently had no helmets or other head protection;  then it was likely the remains are those of lightly armoured ashigaru foot soldiers, and not high ranking (and therefore likely to be heavily armoured) bushi.  Given this, there is no confirmation of a high level of period sword use, skill, or the following of ancient traditions found in these remains.

While the argument continues, most archeologists seem to be following Karasulas’ analysis, as it also coincides with the results of historical research.


The future of bioarcheology?

As a final note, there is continuing development in the field of bioarcheology which may one day be able to present a definitive answer to the question of Kamakura period sword use by bushi.  Continuing Karazulas’ idea of using practical test cutting on animal bones, we find the work of Jason Lewis:  Identifying sword marks on bone: criteria for distinguishing between cut marks made by different classes of bladed weapons (free download).   He conducted a series of tests on cow hind limbs with a variety of weapons to see if specific classes of weapons made identifiable injuries.

The weapons tested were:  katanas, scimitars, broadswords, short sword, machete and knife.  His preliminary results are highly interesting, as certain classes of weapon did indeed tend to create specific bone injuries.  Scimitars left bone fragments in the cut;  machetes created conchoidal flaking (circular, flaking cuts or breaks).  Katanas cuts tended to be clean on one side with rough flaking on the other.  Broadswords created a wide number different cut effects.  The cut effect of each weapon type could be assigned to the physical characteristics of the weapon (weight, sharpness, etc).  Katanas and broadswords created roughly similar cuts, as did short swords and machetes.


Conchoidal flaking


His conclusion was that there was enough evidence  classes of weapons making specific types of damage to warrant further investigation and experimentation, with larger sample sizes hopefully revealing a more exact trends, and therefore better analysis of weapon damage on skeletal remains.

Lewis chart of cut types by weapon

Click to enlarge





Leave a Comment

Your email address will not be published. Required fields are marked *