(Kaogu (Archaeology), vol.5, p.439 558, 1984; translated & interpreted by W. Tsao, Ph.D.; edited by Bryan Gordon)
At the 31st International Symposium of Social Science for Asia and North Africa (Seminar A-2), held in September 5, 1983 at Kyoto, Japan, I presented the topic On the Hemudu Culture. Besides introducing the contents and discovery of the Hemudu culture, I also pointed out in the conclusion: "The elements of the Neolithic culture discovered in Hemudu and the lower Yangtze River that followed could have influences on ancient Japan. The origins of jue (a jade ring with a small segment cut off), lacquerware, rice seedlings in the (Japanese) Shengwen Period and "balustrade style" construction in the (Japanese) Misheng Period are traceable to the lower Yangtze River. Wooden oars and clay boat models were discovered in Hemudu sites and at the same time, similar sites were excavated in the Zhoushan Islands off the coast of Zhejiang Province. This at least proved that seafaring existed at that time. From existing documents, voyages between China and Japan existed in the Han-Wei period (ca. 200 AD), but it is unclear if voyages occurred before it. However, during the Han-Wei period, "balustrade style" construction with long roof ridge and short eave had already disappeared on the lower Yangtze River, so international amenities between the two countries began earlier. Jue, lacquer and rice seedling in the Japanese Shengwen Period apparently had close ties with lower Yangtze River Neolithic culture, making it worthwhile to pursue research on Chinese and Japanese cultural ties as well as the time and route of voyages between the two countries. In any case, we believe Chinese and Japanese ties began much earlier than the Han-Wei period suggested previously." Several Japanese newpapers quoted my talk and reported that Sino-Japanese communication may have begun 5000 years ago. Japanese archaeologists were also interested in the suggestion that rice cultivation was directly imported into Japan by sea-route(1).
As the main subject in my seminar talk was Hemudu culture, I only mentioned briefly early ties between Chinese and Japanese cultures. Although I gave further explanations in the question period, there was no time for in-depth answers. For this reason, I have compared the points at issue from Japanese archaeological research to Chinese archaeological data and made the following additional discussions.
The question whether or not agriculture existed in Shengwen culture in the Japanese Neolithic has been a point of argument for more than 50 years. After World War II, the suggestion of agriculture in Mid- and Late Shengwen stages and Zhaoyelin culture(2) met objections(3). Pre-Shengwen cultivation that was suggested recently(4) implied agriculture started in Japan earlier than suggested. According to the usual classification, the 8000-1000 BC Shengwen period comprises 5 stages: Early, Pre-, Mid-, Post and Late, with Pre-stage starting ca. 5000 BC. As questions of Shengwen agricultural origin are complicated, the following introduces basic content and understanding of various theories.
Supposed cultivated crops recognized at the Pre-stage of Shengwen culture are gourds, green peas, chu orange and lacquer trees, ren oil beans, Zisu (Perilla nankinensis), etc. These crops were suggested to have been imported into Japan from south China via sea. Some even suggested gourd and green pea remains excavated in Wubinbei Tomb in Fujing County were carried to Japan by sea-currents(5). However, there are two questions to be answered. From a botanical viewpoint, it was confirmed that the aforementioned crops were not Japanese local plants. How were they imported? Were they taken from the natural environment or artificially cultivated? There are currently no satisfactory answers. It was suggested that "these plants needed no cultivation and most of them could be harvested when accidentally planted near habitations. Thus, the origin of agriculture may be traced all the way to the Pre-Shengwen stage, which coincided with the beginning of this crop cultivation (in Japan)"(6). This appeared an unlikely solid ground to theorize the origin of agriculture.
Advocates of Mid-Shengwen stage agriculture suggested 18 points(7) to support the theory, but they are inconclusive archaeologically. The growing belt of chestnut trees was considered to coincide with the area where Shengwen culture occurred. Chestnuts, acorns and walnuts are considered cultivated crops because their remains occur in ancient sites. In fact, these nuts can be collected from trees growing under natural conditions and uncultivated. Supposed carbonized chestnuts in the Shenshan sites near Zhongye County provide new evidence for Mid-Shengwen stage agricultural theory. But chestnut remains were later identified using electron microscopy as seeds of ren oil beans and Zisu plants. As no remains of millet, sorghum or other food crops were found, it is therefore too early to prove that cultivation of grain crops existed in the Mid-stage of Shengwen culture.
The discovery of carbonized rice grain, pottery with marks of rice plants and grain, and rice pollen indicate agriculture may have begun in Late Shengwen stage, but there was a delicate connection between the Late Shengwen stage and Misheng culture. Banfu style pottery in Fukuoka sites and classed as Early Misheng artifacts co-exist with Yejiu style pottery of the Late Misheng stage(9). Based on existing archaeological data, Japanese rice cultivation began in Kyushu and then spread east and north. In general, fishing and hunting were major occupations in Shengwen culture and Late stage agricultural practice was limited to areas of Kyushu and minimal elsewhere. Nevertheless, carbonized chestnuts(10) and barley and some pollen of buckwheat varieties in Late Shengwen sites show agriculture likely started in this period.
A definite trait ruled by the natural environment and geographic conditions occurred in Japanese Neolithic Shengwen culture. On one hand, the Japanese islands were rich in natural resources, with fishing and hunting easily supporting daily life. Alternately, they were very isolated from outside influence which limited agricultural development. From available archaeological data, Japanese agriculture may have started in Late Shengwen stage, as there is insufficient data in Pre- and Mid-Shengwen-stages. As proven by archaeology, fishing and hunting were major occupations of Shengwen culture, but plant resources accounted for more than half the economy(11). Collected and stored natural chestnuts, acorns and walnuts exist without cultivation, with chu orange tree fruit used under similar conditions, but ren, Zisu, gourd and green pea did not grow in the wild. There are still no answers on how they were imported and cultivated(12), but gourds at Hemudu(13) and Luojiajiao(14) coincide with their cultivation in Japan, which elicits some interesting questions. Nevertheless, rice cultivation was the major occupation in Misheng culture(15) but not in Shengwen culture. The change from fishing and hunting to agriculture covers not only the import of rice cultivation, but also metal tools allowing large scale forest exploitation and rice paddy development, a situation which stimulated further development of productive resources.
China is a centre of agricultural origin with great Asian influence. Neolithic finds show drought-resistant millet and sorghum were major crops in the yellow loam area of the Yellow River basin and recognized as two of the earliest cultivated crops in China. As seen in Neolithic sites(16), rice cultivation began earlier in Qinlin and the large south area of the Huai River due to warm temperature, humid environment and many swamps, so it is obvious prehistoric Chinese crops depended totally on geographic conditions. Later, increased historic human ability to alter the environment and improve irrigation made it possible to shift rice planting north, but it was still limited under certain conditions and not as popular as other crops.
Neolithic rice remains so far excavated are carbonized grains, husks and straws in red burnt clay, grain marks on sherds and mixed husk and straw in pottery. They occur mainly in Qinlin and the large south area of the Huai River(17). There are 12 sites on the lower Yangtze River: Yuyao Hemudu, Tongxiang Luojiajiao, Hangzhou Shuitianban and Wuxing Qianshanyang in Zhejiang Province; Wuxi Xiannidun, Nanjing Miaoshan, Wuxian Caoxieshan, Shanghai Qingpu and Maqiao in Jiangsu Province; and Qianshan Xuejiagang, Guzheng Haocheng and Feidong Dachendun in Anhui Province. They belong to Hemudu, Majiabib, Songze and Liangzhu cultures and show rice cultivation as the major occupation there in different Neolithic stages. There are also 12 sites on the middle Yangtze River: Jingshan Qujialing and Zhujiaju, Tianmen Shijiahe, Wuchang Fangyingtai, Yidu Honghuatao, Zhijiang Guanmiaoshan, Jiangling Maojiashan and Yunxian Qinglungquan in Hubei Province; Zhechuan Huanglianshu and Xiawanggang in Henan Province; Xiushui Paomaling in Jiangxi Province; and Lixian Mengxi in Hunan Province all belonging to Yangshao, Daxi, Quijialing and Tonglongshan cultures. It is interesting that environmental conditions of the Xiawanggang site of Yangshao culture were like those of the Yangtze and Han Rivers. Rice cultivation was thus the major agricultural practice in this area, completely unlike drought-resistant millet and sorghum on the Qinling and Huai Rivers. Four sites are in the southeast area: Qujiang Shixia and Niling in Guangdong Province; Yongchun Jiudoushan in Fujian Province; and Taichung Yingpuli in Taiwan, with most Late Neolithic or later. Four more sites are further southwest: Yuanmou Dadunzi, Binchuan Baiyangcun, Jianchuan Haimenkou and Dianchi Guandu in Yunnan Province, mainly between Neolithic and Bronze Age. More than 30 of the aforementioned 4500-1000 BC sites cultivated rice, including widely distributed Xia (indica) and Keng (japonica) varieties. Data show Xia or long-grain indica was the first major variety, with increasing later proportions of Keng or short-grain japonica. These finds show rice cultivation in this area and the lower and middle Yangtze River began as Neolithic centres of origin.
Starting in the 5000 BC Neolithic, millet and sorghum were major crops in the yellow loam area centered in the Yellow River basin. More than 20 sites with their crop remains occur(21). While traces of rice cultivation occur, evidence is not overwhelming, suggesting ancient Yellow River rice cultivation was not as popular as millet or sorghum.
In 1921, a potsherd with grain marks was excavated in Yangshao village in Henan Province(22), the grain identified as rice by morphology and other scientific tests(23). As its discovery was unpublished, doubts arose(24), but from recent publications this sherd was oxidized clay and the grain was further confirmed as rice(25), but its dating is unknown. I was fortunate to have the chance to see it when I visited Japan in 1980(26). It was already broken in two, 9.5 x 6.6 x 1.8 (thick) cm for the larger piece and 5.5 x 3.9 x 1.6 (thick) cm for the smaller. The surface of the complete concave sherd of radii 7.58.5 cm was rough and coarse like a piece of clay from a wood-burning stove. As other specific traits are absent and remains of different ages occur in Yangshao village, it is difficult to confirm it as Yangshao culture, so we retain doubt on rice cultivation in this area.
Traces of supposed rice are in Neolithic sites in northern China, such as Huaxian Quanhucun and Huxian Zhangbasi in Shenxi Province; Zhengzhou Dahecun in Henan Province, and Xixia in Shandong Province. But most are only identified morphologically and lack photos or botanical examination. Yangshao culture bei pei (unfired mould for making cups) at Luoyang Xigaoya in Henan Province supposedly was "made of grass and mud marked with rice grain"(28), but its published photo is so unclear that someone suggested the marks are from gu mi (type of wild rice)(29). Mark confirmation await further identification by other scientific methods(30), but the fact remains that rice was cultivated in and after Shang and Zhou Dynasties, with excavated remains discussed in many papers. But, limitations in natural environment and climate make it unlikely rice was cultivated throughout northern China. With advanced agricultural technology, rice cultivation just started to move northeast from northern China, but did not replace local popular crops.
In China, the Yangtze basin is the origin of rice cultivation with the highest density of rice sites. From an archaeological perspective, it was where rice cultivation originated and greatly influenced other Asian areas. That Japanese rice cultivation began in China is a general concept, and there are three hypothetical import routes via north, central and south China(31). Generally, archaeologists tend to accept north China via Hebei and Liaoning Provinces or by sea via Shandong Province and the Korean Peninsula to Japan, while agriculturists and folklorists favor routes from central and south China. While data available is limited for the north China route, this area is not a centre of rice cultivation. Further, millet and sorghum were early major crops in the north Korean Peninsula while rice remains in the Wuwen Pottery culture of the south Peninsula are scarce(32). Recent statistical data shows rice remains in 16 Yusanguo period and 9 Wuwen Pottery sites, much less than the 13 Shengwen and 416 Misheng sites(34). We cannot judge the order of occurrence of these cultures on site numbers, but the many Misheng sites are significant. Rice remains in the Yanxinli site near Seoul, South Korea, according to a Japanese report, date 1030± 70 BC and 970± 70 BC by C-14 (in the Korean report, they are 1260± 70 BC and 670± 70 BC)(35), which is 200-300 years earlier than Japanese rice. But, upper level 8 remains in the Caidian site, Tangjin City, Japan, date 1010± 90 BC (Pre-Misheng pottery) and lower level 8 dates 1280± 100 BC (Late Shengwen pottery)(36). Despite C-14 uncertainty, we believe the origin of rice cultivation began about the same time in Korea and Japan, i.e., ca.1000 BC. This situation may be the reason rice cultivation was brought from the lower Yangtze centre to both Korea and Japan by sea simultaneously (Fig. 1). Besides, jue lacquerware and balustrade style construction in the following discussion are central Chinese traits. Archaeologically, this may be added data to support a sea route for rice from China to both Korea and Japan. Archaeological evidence is not as strong for the same route from south China as that from central China.
Fig.1. Map of prehistoric rice remain distribution in China, Korea and Japan in the Wuwen Pottery period in Korea and Pre-Misheng stage in Japan
·Confirmed sites with rice remains
o Unconfirmed sites with rice remains
Arrow represents sea route for importing rice cultivation
and lacquerware doubtless are two very valuable artifacts when discussing ties between Chinese and Japanese cultures based on cultural remains.
Stone jue is a representative Pre-Shengwen artifact, some believing it unique to the Japanese Neolithic(37), but its Zhou Dynasty jade jue shape suggest a Chinese origin(38) not earlier than Late Zhou(39).
Although Chinese jue was popular in Zhou Dynasty, its origin must be earlier, as archaeological data show jue in the middle and lower Yangtze River and south China Neolithic. 141 specimens are in Hemudu (11 specimens)(40), Jiaxing Majiabing (2)(41) and Wuxing Qiucheng (1)(42) in Zhejiang Province; Qingpu Songze (1)(43) in Shanghai; Suzhou Yuecheng (1)(44), Wuxian Caoxieshan (5)(45), Changzhou Yudun (9)(46), Wujiang Meiyan (6)(47) and Wujin Panjiatang (1)(48) in Jiangsu Province; Beiying Yangyin (37)(49) in Nanjing; Wushan Daxi (43)(50) in Sichuan Province; Yicang Qingshuitan (1)(51) in Hubei Province; Qujian Shanxia (17)(52), Qingyuan Xishan (1)(53) and Shenjiang Donghaibing (2)(54) in Guangdong Province; Taidong Donghe (2) and Taipei Yuanshan (1)(55) in Taiwan. But only 1 Neolithic specimen was found in the Yellow River basin, from the Longsnah site in Mengjing, Henan(56), but many dating to the Shang-Zhou period are in this area. As it can be confirmed that jade jue first appear on the middle and lower Yangtze, it is reasonable to suggest that Pre-Shengwen jue was influenced from the Yangtze basin rather than from the Shang-Zhou period (see Fig. 2 for various jue excavated in China and Japan).
Fig. 2. Jade and stone jue excavated in China and Japan
(2) From Yuyao Hemudu, Zhejiang, China
(3) From Qingpu Songze, Shanghai, China
(4) (9) From Wushan Daxi, Sichuan, China
(10) From Suzhou Yuecheng, Jiangsu, China
(11) From Taidong Donghe, Taiwan
(12) From Taipei Yuanshan, Taiwan
(13) From Fukuoma Beishaoye, Japan
(14) From Fukuoma, Japan
(15) From Henei Guofu, Japan
(16) From Qiutien Beifu, Japan
(17) From Shanxin Chuipu, Japan
Lacquerware is a unique Asian product, with China generally considered its origin. Its oldest is a lacquer bowl with wood inner mold excavated in Hemudu (Fig. 3)(57), but remains of two lacquer artifacts are in the Yyudun site of Majiabing culture(58). That lacquerware was also very popular in Shang-Zhou Dynasties suggests a very long history. Most Japanese lacquerware, including bamboo or wood inner mold (Fig.4) and lacquer bow(59) began in Late Shengwen culture, but some excavated lacquer combs date to Early Shengwen(60). Japanese lacquerware origins appear closely tied to the Yangtze basin.
Fig. 3, Lacquer bowl with wooden Fig. 4, Japanese lacquerware with inner mold excavated in Hemudu basket inner mold (Late Shengwen-stage)
Many archaeological finds confirm balustrade as an aboriginal style of construction in the Yangtze basin and its southern areas(61). Besides wood balustrade posts and hooks in the Hemudu and Luojiajiao sites, similar construction occurs at Wuxin Qianshanyang in Shejiang Province and in Neolithic sites at Danyang Xiangcaohe and Wujiang Meiyyan in Jiagsu Province. Even in sites as late as Bronze Age, such construction occurs at Jianchuan Haimenkou in Yunnan Province, Chechun Maojiazui in Hubei Province and Gaoyao Maogang in Guangdong Province(63), but it is limited to incomplete posts and hooks impossible to restore. But the clay model house at Qingjiang Yingpanli in Jiangxi, brass model houses (Fig. 5) at Shizaishan and the brass house-shaped coffin(64) at Xiangyun Dabona in Yunnan have a balustrade style long roof ridge and short eave. Han Dynasty clay balustrade models are in Guangdong, Guangxi, Sichuan, Guixhou and Hunan. Besides their foundation posts, the modified roof is typically Chinese, the former clay and brass models obviously representing original balustrade. Excavated Japanese engraved tong duo (brass bell with clapper) and tong jing (brass mirror) have balustrade foundation posts, long roof ridges and short eaves (Fig. 6)(65). Balustrade construction engraved on the Misheng sherd from the ancient Nara site, Japan, resembles the basic tong duo style(66). The above data of Japanese balustrade construction since the Misheng period and its depiction on coffins and graves(67) show a close tie with the Yangtze basin. As the Han dynasty clay house model balustrade roof was no longer long ridged and short eaved, Chinese influence on Japanese construction must pre-date Han.
Fig. 5. Ancient Chinese balustrade style construction.
(above: clay model house at Qingjiang Yingpanli in Jiangxi)
(below: brass model house at Jinning Shizaishan in Yunnan)
Fig. 6. Ancient Japanese figures of balustrade style
(above: engraved figure on tong duo; below: engraved figure on "tong jing")
While Japan is separated from the Asian mainland to the west by sea, its islands once were extensions of the mainland via landbridges formed when Ice Age sea-level fell(68). Hence, initial habitation and cultural growth and production were once closely tied to the mainland. Later, the sea not only separated them, but gave a means of transport, making it logical that Japan was continuously influenced by mainland cultures since that time. Archaeological and folkloric theory suggest five routes as factors of cultural influence(69): (1) north route from Siberia; (2) Korean route via its peninsula and across the Tsushima or Korean Strait; (3) from Jiangsu and Zhejiang across the East China Sea to Kyushu; (4) from Taiwan and Fujian via Ryukyu Islands to Kyushu; (5) south sea route from the South Pacific via South China Sea islands to Manchuria. Archaeologically, the main cultural influence was via Korea in concert with minor routes. This is seen in similar jue and lacquerware, rice cultivation, balustrade construction and supposed Shengwen crops along the East China Sea route.
That primitive canoes or wooden rafts sailed across the sea is truly questionable, but would have been aided by Japanese Sea currents, as warm south ocean currents and cold north currents pass through the Sea of Japan. The north current begins at Tatar Strait, moves south along the East Korean Peninsula and mixes with the southwest current, but part of its mixture flows east while another part moves south to Cheju Island to become the source of the East China Sea cold current. The south current flows north via Taiwan Strait past the Ryukyu Islands through Korean Strait. It turns northeast along the Japanese west coast, eventually diminishing in Tsugaru and Uchiura Straits(70), movements seen in floating debris along the Japanese west coast(71). While ancient shipbuilding and voyages were primitive, currents helped voyages from Jiangsu and Zhejiang across the East China Sea to Kyushu, making a reasonable sea route between ancient Japan and China in the South-North Dynasty, when an ambassador was sent to Tang Dynasty China and trade began. This route might have begun before the Han-Wei Period, and would explain how jue, lacquerware, rice cultivation and balustrade construction were taken to Japan.
It is even possible that sea voyages between China and Japan began in Hemudu and Shengwen periods. Although archaeology lacks adequate solid evidence, cultural factors popular in the lower Yangtze basin such as the making of jiatanheitao (a kind of pottery containing fibers), woodware and the habit of tooth extraction(72) often arise in Shengwen culture. This allows us to conclude jue and lacquerware were not isolates, and that ties of Hemudu culture and later cultural periods continued. Thus, jade jue and lacquerware arrived in Japan as early as 4000-5000 BC; rice cultivation was taken to Korea and Japan ca. 1000 BC; and balustrade construction was introduced to Japan before Han Dynasty.
The above examples show close ties between ancient Japanese and Asian mainland cultures, with the route from China to Japan starting likely before the Han-Wei period. Limited archaeological data suggests the above can only be treated hypothetically - a new subject on cultural contact between China and Japan. Further evidence can only be obtained with new archaeological work.