A History of Japan: From Mythology to Nationhood/The Nara Period

The traditional dates for the second Nara period are 710-784 (current era). The first is the date of the move to the new state and the last is the date of the move to Heiankyo. However, it is not difficult to find more meaningful dates. My starting date is 689, the date of the publication of the Kiyomihara Ritsuryo, which was the formal inauguration of the new system of government promised in 645. The end date is the death of Kammu Tenno in 806. This range provides complete coverage of what may be termed the second of five phases in the Japanese experiment with adapting the Chinese system of government to their needs.

The first phase was the Asuka period when the rulers of Japan first aspired to transform the political structure and culture of the country, starting in the time of Prince Umayado and Soga no Umako, continuing with the Taika reform movement of Samari no Oe/Tenchi Tenno, and culminating with the work of Temmu Tenno, leading to the accomplishment of the Kiyomihara code.

The second phase is the time when the Japanese rulers seriously grappled with the task of making the new system work. The Nara period is the high point of the enterprise, and in many ways can be considered a period of heroic achievement. It is impossible to avoid the conclusion that overall the effort was a failure. Japan did not have the economic resources to support a miniature version of the Chinese empire. However, they gave it a very good try. They did not achieve their goals, but they transformed Japan into a very different country from the tribal state of the 6th century, a country that can fairly be called civilized in terms of intellectual and artistic development, a status that would never be lost.

Documentation
The basic text for a history of the Nara Period is Shoku Nihongi. It begins with the accession of Mommu Tenno in 697 and covers 9 reigns through the 10th year of the reign of Kammu, 791. It was begun in 791 by Fujiwara no Tsugutada, Udaijin. Tsugutada died in 796 and the last half of the book, which was already finished, was published that year. The first half was published after further work by Sugano no Mamichi the following year. The two halves of the book differ considerably. The first half covers 61 years and the second half 33 but each is roughly the same number of pages. What is now the first half seems to have been compressed from a planned 30 volumes to 20. Neither author was interested in recording the details of life, but rather moral and political lessons.

We also have a substantial portion of the original administrative codes and there are a considerable number of Heian period compilations of edicts and rulings that include commentary on the codes which have permitted scholars to reconstruct large sections that have not survived.

Kaifuso is a collection of poetry written in Chinese which contains biographical information about the poets. Toshikaden was published by Fujiwara no Nakamaro and records traditions concerning Fujiwara no Kamatari and other early members of the clan. Todaiwa Joto Seiden is a biography of the immigrant Chinese monk Ganshin. Approximately 12,000 items of 8th century government documents still exist. Some 99% of these come from Shosoin and were published in 25 volumes between 1901 and 1942 as Dainihon Komonjo. There are about 20 documents surviving from before the 8th century and some 10,000 from the entire 400 years of the Heian period. Nothing that might be called personal writing survives from before the Heian period. For those who may not know, Shosoin was an endowed foundation set up by Shomu Tennu and affiliated with Todaiji in Nara, a temple Shomu had built, which remains a major tourist attraction today. A wooden storehouse was constructed to hold the personal property of Shomu after his death, a time capsule, in effect, from the Nara period, which survives intact with its contents today. It contains clothing, musical instruments, art work from as far away as Persia, and the thousands of documents mentioned above. No one is sure why the documents wound up there, but we are certainly glad that it happened. A large part of the documents relate to the affairs of Todaiji, but there are also many generated during the normal operation of the government.

From the population registers included among these documents it is possible to estimate a population of 6 million people for Japan in this period. The calculation starts from a note in Shoku Nihongi dated 747 to the effect that the government had established a standard population for a "village" (an artificial unit created for tax purposes) of 330 taxable males, that is, free males between the ages of 17 and 65. This resulted from years of discussion about the proper allocation of tax burdens in the absence of a census. The people were grouped into "households" but these varied in size. It was argued that by grouping 50 households into a nominal "village" the differing family sizes would even out and a fair tax quota could be assigned at the "village" level. A lengthy research project then produced the figure of 330 men per village. It must be clearly understood that this village was a mathematical abstraction that had no necessary correlation with how the population's houses were grouped. Analysis of surviving household registers shows that the taxable males on the average amounted to 23.58% of the whole. That gives a calculated total population of 1399 persons per "village". There is a 9th century book by Minamoto no Shitagao, Wamyo Ruijusho that gives the names of all the "villages" of the nation. Excluding certain special types that were thought to have fewer than 50 households, the total is 4041. Another old book of very uncertain origins called Ritsusho Zanben gives a total of 4021 "villages". If we take 1399 persons per village and 4000 villages we get 5,600,000 persons total. There are other ancient figures for numbers of soldiers raised, areas of land cultivated and the like, that do not conflict with a total on this scale.

One of the most valuable and curious sources of information about this age are 5 surviving books called, collectively, fudoki. They were evidently prepared at the order of the government, and there were, presumably, a great many more like books that have not survived. However the oldest mention of their existence dates to 914. That says that an edict was issued in 713 that the local officials of each district and province were to prepare what we would call gazeteers, describing the districts and villages of the province, describing local products of note, whether grown or manufactured, giving the names of all mountains, rivers, and plains, and also recording any ancient traditions of interest.

We have all or part of books for Harima, Hitachi, Izumo, Bungo, and Hizen provinces. This is a nicely diverse group. Bungo and Hizen are in Kyushu, Hitachi is in the “wild east” near modern Tokyo, and Harima and Izumo are in western Honshu, on the Inland Sea and the Japan Sea sides respectively. Both of the latter were very important places as far back as history and tradition go. There are many doubts about these books despite their obvious value. They were ultimately put together under the direction of provincial governors, who were nobles from the capital serving short terms in places they had never seen before. How much do they represent authentic local traditions, and how much do they simply say what the officials back in the capital wanted to see?

It is possible to put together some inferences about who was responsible for the books. The governor of Harima in 713 was Kose no Oji and his assistant was a well known literary figure, Sakanami no Kochi. There is evidence that the Harima Fudoki was completed in 718 and at that time the governor was Fujiwara no Umakai, a famous poet. Another well known poet, Takahashi no Mushimaro, was one of his subordinates.

The styles of the Bungo and Hizen volumes are identical, and it is believed that they, and presumably the other Kyushu volumes, were put together at Dazaifu with a common editor. It is entirely possible that Fujiwara no Umakai was involved with these also, because he served at Dazaifu after his assignment in Harima. Internal evidence for the dating of these two volumes shows that they were completed earlier than 740.

Izumo Fudoki alone gives its date of publication, 733, and the name of its author, Izumo no Kuninomiyatsuko Izumo no Omi Hiroshima. As far as provincial aristocratic names go, this one is at the top. A hundred years earlier, his ancestors ruled Izumo as a minor kingdom loosely subordinate to the emperor. Now, he was, officially, a mere district magistrate. As shall be seen, it is not at all obvious that the real power of such men had been suppressed nearly as much as the new laws called for. However, for the time being, they were keeping a low profile. It would, however, be an important piece of information if such a rural person, with no known time of residence in the capital region, were actually able to create a well-written book, which this one is. It is notable that Izumo Fudoki is the only one that actually records significant ancient traditional history, which makes sense if it was the only (surviving) one actually written by a local author. Most of the anecdotes found in the other books actually could have been taken directly from Chinese books of a similar character. There is nothing local about them. Stories of spectacular acts of filial piety are the norm. The Japanese could talk the talk, but only the Chinese walked the walk. The Japanese used to dump the old folks on a hillside to starve when times were tough, on the grounds that the future child bearers had to be saved. Not acceptable thinking in China, nor in 4 out of the 5 Fudoki. Some stories in the Fudoki can actually be traced to specific Chinese books.

Outline of Nara Political History
689 (3rd year of Jito Tenno): Jito did not formally ascend the throne until 690, but historians traditionally date events from 687 as her first year. In the 6th month the "Asuka Kiyomihara Ritsuryo" was distributed to government offices in 22 volumes, and in the 8th month the work of implementing the system in the provinces was begun. This was a comprehensive set of law codes specifying the rules for the administration of the government and also laws pertaining to criminal matters and what we would call family law. "Ritsu" is the Chinese word for law as administered by judges and "ryo" is the word covering the administrative regulations. Japanese historians often use ritsuryo system as a shorthand term for the entire Chinese-inspired system.

Prince Kusakabe, Jito's son, whom it is assumed would have succeeded Temmu Tenno but for his bad health, died at the age of 28. This left Kusakabe's son, prince Karu as the preferred heir, but he was too young to be considered eligible at this time. If anything happened to Jito, it is assumed that the most probable successor was now prince Takechi, the oldest surviving son of Temmu Tenno. Prince Takechi would not be considered as a prime candidate because of the low status of his mother.

690 (4th year of Jito Tenno): The enthronement ceremonies were held at the beginning of the year. In the 7th month the central administrative system described in the "ryo" went into effect. This means that the names of offices and official titles were changed over and appointments were made to the new offices. Prince Takechi received the highest office in the system, Dajodaijin, in this period reserved for a prince.

692 (6th year of Jito Tenno): The dedication ceremony for the new capital city of Fujiwara was held. In the 9th month officials were sent to the provinces to supervise the surveying of the rice fields required for the new land redistribution system.

694 (8th year of Jito Tenno): The government moved into the Fujiwara capital at the end of the year.

696 (10th year of Jito Tenno): Prince Takechi died. The post of Dajodaijin was left vacant.

697 (1st year of Mommu Tenno): Prince Karu was granted the title of crown prince at the start of the year, and in the 8th month Jito abdicated in his favor. Miyako, a daughter of Fujiwara Fuhito, was married to the emperor at this time.

700 (4th year of Mommu Tenno): Fujiwara Fuhito and other officials were ordered to prepare a revised version of the ryo. There is no surviving information as to exactly why. However, it seems clear that the new code was a substantial rewrite of the original, though it only took one year to prepare.

701 (1st year of Taiho, a Chinese style reign title which was adopted in the 3rd month in celebration of a report (later found to be false) of the discovery of gold): The revised law code, known as the Taiho Ryo, was published in the 8th month, and there was a lot of effort to make sure that its provisions were put into effect, especially in the provincial and district administrations.

702 (2nd year of Taiho): The retired empress died at the end of the year. Earlier in the year, a project was conducted in the provinces to put together a register of all persons who were eligible to be considered "kuni no miyatsuko". This was now an out of date title, but it was used to determine which persons could become district magistrates in the new system. An edict ordered that any candidate for this post who was of kuni no miyatsuko status was to be given preference.

703 (3rd year of Taiho): A new office substantially identical to Dajodaijin was created and prince Nittabe was appointed to hold it.

705 (2nd year of Keiun, a new reign title): Prince Nittabe died and prince Hozumi replaced him.

707 (4th year of Keiun): In the 6th month Mommu Tenno died at the age of 25. His mother took over the government, and one month later ascended the throne. She is known as Gemmei Tenno.

708 (1st year of Wado): Shortly before his death Mommu had ordered work to begin on a new capital city. His death disrupted the project, but it was now resumed. The site of Nara was selected.

710 (3rd year of Wado): In the 3rd month the government formally moved into the new capital. Most historians begin the Nara period at this time.

715 (1st year of Reiki): In the 7th month prince Hozumi died and was not replaced. In the 9th month Gemmei Tenno abdicated and was replaced by a sister of Mommu Tenno who is known as Gensho Tenno. No real reason was given for this. Some think that the death of Hozumi changed the political balance at court depriving Gemmei of support, others think she simply wanted to retire.

718 (2nd year of Yoro): Fujiwara Fuhito and others were ordered to prepare a further revision of the administrative law code. This work was never completed, but a copy of the code in the possession of Fuhito was discovered many years later and is the only version of the code (substantial) parts of which survive. Heian period commentators did not possess any copies of either the Kiyomihara code or the Taiho code and used the Yoro code. There is much debate over the question of whether the Yoro code is different than the Taiho code or is simply a copy of the Taiho code that Fuhito planned to use as a starting point. I don't believe that anyone has ever shown a significant difference between the Taiho code (as known from mentions in edicts and other sources) and the Yoro code.

720 (4th year of Yoro): Fujiwara Fuhito died in the 8th month and soon after prince Toneri was appointed head of the government. Most historians believe that Fuhito had been functioning as the head of government since the death of prince Hozumi, and those who think that the replacement of Gemmei by Gensho was political think that he was behind it. Fuhito had 4 sons, two of whom were members of the Dajokan (cabinet) at the time of his death, and all of whom were prominent politically.

721 (5th year of Yoro): The retired empress Gemmei died at the age of 61.

724 (1st year of Jinki): Gensho Tenno abdicated and the son of Mommu Tenno ascended the throne. He is known as Shomu Tenno.

727 (4th year of Jinki): A male child was born to the emperor and the daughter of Fujiwara Fuhito. Within a month the child was named crown prince, a title which up to now had always been given to an adult who was functioning as one of the leaders of the government. This represents a transition to the use of the title as meaning the designated successor, which became routine in the Heian period. However, at no time in ancient Japan was there ever a clearly defined path of succession. Many crown princes failed to become emperor when the moment arrived.

728 (1st year of Tempyo): The infant prince died shortly after his first birthday.

729 (2nd year of Tempyo): In the 2nd month two low ranking officials accused prince Nagaya, a grandson of Temmu Tenno who held the post of Sadaijin, second ranking in the government at the time, of having caused the death of the infant prince through sorcery. Within 24 hours the prince was ruled guilty and was ordered to commit suicide along with his wife and sons. Fujiwara Fusasaki, the highest ranking of Fuhito's sons, was in charge of the investigation, and nearly all historians assume that this was planned by the Fujiwara to get rid of the strongest candidate to succeed to the throne if anything happened to Shomu Tenno. Soon afterwards Fujiwara Miyako was formally designated Empress, and it is thought that prince Nagaya had opposed this.

735 (7th year of Tempyo): Prince Toneri died and was not replaced.

737 (9th year of Tempyo): There was a great epidemic, believed to be Japan's first encounter with smallpox, that began in Kyushu and spread over most of the country, killing large numbers of people, including many nobles. All four sons of Fujiwara Fuhito died between the 4th and the 8th months. In the 9th month prince Suzuka became the head of government, though it is believed that Tachibana no Moroe, head of a new clan that had been branched off of the imperial clan during the reign of Gensho, was effectively in charge.

738 (10th year of Tempyo): Princess Ae, a daughter of Shomu Tenno, was designated crown prince and heir. Several women had reigned as Tenno, of course, but this was the first time one was ever designated the heir. It would appear that this was a decision of Shomu Tenno personally. He never showed much interest in routine affairs of government, but was a willful individual who always got his way. At the same time Tachibana no Moroe was promoted to Udaijin, second in rank (allowing for vacant posts) to prince Suzuka.

740 (12th year of Tempyo): Fujiwara Hirotsugu was son of Umakai and grandson of Fuhito. He had been working as a provincial governor but was denied promotion to an office in the central government and sent off to Kyushu. In the 9th month he sent in a memorial attacking the government (there were famines and conditions had remained bad since the epidemic of 737) and specifically recommended the dismissal of two close advisors to Tachibana no Moroe. He also began raising troops. The government responded quickly, mobilized a substantial army and sent it off to Kyushu. In the event, Hirotsugu's rebellion collapsed without a battle because many of his troops refused to fight. He tried to escape to China or Korea but was caught and executed. This event, and also the generally bad condition of the country, seems to have affected Shomu Tenno personally. He left the capital and wandered about for a bit and eventually settled at a palace in Yamashiro province, just south of the modern city of Kyoto.

741 (13th year of Tempyo): Shomu Tenno ordered that all officials of 5th rank and higher had to leave Nara and move to his new palace at Kuni in Yamashiro Province. In this year he also announced the "kokubunji" project. The plan was to construct a small Buddhist monastery and a separate nunnery in every province in the country so as to facilitate the spread of Buddhism among the common people. This turned into a very expensive undertaking. For a long time there were debates about whether this was ever actually done, those who didn't believe in it arguing that the government could not have afforded it. However, modern archeological work has conclusively demonstrated that large numbers of these facilities were built over the next several years following a standard plan. Some of them became the starting points for later temple projects, and in that sense still exist. A significant part of each province's tax revenues was designated for the construction and later support of these temples. In the same year Shomu further demonstrated his commitment to Buddhist principles by forbidding the slaughter of cattle or horses.

742 (14th year of Tempyo): Shomu moved his palace again, to Shigaraki in Omi province.

743 (15th year of Tempyo): Shomu ordered that Shigaraki become the new capital and abandoned Kuni.

744 (16th year of Tempyo): Shomu questioned officials and commoners both as to whether it would be better to have the capital at Kuni or Naniwa. He soon moved to Naniwa, and later in the year designated it the capital.

745 (17th year of Tempyo): Shomu once again undertook an opinion poll, and this time the officials almost unanimously voted to return to Nara. Shomu himself went to Kuni, but the commoners living at Kuni all evacuated and moved back to Nara and Shomu eventually followed and once again designated Nara the capital. I think that at this point it is fair to say that Shomu Tenno was rather unstable mentally. Even after he abandoned the constant moving about, he was subject to sudden enthusiasms and abrupt decisions. He was known, for example, for banishing people without warning and then equally suddenly pardoning them. In the general period between 741 and 745 there was a large amount of experimentation with the system of provinces. It is not known how much Shomu was involved with this. The return to Nara coincided with the start of Shomu's biggest project, the construction of Todaiji and its colossal Buddha statue. In this year prince Suzuka died and was not replaced.

747 (19th year of Tempyo): The construction of the great Buddha began.

748 (20th year of Tempyo): The retired Tenno Gensho died and was cremated in a Buddhist rite.

749: This was a busy year. In the 4th month the reign title was changed to Tempyo Kampo. Then in the 7th month Shomu abdicated in favor of his daughter, known as Koken Tenno. He had no son. This was a very remarkable step to take. Every previous female ruler had served as the protector of a male child or other near relative who was intended to later take the throne. However, once Koken took the throne her status became so high that she could never marry, so she could never have an heir. The state of the imperial clan was such that one could in 749 already foreshadow that the accession of Koken made it possible that the line of rulers descended from Temmu Tenno would come to an end. At the time that Koken assumed the throne the reign title was changed again, to Tempyo Shoho. A special office was set up for the support of Empress Komyo (Fujiwara no Miyako) and Fujiwara no Nakamaro was appointed to run it. The abdication of Shomu and accession of Koken also coincided with completion of major structural work on the great Buddha statue at Todaiji. Shomu took up residence at the temple, by the way, and lived the remainder of his life as a monk.

752 (4th year of Tempyo Shoho): Koken Tenno banned the slaughter of all animals and the "eye opening" ceremony was conducted for the great Buddha statue, the occasion for an enormous public celebration. Only a small bit of the original statue survives today. The present statue (and the hall it is in) is a medieval reproduction.

756 (8th year of Tempyo Shoho): Tachibana Moroe retired from public life and moved to Naniwa (he died early in the next year). Shomu Tenno died in the 5th month, and the Shosoin was established in his memory. Prince Funado, a son of prince Nittabe, was appointed crown prince.

757 (1st year of Tempyo Hoji): Early in the year prince Funado was dismissed as crown prince and was replaced by prince Ota, a son of prince Toneri. In this year Fujiwara Nakamaro discovered the Yoro code manuscript in his grandfather's library and publicized it. In the 7th month Tachibana Naramaro (Moroe's son) was accused of plotting to kill Fujiwara no Nakamaro and he and several other members of the clan were killed. Also, Fujiwara Toyonari, the Udaijin, was accused of involvement and was politely banished through appointment to a relatively minor post in Kyushu.

758 (2nd year of Tempyo Hoji): At the start of the year Koken Tenno abdicated in favor of prince Ota, who is known as Junnin Tenno. Immediately afterward, Fujiwara Nakamaro was permitted to establish a new clan. His name became Emi Oshikatsu. New names were assigned for most higher offices, and Emi Oshikatsu was appointed to the new office of Daiho, the highest office not vacant.

760 (4th year of Tempyo Hoji): Emi Oshikatsu was promoted to Daishi, which was the new name for Dajodaijin. Except for the posthumous promotion of Fujiwara no Kamatari, this was the first time anyone not a prince held this office. It is clear that by now Emi Oshikatsu was virtual dictator.

761 (5th year of Tempyo Hoji): Emi Oshikatsu conceived the idea of invading Korea and he ordered very extensive preparations, including training a number of young people in speaking the Korean language. His plans called for the construction of 393 warships and drafting 47,000 men.

762 (6th year of Tempyo Hoji): At the start of the year Emi Oshikatsu was promoted to senior first rank, a level that was nearly always left vacant. In the 6th month the retired Tenno Koken announced that she was superseding Junnin Tenno and would take over management of the government. A military review was held of the expeditionary force for Korea.

763 (7th year of Tempyo Hoji): Koken Tenno appointed a little known monk named Dokyo to the office that supervised Buddhist institutions.

764 (8th year of Tempyo Hoji): In the 9th month Emi Oshikatsu sent out officials to the provinces near the capital in order to inspect military preparedness. His enemies accused him of using this as a cover to start a rebellion and attacked him. He attempted to flee but was caught and killed. Dokyo was appointed to a newly created high office, and all of the official titles changed by Nakamaro were restored. In the end it appears that the spectacular career of Fujiwara Nakamaro depended on support first from Koken and later Junnin. It is believed that he worked to persuade Koken to abdicate and that he completely dominated Junnin. However, Koken had the clout to reel him in. When the crisis came nearly all of the nobles contributed men to fight Nakamaro and he was caught badly outnumbered. Dokyo was a monk of relatively humble origins who visited Koken Tenno when she was ill and she believed that he cured her and made him an intimate. The exact nature of the relationship between Dokyo and the empress is a matter of great controversy that will be discussed elsewhere. In the 10th month Junnin Tenno was deposed and exiled to Awaji and Koken resumed the throne. She was given a second reign title for this and is known as Shotoku Tenno.

765 (1st year of Tempyo Jingo): Many more of Nakamaro's changes were abolished. Officials and princes were forbidden from keeping weapons. Prince Waki was accused of plotting rebellion and killed. In the 10th month Junnin Tenno died. It is not known whether this was due to natural causes. He was 33. In the inter-calary 10th month Dokyo was appointed Dajodaijin. It would seem that Dokyo was almost exclusively interested in religious matters and, perhaps, nepotism for his extended family. Shotoku Tenno was pretty definitely in charge, as is proven by the eventual fate of Dokyo.

766 (2nd year of Tempyo Jingo): Dokyo's title was changed to one more suited to a Buddhist monk, Hoo. "Ho" is a reference to "the law" in the Buddhist sense, what we would be more inclined to call "the truth". "O" is the Chinese term used for "king" and is in this period the routine word for "prince". The Buddhists also use the word for supernatural protector figures based on Indian and Chinese dieties that fight evil on behalf of Buddhism, for example, the Shitenno "four heavenly kings" figures seen in many Buddhist temples trampling fiercely on demons.

770 (1st year of Hoki): Shotoku Tenno died. A conference of high officials considered various candidates for the throne and selected prince Shirakabe as her successor. Dokyo was not invited to this meeting. Dokyo was not important enough, by himself, to kill and he was merely dispatched to serve as abbot of a temple in a remote province in the east. Prince Shirakabe was enthroned and is known as Konin Tenno. What with one thing and another the line descended from Temmu had died out, and Konin was instead a grandson of Tenchi Tenno. All later emperors were descended from Tenchi through Konin. He was already an elderly man of 62 and had several sons.

771 (2nd year of Hoki): Prince Osabe was made crown prince. Very soon after this prince and his mother were arrested on the accusation of trying to recruit officials to support a coup d'etat. They were never tried, but were imprisoned.

773 (4th year of Hoki): Prince Osabe and his mother died on the same day and immediately afterward Prince Yamabe was made crown prince.

781 (1st year of Teno): Prince Yamabe took the throne before the death of his father later in the year, eliminating all chance of a succession crisis. He is known as Kammu Tenno, famous as one of the strongest emperors. There was never any question who was in charge, and it is thought that he had considerable power while his father was on the throne as well. He immediately named one of his sons, prince Sawara as crown prince.

782 (1st year of Enryaku): There was a minor rebellion resulting in banishments. In the 6th month Kammu dismissed the Sadaijin Fujiwara Uona from office.

784 (3rd year of Enryaku): Kammu began a project to move the capital to a new location. He selected a location in Yamashiro province known as Nagaoka. He himself moved to Nagaoka before the end of the year.

785 (4th year of Enryaku): The official in charge of constructing the Nagaoka capital was Fujiwara Tanetsugu. He was mysteriously murdered, perhaps by bandits, perhaps by enemies. It is not at all clear what really happened, but the crown prince, prince Sawara, was accused of being behind it and was exiled to Awaji, but was killed while en route. There is reason to believe that this episode had a powerful impact on Kammu, who showed many signs of feeling guilty about it. Prince Ate (who survived to ascend the throne) was the new crown prince.

787 (6th year of Enryaku): The government was ordered to transfer to Nagaoka.

793 (12th year of Enryaku): Kammu began to investigate the possibility of moving the capital once more, to a different location in Yamashiro province. Things moved fast this time. Kammu moved to the new location almost immediately.

794 (13th year of Enryaku): The government was ordered to move to the new capital. At the end of the year it was named Heiankyo. It may well be that it was only now that people began to refer to Nara as Heijokyo. It was a characteristic of Kammu's court that the culture was strongly Chinese. Chinese was the routine written language of government throughout the Nara period, but there is plenty of evidence that all of the Chinese office names and official titles had Japanese translations that were routinely used in speech. However, in Kammu's time it was definitely the fashion to pronounce all Chinese words as (the Japanese version of) Chinese. This custom has generally continued down to today. Heiankyo never did have an actual Japanese name. Kyoto is also Chinese. Most likely people usually simply called it "miyako" or "the capital," the Japanese word for the "kyo" character in Heijokyo and Heiankyo. For most historians the move to Heiankyo marks the end of the Nara period. However, Kammu's reign is very important in administrative and also military history (this was the time of the most important fighting against Emishi in the northeast), so I prefer to treat it all in one place. It is widely believed that this second move of the capital was because Kammu grew to hate and fear Nagaoka because of its association with his probably over-hasty execution of his son in 785. There is ample evidence over the next two centuries or so of a widespread belief in Japan in the power of the ghosts of people who were killed unjustly to come back and get revenge. Kammu is known to have put in a lot of effort in religious ceremonies intended to appease the ghost of prince Sawara.

806 (1st year of Daido): Kammu died at the age of 70 and prince Ate succeeded without fuss. He is known as Heizei Tenno. This is the end of my Nara period.

Nara period government
The new administrative system was put into effect in 689 by the publication of what is known as the “Kiyomihara Code.” Kiyomihara was the name of the imperial palace at that time. Nothing survives of this code. It is known that the names of many offices were substantially different than in later times, and it is hypothesized that the system of local administration was much simpler than in later times. However, the total amount of information about ways in which it might have differed from the later codes is small. It was replaced in 701 by the publication of the Taiho Code. This code does not survive either, but it has been largely reconstructed from citations in later books. In 718 the government began work on a third version of the code, but this project was put aside when Fujiwara Fuhito died in 720. A copy of the code remained in his library and was discovered 39 years later by his grandson. At the time of this discovery it seems that no official copy of the Taiho Code survived. The 718 version is known as the Yoro Code after the current year title. During the reign of Saga Tenno in the early 9th century copies were made and substantial portions of some of these survive. Efforts to reconstruct the Taiho Code and the missing parts of the Yoro code from later commentaries have so far failed to identify any differences between the two. It seems fair to hypothesize that the Yoro Code was actually a complete copy of the Taiho Code and that whatever changes were planned in 718 were not incorporated into this manuscript but existed in some other form. Only a small part of the judicial codes have been reconstructed, but the greater part of the administrative code is known. I frequently see statements, even in works in Japanese, that the Yoro Code was enacted and put into effect in 718, but this is wrong. Its existence was entirely unknown until its discovery by Fujiwara no Nakamaro. What happened was that later commentators, starting with the author of Ryo no Gige in 833, used and cited the Yoro Code because no copy of the Taiho Code survived.

It is important to realize that there were no further efforts to modify the code after 718. This, combined with the fact that the official copies of the Taiho Code had been allowed to deteriorate and disappear show that the code itself was no longer considered particularly important after the first two generations or so of the new era. What mattered was the existing structure of government on the ground and the collection of recent government documents including, especially, edicts and decisions. Starting with Ryo no Gige officials put together series of compilations of these materials to serve as guides for judges and administrators. New series were constantly needed because the actual practice of the government was constantly evolving during the period of the 8th, 9th, and 10th centuries. No one ever seems to have felt that it was necessary to stop and construct a complete new administrative code that would show exactly how things were done at a particular point in time. It would appear that practices were simple enough to keep straight in a more informal way.

If the central government was fully staffed as called for by the Taiho/Yoro Code there would be 8117 officials from 3rd rank through 9th rank. In addition, the code called for the assignment of 2 persons per village as laborers at the capital (there were 4000 villages). The first and second ranks were almost never awarded on the grounds that only rarely would a person appear who was worthy. At the start of the Nara period there were probably were about 8000 officials receiving salaries, but as time went on the number steadily dropped, and by the middle of the 9th century many offices in the table of organization had effectively ceased to exist. It seems that the original code was over ambitious and too closely modeled on the administrative code of the T'ang dynasty. China was, of course, much bigger than Japan and was also much more complex economically and socially. Japan's actual need for government was satisfied with less than full implementation of the code. There is also the factor that from the very start the government failed to reach its expectations as to the amount of tax revenues that would be received. This matter was a constant topic during the Nara period. The Japanese could not afford to pay 8117 officials and had to settle for less. It is also true that Japan's geographical isolation meant that the country could get along with extremely small military forces. The military represented a very large portion of the total expenditures of China and Korea, but this was not true in Japan. There were always active forces on the northern frontier, but these were essentially local militia that fed themselves. The government paid them by giving them tax breaks. In the early days two or three thousand conscripts from eastern Japan were sent to serve as coast guards in Kyushu, but this was abandoned rather early in the period. The only regular armed forces were the guards units maintained in the capital, and these probably numbered only one or two thousand men. When troops were needed, as in the case of Hirotsugu's rebellion in 740, they were conscripted from the rural population. Seventeen thousand were raised and marched to Kyushu on that occasion in quite a short period of time, as the entire rebellion lasted only four months.

The three most important parts of the government were the staff that actually operated the imperial palace and the government offices, the organization that supervised the provinces and collected and managed the tax revenues, and the provincial and district administrations, which were responsible for maintaining order and generating the tax revenues. Most of the revenue that came in was used for official and rank salaries and the imperial palace. There was also from time to time considerable spending on building and equipping temples, shrines, palaces, and entire new capital cities. During the Nara period there seemed to be plenty of money to spend despite all complaints about revenues. When the great epidemic of 737 hit agricultural work was heavily impacted and harvests failed around the country. The government had sufficient reserves of rice in its storehouses to permit the cancellation of taxes and provide relief for the people. There were complaints by 740 that the reserves had been mostly used up, and financial reforms were undertaken to attempt to get them built up again.

The taxation system included three separate components. There was a relatively small crop tax paid in rice which was stored in government warehouses, mostly at the provincial capitals. Most of this was used to feed laborers and soldiers, but some was sent to the capital and paid out to officials as part of their salaries, and a significant proportion was earmarked for the reserves already mentioned. Provincial and district officials also received salaries, of course. There was a second tax calculated on a per capita basis that was paid in handicraft items. The largest amount was paid in cloth, but the governors had the power to substitute other local products as needed. Among the items known to have come in were metal tools of all kinds, raw materials for lacquer and other luxury items, and ceramics. Many items were consumed by the palace, but cloth and metal tools were also important parts of the official salaries. It was the general custom that these items, especially cloth, for which there was a more or less official exchange rate, could be traded at markets for food or other products. The third tax was a labor tax, which had many complexities.

The labor tax was used for public works projects, to raise military and police forces, and to provide menial staff for palaces, government offices, and officials' private residences. Most men who were drafted worked locally and lived at home. If they were drafted in winter months they were liable for up to 50 days per year, or if drafted during times of the year when farming was active for 30 days per year. Often they worked less than the full amount because they were not needed. If people were dispatched from a rural area to the capital or some other distant location, they were often kept for a period of years. During that time the work they did was credited to their household and their village so that others were exempted from labor tax as a result, and their family would get tax adjustments to pay for their excess labor. This was because it was expensive and difficult to move numbers of men around the country. However, it was very unpopular, and eventually the government was forced to keep men for no more than one year before allowing them to return home. Shosoin documents make it clear that at least the men working in Nara were fed by the government and paid a quantity of cloth every 10 days which they could trade in the market for other necessities. They worked from dawn to dusk with no days off, but got a 2 hour break in the middle of the day during the heat of summer. If it rained there was no work, but rations were cut in half. There are constant complaints that men escaped from labor details and snuck home. The districts were responsible for this and had to send replacements. There were experiments with allowing people to buy their way out of labor service, but these were quickly abandoned on the grounds that they were too easily abused. There was no regular way possible in ancient times to hire casual labor. Everyone was bound to service somewhere. The appearance of mobile workers paid wages is one of the main markers for the transition from "ancient" to "medieval" conditions.

This makes a good point to say something about Marxism as it applies to Japanese history. Long before Karl Marx became a man of political renown he was an historian, and in Japan in the period after the Second World War it is fair to say that the majority of historians have been Marxist. This does not mean that they have necessarily been admirers of Joseph Stalin, Mao Tse-tung, or Kim Il-Sung, but that they accept the basic position of Karl Marx on the pattern of evolution of human society, and tend to use terms and phrases derived from Marx in their writings.

In the Marxist view human society has since the development of civilization gone through three successive phases which Marx called the slave society, the feudal society, and the capitalist society. He went on to predict that there would be a fourth phase, communist society, but the fact that he was wrong about that does not mean that he was necessarily wrong about the others. As the man said, prediction is difficult, especially about the future. For political reasons almost no American and relatively few European historians would admit to being Marxists, but most who study the evolution of political and economic systems would agree (quietly) that Marx was an important historical thinker and that not everything he said was wrong.

The "slave society" phase refers to the time when economic structure focused upon the control of labor as the primary form of wealth. By modern standards the ancient world was very thinly populated. There was farmable land everywhere, but only so many men available to farm it. Ancient Greece and Rome dealt with this problem by relying on formal slavery with markets where slaves could be bought and sold and this is why Marx used that term. However, most of the world did things differently, and Japan, Korea, and China relied on labor taxes. There were slaves, but they were a small part of the population and they were not bought and sold. A constant problem facing the Nara government was that farmland went out of production because there was no one to farm it. For several provinces we have records showing that 30 or 40 percent of the developed land had been abandoned. Once land was abandoned, the associated irrigation works started to collapse, so it was not always easy to get it back into use.

Looking ahead, the "feudal phase" is a period in which population growth had reached a point where essentially all of the good land was farmed and there was a surplus population that must be fed out of surplus food production, which required the development of a market system so that food could be moved to where it was needed and sold. It also led to an interest in applying technology to increase the productivity of farm land. Slaves disappeared and were replaced by hired laborers, who did not have to be fed when they were not needed, and who moved about the country as different crops required temporary extra manpower, or moved into the cities to work in emerging industries. In the feudal stage there are more than enough people and the focus for the elite shifts to the control of land as the primary source of wealth. Marx called this feudal because Western Europe had a feudal system of government during this stage, but that is not a requirement. Marxists are quite happy to call T'ang dynasty China a feudal society because it was a land owning society, despite the fact that there was nothing "feudal" about its governing structure. Still further off, the transition from feudal society to capitalist society means that the focus shifted again from the control of land to the control of money, because a steadily increasing portion of the total national wealth came from activities other than farming, and the "surplus" population was beginning to become larger and wealthier than the farming population.

In the case of Japan, the Asuka, Nara, and the first century of the Heian period represent the "slave society" phase. The latter part of the Heian period and the Kamakura period are a very slow transition towards a landowning "feudal" period, though even in the early Kamakura period only about half of the farmland in the country was owned by nobles or temples. The Muromachi period covers a rather rapid transition in which on the one hand you see the development of full-fledged political feudalism, very similar to the European variety, accompanied by the development of consolidated landownership in which every piece of farmland had one and only one specific owner, and simultaneously the beginnings of capitalism, in the form of banking and organized industry and real cities. Essentially, changes that in Europe took 500 years to develop, from 1100 to 1600, took only 300 hectic years in Japan, from 1300 to 1600. There was an awful lot going on during that time. Then things slowed down again. Technically, in Marxist terms, western Europe and Japan were both infant capitalist societies by 1600, but Japan in its northeast Asian isolation changed much more slowly than Europe after 1600 and had fallen far behind the pace of western development by the time that its isolation was forcefully ended in 1853. This put modern Japan in a very stressful situation which it could have handled better than it did. However, it has survived to become one of the main capitalist societies. Despite the assumptions of the westerners who first forced Japan to end its isolation, it was never a "third world country" but already a modern capitalist society, and by 1895 or so most who paid attention were aware of this.

However, our present concern is with the efforts of the Nara period government to monopolize as much of the national wealth as possible. Their main problem initially was to establish a balance of power and interest with the rural nobility. Before the Asuka period the court and its associated nobles lived entirely off of Yamato and nearby provinces, especially Kawachi province to the west. The surplus wealth of the remainder of the country went almost entirely to the local nobility. The only real exception seems to have been the "miyake" institution which was used to finance military operations in Korea. If the rulers were to move armed men from Yamato to Korea they had to be fed along the way and the miyake were granaries that were dedicated to that purpose. The new system called for wealth to be shifted from every province, however remote, to the capital or (in the case of Kyushu) to a location under the control of the capital (Dazaifu). In the original Taiho code every province was expected to transport 10% of the rice tax to the capital, but the extreme expense of doing this forced an almost immediate change. As before, the capital got all of its rice locally and rice collected in remote areas was spent in remote areas. The portion of the rice reserved for the central government was kept in storage and was available for use during times of famine, when taxes were normally forgiven. Some of the taxes in kind and often quite a large proportion of the labor taxes were consumed in the capital area. Major construction projects brought very large numbers of workers to the capital, and in those cases the provinces not only had to transport the workers but they had to feed them as well. The Taiho Code seems to have assumed that a party of workers being marched to the capital would be provided with all the food it needed which would be carried along. However this did not work and it was necessary to change the rules so that each province was required to feed parties coming through or returning home. The men carried documents showing that they were eligible (and travelling legally), a few of which have survived. Provinces close to the capital were required to provide extra food to the construction sites because Yamato alone would not have enough to handle the demand.

The challenge in all of this was to force or persuade the rural nobility to obey instructions from the center. Why would they do this? If the center gets more, they must get less. The answer seems to have been that they used the political structure to create a steeper hierarchy in rural areas than would have otherwise existed. That is, the district magistrate, known as dairyo, was paid generously and given considerable local power. Also, it is clear from many sources, that members of his family could go to the capital and find minor positions in and around the government. There are several cases in the record of men who petitioned the government for appointment as dairyo in their native district who had spent the previous 10 or 20 years working at low levels in the central government. Dairyo were appointed by the central government from a list of men recommended by the provincial governor. During at least one period they were required to go to the capital for a personal interview before they could be appointed. The post was clearly valuable and eagerly sought. If the provincial nobility as a whole had to take less, the dairyo and his family were not going to suffer because of it.

The other thing that the government did in this connection was keep the provincial and district administrations very small. Depending on its size, a district could have as few as two or as many as eight salaried officials. Everything else was handled by locals serving out their labor tax.

The taxation system depended completely on the national census that was supposed to be taken every six years. From the very beginning this schedule was almost never kept and the government at least investigated the possibility of skipping the census and simply assigning tax quotas to provinces, districts, and villages, but it continued regularly throughout the Nara period and beyond. The census was based on a report submitted by the head of each household, which was pulled together and validated (he was supposed to verify its accuracy) by the village head who passed it on to the district, where it was consolidated again and passed to the provincial governor, who sent it to the Dajokan. Officials in the capital were supposed to compare it carefully with their copy of the previous census and if there were any apparent discrepancies they would report this to the provincial governor, who had to investigate and provide corrections. Many of the surviving population registers show erasures and corrections. All in all the process of completing the census took the better part of a year.

Then the district magistrates were responsible for completing a survey of all farmland in the district listing the exact acreages and including maps. Several of these survive. They listed the status of all of the land, whether it was public land subject to redistribution, exempt from redistribution for one of many possible reasons, or abandoned. These were used in the capital to calculate the expected rice tax for each province and the provincial governor was informed as to his total quota. The district magistrate also prepared a tax register which was essentially the same as the census register, except it was intended to report all cases where a person listed in the census was not going to pay tax (perhaps because he was absent from the area and thus could not pay labor tax, or because he had died since the census was taken). These registers were used to calculate the head tax expected and the amount of labor available for corvee. Again, there was a cycle of checking the registers and demanding corrections if there seemed to be any discrepancies. The tax registers had to be done every year because they were the basis for the calculation of the national budget.

As a result of all of this work, the actual redistribution of land was performed three years after the census was begun. This delay was built into the system. Farmland was allocated for persons who were children at the time the census was fixed but who would be old enough to help in the fields by the time the land was allocated. Every official household was allowed to own permanently the land on which its houses and other buildings were located and a surrounding kitchen garden for vegetables, which was not taxed. Any land that was used only for farming crops other than rice was not included in the redistribution system. However, there does not appear to have been nearly as much reliance on wheat or millet as supplementary crops as in later times. Each household was supposed to be assigned land according to very precise rules with respect to the legal status, sex, and age of every member. If there was not enough available (actively farmed, with working irrigation) land, the allotments were reduced proportionately. The land was not allotted as coherent farms, but as parcels scattered around the area. Otherwise a redistribution system would probably have been impossible. One would assume that unless a household drastically grew or shrank it would receive mostly the same collection of parcels each time. Because of the fragmentation of parcels it was impossible for each farmer to be individually responsible for irrigation. That was a public responsibility and the work was arranged through the labor tax. The government constantly pressured the provincial governors and district magistrates to do what they could to expand the farmable land and construct new irrigation facilities, in hopes of increasing the total revenues. However, as noted, due to the limitation on the amount of labor available, opening new land in one place nearly always meant the abandonment of land elsewhere. This is a main reason why the land registers always show a significant fraction of "abandoned" land.

There were definitely cases where it was difficult to find enough land to assign. We know of farmers who were assigned parcels in a neighboring district and even in another province. There was a provision to cope with this because a farmer who had land he could not conveniently reach was permitted to apply for permission to rent it out.

Farmers cannot farm without seed rice. A farmer could keep back a portion of his crop for this purpose, but there was from the beginning an assumption that most, if not all, farmers would borrow seed rice at the time of planting. This meant that they did not themselves have to build facilities to store and protect it. The original system provided for official loans at fixed rates run by the provincial governor, but also permitted private loans. These would, naturally, be made by local nobles. There was definitely no class of businessmen or "rich peasants" who could make such loans as in later times. It seems that in practice private loans were usually made by the governor and his staff or the district magistrate. They were apparently quite profitable, and relatively soon the government abolished private loans. The anticipated revenue from public loans was built into the budget, and very soon the governors were given quotas, and eventually the loans were made compulsory to the farmers. It is possible that the revenue from the loans was larger than the revenue from the crop tax and was mainly responsible for the reserves that were kept to deal with famine conditions.

Back in the capital, one of the ministries, the Monbusho, was responsible for gathering all of the tax registers every year and putting them together so that it could present the Dajokan with a figure for the total revenues available. All government agencies had to draw up a proposed budget showing their needs, and these also went to the Dajokan. The Dajokan drew up a budget and the Monbusho would then administer it, apportioning out the funds during the course of the year. The budget was probably not an enormous amount of work most of the time, as nearly all of the regular revenue went to salaries, which must have been pretty steady year to year. It was definitely the assumption that in a good year reserves would be built up for future contingencies, whether emergencies or large scale projects. In difficult times it was common for taxes to be reduced or even cancelled, so in most years there must have been a significant surplus.

In 728 an edict was issued dividing the body of officials into "inner" and "outer" categories. Outer officials were the rural nobles who held posts in the districts and in the provincial administration. The highest rank permitted was senior fifth rank upper grade, which meant that an outer official could never be appointed to an important office in the capital. Before 728 there were a number of cases where an official had made the transition from district magistrate to an important central position. Moreover, even many lower level jobs were marked for "inner" officials only. The intent was obviously to try to guarantee jobs for the offspring of "inner" officials, but it had the unintended effect of forcing the provincial nobles to find other career paths outside officialdom, and in the long run this worked against the survival of the system.

Even among the inner officials there were sharp status distinctions. Ninth rank positions were essentially menial and for most purposes their holders were legally treated as identical to ordinary commoners. Like commoners, they could be sentenced to flogging for minor crimes, and they were subject to some taxes. Men from noble families were never appointed to any rank lower than 8th. There was a clear distinction between families where it was assumed that the entire career would take place between the 8th and the 6th ranks with the possibility of reaching the 5th rank at the very end of an especially distinguished career, and those where if a young man were forced to take a rank as low as eighth he would be rapidly promoted to at least 6th, where his career might actually begin. Noble youths who were appointed to low ranks usually did not actually take up low ranking offices, but waited until they were 25 or so and high enough in rank to get something worthwhile.

This is clearly shown in the compensation system. There were five separate components of the system, a basic salary paid annually, a seasonal salary paid twice a year, the allocation of servants drawn from the corvee tax pool, the allocation of the tax revenue from specific households, and the allocation of rank land. Not everyone received all types. Sixth through nineth ranks received only the seasonal payments, sixth rank receiving three times as much as nineth rank. These were payments of rice, two kinds of cloth, and iron farm tools. All ranks received this income. Third rank, the highest normally received, was paid 14 times as much as nineth rank. Only fourth and fifth rank officials received the additional salary that included a significant amount of silk cloth. The assignment of servants from the corvee labor pool started only at the fifth rank. Junior fifth rank got 20, senior third rank got 100. Only fifth rank and higher received land. Junior fifth rank received 8 cho, about 20 acres, and senior third rank received 40 cho. Only fourth rank and higher received the taxes of assigned households. Junior fourth rank received 80 households and senior third rank received 250. (This part of the income was as of 706, before then it was significantly lower and it only started at third rank.)

5th and 6th rank was the zone of middle management. From the point of view of the higher nobility, this was the zone where a young man would start his career. All provincial governors came from this pool. There was again a clear distinction between men who would ordinarily spend most of their working lives in these ranks with the chance of appointment to a higher one as a reward for long and effective service, and those who passed through rather quickly on their way to better things. The lower group comprised men who spent their careers in the ministries, or who were career provincial governors, transferred from province to province every few years. The upper group would quickly rise to the governor level, work in one or two provinces, and then be promoted to a 4th rank position in the capital.

The real aristocracy may be defined as that group where a successful career would require promotion to the 4th rank at a fairly young age, with the prospect of promotion to 3rd rank and a seat within the Dajokan as soon as one opened up. There was a restricted number of such seats and not all of them were always filled. The general rule throughout the Nara and Heian periods was that each politically important clan would have just one man in the Dajokan at a time, though the Fujiwara were frequently able to violate this rule and have several. As I believe I have mentioned, the second rank was very rarely awarded except to princes, and the first rank was only ever taken, two or three times, by egotistical dictators. The office of Dajodaijin, the highest in the formal bureaucracy, was a second rank office. Only 5 or 6 men ever held this post after the middle of the Nara period through 1600 and all of them had extraordinary political power and used this office to demonstrate it. Even in the early Nara period this office was frequently left vacant, on the theory that only rarely would there be a person worthy to hold such a lofty position.

The Dajokan was a committee of officials, including a senior group who had no routine administrative duties plus very senior ministers of departments, and may be considered the equivalent of a modern cabinet. They were the primary channel of communication between the government and the population at large, in that all petitions, memorials, reports, and the like were received by the Dajokan, which in turn published all routine orders and edicts. The Dajokan had a staff that handled this flow of documents. The Dajokan met as a body only when necessary and deliberated in secret. The emperor was not a member and the Dajokan and the emperor communicated by sending documents back and forth. Its primary tasks were to advise the emperor and to see to the execution of his commands.

Next below the Dajokan were two large bureaus of secretaries, designated as "left" and "right". The heads of these two were ex officio members of the Dajokan. Each of these offices supported four of the eight ministries. The individual ministers were also Dajokan level officials. The maximum size of the Dajokan if all offices were filled would (normally) consist of the 8 ministers, 2 secretaries, 3 chunagon "counsellors", 2 dainagon "senior counsellors", the "great minister of the right," the "great minister of the left," and the Dajodaijin. At times there was also a "great minister of the center." It was normal for some of these posts to be vacant. The position of commander of the Dazaifu in Kyushu was considered a Dajokan level post, and at some point it became routine for it held by an official who stayed in the capital and held some other post concurrently. The vice commander of the Dazaifu was the man in charge on the ground. That office had approximately the same prestige as the governor of a wealthy province. Dazaifu was also used as a high-level prison camp for officials in disgrace, who would be assigned a post there with no actual duties.

The Attempt to Start a Monetary System
Money was invented, according to the ancient Greeks, in the kingdom of Lydia sometime before 700 BC. It appeared in China 300 or 400 years later than that. Early money derived from the use of standard weights of precious metals as a medium of exchange. Coinage eliminated the necessity to weigh everything for routine transactions. The idea of developing a system of coinage implied a moderately well developed system of markets. There is little point in having money if there is no place to spend it. Throughout ancient history in western Asia, the main use of large denomination coins of gold or silver was to pay armies; a large amount of wealth was highly portable if it were in the form of precious metals. However, low denominations, usually of bronze, were widely used by ordinary people in cities, where most people pursued specialized occupations and could not produce everything that they needed, but must be able to buy many things on a regular basis.

There is a lot of evidence that during the Nara and Heian periods such domestic commerce as occurred was mainly financed through the exchange of standard bolts of cloth as produced through the poll-tax system. There was limited trade with China and Korea and Chinese coins certainly entered the country. Nihon Shoki records edicts of Temmu Tenno and Jito Tenno forbidding the use of silver coins and permitting only bronze coins. The Nara government ordered the establishment of mints and the production of both bronze and silver coins in 708. The bronze coin was of the type commonly referred to as ‘cash’ when produced in China, and was modeled after a T'ang dynasty issue of 621. It was a round coin with a square whole in the center, permitting quantities to be conveniently bound into bundles using a string. Many examples still survive. It was 2.4 centimeters wide and weighed 3.75 grams. The silver coin was twice the weight. Production of the silver coin was abandoned in 709, but the bronze coin was made regularly until 760, with a final run in 765 commemorating the thousandth anniversary of (the mythical) empress Jingu Kogo. The coins have been found all over Japan, and also in Manchuria, in the 8th century capital of Po-Hai (which had direct diplomatic relations with Japan across the Japan Sea). The coins were made in a number of locations ranging from Omi province (not far north east of Nara) to Dazaifu in Kyushu.

There is no information about what measures the government may have taken in 708 to place coins in circulation. The earliest recorded edict on the subject is dated 711. It established an official exchange rate between the bronze coin and rice. Later in that year a portion of official salaries that had been paid in rice was switched to payment in coins. At this time the people were instructed as to the purpose of coinage: “One who has coins can exchange them for things and thus obtain what he requires. Up until now the population has followed old custom and knows nothing about this method. If there is occasionally someone who wishes to use money to buy something, there is no one who is prepared to take it. Those who present money they have taken in exchange for goods to the court will be promoted in rank at the next promotion cycle.” Indeed, some persons took advantage of this and used coins to obtain rank. Then, in 712 the government began giving corvee laborers coins instead of feeding them, and it published an official exchange rate between the bolts of tax cloth and the coins, and permitted provincial authorities to substitute coins for in-kind items of the poll-tax. The first actual payments did not occur until 722, however, and all known payments were from the 8 provinces closest to the capital. This 8 province area is exactly the zone where coins most commonly occur in archeological excavations. There are surviving tax registers from Yamashiro province (immediately north of Nara) for 726 and 733 which show that there were tax receipts in coins amounting to an average of 9 coins per person in the registers. There are also 733 registers for the city of Nara itself showing that all receipts were of coins. There are entries in these registers that have proven difficult to understand that suggest that people were permitted to buy their way out of the corvee labor tax obligation. There is another theory that these amounts were fines assessed against families that evaded the labor tax by having people hide when the assessments were made. This is suggested by the fact that the amounts were very high given that the official exchange rate implied that one coin was worth one man-day of labor. In any case an edict of 737 forbade the substitution of coins for either the poll-tax handicraft items or the labor tax, and there is no later mention of such a practice. There is a complaint of 734 that people were registering to pay coins in return for exemption from tax but then failing to pay up.

There is no evidence that the use of coins ever spread beyond its issuance as part of official salaries and receipt as a component of taxation. There was an edict ordering the provinces to accept coins from corvee laborers who were returning home as payment for food. This may have been the main means whereby the provinces would obtain coins to then send back as a component of taxation. At some point the government began to deliberately take in more coins than it paid out so as to melt them down and use the bronze for the casting of Buddhist statues. Copper for bronze was always in short supply in ancient Japan. The regular production of coins ceased from 760, though, as noted, there was a special issue in 765 as part of a national celebration. There were a few sporadic issues of coins during the Heian period, but Heian period coins are rare in archeological sites and limited to only 5 provinces. As late as the 11th century Japanese fiction makes it plain that the primary currency in use was still the standard bolt of cloth. Coins are seldom mentioned. However, in the 13th century organized periodic markets began to spread into the countryside, and it was then that Chinese coins came into common use, as being far more convenient to carry for significant distances.

Overall, the history of Nara coinage provides solid evidence of the fact that the Japanese domestic economy was still primitive by Chinese standards, which demonstrates in turn that the administrative system that was adopted was much more complicated than would have been required by the demands of Japanese society alone, and could scarcely have evolved without the impulse to attempt to raise Japan to the level of civilization attained by Korea and China. However, once the system had been created and put into effect, we certainly see in everything that happened down at least to the accession of Shomu Tenno a steady determination to make it work. The attempt to establish a monetary system is but one aspect of this ambition.

Policies Respecting Private Landownership
It was a fundamental principle of the system announced in the 645 reform edict that all land was to be the possession of the ruler. The administrative system of the Taiho code called for the periodic redistribution of all irrigated rice land among the tax-paying commoners. However, there was always private property as well. First, the commoner households all held some lands permanently, the lands on which their houses and outbuildings were located and also vegetable gardens located there. Also, any farmland that was not irrigated rice paddy was not included in the redistribution system and must have been owned by somebody.

Then, all high ranking officials were granted lands in three different modes as part of their salaries. There were rank lands, tied to the court rank system, office lands tied to particular jobs, and then households whose tax revenues were assigned directly to an official. There are no clear statements about the details of how any of these were administered, but everyone seems to agree that the rank lands and office lands were possessed and administered by the clans. There is definite evidence that rank lands assigned to Fujiwara Fuhito were retained by the family after his death. If this was normal, then the common practice of giving a high official a posthumous promotion resulted in an increased pension to his survivors. This information comes from the fact that after Hirotsugu's rebellion the Fujiwara donated 5000 cho of Fuhito's office lands to the emperor, who returned 2000 cho but kept 3000 for financing kokubunji temples. If this is true of other clans, then rank lands, at least, were not redistributed, and any clan would permanently own land corresponding to the highest rank attained. It is assumed that these lands and the office lands were managed in the manner of a private estate and that they were exempt from normal taxation. The "assigned household" category would have been more of an accounting device. The main difference between that and normal salary would presumably have been that the tax revenues would vary from year to year depending on the number of taxpayers on the books and the quality of the harvest. There are no internal family documents from the Nara period and there was no narrative fiction either, so that we have no information at all about the domestic lives and economies of either aristocrats or commoners.

The realities of actually running the new system soon confronted the government with problems that required them to consider new categories of property. It has already been mentioned that this was an era in which the primary unit of wealth was a worker. The tax system was firmly linked to the census and the amount of revenue received from each district was directly proportional to its working population, not its geographical extent. Land without workers generated no revenue. The government was constantly concerned to ensure that as much land as possible was being farmed and taxed. It soon began to turn to "the private sector" in an effort to accomplish this.

As soon as we start to examine particular cases, it becomes clear that there was one category of wealth that was not fully accounted for under the system, which was land brought under cultivation through the private efforts of members of the provincial nobility. These were people who had rank and office and who therefore had rank and office lands. If they could bring new land under cultivation, they could increase their wealth outside of the official system as they would use labor that was not on the tax registers. They were remote from the capital and only loosely supervised by the very small staff of the provincial governor. The government attempted to deal with this anomaly by offering inducements to these people to develop land that would eventually be added to the taxable lands subject to the redistribution system. In order to do this they had to make sure that the land was recorded so that it could not be kept off of the tax registers.

In 722 there was a major conference of senior officials on the question of raising revenues by adding new farmland. This has occasioned a lot of discussion and it is not clear that we understand what was actually being attempted. The decision appears to state that the government would provide enough food and equipment to draft the entire farming population for 10 days per person with the view to opening up 1 million cho (2.45 million acres) of new farmland. This statement is hard to accept on its face. It has been pointed out that in modern Japan about 3 million cho is being farmed to feed a population 20 times the size of that in the Nara period. According to Wamyo Ruijusho there was a total of about 730,000 cho on the books in the early Heian period, so this project would have much more than doubled the total amount of farmland. It was, of course, entirely impossible to find farmers for such an increase. This leads to the theory that it was being assumed that the successful incorporation of the barbarian areas of the northeast would enable adding that much land with a population sufficient to farm it. If so, they were seriously mistaken about the population levels up there and unaware that much of the land is too far north and too cold for rice farming. However, given their remoteness from the area, there are scholars who think this is a more reasonable interpretation of what we are given by Shoku Nihongi. There is no further mention of such a project, by the way, if it was actually a project.

One year later, in 723, there was an edict announcing that anyone who opened new farmland could retain it in possession of his family and outside of the land redistribution process for 3 generations, and anyone who reclaimed land formerly registered but abandoned could keep it for one generation. There is no further information about what may have happened as the result of this law.

There is one local aristocrat about whom something is known, named Ikue no Omi Azumahito from Asuha district in Echizen province, because he had a close connection with Todaiji and several documents in which he is mentioned are part of the Shosoin collection. It is known that the dairyo of this district In 731 was Ikue no Omi Kanayumi and in 749 it was Ikue no Omi Yasumaro. Azumahito first appears in history in 749 as a clerk in the office for the construction of Todaiji. Sometime after this service, it is not known when, he returned to his district. There, at his own expense, he constructed an irrigation canal some 9.5 kilometers long and opened up about 100 cho of irrigated farmland. This he then gave to Todaiji. He is next heard of in 755 when he was dairyo of his district and worked in conjunction with representatives of Todaiji to put in place the administration for Kuwabara Sho, an estate of the temple within the district. In 768 he was awarded the extremely elevated rank (for a provincial) of (outer) senior fifth rank lower grade. Ordinarily only men who had distinguished themselves in military activities got such rank. He was almost certainly the administrator of the temple estate, which was in turn probably put together around the core of his gift. The scope of his gift makes it plain the he was a wealthy and powerful man. The decision to provide Todaiji with an endowment in the form of estates that were exempt from taxes and the redistribution system is the inauguration of the system of landholdings known as shoen that eventually came to cover about half of the farmland in the country.

The records concerning Kuwabara Sho are interesting. It was purchased in 755 from an aristocrat and was nominally 100 cho in size, but only 9 cho were irrigated. After the purchase Azumahito developed an additional 23 cho. However, by 757 much of this land was abandoned and replaced by 10 cho of new land. In 758 Azumahito reported that the original irrigation system had not been competently designed and that water could only be delivered to a small part of the potentially farmable land. He proposed a project, with a budget, to build a new set of ditches and ponds that would permit opening up a large area. The project would have to be coordinated with the provincial government because the ditches would have to take over 1.8 cho of public land which could not be done without finding land elsewhere to replace it. Such an estate required constant maintenance. A later record reveals that a few years later it had been abandoned. This estate may have been the equivalent of buying Florida swampland, not actually suited to the intended purpose. We cannot say whether the experience was typical.

The Northeast Wars
The people called Emishi have already been discussed in other contexts. To recapitulate briefly, they were inhabitants of the far northeastern part of Honshu, an area where the climate was significantly colder and less hospitable to rice farming. There has always been a line within this area north of which attempts to grow rice were not worth pursuing. The early centers of higher culture in Japan were in northern Kyushu and then in central Honshu and these lands were remote and backward. No one can say for sure whether the Emishi were racially different than the Japanese population further south. My guess is that they were not. I do not recall ever seeing mention of an Emishi language and must assume that they spoke a dialect of Japanese (even in recent times an inhabitant of Kyushu who spoke only the Kyushu dialect would not be able to understand a person from the northeast who only spoke that dialect). Some Emishi were rice farmers, some grew other crops, and some relied mostly on hunting and fishing. The main thing that distinguishes them is that they were mostly outside of the normal political structure established in the rest of the country. Emishi are found serving as soldiers and guards and parties of Emishi leaders occasionally visited the Japanese court. However, all references agree that they were relatively barbarian, they were loosely organized, and they lived simpler lives with little to defend, so that if attacked they could simply melt away into the mountain forests. Centuries later there are records of people who lived in the mountains and did not farm and followed customs and religious practices that were quite different from the ordinary Japanese. They often worked as miners and loggers. It is entirely possible that these were direct descendants of the Emishi.

The surviving accounts of the Emishi wars are very unsatisfactory from the point of view of those interested in technical military history. There are very few descriptions of battles and there is essentially nothing about the technical details of maintaining armies in remote locations. The main topic is the question of exerting imperial authority. There are two separate aspects of this which were emphasized at different times. Actually, I will correct that to say that there were three. At the very beginning of the process the government thought that it could take over new territories by importing thousands of settlers and simply ignoring the fact that Emishi were present. One assumes they thought that the Emishi that were affected would simply move away. Much later (and this is why I expanded the list to three) the government brought in large numbers of settlers specifically to grow crops for and provide other support services for existing military bases. However, at other times the focus was on recruiting, if you will, Emishi to become normal Japanese subjects. There were two aspects to this also. In the earlier period they tried to deal with chiefs, on the supposition that the chiefs could make their followers become Japanese. Later it was more a question of focusing on Emishi who were already rice farmers and putting off dealing with the Emishi whose life styles were differently structured. It is notable that in the later wars there were always Emishi who supported the Japanese, probably because they realized that incorporation into the Japanese state was inevitable. At any rate there is no doubt that a distinct Emishi line of descent was recognizable for centuries, down to the eleventh century at least. As in the American west, the Indians were not killed off completely, some survived to add flavor to the population.

The letter sent to the Chinese by the Japanese "King Wu" in 478 claimed years of fighting against the Emishi barbarians. We don't have any information about this from Japanese sources, which only seldom mention the Emishi. However, when the Japanese governments began keeping records (after about 600) the amount of information picks up. Essentially, all governments from that time forward took the position that the Emishi were or ought to be Japanese subjects, organized into provinces and districts and paying the same taxes as everyone else. The Emishi resisted this idea and the northeast wars were the result. The wars continued in some sense into the 11th century, but by that time they had evolved into political wars within the larger Japanese community. People were aware that some northeastern leaders were of Emishi descent, but they were not really a different people in any significant way. However, the main wars were mostly restricted to the Nara period and the most intense fighting occurred at the end of the Nara period during the reign of Kammu Tenno.

When we get into the time for which detailed information is available the southern limit of Emishi territory on the Japan Sea side was at the modern city of Niigata in the historical province of Echigo and on the Pacific Ocean side the modern city of Sendai in Mutsu province. The division of this region into provinces has varied a lot over the years. Much of the time the bulk of the area was covered by only two provinces, Dewa on the west (north of Echigo) and Mutsu on the east, but at times these have been divided into several smaller provinces. I will pretty much stick to Mutsu and Dewa. Throughout the main part of the wars the primary Japanese base was Tagajo, on the edge of modern Sendai. The "jo" character has been variously used. In Chinese it is the standard word for a walled city and this was also the normal use in Korea. Before the 16th century Japan didn't have walled cities, and the word was taken over to mean "castle." Given the nature of the wars, it seems appropriate to copy the Indian wars of the United States and use "fort." That is because these were not cities but they weren't castles either. Archeology shows that they were large rectangular enclosures with walls, like the Roman forts along the Rhine or the Saxon Shore. The walls were not intended to stand a siege, but to force people to come and go through guarded gates and to prevent surprise attacks. The Japanese also used a Chinese word that means "stockade". However, their walls were not made of bare poles, but poles that were covered with plaster to make smooth walls. To get an idea of what one looked like, Google Sawajo which has been restored as a tourist attraction. It has a web site which is in Japanese only, but with good pictures. By the way, Google offers to translate web sites to English. Do not bother with Japanese sites as the results are not comprehensible.

As a general rule one finds references to provincial notables named as "kuninomiyatsuko" south of the Emishi areas only. These were people who had some relationship with the ruling clan, either as branches of the clan or as early followers who were assigned to rule a district. Whatever process of conquest set up the polity covered by the earlier parts of Nihon Shoki never extended into the Emishi districts. On the other hand, there were kofun tombs within the Emishi zone, almost entirely on the southern boundary and the western side, around Niigata and Yamagata further inland. There was definitely some cultural overlap. There is also a late mention (837) that the Emishi fought as mounted archers exactly the same as aristocratic Japanese. In the context, the discussion was of the problems in teaching conscript commoners fighting as foot soldiers to deal with them. The solution was to give them Chinese crossbows that were more powerful than bows that could be used on horseback. Iron trigger mechanisms from such bows have been found in archaeological sites. In 789 there occurred a battle won by the Japanese and the report states that all 245 men killed or wounded on the Japanese side were struck by arrows. Japanese aristocrats back in the capital, by the way, were eager to get Emishi horses which were considered particularly excellent. There was a regular trade in horses. During the Nara period the government records distinguish two kinds of Emishi, those who lived a traditional tribal life style and those who had settled down as farmers and whose lives were little different from Japanese. The wars were all directed against the former.

Surviving records name 21 military bases used in these wars. The larger ones tend to be called forts and the smaller stockades. Some of them are called forts sometimes and stockades other times. The locations of 19 of them are known. Two were in Echigo province, five were in Dewa province, and twelve were in Mutsu province. The heaviest fighting was all in Mutsu. The dates of the references run from 647 to 812 and only a few of them were in active use at any particular time. The earliest expeditions were to Echigo, in 647 and 648. After that only locations in Dewa and Mutsu are mentioned.

In the main period of fighting there were 17 major Japanese expeditions to the north east, the first in 709 and the last in 813. The normal procedure was for the government to proclaim an expedition and appoint a general to command it. The highest level general was always a very senior official and it was not thought necessary that he have military experience. He was always assigned assistant generals to handle the technical side of the campaign. The emperor ceremonially presented a sword to the commander and then when the expedition was wound up, the commander returned the sword to the emperor. The sword was,in effect, the baton of office showing that the general was authorized to exercise powers normally restricted to the emperor. There is no mention of such a practice at the time of the war in Korea in 663 nor in the Jinshin War. It is believed that this practice was instituted as part of the administrative codes of 689 or 701. Swords were awarded also to ambassadors to China, presumably giving them the power of life-or-death over the members of the embassy as a general had over his soldiers. More practically, a sword bearing ambassador-designate had the power to requisition men and ships as he needed.

The authority given the general was great, which is why only a very senior man was ever appointed. The Japanese military normally had no generals on the books in peacetime. They were always appointed when needed and surrendered their rank when the expedition was ended. The field commanders tended to be military specialists, a few of whom have remained famous. Most of these men were from relatively minor clans and would spend their careers in the 6th rank and 5th rank zone. A few particularly successful commanders were rewarded with higher rank late in life. A handful reached sangi level, particularly after the smallpox epidemic of 737 killed numbers of high ranking nobles. The titles given generals were all either standard Chinese titles or modeled after them. They varied quite a lot, there were no real rules. The title "sei-i tai shogun" used by the later feudal rulers was one of them. It was used for the last four expeditions of 801, 804, 811, and 813 and this is why it was chosen almost 400 years later for a new role. It was a standard Chinese military title. In the Yoro code the three provinces of Mutsu, Dewa, and Echigo constituted a special case. The provincial governors were given special military powers and there was a lower ranking general, a "chinju shogun," attached to their staffs on a permanent basis. Since this comes from the Yoro code and not from reports about events in the northeast, we cannot be sure whether this was ever put into effect or whether it remained in effect for long.

Five of the expeditions (720, 724, 774, 780, and 813) were officially in response to Emishi "rebellions". The others were all attempts to extend the sphere of Japanese control. Mainly, the plan was to clear armed Emishi out of a specific area so that new forts and stockades could be constructed. For example, three of Kammu's expeditions were all directed at two districts known as Isawa and Shiwa, and as a result Isawajo and Shiwajo were established. Three administrative districts were set up under the protection of Isawajo and three more under Shiwajo. Other expeditions were based on forts already established and were intended to relieve pressure on them from hostile Emishi. There was also, inevitably, lower level conflict outside of the context of the major expeditions and extending earlier than 709 and later than 813.

The first major expeditions were those of Abe no Hirau between 658 and 660, but those were not directed at expanding the Japanese administrative system (which was still under construction) but more exploratory in character. Abe met with Emishi chiefs, awarded them Japanese titles, and encouraged them to send tribute to the court. His only battle was not with Emishi but with intruders from Hokkaido or Siberia that were raiding Emishi. I have discussed these in the Asuka section. He mainly operated in the area of the modern city of Akita, but the Dewa stockade (later renamed Akitajo) was not established until much later, in 733. At the other end, one may look at the rebellion of 878. This was in the area just north of Akita and was caused by misrule on the part of the local government. The response was not a military expedition, but the appointment of a known "good official", Fujiwara Yasunori, to temporarily take over Dewa province. He was able to calm things down and persuade the rebels to put down their arms without fighting. The "northeast wars" were over by this time.

The armies used were based on the regular military specified in the codes derived from conscription as part of the labor tax. That means that the soldiers were ordinary peasants, not specialist warriors. The Japanese aristocracy was still nominally a warrior aristocracy and everyone possessed armor and practiced with weapons. Rural aristocrats still had a genuine military role, but it was outside the codes for the most part. One of the regular guard units in the capital was manned by "toneri" as in early times and another used "kondei". These were traditional forces made up of young members of rural clans. However, this usage was intermittent; for quite a long time in the middle of the Nara period both the toneri and the kondei were suppressed. It would seem that the forces obtainable in this way were too small to make up a serious army.

The code set the basic military unit in the regular army as a regiment of 1000 men. According to the text of the Yoro code every household that had three eligible men (free and aged between 21 and 60) could be ordered to supply one for the army. One assumes that households with only one or two qualifying males were exempt, and that those with six could be asked for two men. Nowhere near that many men were actually taken (it would imply a total pool of 300,000 or 400,000 men based on approximately 1,250,000 free tax paying males). Normal military service lasted for 60 days in increments of 10 days with time off between and mostly took the form of military training at the provincial headquarters. The provincial governor was entitled to keep a force of about 20 men to guard the government storehouses and offices and keep watch on the roads to make sure that only people legally entitled to travel were passing through. The men assigned to this service were fed by their families and apparently also provided their own weapons and other gear. Throughout most of the Nara period 2000 or 3000 men were sent from the east to help guard Kyushu and about 1000 ordinary draftees were sent to the capital to serve as guards and police. In the Inland Sea area some of the local conscripts manned stations with signal fires so that news of a foreign invasion of Kyushu could reach the capital with maximum speed. There was also a network of post stations around the country where horses were provided for high speed messengers which were manned through conscription. Messages routinely passed back and forth between the capital and Dazaifu at very respectable speeds using these riders. This shows clearly in the detailed account given in Shoku Nihongi concerning communications back and forth at the time of Hirotsugu's rebellion (740).

There is not much information at all about the details of the operation of the military system in peacetime. It seems that a group of 2 to 4 districts would be responsible for 1 regiment. This suggests to me that a small province might have one regiment and a large one two, so that there might be as many as 40,000 men receiving training each year, but only if the training program was the size of a full regiment. It is possible that only a portion of each regiment trained in a normal year. We do know that the expeditions to the northeast normally consisted of between two and seven regiments. We also know that in the 9th century one regiment was routinely stationed in Mutsu. Nothing is known about the situation in the 8th century, but because the intensity of the conflicts varied a lot, it is likely that the size of the regular forces did also.

It is definite that during the Nara period the normal word for soldier, heishi, was used only for conscripts serving in regular regiments in their province. The units used for the expeditions were not regular regiments but were made up of detachments of soldiers from several regiments, presumably men with more than average military experience. When talking about the expedition armies the term "gunshi" was used exclusively. This should be translated as "expeditionary soldier."

The administrative code specified the organization of the staff of an expeditionary army. If the army had more than 10 thousand men a staff of 13 officers was specified, if the army was between 5 and 10 thousand there were 9 officers, and if the army was between 3 and 5 thousand there were 7. One commanding general could manage three armies on this scale. However, there are examples known where these relationships were not maintained. The code divided the officers into four ranks, just as a bureaucratic office had internally four ranks of officials, and this appears to have been consistently maintained.

The 709 Expedition and the Establishment of Dewa Province
I focus on this expedition because there is an unusually large amount of information about it.

Eight years after the enactment of the Taiho Code, the government undertook to project its authority into a new area for the first time. The target was what was then the northern part of Echigo province. In 702 there had been fighting with the Hayato people of southern Kyushu, but that was handled by Dazaifu and no special general was appointed. The Echigo expedition was not a response to anything done by Emishi, but was an outgrowth of the establishment of a new district called Dewa in Echigo province in the 9th month of 708. This area was the Shonai plain of the Mogami River and was inhabited by Emishi. It is assumed that the stockade at Dewa was constructed at this time, though it is first mentioned 7 years later. Several years later it was moved to a site at modern Akita, but its original location is unknown. In the Asuka period the settlement of Echigo province had reached only the area of Niigata. The province only had two districts before the establishment of Dewa. The intention was to move in a lot of new settlers from elsewhere, but the Emishi resisted, and it was decided to send in any army to establish control. It was ordered that troops be assembled from 7 provinces and three generals were appointed, headed by the Sadaiben (a member of the Dajokan) senior 4th rank lower grade Kose Ason Maro. The second in command was the vice-minister of the Monbusho but a member of a traditional military family, Saeki no Sukune Iwayu, and the third was a 5th rank man, Ki no Ason Morohito. Morohito's daughter was married to prince Osakabe and immediately after his return from the north Morohito became a grandfather. His grandson was to become Konin Tenno. In the 3rd month of 709 messengers went out to the designated provinces to raise troops. All of the provinces were in the east. The labor resources of the inner provinces were presently occupied by the construction of Nara.

The appointment of two senior generals was because there were to be two thrusts into the north, one into Echigo under the command of Kose and one on the Pacific Ocean side into Mutsu under Saeki. Clearly, this was not so much to suppress whatever petty resistance was being mustered in Dewa but to generally impress the Emishi with the futility of resisting at all. There were later occasions handled the same way, but this is the only one where the senior commander was on the Japan Sea side. All later invasions featured the Pacific Ocean side. It is theorized that the supporting force was intended to prevent the Emishi from reinforcing the other side of the island.

On the 13th day of the 7th month 100 boats from four Japan Sea coast provinces arrived at the Dewa stockade. There are no details about the fighting which took up the last two weeks of the 7th month and the first two weeks of the 8th month. By the 25th day of the 8th month both Kose and Ki were back in the Fujiwara capital and their appointments were terminated. In the 9th month all soldiers from 10 named provinces that had participated for at least 50 days were granted 1 year's exemption from labor tax. There is no information about casualties, nor about Mutsu. However, the expedition was deemed successful enough to permit the elevation of Dewa to the status of a province. This was done in 712, after the change of capitals from Fujiwara to Nara was completed. With the loss of Dewa Echigo was reduced to two districts, but it eventually expanded to 6 with a steady stream of settlers from elsewhere in Japan. Dewa became a barrier that enabled the development of Echigo to continue without further significant problems with Emishi. With the exception of the major Emishi rebellion of 720, the focus of the wars shifted from the Japan Sea side to the Pacific Ocean side. The reason is not hard to find. The northern part of Dewa province is largely uninhabited even today. The valley of the Mogami River is the only valuable part. Two inland districts were transferred from Mutsu province to make Dewa big enough pay a governor's salary. It should perhaps be noted that Dewa and Mutsu are modern contractions. The original names were Idewa and Michinooku.

The Emishi rebellion of 720
In 715 one thousand households of new settlers were brought into Mutsu from 6 provinces. That would be approximately 28,000 people if the households were typical, but for such a project they may not have been typical, excluding the usual extended family elements. They were settled in the Yamazaki plain, more or less due east of the Mogami plain on the other side of the island. In 719 the government appointed inspectors to visit all provinces and report on the state of affairs. In 720 word reached the capital that there had occurred an Emishi rebellion in Mutsu and that the inspector visiting Mutsu had been killed. The court responded by appointing a general to suppress the rebellion the day after this message was received.

The commander was senior 4th rank junior grade Tajii Agatamori, who was currently the inspector for Harima province and his assistant was junior 5th rank lower grade Shimotsukeno Iwashiro. There was also appointed a "pacification general" junior 5th rank upper grade Abe no Suruga. Tajii had previously carried a sword as ambassador to China, but had no known military experience. However, in 733 he wrote a military textbook. It took two months before it was possible to send out the orders to pull the expedition together, involving 9 provinces for men or supplies. By this time it was the 11th month. Tajii and Abe returned to the capital at the beginning of the 4th month of 722. There was a simultaneous rebellion of Hayato people in Kyushu and the victorious generals of the two expeditions were publicly rewarded the same day. Tajii brought with him many Emishi who had supported him, and early the next year the governor of Mutsu rewarded 52 other Emishi leaders who had supported the Japanese in the fighting.

The Hayato rebellion had actually started first, in the region that later became Satsuma province. It was not at this time organized as a province and was under the supervision of Osumi province to the east. At the start of the rebellion in early 720 the governor of Osumi was killed. Osumi itself had only become a province in 713. In 714 200 households of people from northern Kyushu were moved to Osumi to "educate" the Hayato. It looks like the two rebellions had similar causes. In Kyushu there was some hard fighting and when the generals returned they brought with them the heads of dead enemies and living captives totalling 1400. No breakdown is offered.

Political History
The political history of the Nara period naturally falls into three periods.

First is the period from the promulgation of the Kiyomihara Ryo in 689 through the abdication of Gensho in 724. During this period there were no major political upsets and the government was mainly devoted to attempting to make the new system work properly. There is not a lot of information about the personalities in government, and it is presumed by most that the earlier part of the period was dominated by Jito Tenno and the latter part by Fujiwara no Fuhito. The death of Fuhito in 720 does not seem to have caused a major problem and his son Fusasaki seems to have assumed the leading position in government without resistance.

The second period was dominated by more flamboyant personages, Shomu Tenno and his daughter Koken Tenno/Shotoku Tenno, and also Fujiwara no Nakamaro, alias Emi Oshikatsu, and the monk Dokyo. This was a time of significant political turmoil, and it is possible to argue that it had a powerful effect on future political history, as the practice of frequent resort to female rulers was abandoned after almost 200 years, not to reappear for another 900 years in a very different political environment.

The final period was dominated by Kammu Tenno. He had strong ideas and attempted to establish a structure for permanent rule by the imperial clan which did not, as it happens, survive him for very long. There were significant changes to the working of the central government and many attempts to deal with problems in the provinces during his time.

Because Nihon Shoki and Shoku Nihongi provide very little information about political incidents, the first period has really been covered already in the section on the workings of the new administrative system. There seems to be little doubt that the main problem faced by the government throughout was the constant recurrence of problems resulting from poor harvests and epidemics of disease, which were not unrelated because any time a large percentage of the population was sick, work in the fields suffered. On the whole they managed it successfully to the extent that in good years the grain taken in by the taxation system was sufficient in quantity to provide a reserve for the bad years, when there were normally remissions of taxes.

Another constant throughout the entire period was that relations with the Korean kingdom of Silla were bad and war was threatened on several occasions, though nothing ever happened. Nowhere do the sources explain exactly what the issues might have been. It is expressed in Shoku Nihongi in terms of diplomatic slights by Silla, not offering enough respect to Japan. There seems to have been a contest between the two to try to get T'ang dynasty China to recognize one of them as the "ruler" over the other, without success. The Chinese clearly did not want to get involved in anything, having their hands full with Turks. Late in the period Japan entered into direct diplomatic relations with the Manchurian kingdom of Po-hai, which claimed to be a successor state of Koguryo, and which controlled much of what is today North Korea, so that it could communicate directly with Japan by ship across the Japan Sea. This was natural as both were hostile to Silla. Ancient Korea is a particularly thorny topic because of the non-existence of good information about it, but it does appear that Silla was different from Paekche and Koguryo culturally, religiously, and probably linguistically, which meant that there was less sympathy between Korea and Japan than if one of the other kingdoms had accomplished the unification.

This general pattern of dislike has persisted down to the present day, even though periods when one side was actively annoying the other were few and far apart. The "Mongol Invasion" of Japan of the 13th century was largely actually a Korean invasion, then Japan invaded Korea in the 1590s and dominated it between 1905 and 1945. This doesn't even come close to the constant irritations imposed on each other by the French and the English over the centuries. In the Nara period it is notable that there continue to be frequent mentions of migration of Koreans into Japan, presumably persons from former Paekche and Koguryo territories who were unhappy with rule by Silla. It is easy to imagine that the government of Silla may not have been pleased about the constant leakage of taxpayers. We don't know enough about ancient Silla to know whether there were unhappy Japanese going the other way. We do know that there was a law against Japanese leaving the country except for diplomatic missions (which remained on the books for centuries, and led to an interesting court case in the Heian period we will come to later).

Shomu Tenno
Shomu Tenno assumed the throne in 724 and ruled until his abdication in favor of his daughter in the 7th month of 749. He lived for another 7 years until 756. So far as Shoku Nihongi tells us, he was not associated with any specific approach to the problems of governance, nor is there any indication that he took much interest in the topic. He definitely does seem to have been interested in court ceremonial and he frequently appeared in public ceremonies before an audience of commoners. Earlier rulers functioned mostly in seclusion and ceremonies were observed only by nobles. Even when the ruler was present at a ceremony, he was often invisible behind a screen. There can be no doubt whatever of his strong interest in Buddhism and his devotion to its spread among the Japanese population. He was responsible for the very expensive "Kokubunji" project that sought to establish a monastery and a nunnery in every province, and the even more expensive project to construct an enormous temple at Nara, Todaiji, and provide it with a colossal bronze Buddha statue, one of the wonders of the ancient world and still very impressive in its modern incarnation. Then, there is also the vital importance of the Shosoin treasury, established on his death to preserve a selection of his personal possessions, including, more or less by accident, many thousands of original government documents of his reign.

Shomu Tenno was on the throne during the great smallpox epidemic of 737/738 and there is evidence that he accepted the Chinese theory that disasters of that sort were a reflection upon the moral worthiness of the ruler. This was clearly a major factor in his efforts to promote Buddhism. The main government response to the problem was a pragmatic one; it dispatched medicines and sent provincial officials copies of Chinese medical writings, some of which are quoted in Shoku Nihongi and were quite sensible. For a long time there had been a system called suiko for loans of seed rice to farmers. There were two types, public loans from unused tax revenues that were paid back with 50% interest and private loans using wealthy families unused funds that were paid back at 100%. At some point the loans became compulsory and the interest became a regular part of the revenue. In 737 Tachibana no Moroe’s government outlawed privately funded loans on the grounds that they were oppressive to the people, and in 738 the provincial governors were specifically forbidden to make loans to farmers out of their personal funds. Also, the conscription of men for military service was temporarily halted, to be resumed only early in 740 for normal conscripts. There had been two special categories of conscripts, "toneri" who were upper class warriors from eastern provinces with a special connection to the imperial clan, and "kondei" who were sons and other relatives of men of the district magistrate class. The recruitment of toneri did not resume until 746 and that of kondei only in 762. In addition to these practical responses to the crisis, Shomu organized a series of large scale Buddhist ceremonial responses as well.

The epidemic inevitably had a powerful effect on politics in the capital as a large number of high ranking nobles died, including all four of the sons of Fujiwara no Fuhito. One member of the next Fujiwara generation, Toyonari, was immediately promoted to the Sangi, but Tachibana no Moroe received the highest rank. Three men from minor clans were made Sangi, something that had never happened before and would never happen again. Fujiwara Hirotsugu was another grandson of Fuhito who, it seems, had been promoted to 5th rank in 737 on the death of his father, Umakai, and in 738 was made governor of Yamato province, a high prestige job. However, when the new Sangi was put together he was left out and instead transferred to a position at Dazaifu in Kyushu. At the beginning of 740 he sent in a memorial blaming the government for everything that had gone wrong, and claiming that it was necessary to remove two specific persons from positions of influence if the catastrophes were to come to an end.

The two attacked by Hirotsugu were comparatively low ranking men who had become influential when they returned after many years in China. One was a priest named Genbo, who had been sent to China in 717 and remained there for 18 years. When he returned he brought with him 5000 rolls of books which led to the establishment of several new sects of Buddhism in Japan. In 737, at the height of the smallpox epidemic, he became the head of Shomu Tenno's personal chapel in the imperial palace. The second man was a provincial noble from the Kibi region named Shimotsumichi no Kibi no Masabi. He was a companion of Genbo in China and returned at the same time. He was promoted to high rank (for his class) and made assistant director of the imperial academy, responsible for the education of young noblemen. By 740 he had very elevated rank for a person from a provincial clan and was working in the imperial palace where he was on intimate terms with the emperor.

Hirotsugu's rebellion didn't amount to much as the army he raised refused to fight for him when a powerful imperial army arrived and was wrapped up in the 1st month of 741. 26 persons were sentenced to death, 5 were sentenced to slavery, 47 were sentenced to exile, and 209 to lesser punishments. All of Hirotsugu’s brothers were included among the exiles.

In the 10th month of 740 the emperor along with his mother and his empress and all of the senior officials left the capital along with 4,000 troops commanded by Fujiwara no Nakamaro, and over the next two months proceeded to tour four provinces east of the capital before winding up at a palace in a place known as Kuni in Yamashiro province. About 10 days before the arrival the senior minister, Tachibana no Moroe, had left the main party and hurried to the final destination to supervise arrangements for the party’s arrival. The palace had previously been visited by the emperor’s mother (empress Gensho) and empress Gemmei on excursions of their own. The location is at the southern end of Yamashiro very close to Nara (probably 2 train stops today). Modern scholars have theorized that this was to get the emperor out of the capital just in case there was any follow up to Hirotsugu’s rebellion on the part of other descendants of Fujiwara no Fuhito. They stayed there through the punishment of the rebels. Immediately afterwards emissaries were sent to all major shrines to report the intention to move the capital. Throughout 741 a series of steps were taken to execute the transfer of the capital from Nara to the new location.

5500 workers were conscripted to work on the new palace and city. The site presented problems that Nara did not. There was a small mountain in the middle of it and a river running through it. Building bridges over the river turned into a major project. Transfer of all government functions had not yet been completed when in the 8th month of 742 Shomu picked a new location 30 kilometers to the northeast at Shigaraki in Omi province. He moved there immediately. The official transfer was not proclaimed for some time, so the Kuni palace, as it came to be called, was officially the capital for 3 years and 2 months.

The decision to order the construction of the kokubunji was made during the Kuni period. The decision to construct the Great Buddha at Todaiji was made during the Shigaraki period.

In the first month of 744 Shomu announced that he was moving to Naniwa. Naniwa had been the site of imperial palaces from time to time for more than 300 years. Then, a few weeks later, Shomu assembled all of the officials in the Kuni palace and asked them to vote on whether the capital should be at Kuni or at Naniwa. Among the officials of 5th rank and above 23 voted for each place. Among the officials of 6th rank and below 157 voted for Kuni and 130 for Naniwa. Hardly a landslide result. Three days later officials went to the market in Kuni and polled the people there. All but 2 voted for Kuni. One person voted for Naniwa and one, thinking outside the box, voted for Nara.

The emperor thought about it for a week and moved to Naniwa. In the second month the edict came ordering everyone to move. While traveling with his father to Naniwa Prince Asaka fell ill and died at the age of 17. Before the end of the month the emperor was back at Shigaraki. Empress Gensho and the senior minister Tachibana no Moroe did not go. They stayed in Naniwa. Two days later Tachibana no Moroe published an imperial edict saying that although the Emperor was going to be in Naniwa, the common people could freely choose whether they wanted to reside in Naniwa, Kuni, or Nara. In the summer of 745 serious forest fires broke out in the area of the Shigaraki palace, and the emperor evacuated. He held another meeting of officials to decide what to do and this time they were unanimous: go back to Nara. The emperor agreed and immediately returned to the palace there.

Moving the palace was expensive. Shoku Nihongi reports financial problems for the government for 734. When the decision was made to return to Nara the palace had to be cleaned, and the monks of the area and also farmers from nearby villages were drafted for the work. The official in charge dismissed the farmers immediately on the ground that it was a busy time of the year in farming. There were complaints among the nobles that the government was not handling things well. There were two young nobles who began to become prominent about this time. One was Tachibana no Naramaro, son of Moroe, who made no secret about his dissatisfaction with the way things were going. The other was Fujiwara no Nakamaro, who was more discreet.

Only three months after returning to Nara the emperor left again and went to Naniwa. While there he fell seriously ill. He summoned all living descendants of Tenchi or Temmu to Naniwa. There is no evidence as to exactly why, however one topic has to have been the succession to the throne. The senior member of the Tenchi lineage was Prince Shirakabe (future Konin Tenno). He was then 38, held comparatively low rank, junior 4th rank lower grade, and is said to have been a serious drinker with little interest in government. Shomu had 7 years previously established Princess Ae, daughter of empress Komyo, as the official successor. Never before had a princess received such designation. The previous two reigning Empresses had clearly been placeholders. This was very different. Nothing happened at this point because the emperor recovered and returned to Nara. However, he abdicated in favor of Princes Ae, Koken Tenno, three years later on the occasion of the auspicious discovery of gold in Mutsu province.

At the beginning of 741 the Fujiwara family offered to return to the court the office lands amounting to 5000 households that had been assigned to Fujiwara no Fuhito. This was immediately after Hirotsugu’s rebellion and was presumably a sort of apology. 2000 households were given back to Fuhito’s estate but 3000 households were retained by the government. At the request of former empress Komyo (Fuhito’s daughter), it is said, this revenue was assigned to support the construction of kokubunji temples. This project was announced one month later. It was presented as a pious act intended to help protect the nation from a repetition of the recent disasters. Up to now formal Buddhism had been pretty much limited to the immediate vicinity of the capital (and Dazaifu). No serious effort had been made to convert the common people or even the rural nobility. The kokubunji project called for the construction and subsequent support of two temples, one a monastery and one a nunnery, located at the provincial capital of every province. Archeological evidence has been found for quite a number of them, and some were the starting point for temples that survive today. This was actually the culmination of a process that started in 737 when an edict called for sending one 6 foot Buddha image to every province, to be placed in a suitable Shinto Shrine. Then in 740 it was provided for the construction of a seven story pagoda in each location.

Building a temple was a big project and it required specialized craftsmen who were not likely to be found out in the countryside. There were at the time 62 provinces, so the revenue donated by the Fujiwara amounted to 50 households per province. This was to be used for construction. Each of the two temples was to be assigned 10 cho (24.5 acres) of rice land for its ongoing support and the monastery temples were to have 20 monks and the nunnery temples 10 nuns each. There were definitely problems executing all of this. In 744 the government cut the primary grain tax quota for each province in half, ordering that the other half be used to forward the construction of the temples. In 747 it noted that many of the temples had not yet been completed and that this was causing natural disasters to continue. It increased the endowment for the monasteries to 220 acres and the nunneries to 100 acres, and urged provincial governors and district magistrates to press the construction. The district magistrates were offered rewards for finishing up within three years. Despite these problems it seems that by the end of the 8th century the project was generally complete. Recent studies indicate that all were constructed to the same plan, a miniature version of the plan used for Todaiji in Nara.

The government of Shomu Tenno also undertook the colossal project of the construction of the Great Buddha hall and statue at Todaiji. The emperor later reported that he made a vow to accomplish this in 740. It is theorized that the idea came from the spread of the Kegon sect of Buddhism at this time. Its scriptures are noted for utilizing very large numbers ("billions and billions" like Carl Sagan) in descriptions intended to inspire awe. The image was intended to represent Vairocana, heretofore unknown in Japan. The construction office was established in 743, and work began one year later on the framework for the statue. The site chosen was the kokubunji temple for Yamato province, just outside the city, which was remodeled for the purpose. According to Todaiji’s own history, ground was broken by the emperor personally in 745. The man in overall charge of the statue portion of the project was Kuni no Kimimaro, a member of a family that came to Japan from Paekche in 663. There can be little doubt that a large proportion of the technical specialists on the project would have been of Korean origin. However, Kimimaro’s three chief assistants all had names indicating descent from the minor nobility of Yamato province. Two were from a family that had been rewarded for fighting for emperor Temmu in the Jinshin War, and the third was from a family that had a Manyoshu poet. According to Todaiji, 51,590 men worked on the statue, supported by 1,665,071 laborers, and 372,075 men worked on the buildings, supported by 514,902 laborers. The labor forces were managed by 10 members of the local aristocracy. The most intense part of the project ran from 747 to 753. One assumes that the numbers are "man days" of labor, not of individual workers.

In 749 news reached the capital that quantities of gold had been discovered in Mutsu province in the northeast. Up to then there had not been enough gold in the country to gild the statue according to Shomu’s desires. This event was treated as a miracle, and a few weeks later the emperor assembled all of the officials at the construction site to announce the news.

Koken Tenno
Retired emperor Shomu included in his will that after Koken Tenno prince Funado should reign. Funado was the son of prince Nitabe and a descendant of emperor Temmu. When Shomu Tenno (756) died the palace guards were put on alert and barriers on the main roads were closed. This appears to have been routine at the death of a ruler. However, two days later two men were arrested on charges of “slandering the court”. They were questioned, but after three days were released and permitted to go home. One of them was married to a sister of empress Komyo. Shoku Nihongi says that he had committed no slander but was falsely accused by Fujiwara no Nakamaro.

In 757 Tachibana no Moroe died at the age of 74. He had resigned his office a few months previously, perhaps ordered to do so. Two years previously one of his secretaries had been suddenly executed. It was said that Moroe got drunk and said slanderous things about the court at that time. A bit later it was discovered that during the mourning period for Shomu Tenno, Prince Funado had been dallying with girls and gossiping about court secrets. Empress Koken then revealed Shomu’s will to the courtiers saying “maybe we should be a little careful”. The Sangi, now led by Fujiwara Toyonari, decided to leave the will in force. However the Empress the same day demoted Funado from "Near Prince" (the group eligible for the succession) to ordinary prince and cancelled permission for him to live in the palace. He left to go to his own house in the city. Five days later he was dismissed as crown prince. Five days after that the Empress convened the Sangi to discuss the succession. The two senior ministers Fujiwara Toyonari and Fujiwara Nagate recommended prince Funado’s younger brother prince Shioyaki. Two other officials, Fumuya no Chinu and Otomo no Komaro recommended prince Ikeda, son of prince Toneri and also a grandson of Temmu. Only Fujiwara no Nakamaro declined to offer advice, deferring to the empress. Koken Tenno rejected both Shioyaki and Ikeda as unsuitable and nominated a brother of Ikeda, prince Oi, later Junnin Tenno. The Sangi agreed. Shoku Nihongi says that previously Fujiwara no Nakamaro had moved Prince Oi into his own mansion. Prince Oi was 25 at this time. His mother was said to have been Tajima Yamashiro, a family whose “kabane” name Mahito indicates that it was considered an offshoot of the imperial clan.

An extraordinary fuss was made over the nomination of Prince Oi. There was a general amnesty for convicted criminals, there were remissions of taxes for select groups, bonuses on salaries for others, and permanent tax cuts by shifting special (lower) tax rates for men aged 17 through 21 to apply to ages 18 through 22 instead.

Shortly afterwards the empress temporarily moved into the mansion of Fujiwara no Nakamaro so that the imperial palace could be remodeled. It was at this time that the text of the Yoro revision of the Taiho administrative code that Fujiwara Fuhito had been working on at the time of his death was discovered. As discussed previously this is the only version of the code that survives (in part) today.

In the same year there was a general round of promotions for the senior officials, including princes Shioyaki, Ikeda, Funado, and Shirakabe. Fujiwara Nakamaro was appointed to a new office called Shibi Naisho. This renamed and elevated the status of the office that ran the palace of the empress, placing it on the same level as Dajodaijin, heretofore the highest office of all. This was done on the first anniversary of the death of Shomu. On the 6th month 9th day of 757 there was published an edict that forbade clan chiefs from overseeing public affairs or gathering clan members for meetings, and it also forbade families to accumulate horses and military weaponry or keep more than 20 armed guards at their mansions. On the 19th day there was a major shakeup within the government, especially affecting the offices with military functions. The minister of war, Tachibana Naramaro, was shifted to a civilian job and replaced by a close associate of Fujiwara Nakamaro, Ishikawa Toshitari, and two generals were appointed, something usually only done when actual military action was expected. A very considerable amount of information exists about what happened next. On the 28th day prince Yamashiro (a son of Nagaya) came to the temporary palace and met secretly with the empress and Nakamaro. He claimed that Tachibana Naramaro had gathered troops and was planning to attack the palace. On the 7th month 2nd day the empress publicly announced that she had heard that dissident members of the nobility were planning a coup. She promised an investigation of the rumors. Then she gathered the senior officials and told them that she was counting on them to continue their long tradition of loyalty to the throne. That evening a member of one of the guards units came to Nakamaro and claimed that that afternoon he had been approached by Ono no Azumabito (the commander of the force that handled Hirotsugu's Rebellion) to join a plot to assassinate the crown prince and Nakamaro. The leaders were said to be prince Kibumi, prince Asuka, Tachibana Naramaro, and Otomo Komaro. The next day Nakamaro stood beside the empress in the palace and read a statement she had written. It said that there was evidence that prince Shioyaki and the four named above planned to commit treason. She could not imagine why her near relatives would hate her. Had she not appointed them all to high office? They were not going to be punished. She urged them to reconsider their ways. They immediately left the palace, and outside the gate, bowed and apologized. This method where a leading official reads the edict of the emperor was routine, by the way, and had nothing to do with the fact that Koken was a woman.

The next day interrogations began, with the senior minister Fujiwara Toyonari in charge. Ono Azumabito confessed. He said that the plotters had held three meetings during the 6th month. He named a number of people and said there were others whom he never saw. The attack was supposed to be that very day. Nakamaro was to be killed, the crown prince deposed, and the empress also. Afterward a new emperor would be chosen from among princes Shioyaki, Funado, Asuka, and Kibumi. All of the named persons were arrested and interrogated in turn. Shoku Nihongi quotes prince Asuka’s confession and Tachibana Naramaro’s interrogation in question and answer format. Naramaro was asked why he rebelled and said it was because of bad government. What was bad about the government? The colossal expense of the Todaiji project had ruined the common people.

Another of the plotters said that Naramaro had spoken to him as early as 745, when emperor Shomu was ill at Naniwa, complaining of the ruinous government under Shomu, and said that many people agreed with him. He said he had similar discussions with Naramaro in 749 and 750 also.

Ten days after the incident the empress held an audience with an unusually large number of officials, including many local officials from the region who were brought in. Her statement which was read out said that the law requires that the traitors be executed. However, out of benevolence she was reducing the seriousness of the charges by one degree. Their names and ranks were to be changed and they were to be exiled. That was actually a lie.

Their “kabane” titles were certainly changed, essentially stripping them of aristocratic status. However, a number of them were then executed. Shoku Nihongi names six men including prince Kibumi, prince Funado, and Ono Azumabito, but not Tachibana Naramaro or several others prominently named. This has led some to conclude that those men were actually exiled, but none were ever heard of again. A granddaughter of Naramaro later married emperor Saga, so his name must have been restored then or earlier. There is an official statement dated 770 that a total of 443 persons were executed or exiled as a result of this incident. One of the confessions said that 300 soldiers were to be used. They would seem to be included in this total.

Fujiwara Ototada, third son of minister of the right Fujiwara Toyonari, was judged to have been involved and sent to Himuka province in Kyushu. Toyonari himself was judged to be unreliable and appointed to an “ingai” position at Dazaifu. This meant that it was outside of the official table of organization and received no salary, but required him to travel to Kyushu.

What are we to make of Naramaro’s complaints? There are no documents to show whether the nation’s finances were or were not ruined at the time, or whether the common people were particularly badly off. All of the surviving financial documents date to earlier in the period. There is a book written by Miyoshi Kiyoyuki about 160 or 170 years later. Miyoshi says that the enormous spread of the influence of Buddhism at that time did cause considerable strain. In particular the campaign to construct the kokubunji temples sucked up 50 percent of the national wealth.

A land distribution took place in 740 and three years later the government decreed henceforth the lands assigned to officials as part of their salaries were to be privately owned. We have seen that the Fujiwara family had retained the lands that had been assigned to Fuhito, and since office holding was essentially hereditary, baring political disaster a family was going to remain at roughly the same rank all the time, so it perhaps made sense to make this change. This would have been of particularly great benefit to the district officials it is thought. Once you can legally own land, many possibilities are opened. This is the starting point for privately owned shoen.

In 744 it was decreed that the profits from suiko (crop loans) were to firstly go to make up any deficit in the local provincial budget, then to make up any deficit in the taxes forwarded to the central government, and then anything left over went to the governor. The rice available for suiko loans came from surplus tax revenue over the quota the province owed the central government, and a governor who used it aggressively could make quite a profit out of it. Some studies have indicated that the total amount that was available to divide among all of the officers of the province (there was a rule about how to split it up) was about equal to the annual tax quota. Some scholars believe that already about 50% of the total revenue was allocated to local officials, and that the era is already starting where appointment to provincial office became a form of property unrelated to any work done.

Up to now the laws had been coming thick and fast, but for the 8 year period from the abdication of Shomu Tenno to the destruction of Tachibana Naramaro there was nothing at all. Fujiwara no Nakamaro had been born in the Fujiwara capital in 706, and so was 5 when Nara was occupied. From an early age he was noted for intelligence and also physical vigor. As was typical at the time, his youth was spent at the house of his maternal grandfather Ae no Sukunamaro. At the age of 15 he became a page in the palace, as his father had before him. His first office was in the university. This was at the time of the death of prince Nagaya and the establishment of his aunt as empress. At the age of 29 he was elevated to 5th rank, making him a member of the higher aristocracy. His older brother Toyotsugu had reached this rank at the age of 21. He passed through the smallpox epidemic without catching the disease. He was probably living in a house of his own by this time, not with his father. When Shomu Tenno evacuated the capital at the time of Hirotsugu’s rebellion, Nakamaro held the rank of general and commanded his armed escort. He was promoted to 4th rank and elevated to the Sangi as minister of the Monbusho at the age of 38. The law permitting private ownership of konden was issued by his ministry 20 days later. He was also minister at the time of the law regulating the disposition of profits from suiko loans. It is not known what kind of relationship he had with his cousin, 12 years younger, who became empress Koken. He was promoted to 3rd rank and minister of justice the day of her enthronement. He was now the third ranking member of the government after Tachibana no Moroe and Fujiwara Toyonari, his older brother. He concurrently held the post of commander of the palace guards.

We have already seen that Nakamaro played a conspicuous part in overcoming the conspirators led by Naramaro. His office as head of the empress’ palace staff meant that he was her spokesman at all public gatherings, standing at her side, reading her instructions aloud to the assembled officials. One month after the incident, the era name was changed, the main crop tax was cut in half, and the interest on suiko loans was forgiven. At the beginning of the next year officers were dispatched to tour the provinces and collect the grievances of the common people. Later in the year, as the result of these inquiries, the tax brackets were adjusted with adults defined as ages 22 to 59 and elders as age 60 and higher. In 759 all officials of the 5th rank and higher plus all monks of ranks that could be considered equivalent were solicited for opinions on what could be done to reform the government. The results were published a month and a half later.

Ishikawa Toshinari, 3rd rank, said that “the procedures needed to properly carry out the administrative laws have not been fully worked out. Additional procedures should be prepared.”

Fumuya no Chinu, 3rd rank, said that “every year during the first month there is a festival carried out at all temples from the 8th day to the 14th day when meals are served to the monks at government expense. Monks who happen not to be at the temple are given cloth worth 7 days meals in compensation. There are those who collect this from more than one temple. This distribution of cloth should be stopped.”

Prince Shioyaki, 3rd rank, said that “princes 3 and more generations removed from an emperor receive their stipends twice a year, in spring and autumn. The same as with officials and their salaries, on the day of payment it is necessary to present proof that the person is actually eligible for the payment. Since the stipends of the imperial family are a permanent entitlement, these examinations should be stopped.”

Yamada Komaro, 6th rank said that “commoners who have a large number of children to raise find it very difficult to pay their taxes. Fathers who have 5 or more children should be exempted from the labor tax.”

A monk who had been in China said that the roads should be lined with trees to protect travellers. Nakamaro actually did order the planting of trees along the main roads.

Junnin Tenno
In 758 empress Koken abdicated and prince Oi became emperor (Junnin Tenno). The emperor then awarded Nakamaro a new name, Emi Oshikatsu. Back in the 7th century Nakatomi Kamatari who had played a vital part in establishing the regime was granted a new name, Fujiwara, at the time of his death. This established his descendants as a new clan, entirely separated from the Nakatomi clan. Now his great-grandson Nakamaro was given the same privilege. He was granted the tax revenue of 3000 households and 245 acres of office land plus other entitlements. At the same time the names of most government offices were revised to match the names used by the T’ang dynasty. Nakamaro received the title Taiho, formerly minister of the right. He was then promoted at two year intervals. In 760 he was raised to junior 1st rank and Taishi, equivalent of Dajodaijin. This title had heretofore only been held by princes, and that rarely. In 762 he was made senior 1st rank. He couldn't be promoted any higher, but now his sons were given high rank. At the start of 762 his son Masaki was elevated to Sangi grade, and at the end of the year Kusamaro and Asakari were also given rank and office. The fourth son Masako was given rank but no office. All of this was without precedent.

However, it seems that 760 marked his high point and then things started to go wrong for him. It began with the death of Komyo in that year. She had always been a powerful ally. It seems that after the death of her mother, former empress Koken began to turn against Nakamaro. It was at this time that she began to become intimate with the monk Dokyo. In that year Nakamaro began the construction of a new capital at Lake Biwa and in the winter of 761 it was named Hokukyo (northern capital) and Koken and Junnin were moved there with a view to reconstructing the Nara palace. Koken became ill during this move, which was the event that brought her into contact with Dokyo. Nakamaro sent her a letter complaining about this through Junnin and Koken reacted angrily. In the summer of 762 she left the northern palace on her own initiative and returned to Nara, where she moved into a temple. Junnin soon also returned, going to the palace. In the 6th month Koken assembled all high level officials at the palace and announced “Although a woman, because there was no other successor I served as emperor. Since I resigned that office in favor of the current emperor, he has not been respectful, and has said things that ought not to have been said and done things that ought not to have been done. Since these are not things I should be forced to bear, I have taken up residence in a separate palace. From now on the present emperor is to be limited to ceremonial matters and small things, but I will handle all important matters and judgments.”

The government had a lot of problems at this time. There was still much military activity in the northeast, and in 759 orders had been sent to Dazaifu to begin planning for an invasion of Korea (this seems to have been Nakamaro's project). Relations with the kingdom of Silla had been bad essentially since it unified the peninsula, destroying Japan’s ancient ally Paekche in the process. It is thought that this project was the result of the widespread rebellions that were then raging in China. It was either thought that Korea should be conquered to provide a buffer against danger, or just because the weakness of China meant that Silla would receive no assistance. There is no evidence as to just what the government was thinking. They had clearly been thinking about it for some time because they had been promoting the study of the Korean language for some years. In 761 preparations began to send 394 ships, 17,360 sailors, and 40,700 soldiers.

Nothing much seemed to happen until 764. In that year two of Nakamaro’s close allies died, Ishikawa Toshitari and Fujiwara Nomitate. Nakamaro responded by placing a number of his supporters in key positions, including putting his 6th son in charge of one of the palace guard units, and making his 7th son governor of Echizen province and his 8th son governor of Mino province. In the 9th month emperor Junnin appointed Nakamaro to a post giving him military authority in 10 provinces around the capital. His specific mission was to gather 20 men from each of the 10 provinces and assemble them at Nakamaro’s mansion within 5 days. Nakamaro informed Takaoka Hiramaro, a secretary of the cabinet, and he in turn rushed off to tell the retired empress. She immediately instructed prince Yamamura to go to the quarters of Junnin in the palace and arrest him, and replace all of the guards in the area. The assumption, not spelled out clearly, is that Junnin Tenno and Nakamaro were planning to stage a coup that would remove the retired empress from power. When Nakamaro heard of the arrest of Junnin Tenno he ordered his son Kusumaro to attack prince Yamamura, but Kusumaro was defeated and killed. Meanwhile Nakamaro’s mansion came under attack by Ki no Funamori, an old enemy. One of Nakamaro’s men charged out wearing armor and on horseback to fight these forces, but was shot down by arrows. After this Nakamaro escaped his mansion and fled into Omi via Uji.

The empress rallied substantial forces from guards troops and also from Korean descended families that had served the imperial clan for many generations. The other branches of the Fujiwara clan also rallied around and provided forces of their own. These troops fanned out over the eastern districts looking for Nakamaro. Some also went to Echizen and attacked the governor, Nakamaro’s son. Eventually Nakamaro was cornered at Lake Biwa, and there was a fierce fight which he lost. He was captured and immediately executed. Some 30 or 40 other persons were also cut down on the battlefield. This was 10 days after he began to raise troops for Junnin and 8 days after the empress began taking action against him. Immediately afterward the empress’s friend Dokyo was raised in rank.

Shotoku Tenno
The empress resumed the throne. She was assigned a second reign title, Shotoku Tenno, following the precedent of Kogyoku Tenno and Saimei Tenno. These reign titles were all invented later, during the Heian period, in imitation of the Chinese practice. They were not used during the Nara period. Junnin Tenno did not abdicate, he was deposed, for the first time in history. He was then exiled. He died one year later at the age of 33, and it is widely thought this may have been as assassination. Shotoku Tenno ruled for 7 years relying heavily on Dokyo and taking a strongly Buddhist point of view on all issues. In the Heian period it was widely believed that Dokyo was a descendant of emperor Tenchi via prince Shiki. However, there is no basis for that. He came from a family called Yuke in Kawachi province. The Yuke name means bow-maker and they were connected with the Mononobe clan, the traditional providers of military equipment back in the barbarian period. When Temmu Tenno reformed the kabane names the Yuke were assigned sukune, a low ranking title. They lived in what is now part of Osaka and appear to have been involved in a military unit connected to prince Shiki. Dokyo's age is not known. However, he studied with a monk named Gien who died in 718. Nothing to speak of is known about his career before he came into the palace and became trusted by the empress. Bringing a monk in to pray over a sick person was routine practice, so it is easy to explain how he made his initial contact with her.

In 765 the empress and Dokyo took a trip to a scenic spot and on their way back to the capital stopped at Yuke in Kawachi, Dokyo's ancestral home. There, in the local temple, she promoted him to the office of Dajodaijin. The next year Dokyo was granted a new title Hoo where the final “o” is the Chinese word for king, used to write the Japanese prince, and the first element "Ho" means law, which in a Buddhist context means "doctrine", so that Hoo was a name better suited to a monk as an equivalent for Dajodaijin. In the same year two other monks close to Dokyo were raised to the Sangi. This was all very strange, of course, and the court was also unsettled by the fact that Shotoku had no successor. Junnin was dead. His brothers prince Funado and prince Ikeda were exiled. Junnin had a son, prince Wake, who had been hung (or strangled, the words are the same). It is said that at first he was to be exiled, but minds were changed and he was killed while en route to his destination.

What was different about the regime of Shotoku Tenno and Dokyo? It is hard to say. The total number of new people who were brought in was very small. The same nobles as always, Fujiwaras conspicuous among them, ran most of the government as always. Dokyo and his allies took positions at the top of the hierarchy. There was little additional movement of new people below, except that 14 members of Dokyo’s family were given jobs. There is no indication that either Shotoku Tenno or Dokyo had much interest in the daily business of government. Their main accomplishment was the construction of a new temple, Saidaiji, "Western Great Temple", to match Todaiji, "Eastern Great Temple" constructed by Shomu Tenno. To my mind the odd thing is that there was no sign of any opposition to Dokyo among the nobles. It shows, as had several oddities earlier, such as the behavior of Shomu, that there was an enormous amount of respect for the emperor, no matter how odd the emperor might turn out to be. An outsider like Nakamaro seems not to have had a chance in a real showdown with a ruling member of the imperial clan.

However, things continued to become more bizarre. In 769 it was reported that an oracle had been heard at the Usa Hachiman shrine in Kyushu saying that “if Dokyo were to be emperor, the world would be at peace”. The Usa shrine was the headquarters shrine for all the Hachiman shrines all over Japan. Oracular messages were not unexpected. When Shomu wanted encouragement that his project to build the Great Buddha statue could succeed, he asked for one. However, whatever this may or may not have been intended to evolve into, the empress became ill in early 770 and died in the 8th month at the age of 53. Dokyo immediately lost all power. It is noteworthy that no one thought it necessary to execute him or accuse him of crimes. He was appointed as abbot of a temple in eastern Japan. His power was merely a reflection of the wishes of the empress. The same had also been true of Nakamaro. He had no power of his own, but flourished when supported by Koken and Junnin, and then crashed when support was removed. He was only unique for his gluttony for extravagant titles.

Konin Tenno
When Shotoku Tenno died a conference of leading officials was called to choose a successor. Dokyo was not invited. There was no obvious candidate. The udajin, Kibi no Makibi, proposed selection of either Funya no Kiyomi or his younger brother Funya no Ochi. They were grandsons of emperor Temmu but had earlier been officially removed from the imperial clan and assigned to the new clan of Funya. Fujiwara Nagate, the sadaijin, supported by his cousins Fujiwara Yoshitsugu and Fujiwara Momokawa opposed this and suggested instead prince Shirakabe, a grandson of Tenchi Tenno. Shirakabe was a member of the Sangi, but had had a rather quiet career up to this point. Shirakabe was selected and is known as Konin Tenno. His son prince Osabe was immediately designated successor, and Dokyo was now removed to a remote province and Kibi Makibi was forced to retire. The new emperor was 62 years of age. The circumstances are not explained very well in Shoku Nihongi, but Konin Tenno's wife, who was a daughter of Shomu Tenno, appears to have been distressed by the amount of power held by Fujiwara Nagate and approached Fujiwara Momokawa to see whether he could be removed. Momokawa immediately reported this and Konin Tenno was enraged. The empress and prince Osabe were not tried, but were imprisoned at a rural location. After three years they both died on the same day, which is not likely to have been a coincidence, and immediately after another son, prince Yamabe, was made the official successor. Yamabe was an active member of the government already at the age of 37 and was minister of the department that managed the imperial palace. In the same year, 773, the government conducted a land redistribution.

The most notable event of the reign of Konin Tenno came near the end, in 880. An edict was issued that aimed at reducing the cost of the government considerably in order to relieve the burdens on the common people. Several offices outside of the official code that had been added over the years were abolished, and some regular offices were consolidated into a smaller number. Conscription for regular military service was abolished. To judge from the way that Hirotsugu's rebellion and the fighting that accompanied the fall of Fujiwara Nakamaro were handled, regular military forces had not been a serious factor for some time. There were supposed to be organized bodies of men available to the governor of each province, but there are a number of references that show that when forces were needed at the time of the overthrow of Fujiwara Nakamaro/Emi Oshikatsu, officials simply went out into the fields outside Nara and rounded up farmers and handed them weapons. Afterward these men made claims for compensation in the form of reduced tax liabilities, which were granted.

Kammu Tenno
Early in 781 Konin Tenno abdicated because of poor health and prince Yamabe ascended the throne. The retired emperor died at the end of the year. Early in 782 an armed man attempted to break into the palace, but was arrested. He supposedly claimed to have been operating on behalf of a son of prince Shioyaki. Kammu Tenno responded to this by a major purge of officials. Among the losers was Fujiwara Uona and his immediate relatives. This had the effect of permanently removing one of the four branches of the Fujiwara (descended from the four sons of Fuhito) from participation in high level politics. Kammu Tenno was an unusual emperor because his status within the imperial clan was quite low in conventional terms. Kammu had perhaps the most humble mother in the entire history of the imperial clan. She was not a princess, but a member of a minor aristocratic family of Korean origin. Kammu was thus not raised up in the expectation of elevated rank, and he did not behave like a normal prince or emperor. When Konin took the throne his first designated successor was Prince Osabe, whose mother was a daughter of Shomu. Only the disgrace and eventual murder of Osabe opened the way for Yamabe. There are many instances during his lifetime to show that Kammu Tenno was sensitive about this and was extremely suspicious and on the watch for signs that people were not giving him enough respect. This round of purges in 782 was only the first of several similar events.

Kammu, naturally, received many high status women into his palace. Three of his wives were Fujiwaras. Three of his sons became emperor. Two were grandsons of Fujiwara Yoshitsugu and the third was the grandson of Fujiwara Momokawa. He had many other wives and 30 children whose names are known. It was to be a long time before the dynasty had again to worry about a shortage of possible successors to the throne.

Upon taking the throne Kammu appointed no new high officials for a long time. Many posts were vacant, and he made it clear that he was going to be the day to day ruler, not merely the man who signed off on things. His enthronement proclamation made many references to Tenchi Tenno, who ruled the same way. Eventually he created a new office known as Kurando or Kurodo (both pronunciations are legal) which was a private secretariat outside of the administrative structure specified in the codes. It was a private office of the emperor and located inside his apartments in the palace. The chief of the Korodo became a very important official. This office took over some of the functions of the Dajokan, in particuler receiving and sending documents throughout the government. It was temporarily suppressed after his death, but was soon revived and served as a model, for during the course of the next few decades all of the major clans created offices of the same sort for themselves. Eventually large amounts of official business came to be conducted within private offices located in the mansions of the senior officials, offices which sent messages to each other and to the Kurodo, so the importance of the official ministries and bureaus declined. From its very start, the Heian period had a different political syle than the Nara period.

It may also be noted that the post of Dajodaijin was essentially abolished. It had been held by Dokyo, but now was left vacant for many generations. In later times it was claimed very rarely by men who held dictatorial power and were not content to rule with one of the normal titles. This post had originally had the purpose of ensuring that a member of the imperial clan had the highest ranking position in the government. From this time the participation of princes in the bureaucratic administration declined and soon ended completely. Already the Tachibana and Funya clans had been created to move a prince and his descendants from the imperial clan into the civil aristocracy, and this came to be repeated several times during the Heian period. Princes who were not immediately descended from an emperor (and thus plausible candidates for the succession) became irrelevant.

Almost the first event of the reign of Kammu Tenno was a census, and he applied pressure to make sure that everyone was counted, including people who were trying to hide, so that they could be forced to accept land and pay taxes. He demanded that officials refuse to accept "in kind" tax items of inferior quality and that fines were to be assessed for items turned in late. He was successful enough in these efforts to eventually provoke resistance. I am treating it as a separate topic, but when Kammu came to the throne the wars against the Emishi "barbarians" in the northeast were running hot and they consumed much of his attention and many resources. At the beginning of his reign he clearly thought that the men in charge under his father had not been performing well enough and there was a big shakeup. He decreed that the districts that were providing most of the soldiers were to be exempted from taxes for three years and appointed some very experienced men to take command.

Very soon after taking the throne Kammu began to investigate moving the capital. This should hardly be a surprise. The capital consisted of the emperor and the officials surrounding him and then a relatively small number of commoners who provided services to the aristocrats. It had been moved many times and would be moved again. However, by now, it had to have been clear that the government had become complicated enough that moving the capital was a major enterprise. I believe the ancient aristocratic tradition that when a man died his mansion was burned down as part of the funeral observations and his heir built a new one in a new location, not just across the street, was still a factor. This was probably not unrelated to the kind of consideration that kept most medieval European rulers constantly on the move from castle to castle so that they and their entourage did not totally impoverish the districts they stayed in. However, Kammu was almost the last to do it. There was one attempt to move the capital at the end of the 12th century which failed, Hideyoshi allegedly intended to move it to Beijing in China in the 1590s, and then it was moved from Kyoto to Tokyo in the nineteenth century. This again tells us that the reign of Kammu was at a point of transition from one mode to a new one.

Older histories often state that the move from Nara was to escape the excessive influence of the Buddhist temples. The writers were thinking about Dokyo. However, there is no reason at all to believe this. Most of the Nara temples were dismantled and re-erected in Hieiankyo, and many new ones were built. Some modern Japanese historians think that Kammu was motivated by rational concerns. Yamashiro province, where both of his new capitals were located, was better situated for communications with the east, where the wars were going on. Yamato province was the ancient heartland because it was big enough and rich enough to support many aristocrats, but relatively isolated because it was surrounded on all four sides by significant mountains, so it was easy to defend. Yamato means "within the mountains." However, supplies for the frontier wars were not being shipped out from the capital, they were all derived locally. The capital was a point of communications because messages went there, and Nara was not fundamentally inferior to Kyoto from that point of view. What made Yamashiro better than Yamato was the fact that by this time it was the richer province, with much more land suitable for rice farming. What made Kyoto the permanent capital was the evolution of ideas and society. The capital became more than just the fringes surrounding the palace. The emperors moved their palaces constantly, except that later they moved them within Kyoto. Many of the Heian period emperors' reign titles are actually the names of their palaces. The same applied to the senior aristocrats. When one died it often happened that his mansion was turned into a temple dedicated to his memory and his successor built somewhere else. Some mansion sites served for many generations, but often for men from different clans, for a man was more likely to inherit a palace from his mother than from his father. Men's palaces were often inherited by daughters.

Returning to Kammu Tenno and his ambition to move the capital, in 784 Fujiwara Tanitsugu recommended to the emperor that a place called Nagaoka should be selected. This was in the traditional territory of the Hata clan and Tanitsugu's mother was a Hata. It was located at the point where three rivers (the Uji, the Katsura, and the Kitsu) come together to form the Yodo river which flows to the Inland Sea through modern Osaka. The spot is slightly southwest of Kyoto, and also not very far from Naniwa, a frequent site of imperial palaces. The emperor accepted this choice, and assigned the tax revenues of a number of provinces to the project. Tanetsugu was put in charge of the construction office. Work began in the 6th month of 784 and the emperor moved to the site in the 11th month. The emperor personally appealed to wealthy persons to contribute. One man native to Omi province named Sururi no Ekimaro provided 30,060 man days of labor over a period of 8 months. This shows something of the scale of the wealth of provincial nobles.

In the 4th month of 785 Shoku Nihongi says that an official reported that a red sparrow had flown into the empress's palace. This was taken as a favorable omen and it was publicly recognized. This was immediately followed by a change in crown prince from prince Sawara to prince Ate. In the 8th month the emperor staged a hunting party, with prince Sawara, Fujiware Tanetsugu, and Fujiwara Korekimi participating. Late during the 23rd day of the 9th month Tanetsugu was shot by an unknown person. His body had two arrows in it, so it was unlikely to have been a hunting accident. An official investigation was started and several men were arrested. Saegi no Imaemishi, who had reported the auspicious sparrow, was reported to be one of the ringleaders. A member of one of the guards units was accused as the "hit man". The plan was allegedly to destroy the Fujiwara clan and bring the Saegi and Otomo clans to power, and the arrested men testified that prince Sawara was part of it. Prince Sawara was arrested and allegedly refused to eat. The emperor decided to exile him to Awaji (a province that was a small island near Osaka and a popular destination for exiles) but he died along the way. The officials in charge continued anyway and buried him on Awaji.

This event has attracted a lot of attention. Someone shot Tanetsugu and it was probably one of the guards. Several of them were arrested and tortured and the case grew out of that. The main result in the long term was that Kammu came to feel guilt about his hasty decision in respect to Prince Sawara. There is ample evidence that this bothered him, and it is alleged to be the main reason why he eventually abandoned Nagaoka for Kyoto, hoping that the ghost of prince Sawara might not be able to find him if he moved. There are numerous references to vengeful ghosts during this era. A person who was put to death unjustly was believed to be able to exact revenge. There is no doubt at all that Kammu took it seriously. There were many religious ceremonies intended to appease Sawara's ghost. Work on the Nagaoka capital continued despite the murder. 3,014,000 man days of conscript labor had been arranged and could not be diverted. Converted into actual men working 30 or 50 days this meant 10,000 to 16,000 men. This particular item proves that when numbers of men are stated for these projects man-days are meant, because 3 million men would be half the total population of the country including women, children, and old people, and we know from the analysis of census documents that only 25% of the population consisted of taxable adult males.

Shoku Nihongi records an edict of 788 on the construction project that says that many government offices that were supposed to move had done nothing about it. It also said that there were reports of popular unrest over the conscription of labor. Since there was no alternative to the corvee labor, the government decreed that the people were to be relieved by being forgiven a percentage of the interest due on the seed tax loans. In this general context we find a mention of an order to the governors of 7 provinces to provide labor for dismantling all of the gates of the Nara capital and reconstructing them at the site of the Nagaoka capital.

In 791 the crown prince fell seriously ill and diviners claimed that prince Sawara's ghost was responsible. Soon after there was a big earthquake. This seems to have tipped the emperor over to the idea of abandoning Nagaoka for a new site. The location was a site that had been an estate of Fujiwara Tanetsugu. Kammu spent four months there hunting and then decided that it would become the new capital. Nara still existed and most of the government probably still worked there, and workmen began immediately dismantling the buildings at Nagaoka to reuse the materials for Heiankyo.

The new capital site included 107 acres of farm land, and the emperor paid the affected farmers 3 years profit for the inconvenience. They would be allocated other lands elsewhere, but would have to buy land and build new houses and barns. He also tried to mollify the officials by offering all fifth rank and higher corvee laborers to help them move their residences.

794 saw both the second Mutsu expedition and all out work on the new capital going simultaneously. The new capital officially opened and the emperor moved into the palace in the 10th month. I have never found a discussion of the fact that the site of the Nagaoka capital was much smaller than Nara, but the new site of Heiankyo was larger than Nara (which was half empty). The basic layout of the new city was identical to the old one, both being modeled after the T’ang capital at Ch’ang-an. The palace is at the northern end facing south with a grand avenue proceeding all the way to the main city gate in the center of the southern wall. This avenue divided the city into left and right halves, each with its own mayor and market. The leading nobles were awarded large plots for their mansions, which they had to build at their own expense. The basic layout of the main streets survives today, though the palace and public buildings are all long gone. Only a handful of structures survive from the Heian period due to numerous fires, plus a few wars. The present imperial palace is in a different location and is much smaller than the original. It was set up in its current location in the 14th century and the present buildings date only to 1855 thanks to one of the fires. They are quite nice in an ultra-modern style (for 1855) that influenced European art deco.

The city is surrounded to the west, north and east by mountains. The Kamo River flows south. The present course of the river is east of the old city, but it is believed that originally the river followed Horikawa street in the “left” city a short distance east of the palace. The present river course was formerly the Takano River. The Kamo was diverted to merge with this river north of the city as part of the construction project. Along the west side of the city is the Katsura River. The Kamo and Katsura rivers merge just south of the city and almost immediately flow into the Yodo River (very close to the Nagaoka site). As a result of this the southwest section of the city was frequently flooded. The entire “right” half of the city remained mostly undeveloped in ancient times. On the other hand, the city early expanded to the east across the Kamo River into the Higashiyama (eastern mountain) district, now the touristic center of the city with numerous shrines and temples and the national museum. In the later Heian period the western part of the city was the slums, where gangsters and beggars lived. This shows up in anecdotes over the next several hundred years. Eventually, in the 14th century, it became fashionable in its turn and numerous buildings from the Ashikaga period remain there, including most of the famous zen temples. The south is the zone that never moved up. Poor people lived there in the 9th century and they still live there today.

It has been argued that between the construction projects and his wars Kammu needed to maximize revenues and tolerated many deviations from the administrative rules as long as they worked. The Kurodo office is but one example. This is alleged to have begun the trend that eventually led to the abandonment of all salaried offices and the absorption of such public functions as were still performed into the household offices of the leading nobles. It would be a mistake, however, to exaggerate the speed of this transformation. Kammu and several of his successors continued to fight to preserve the power of the central government to control things, primarily through the provincial governors.

The law codes include a rule that a governor at the end of his term could not leave his province until his books had passed an audit by his successor. This law had been allowed to lapse at some point during the reign of Shomu. Kammu revived it. He ruled that the incoming governor had to complete the audit within 120 days or his salary would be stopped. Kammu also took official notice of the fact that warehouses full of official rice burned down at a suspiciously high rate compared to privately owned warehouses. This was always attributed to lightning. Kammu first tried increasing the penalty for arson, and when this didn't work, changed the system so that the farmers were to keep the unused tax rice and some corvee labor would be allocated to build and maintain the storage facilities they would need. This indicates the Kammu believed that most fires were either fictitious or set to cover up embezzlement by the governor.

Kammu also enforced rules that prevented a governor from retiring in a province he had formerly ruled. Even aristocrats were registered and had an official home and retired officials were required to go there and to sign in with the local government when they did so. District magistrates were different because they worked only in their native district. In the system up to Kammu's reign there was a custom/law that said that only members of the ancient kuninomiyatsuko families could be magistrates. Kammu disregarded this and tried to appoint people who had shown signs of ability.

At one point Kammu sought to revive the institution of circuit inspectors who were supposed to travel a standard route checking out how effective the provincial and especially the district officials were. However, this was abandoned before it was implemented, and instead he carried out a nationwide survey in 799. He sent out officials whose mandate was to interview ordinary subjects to find out what they thought. Emperor Tenchi had done something similar. We have no details of the results except that it led to an edict prohibiting provincial governors from hiring armed guards.

Over the years Kammu paid a lot of attention to "vagrants", which is to say people who managed somehow to live without getting registered as taxpayers or participating in the land redistributions. In 783 there was a campaign in 8 eastern provinces in which local men who held aristocratic rank but no office were ordered to round up "vagrants" and then lead them to join the northern war as soldiers. The Dajokan estimated that 52,800 men could be found that way and took steps to assemble equipment and supplies for such a number. Late in the reign of Konin there was a special study of "vagrants", based on a close look at Ise province. The provincial governor reported that many people evaded the labor tax by hiding and other taxes by falsely claiming death. Also, there were people who, when the labor tax levy was being conducted, falsely claimed to be slaves. In 797 Kammu decreed that any one who had one parent who was free was free and subject to all taxes. The special study in Ise had identified about 1,000 tax evaders. Other people evaded taxes by leaving public lands and moving to estates owned by temples or aristocrats. There are also surviving records that seem to indicate that people who were unable to repay the suiko loans resorted to "disappearance." In an edict of 785 Kammu observed that when the time comes to distribute land families are large, but when the time comes to pay tax they are small. He was also aware that people who had disappeared from the registers were not necessarily hiding, they may have moved to a different area. Officials were urged to make sure to capture new arrivals and register them properly. A commoner who settled on an estate didn't own the land and was not subject to crop tax. However, he was still subject to the head tax and the labor tax. This was a major area of conflict over time with the owners of estates as they tried to hide their people from the tax collectors so they could use the labor themselves.

In 797 an edict of the Dajokan noted that a great many people had been moving into estates controlled by princes and nobles in order to escape taxes. It ruled that once a year the provincial authorities could enter such domains searching for absconders, and if they found them, could put them back on the rolls of taxpayers as far as the corvee and head taxes were concerned. They needed to first prepare a list of persons they were looking for. Any estate manager who resisted was to be severely punished. The problem here was that the officials had to have the names of the people they were looking for. Someone who had moved from a different province would be safe because the local officials would not have his name available for their lists.

The conducting of censuses and redistributions had become sporadic. It was supposed to happen every six years, but in fact there were distributions in 742, 755 (12 years later), and 773 (18 years later). This means that three out of six cycles were skipped. During Kammu's reign there were three censuses and two redistributions. In 801 he ordered that the cycle be set at 12 years since it had proved impossible to keep to the 6 year cycle. In 800, by the way, it was mentioned that a complete redistribution was conducted for the first time in Satsuma and Osumi provinces at the southern end of Kyushu. Formerly these had been "barbarian" terrorities only loosely controlled by the Japanese government.

In 802 there is the first recorded popular petition to the government. There are earlier mentions of petitions, but this is the first with a text. “When we assemble to prepare the taxes in kind, the officials force us to work as laborers to take everything to the warehouses. Everyone, male and female, old and young, is made to do this work. Because this happens during the 6th month in between harvests, there is not much food and yet people are forced to work hard. The work takes 10 days. We think that we should be fed while doing it.”

Through 804 Kammu remained vigorous as shown by the fact that he spent a lot of time hunting. However, at the end of the year he fell ill and never completely recovered. In the 4th month of 805 he assigned administrative responsibilities to the crown prince for the first time. He died in the 3rd month of 806 at the age of 70. Prince Ate was enthroned one month later without incident. The next day his younger brother Kamino was named crown prince.