倭人考 6On the Wajin, Part 6

The image of antiquity inevitably fades through the accumulation of scribal errors and misprints across history. This has been an inescapable problem ever since the idea of committing records to writing was born. At the same time it is an important path toward understanding the deeper strata. The question of where one begins to discern those strata therefore becomes critically important.

On the Chinese side, where writing and documentary techniques were acquired early, the imperative of knowing one's neighbors grew steadily in importance as a means of maintaining a stable regime anchored in the dynasty, given the reality of a vast territory open on all sides. Out of this imperative the concept of the Four Barbarians eventually emerged, and the world of ancient Japan came to be understood as one of the peoples of the Eastern Barbarians. The earliest documentary source for this understanding is the text now known as the Gishiwajinden. Yet if one attempts to gain insight into the Wa people of the Wa kingdom as the true image of ancient Japan, or into Himiko the queen of Yamatai, a single lifetime will not suffice. The historical confusion involved is so intense and so deeply entangled with the broader history of East Asia that no single interpretation or definitive answer can be clearly offered. Still, with that impossibility as a premise, the journey of cutting through layer upon layer of extraordinary complexity to bring the image of ancient Japan to the surface can be continued on one's own terms.

The subject of this piece is the relationship between the Wei envoys who visited the Japanese archipelago in the 240s and Himiko. In fact, even the description of Himiko as queen of Yamatai conceals a number of premises that must be questioned. Let me begin with the dimension of sound.

As I suggested at the end of the previous piece, the current practice of reading [邪馬台国] as Yamatai is a convention that became fixed over time through the accumulation of scribal errors, misprints, and the upheavals of successive Chinese dynasties that held the source documents. The important point is that reading it as Yamatai is wrong. The correct modern reading is Yamato. But it is worth examining how the discrepancy between the sounds Yamatai and Yamato arose.

The Gishiwajinden was transmitted over the centuries as a section of the Records of the Three Kingdoms. This work was largely completed around the 280s by Chen Shou (233–297), a bureaucrat in the service of the Western Jin, but because Chen Shou died in 297 the text we have is based on copies made by officials sent to his home around the year 300 under imperial edict. This was the first step by which the Records of the Three Kingdoms as we now read them were transmitted to the world, though at that stage the text served primarily as a historical source and was not yet widely read. Roughly a hundred years later, in 429, the politician and historian Pei Songzhi (372–451) took one of the transmitted manuscripts of Chen Shou's work as his base text and produced an annotated edition. The reinforcement provided by Pei Songzhi's commentary earned high praise, and it was at this point that the Records of the Three Kingdoms entered an era of broad readership.

From here it can be assumed that countless manuscript copies were produced, though the details of that transmission are entirely unknown. When the era of the Northern Song dynasty (960–1127) arrived, however, the documentary history of the Records of the Three Kingdoms entered a new phase. That phase was the introduction of printing technology. Full-scale printing technology was, according to the standard account, invented by Bi Sheng (972–1051) of the Northern Song, and is regarded as the beginning of movable-type printing. Its peak came in the first half to the middle of the eleventh century. Yet already around 1002 to 1003 a printed block edition of the Records of the Three Kingdoms had been produced under the editorial supervision of the Northern Song's central educational administration. This development had its roots in the fact that toward the end of the preceding Tang dynasty (618–907) a period of upheaval had caused the dispersal of countless transmitted texts. Beginning around 932 the newly established educational administration undertook a project spanning roughly twenty years to retransmit those works. This momentum fed into the emergence of printing technology, and by the early eleventh century the Records of the Three Kingdoms had been printed as well.

From this point the text diverged into multiple printed editions as Chinese dynasties succeeded one another. The details are complex and I will not go into them here, but for the purposes of the historical study of the Gishiwajinden that is the focus of this series, approximately five editions are used. The earliest Northern Song editions have been lost. The oldest surviving editions date from around the middle of the twelfth century. Research proceeds on the basis of the various editorial lineages printed between that point and the early eighteenth century, but in the long process from manuscript to printed edition, discrepancies arising from unmanageable scribal errors and omissions have accumulated. The situation is extremely tangled, so let me proceed carefully by comparing the notations and surveying the overall picture.

First, in the printed editions of the Records of the Three Kingdoms described above, the name of Himiko's kingdom appears as [邪馬壹国] and [邪馬一国]. The character [壹] is now written [壱], but in older texts it was written [壹]. To establish the baseline readings, [邪馬台国] is read Yamataikoku, with [邪馬] as Yama, [台] as tai, and [国] as koku (since koku simply means kingdom I will set it aside). In the printed editions, however, the forms are [邪馬壹] and [邪馬一]. Both [壹] and [一] are read ichi, so the result is not Yamatai but Yamaichi. The reading Yamaichi is itself incorrect, and the deeper problem is that on the character level every printed edition from the very start contains the error. This is presumably the result of a misprint in the base text that the Northern Song printed in the early eleventh century, though it is also possible that the base text itself already carried this notation. In other words, the erroneous reading Yamaichi had been perpetuated through misprints and scribal errors over a very long span of time.

The reason such an error could occur lies in the confusingly similar character forms of ancient Chinese. The original notation was [邪馬臺国]. At a glance the difference is almost invisible, but place [邪馬臺] and [邪馬壹] side by side and the divergence becomes clear. It is the difference between [臺] and [壹]. This was no doubt the cumulative result of the long manuscript era compounded by linguistic shifts accompanying changes of dynasty, but the evidence that the post-Northern Song printed editions are wrong is straightforward. Every edition of a text written between the early fourth century and the late tenth century, the manuscript era of the Records of the Three Kingdoms, uses [邪馬臺]. For example, the Hou Han Shu (Book of the Later Han), composed in the first half of the fifth century, writes [邪馬臺] and not [邪馬壹]. The historical period this work covers is older than the Wei era in which the Gishiwajinden is set, but the text itself was completed after the Records of the Three Kingdoms. This means that the version of the Records of the Three Kingdoms consulted by the authors of the Hou Han Shu contained the notation [邪馬臺].

The Liang Shu (Book of Liang), completed in 629 and recording the history of the Liang dynasty of the Southern Dynasties (502–557), also uses [邪馬臺]. Surveying this evidence, it becomes clear that at least through the first half of the sixth century the notation [邪馬臺] was transmitted through Chinese manuscripts as the correct form. One more source is indispensable for investigating this problem. That is the Sui Shu (Book of Sui), completed around the mid-seventh century. The Sui Shu contains a passage recording the location of the Wa kingdom's capital.

The original reads [都於邪靡堆、則魏志所謂邪馬臺者也]. This states that the capital is at [邪靡堆] and that this is the place called [邪馬臺] in the old Gishiwajinden. For those unfamiliar with Chinese characters this may be cumbersome, but [靡] in [邪靡堆] is interchangeable with [摩], and [堆] can be understood as phonetically close to [推]. This connects to the fact that when annotations on the Wa kingdom were added to the Hou Han Shu around 676 to 680 the form had shifted to [邪摩推], and the Tong Dian, compiled under the Tang between 766 and 801, also uses [邪摩推]. What the Sui Shu passage demonstrates clearly is that in the manuscripts of the Gishiwajinden circulating in the early seventh century, the notation [邪馬臺] had been correctly preserved.

Various manuscripts and editions continued to be produced in various places after the seventh century, and some do preserve the older accurate notation. Setting aside the full complexity of these discrepancies, what we can grasp at a minimum is that the original text of the Records of the Three Kingdoms wrote [邪馬臺]. The earliest printed edition by the Northern Song, however, already had [邪馬壹].

We can now establish four basic variants of the term surrounding the location of Queen Himiko: [邪馬臺], [邪馬壹], [邪靡堆], and [邪摩推]. These serve as the raw material for tracing the mutation toward [邪馬台], the Yamatai in general use today. The next question is how to read them.

Since ancient sounds cannot be recorded, it is of course impossible for us to know them precisely. Throughout this series I have written that the phonetic root behind terms such as Wajin [倭人] and Wakoku [倭国] was something like Wan, but even that is only what the sound amounts to at the level of modern understanding. How the Wa people actually pronounced it is unknown. That said, regarding the sound of this ancient Japanese capital, the characters [邪馬臺] are estimated to have been pronounced in the Chinese of that era as something in the vicinity of <hjɐg-mɐg-dəg> or <iɐ-mɐ-d'əi>. Rendered in notation this becomes something like <ia-ma-deo>, but the actual sound it represents is <yamadö>. The <yama> component is immediately recognizable as the Japanese word for mountain. The problem is <dö>. The explanation I find most convincing is that <dö> belongs to the category of voiced sounds in Japanese, which in ancient speech could arise as a muddied sound following a nasal syllable, among other reasons. This corresponds to the <tö> reflected in later Sui and Tang-era sources that rendered the final syllable as [堆], read as <tuəi>. In other words, the location of Himiko was Yamato.

The reason Yamato became Yamatai is thought to lie in phonological changes in the Chinese language during the Tang dynasty. This falls within the domain of Chinese historical phonology and is beyond my competence to treat in detail, but the people of the Tang era may have perceived [邪馬臺] as something close to <yama-ta> while reading [邪靡堆] and [邪摩推] as <yama-tö>. The reason for this level of complexity is that the sounds of ancient Japanese were being approximated using Chinese characters. To consolidate the point, the capital function of the Wa kingdom where Himiko was based was not Yamatai but Yamato. A place or country called Yamatai never existed in Japan. And this Yamato later came to be written on the Japanese side as [大和], a designation referring to the southern Nara Basin region. Himiko once lived in that land.

That yama is the Japanese word for mountain is self-evident. But what tö represented becomes a new question. This too leads into complications. The long-standing debate within Japan between the Kinai theory (placing Himiko's kingdom in southern Nara Prefecture) and the Kyushu theory arose precisely because the analysis was conducted with these historical layers in a state of confusion. What it actually indicates is that the leading faction that would later establish the Yamato kingdom in Nara was originally based in ancient Kyushu. The evidence is still inscribed in the place names of Kyushu. In what is now the southern part of Fukuoka Prefecture there was once a district written [山門]. It is read Yamato. Further south, in the mountainous eastern part of Kumamoto Prefecture, there is a town still written [山都] that is also read Yamato, though this name was adopted in the early 2000s. The origin lies in the Fukuoka-side [山門] read as Yamato.

From this notation a straightforward inference can be drawn. For the ancient Japanese the primordial sound Yamato referred to land that opened at the gate of the mountains. This is because [山] means mountain and [門] means gate. This would seem to be a word belonging not to the Jomon people, who balanced settlement with hunting and gathering, but rather to the Yayoi and later populations whose way of life was grounded in agriculture. Yet even here caution is needed. I am inclined to think the ancient form of Yamato may in fact derive from the Jomon, and that it originated not as a word of farming peoples but as a word of maritime tribes. The maritime peoples of ancient Japan used a technique called mountain sighting, to determine their position at sea. In keeping with the island geography of Japan they used mountains as landmarks to locate fishing grounds and other sites. I believe these traces came to be identified with [山門], the prototype of Yamato. Later, when migrant groups arrived carrying agricultural technology, the primordial meaning would have shifted closer to land suitable for farming. Alternatively, a tribe that had once inhabited land called by a sound close to Yamato by certain Jomon groups may have been the original core of the force that ultimately conquered the southern Nara Basin and established [大和].

The scholar who made the most important observations about this elusive and invisible history was the folklorist Orikuchi Shinobu. He argued that the reason all of Japan eventually came to be called Yamato was that the [山門] faction originally based in southern Fukuoka Prefecture carried the name with them when they undertook a great migration eastward at some point in history. Some readers may already see where this leads. The myth that transmits the history of this eastward tribal migration is the Eastern Expedition of Jinmu. The Jinmu expedition begins in Kyushu, advances eastward, and culminates in victory at what is now the city of Kashihara in Nara Prefecture, in the southeastern part of the Nara Basin, where the first emperor is enthroned. The name of the tribe's place of origin traveled with them, and over time the southern Nara Basin itself came to be called Yamato.

The debate over Himiko's location as it is generally understood is a distortion produced by shifts within the complex and comprehensive movements of East Asian history. No one can treat the subject in its totality. But what can be understood at a minimum from this article is that Himiko was not the queen of Yamatai. She was the queen of Yamato. Please take this as the working premise going forward. In the next piece I will build on everything presented so far and examine the possibility that the Wei envoys may never have had an audience with Himiko at all. This will suggest how powerfully the circumstances of China during the chaotic Three Kingdoms period influenced the whole of East Asia, and at the same time it will bring into view the extent to which the image of Japan recorded in the Gishiwajinden failed to reflect the actual reality.