Kamakura Era: When the Warriors Took Over and the Mongols Tried to Stop Them

On the night of 15 August 1281, a typhoon hit the western coast of Kyushu and tore apart roughly 4,000 vessels of a Mongol invasion fleet that had been sitting offshore for almost two months. Estimates of the dead among the Yuan dynasty’s combined Mongol, Korean, and Southern Song forces run from 60,000 to over 100,000. The Japanese called the storm kamikaze (神風), the divine wind, and the word stuck for the next 664 years until it was repurposed for naval suicide pilots in 1945. That single weather event, hitting the second of two failed Mongol invasions, is the most famous moment of the period I want to walk you through here, the Kamakura era (鎌倉時代) of 1185 to 1333.

What changed in 1185

For four hundred years before this, Japan had been run from Kyoto by the Heian court (平安京), an aristocracy of poetry-writing nobles whose chief skills were calligraphy, perfume blending, and recognising eight-syllable allusions to Tang dynasty verse. The warrior class existed, but it served the court as bodyguards and provincial enforcers. Then in five years between 1180 and 1185, a civil war called the Genpei War (源平合戦) ended that arrangement and put the warriors in charge for the next 700 years. The Kamakura era is the start of warrior rule and the genuine end of the classical aristocratic period in Japan.

I always tell people who are trying to get a handle on Japanese history that the country has had three big political pivots, and 1185 is the first of them. Before Kamakura, the emperor and the Fujiwara regents in Kyoto ran everything in name and largely in fact. After Kamakura, the emperor still existed in Kyoto, but real military and tax power lived 460 kilometres east in a coastal town that until 1180 had been a fishing village. The man who built that move was Minamoto no Yoritomo (源頼朝, 1147 to 1199).

How the Genpei War set up the bakufu

The Genpei War was a five-year fight between two warrior clans, the Taira (平氏) and the Minamoto (源氏), both descended from imperial princes pushed out of the line of succession. The Taira had been running Kyoto since 1160, and their head Taira no Kiyomori had married his daughter into the imperial family and put his two-year-old grandson on the throne. The Minamoto had been almost exterminated in 1160, with the young Yoritomo exiled to the Izu Peninsula and his half-brother Yoshitsune sent to a monastery north of Kyoto. In 1180 a discontented prince called Mochihito issued a call to arms against the Taira, and the surviving Minamoto answered.

If you want the easy version: Yoritomo, exiled in eastern Japan, raised the eastern warrior clans against the Taira’s western base. His cousin Kiso no Yoshinaka and his half-brother Yoshitsune did most of the actual fighting. The Taira fled Kyoto in 1183 with the boy emperor Antoku and the three sacred imperial regalia, and the war ended in 1185 at the sea battle of Dan-no-Ura in the Kanmon Strait between Honshu and Kyushu. There the Taira fleet was crushed, the seven-year-old Antoku drowned in his grandmother’s arms, and the imperial sword Kusanagi-no-Tsurugi was lost in the sea, an event I get into in more detail in my piece on Japan’s three sacred treasures.

Why Kamakura, not Kyoto

What Yoritomo did next is the move that defines the era. Instead of marching to Kyoto, taking a court title from the emperor, and ruling like a Heian noble, he stayed in Kamakura. The town sat on the Pacific coast in what is now Kanagawa Prefecture, hemmed in by hills on three sides and water on the fourth, with seven narrow passes called kiridōshi as its only land entrances. Yoritomo had picked it as his base in 1180 because the geography made it easy to defend and because his ancestor Minamoto no Yoriyoshi had built a Hachiman shrine there in 1063.

The deeper reason was that Yoritomo wanted distance. He had grown up in Kyoto until age thirteen, watched his father executed, and spent his exile reading. He knew exactly how the court culture worked, and he knew that anyone who tried to rule from inside it ended up either co-opted, married to Fujiwara women, or stabbed in a sliding door corridor at three in the morning. Building an entirely new capital in the east was his way of refusing to play that game.

In 1192, three years after the imperial regent in Kyoto stopped resisting, Yoritomo accepted from Emperor Go-Toba the title of Sei-i Taishōgun (征夷大将軍, “barbarian-subduing great general”), originally an eighth-century rank for generals fighting the Ainu in the north. From that point onwards there was a parallel government in Kamakura called the bakufu (幕府, “tent government”), and the era is named for the town. The shogun was the formal head, but the apparatus underneath him was a council of warrior-class administrators called the Mandokoro, the Samurai-dokoro, and the Monchūjo, dealing with finance, military discipline, and law respectively. None of this had existed in 1184.

Yoritomo’s death and the Hōjō takeover

If you read the popular history, you might think Yoritomo lived a long satisfying life ruling his new bakufu. He did not. In December 1198 he attended the dedication of a new bridge over the Sagami River, fell off his horse on the way home, and died early in 1199 at the age of fifty-one. The chronicles are coy about whether he had a stroke, an aneurysm, or simply broke his neck when the horse threw him, but he was dead inside two months.

His sons Yoritomo’s heir Yoriie was eighteen and Sanetomo was eight. Neither was capable of running a warrior government, and Yoriie spent more time playing kemari football with his pages than reading the Mandokoro briefs. So the actual government was taken over, almost immediately and very quietly, by Yoritomo’s widow Hōjō Masako (北条政子, 1156 to 1225) and her father Hōjō Tokimasa.

Masako is the figure I find most consistently underrated when people talk about the Kamakura era. She was the daughter of one warrior chieftain and the wife of another, and after 1199 she ran the bakufu through every generation of weak nominal shoguns the Hōjō clan kept installing. When her son Sanetomo was assassinated in 1219, the line of Minamoto shoguns ran out, and Masako personally went to Kyoto and negotiated to bring a two-year-old Fujiwara child east as the new puppet shogun. The Hōjō clan, her birth family, then formalised a position called shikken (執権, “regent for the regent”) and ran the government from there for the next 114 years.

The 1232 Goseibai Shikimoku and warrior law

One of the things the Hōjō shikken did, and one of the genuinely innovative things to come out of this period, was write down samurai law for the first time. In 1232 the third shikken Hōjō Yasutoki promulgated a fifty-one-article code called the Goseibai Shikimoku (御成敗式目), or “Formulary of Adjudications.” It was meant to apply only to the warrior class and only on bakufu-administered estates, but it was the first written body of secular Japanese law.

The articles cover what you might expect and a lot you might not. Article 18 says that women may inherit and hold land in their own name, which would be revoked by later samurai codes but was the rule for the whole Kamakura period. Article 30 says false accusation in a lawsuit is punished by exile, regardless of the accuser’s rank. The code was so respected that it was still being used as a private contractual reference text in the eighteenth century, almost six hundred years after it was written.

Reading the Shikimoku now, the part that always strikes me is how much it reads like a working office manual rather than a piece of imperial pomp. The Heian court’s legal codes had been adaptations of Chinese Tang dynasty law, with grand statements about the cosmic mandate of the emperor. Yasutoki’s code is short, in plain language, and worried mostly about whether a vassal who has not paid his rent should be allowed to stay on his estate. That practicality is the warrior class showing what it cares about.

New Buddhism: when religion went mass-market

The other thing that happened during the Kamakura era, and that I think gets under-emphasised compared to the political stuff, is a wholesale reorganisation of Japanese religion. The Heian period had built two giant aristocratic Buddhist schools, Tendai and Shingon, both based on enormous mountain temple complexes that you needed wealth and family connections to enter. Between roughly 1175 and 1253, six new schools of Buddhism appeared, all of them aimed at people the older schools did not really serve.

The first wave was Pure Land. In 1175 the monk Hōnen (法然) broke with the Tendai school on Mount Hiei and started teaching that anyone, regardless of class or learning, could be reborn in Amida Buddha’s Pure Land just by reciting the chant namu Amida Butsu, a practice called nembutsu. His student Shinran (親鸞) took it further, founded a new sect called Jōdo Shinshū (浄土真宗), and broke the rule that priests could not marry. Shinran’s was the first married Buddhist priesthood in East Asia, and Jōdo Shinshū is still the largest single Buddhist denomination in Japan today.

The second wave was Zen. In 1191 the monk Eisai (栄西) returned from a study trip to Song-dynasty China and brought back two things: the Rinzai school of Zen, and the seeds of green tea. He planted the tea on the grounds of the Reisen-ji temple in Kyushu, and the practice of drinking powdered green tea spread up through the Kamakura warrior class along with Zen meditation. This is the actual origin of what later became the codified Japanese tea ceremony, which I get into in more detail in my piece on the tea ceremony.

Eisai’s student Dōgen (道元) went to China himself in 1223 and came back in 1227 with the Sōtō school of Zen, which emphasised silent seated meditation called shikantaza (“just sitting”) rather than the koan riddles of Rinzai. Dōgen rejected the city and built his monastery deep in the mountains of what is now Fukui Prefecture. His writing is the part of Kamakura Buddhism that has had the strongest reach outside Japan in the twentieth century, with translations of the Shōbōgenzō (正法眼蔵) influencing everyone from the Beat poets to Heidegger.

The last of the new schools was the most aggressive. In 1253 a Tendai-trained monk who renamed himself Nichiren (日蓮, “Sun Lotus”) declared that all the other schools of Buddhism, including the brand-new Pure Land and Zen schools, were leading Japan to ruin. He taught that the only valid Buddhist text was the Lotus Sutra, and the only valid practice was chanting its title, namu myōhō renge kyō. He was exiled twice, almost executed once at Tatsunokuchi in 1271, and his prediction that Japan would be invaded by foreign powers seemed validated three years later when the first Mongol fleet arrived.

The Kamakura Daibutsu and Hase-dera

The single most photographed object from this entire era is the Kamakura Daibutsu (鎌倉大仏), the enormous bronze Amida Buddha at Kōtoku-in temple. The figure was cast between 1252 and 1264, weighs about 121 tonnes, and stands 11.4 metres tall, making it the second-largest bronze Buddha statue in Japan after the one in Tōdai-ji in Nara. Unlike the Nara Daibutsu, which sits inside the world’s largest wooden building, the Kamakura one has been outdoors since 1498, when a tsunami took down the Great Buddha Hall and nobody had the budget to rebuild it.

What I always notice when I sit in front of it is the slight forward tilt of the head, deliberate so that the face appears to look down at a worshipper standing a few metres away. The Pure Land iconography here is doing real work. This is Amida Buddha as imagined by a population that had been told for the last seventy years that all they had to do was chant his name, and the statue is sized and positioned to make that promise feel solid.

Walk fifteen minutes downhill from the Daibutsu and you reach Hase-dera (長谷寺), the Kannon temple on the slope above Yuigahama beach. Hase-dera was founded in 736 according to its own records, but the entire current Kannon-dō hall and most of the temple complex were rebuilt during the Kamakura period under Hōjō patronage. The temple’s claim that the eleven-headed wooden Kannon was carved from the same camphor tree as the famous Kannon at Hase-dera in Nara is almost certainly a later legend, but the carving itself is one of the largest wooden Buddhist statues anywhere in the country.

Tsurugaoka Hachimangū and the city of Kamakura

If the Daibutsu is the era’s signature religious image, Tsurugaoka Hachimangū (鶴岡八幡宮) is its political one. The shrine sits at the head of the long avenue that runs through the centre of Kamakura down to the sea, an avenue called Wakamiya-ōji that Yoritomo himself laid out in 1182 as a north-south axis for his new capital. Hachiman (八幡神) is the Shinto deity of war and the protector of warriors, the guardian kami of the Minamoto clan specifically.

The shrine’s full name is Tsurugaoka Hachimangū-ji (鶴岡八幡宮寺), and until 1868 it was a shrine-temple complex with both Shinto and Buddhist halls on the grounds. The Meiji government’s forced separation of Shinto and Buddhism in 1868 stripped out the Buddhist halls and renamed the institution as a pure Shinto shrine, but if you look carefully at the architecture you can still see the temple-style gables on some of the older buildings. The yabusame mounted-archery festival held every September is a direct continuation of the warrior martial arts that Yoritomo himself sponsored at the shrine in 1187.

The Mongol invasions: 1274 and 1281

Now I want to walk you through the part of the era that everyone has heard about, even people who have never heard of Yoritomo. In 1268 the Mongol emperor Kublai Khan, having finished conquering the Jin and Western Xia dynasties and most of the way through finishing off the Southern Song, sent a letter to Japan via the Korean kingdom of Goryeo. The letter was politely worded but the substance was: become a Yuan dynasty tributary, or face invasion. The Hōjō shikken Tokimune declined to reply at all.

In November 1274 a Yuan-Goryeo combined fleet of about 900 ships and 40,000 troops sailed from Goryeo, took the islands of Tsushima and Iki almost without resistance, and landed at Hakata Bay on the north coast of Kyushu. The fighting that followed, known to the Japanese as the Battle of Bun’ei (文永の役), was a one-day bloodbath. The Mongol forces had massed crossbow volleys, ceramic exploding shells called tetsuhō, and tight infantry formations. The Japanese samurai had individual mounted archery and a tradition of riding forward and announcing your lineage before single combat.

Suenaga, the samurai who commissioned that scroll, fought in both invasions, and his text is one of the rare first-person warrior accounts to survive from the Kamakura period. He describes the Mongol forces as fighting in tight units that did not break, using an alarming overlapping shield wall, and shooting volleys of poisoned arrows. He also notes, with a kind of professional irritation, that the Mongols would not stop and fight one-on-one no matter how loudly he announced his name and lineage.

By the end of the first day of fighting, the Japanese had been pushed back from the beach into the inland defensive line at Mizuki. That night the Mongol fleet, perhaps worried about a counterattack and definitely worried about a brewing storm, withdrew offshore to anchor. A typhoon then hit the fleet on 19 November 1274 and is estimated to have sunk about a third of the ships and drowned over 13,000 men. The survivors retreated to Korea, and the first invasion was over in five days.

The Genkō Bōrui and the second invasion

The Hōjō government’s response to surviving 1274 was to assume, correctly, that it would happen again. From 1275 onwards Tokimune ordered the construction of a defensive stone wall along the inside of Hakata Bay, the section of coastline where any second Mongol fleet would have to land. The wall was called the Genkō Bōrui (元寇防塁) or “Bun’ei-Kōan defensive wall,” and it ran for about twenty kilometres in segments, two metres tall and three metres thick, made of beach stones piled and faced with cut stone.

The second invasion came in summer 1281, and it was on a different scale entirely. A southern fleet sailed from the conquered Southern Song coast at Ningbo with about 100,000 men in 3,500 ships, and an eastern fleet sailed from Korea with 40,000 men in 900 ships. Combined, that gave Kublai roughly 140,000 troops, which made it the largest amphibious operation ever attempted up to that point. The two fleets were supposed to rendezvous off Iki Island and land together at Hakata.

The Mongol commanders found that the wall worked. The Korean-launched eastern fleet arrived first and was unable to find a beach to land on, with the Genkō Bōrui making every plausible site impossible. They captured Shika Island as a temporary base and waited for the southern fleet, which was several weeks late because of supply problems in Ningbo. While the eastern fleet was waiting, samurai units made night raids on the anchored ships in small fast boats, boarding and burning at random.

By 12 August the southern fleet had finally arrived and the combined force of 4,400 ships was anchored off Takashima Island, preparing for what would have been the actual landing. Then, on the night of 15 August, the typhoon hit. The two fleets had no time to make for open water and were caught at anchor along the Imari Bay coastline.

What followed is the kamikaze of legend, and the actual numbers are genuinely staggering. Modern Japanese estimates put the Yuan losses at between 60 and 90 percent of the men in the fleet, with one estimate from the Yuan court’s own records counting only 19,397 of the 142,000 men returning home. The samurai units mopped up the survivors who reached shore over the next six days, killing or capturing tens of thousands more. Kublai Khan would plan a third invasion, but he never managed to launch it before his death in 1294.

The land grants the Hōjō could not pay

Here is the bit that I think is genuinely under-told in the standard story. The samurai who fought in both invasions had been promised land grants in compensation for their service, the way every successful Japanese war up to that point had been settled. The problem is that the Mongol invasions had not produced any conquered territory. The Hōjō shikken had no Korean estates to hand out, no Yuan provinces to assign as fiefs, nothing.

So roughly 30,000 samurai found themselves with formal claims for land and money, decades of garrison duty along the Hakata coast on top of those claims, and a government that could only pay them in promises. The bakufu issued debt-cancellation edicts called tokusei (徳政令) in 1297 to wipe out warrior debts, but the edicts were so disruptive to the credit economy that they had to be revised within a year, and they cost the Hōjō clan significant political legitimacy. By 1300 the warrior class that had spent the 1270s defending Japan was a class of bankrupt complainers, and the bakufu’s grip on its own samurai started to slip.

The other slow-burn problem was that the Hōjō clan had concentrated power inside its own family. By 1300 the shikken was a Hōjō, the deputy regent was a Hōjō, the chief retainer of the Hōjō was a Hōjō, and most of the senior judicial and military offices were filled with Hōjō relatives or clients. Other warrior clans, especially those with claims still unpaid from the Mongol-defence service, had no path forward inside the system. That is the structural setup for what happens in 1331.

Go-Daigo, the Kemmu Restoration, and 1333

In 1318 a determined and unusually capable emperor named Go-Daigo (後醍醐天皇) ascended the throne in Kyoto. Unlike most of his recent predecessors, he had no intention of being a ceremonial figurehead while a retired emperor and the Hōjō shikken ran the country. He spent the 1320s building a network of monks, court nobles, and discontented warrior families, and in 1331 he openly raised an army against the bakufu.

Go-Daigo was caught and exiled to the Oki Islands in 1332, but his cause was picked up by warriors the Hōjō had alienated. The decisive figure turned out to be a man called Ashikaga Takauji (足利尊氏), head of the Ashikaga clan, traditionally a Minamoto cadet branch loyal to the bakufu. In May 1333 Takauji was sent west by the bakufu to suppress an imperial loyalist army in the Kyoto area, switched sides en route, and attacked the bakufu’s western military headquarters at Rokuhara instead.

Simultaneously another Minamoto-line warrior called Nitta Yoshisada raised the Kantō region and marched on Kamakura itself. On 4 July 1333 his forces fought their way through the seven mountain passes and into the city, and the last Hōjō shikken Hōjō Takatoki retreated to his clan’s family temple of Tōshō-ji on the eastern edge of town. There, on 4 July, Takatoki and approximately 870 members of the Hōjō clan and their senior retainers committed mass suicide. The Kamakura bakufu, after 148 years, ended in a single afternoon at a temple that no longer exists, with the entire ruling clan dead.

What came after, and what 1333 actually meant

The standard textbook version is that Go-Daigo returned from exile in triumph and started the Kemmu Restoration (建武の新政) of 1333 to 1336, in which the emperor would supposedly govern directly without a shogun. The reality is that the Restoration lasted three years and collapsed because Go-Daigo systematically failed to reward the warrior families who had put him on the throne. By 1336 Ashikaga Takauji had betrayed Go-Daigo, set up a new shogunate of his own in Kyoto, and started what is now called the Muromachi period and the Northern and Southern Courts schism.

What 1333 really meant, in retrospect, was that the political question raised by Yoritomo in 1185 had been answered. Could the warrior class actually run Japan? Yes, it turned out, but only if the warrior government was willing to share the spoils with the warrior families that fought for it, and only if it kept enough distance from court culture to remain a warrior government. The Kamakura bakufu had failed on the first point and was starting to fail on the second by the time the typhoons of 1281 had finished blowing.

The Ashikaga successor government would learn from the failure on the first point and lose on the second, becoming so absorbed in Kyoto culture that it could not control the provincial warrior houses, and the country slid into a hundred years of civil war. Only with the Tokugawa unification of 1600 did the warrior class manage to thread the needle. But the model they used was Yoritomo’s: a warrior capital outside Kyoto, a separate warrior administration, and a strict refusal to be absorbed by court culture.

What survives, and where to actually go

If you want to walk the era yourself, Kamakura the city is the obvious place to start, and a single day will not be enough. Tsurugaoka Hachimangū sits at the head of Wakamiya-ōji and is two minutes from the JR Kamakura Station. The Daibutsu and Hase-dera are a fifteen-minute Enoden tram ride west. Engaku-ji, Kenchō-ji, and the Zen temple cluster of Kita-Kamakura are one stop further out on the JR line. You can do all of these in a single twelve-hour day if you start early, but I would not recommend it.

For the Mongol invasions, the place to go is Fukuoka in northern Kyushu. Sections of the Genkō Bōrui still survive along the Hakata Bay coastline at Imazu, Nishijin, and Iki-no-Matsubara, and the Imazu site has interpretation panels and a partial reconstruction. The Genkō Memorial in Higashi Park has a giant statue of Hōjō Tokimune, and the small island of Takashima off Imari Bay has a Mongol Memorial Park where the second invasion fleet came ashore. Underwater archaeology in Imari Bay since 2011 has recovered substantial wreckage and is still ongoing.

For Zen architecture as imported in this era, Kenchō-ji and Engaku-ji are the obvious choices, but if you want the Sōtō school’s flagship, you have to go to Eihei-ji deep in the mountains of Fukui. Dōgen built it there in 1244 to be inaccessible, and the train and bus journey from Tokyo still takes most of a day. The discipline of the place has not slackened in the intervening 780 years, and the morning meditation in the founder’s hall is one of the things on Earth I would tell anyone interested in this era to make a pilgrimage for.

What it left in the cultural DNA

I think the single most important thing the Kamakura era left behind is the warrior aesthetic, which the Edo period would later refine into the codified bushidō but which is fundamentally a Kamakura invention. The Heian court had treated combat as a thing professional warriors did somewhere far from the capital, the way a Roman senator treated soldiering. The Kamakura warrior class made combat a matter of personal honour, family lineage, and aesthetic restraint, and it imposed those values on its own descendants for the next 700 years.

That aesthetic shows up everywhere downstream. The cavalry sword called the tachi (太刀) becomes the samurai standard during this era, an angled long-sword designed for cutting from horseback rather than the straighter chokutō of earlier centuries; I cover that evolution in my piece on Japanese swords katana and tachi. The hakama (袴), the formal wide-legged trousers worn over the kimono, becomes standard warrior dress, and the construction techniques get codified in this era; I get into that in samurai and the hakama. The Heian period the Kamakura era replaced is itself a story I cover at my piece on the Heian era.

The Kamakura era also left the Zen temple garden, which would mature into the dry-landscape karesansui form during the late Kamakura and early Muromachi periods under the monk Musō Soseki. Musō’s gardens at Saihō-ji in Kyoto and Tenryū-ji are the prototype of the form everyone now thinks of when they think “Japanese garden,” and I get into that history at length in my piece on Japanese gardens and on Japan’s greatest gardens. The Kamakura warrior also built the first true mountain fortifications, the precursors to the great stone keeps, which I cover in Japan’s twelve original castles.

And the unarmed-combat traditions that would later become karate and aikido have roots in this era’s warrior training as well. The Kamakura samurai trained in jūjutsu-style grappling for moments when the sword was lost or unavailable, and that body of practice is the starting point of aikido and indirectly of karate centuries later.

The summary I keep in my head

If I had to compress 148 years into a paragraph: in 1185 a careful warrior named Yoritomo built a parallel government on the Pacific coast and refused to play the Kyoto court’s game. He died young in 1199, and his wife Masako and her birth family the Hōjō ran the bakufu through nominal child-shoguns for the next 130 years. In 1274 and 1281 the Mongols came, twice, and were beaten by a combination of stone walls, samurai night raids, and two genuine typhoons. The financial cost of those defences bankrupted the warrior class the Hōjō relied on, and in 1333 a coalition of disaffected warriors and a determined emperor brought down the entire regime in one afternoon.

The era is in some ways the foundation of everything that follows in Japanese political history. The Tokugawa Shogunate of 1600 is doing Yoritomo’s playbook, fifteen generations later. The Imperial Restoration of 1868 is undoing it, by reversing the basic Kamakura settlement. Even the constitutional arrangement of post-1947 Japan, with a ceremonial emperor in Tokyo and a real government also in Tokyo, owes something to the original Kamakura split between a Kyoto throne and a Kanagawa-coast cabinet.

I will say what I always say at the end of these era walkthroughs, which is that the best way to get a feel for any of this is to put your boots on the ground. Kamakura is an hour from Tokyo on the JR Yokosuka line, the Daibutsu charges 300 yen for entry, and Engaku-ji opens at 8 in the morning with the meditation hall taking visitors. If you can sit on the steps below Tokimune’s gate and watch the cedar shadows move across the courtyard, you will understand more of the era in twenty minutes than I can put across in 7,000 words here.