Where is the oldest pottery in the world? Perhaps surprisingly, the answer is, Japan. The discovery that Japanese pottery goes back a long way is not, in fact, new. In the 1960s, excavations in the Fukui cave in the southwestern sland of Kyushu produced what appeared to be remarkably early pottery. When some of the earliest radiocarbon dates were obtained from the site, they came out at around 12,000 uncalibrated years ago. Ever since then, one of the big questions of Japanese archaeology has been: just how old is pottery in the archipelago?
The pottery concerned is known to the Japanese as ‘Jomon pottery’. The word ‘Jomon’ means cord-marked, and the Jomon period takes its name from this type of pottery. Jomon pottery was first recognised by the American zoologist
Edward S. Morse, who in 1877 undertook what is widely recognised as the first scientific excavation in Japan, at the shell mounds of Omori, a short distance west of Tokyo in the modern city of Yokohama. Morse considered that the
pottery he had discovered was no more than a few thousand years old. In the 125 years since Morse’s investigations, many tens of thousands of sites dated to the Jomon period have been investigated, and it is now widely accepted that pottery was indeed being made in the archipelago from before 10,000 years ago. Over 70 major regional styles and in excess of 400 minor styles have been indentified and the Jomon period is now divided into six subperiods (Incipient, Initial, Early, Middle, Late and Final), running from the Palaeolithic down to the appearance of rice agriculture towards the end of the 1st millennium B.C.
The earliest appearance of this pottery is usually associated with the major environmental changes at with the end of the Pleistocene and the start of the Holocene. (Or, in British terms, with the end of the Palaeolithic and the beginning
of the Mesolithic.) The last glacial maximum (c. 18,000 years ago) saw sea levels up to 100m lower than today, and the present-day Japanese archipelago was connected to the East Asian continental mainland by landbridges, joining the southwestern island of Kyushu with the Korean Peninsula and the northern island of Hokkaido with Sakhalin and on to the Siberian coast. The area of the Japanese archipelago was not covered by ice sheets, and Late Paleolithic hunters were active before 30,000 years ago: it is normally considered that they led mobile lifestyles, hunting the large Pleistocene fauna including Naumann’s elephant and giant deer.
However at the end of the Pleistocene, when the climate suddenly went from ‘Ice Age’ to our present warm spell, major environmental changes took place. Tundra was replaced by forest, in particular beech and oak. The large animals were replaced by smaller creatures, including Japanese sika deer, wild boar, and bear. These animals were best hunted using different techniques: whereas spears are required for hunting large mammals, once they were replaced by smaller animals, the first arrowheads appear, presumably implying the development of archery. The new forests provided a different range of foodstuffs, in particular plants and fish. Pottery provides useful containers in which to boil vegetal foods and make them edible.
Perhaps the best newly discovered example of a Jomon settlement right at the beginning of this new warm era is the site recently excavated at Oushikakubo. Oushikakubo is located on the western slopes of the Habuna range of hills in Shizuoka Prefecture, located to the southwest of Mount Fuji, one of the most evocative landmarks of the contemporary Japanese landscape. In line with the laws for the protection of cultural properties in operation in Japan, the site was excavated prior to the development of agricultural land by the local Board of Education. The excavations produced in excess of 20,000 worked stone objects, including tools such as points, endscrapers and flakes. There were also pottery sherds. These were variously decorated with thin appliqué bands of clay (so-called linear-relief pottery), short curved incised patterns reminiscent of finger-nails (the finger-nail pattern pottery) and pottery bearing the impressions of twisted cords (known as Jomon pottery in Japan).
The site was an extensive one – more a village than a single house. At least ten round pit buildings whose floors had been dug into the ground, were discovered with sherds of cord-impressed pottery. Around the outside of each pit, associated post-holes were found, which would have held posts supporting the roof. There was also evidence for the processing of plant foods -perhaps nuts. Within each structure was found a flat stone and a grinding stone, and some houses contained traces of burnt earth hearths. A short distance to the east of these buildings were examples of two other categories of feature, stone clusters and stone arrangements. However, it is difficult to be sure how many of the buildings were in use at any one time. Oushikakubo may have been a ‘village’, and suggests a higher degree of residential stability than previously thought for the Incipient Jomon.
Back into the Pleistocene?
But was there any pottery earlier than this, going back right into the Ice Age? In July 1998, a team led by Taniguchi Yasuhiro of Kokugakuin University in Tokyo, was invited by the Kanita Town Board of Education (in Aomori Prefecture, on the northern tip of the largest island in the Japanese archipelago, Honshu) to conduct excavations prior to the construction of some houses at a site called Odai Yamamoto. Here Late Palaeolithic stone tools had previously been discovered, belonging to an assemblage recognised as part of the so-called Chojakubo culture. Hitherto the Chojakubo culture had been dated by its stratigraphic relationship to a volcanic tephra, the Hachinohe Tephra, which has been dated by radiocarbon to 12,500-13,000 years ago. The Chujakubo assemblages predate this tephra.
Taniguchi and his team excavated 148 m2 in total. Artefacts were recovered from two strata. Stratum III was a yellowish tan volcanic ash loam, and Stratum IV was a light yellowish brown sandysilty loam. 262 stone artefacts and 46 pottery sherds were discovered in an area about 10 metres in diameter, described in the report as ‘a very clear, single point-in-time assemblage’.
The 46 sherds of pottery from Odai Yamamoto all come from the same pottery vessel. Analysis of the clay demonstrated the presence of local rock materials, including quartz, biotite and felsitic white groundmass, evidence that the vessel was made from locally available clay. Although it was not possible to completely reconstruct the vessel, it was clear that it had a flat base and a straight rim. The walls of the vessel were 7.6 mm thick. Apart from some narrow engraved lines on nine of the sherds, the outer surface of the vessel does not appear to have been decorated. 30 of the 46 sherds had carbonized accretions, some of which formed a water-line along the inner surface of the vessel, suggesting that the vessel had been used to boil up foodstuffs.
These pottery fragments were found in association with the a lithic assemblage comprising 1 core, 65 blades and blade flakes, 2 anvils, 2 axes, 2 arrowheads, 1 spear point, 1 sidescraper, 11 end-scrapers, 4 gravers, and 11 combined gravers/endscrapers/sidescrapers. Over 90% of these stone tools were made of locally available hard shale and a further 5% were made of obsidian which originated in a source some 70 kilometres away from the site. Of particular interest in this assemblage are two arrowheads made of hard shale, just under 3 cms long, and not quite 2 cms wide and with rounded convex bases. Previously, it has been thought that arrows and their associated technology of archery came into usage around 11,000 years ago, but these examples from Odai Yamamoto suggest that archery, along with pottery, was being practised long before that.
But just how old was the site? A recent survey of radiocarbon dates from the Late Pleistocene from the Japanese archipelago in Radiocarbon (Ono et al, 2002) suggests that there are some 4500 Pleistocene sites in Japan, and that about 100 new sites are being investigated each year. The same survey lists 429 radiocarbon dates from about 100 of these sites. Ono Akira from Tokyo Metropolitan University and his colleagues, who produced this useful survey, concluded that the Japanese Upper Palaeolithic should be divided into three stages: an early Upper Palaeolithic (from before 34,000-26,000 years ago), a late Upper Palaeolithic (from c. 25,000 years ago to 15,000 years ago) and a Final Upper Palaeolithic and Incipient Jomon period (from c.14,000-12,000 years ago).
Some of the sites they mention, however, suggest that there was more to Upper Palaeolithic life than chasing after large mammals. At Hatsunegahara, also in Shizuoka Prefecture, 56 pit traps, up to 1.5m deep, have been discovered sealed beneath one of the most important of all the Japanese Pleistocene volcanic tephras, the AT horizon, which resulted from the massive eruption of the Aira Caldera, whose remains can be seen around Kagoshima Bay and present day Sakurajima in southwestern Kyushu. This tephra is well-dated by AMS dating to 23-22,000 years ago. The finds from Hatsunegahara suggest that medium-sized animals were being hunted in an organised planned fashion prior to 25,000 years ago. At the remarkable waterlogged Palaeolithic forest floor at Tomizawa in Miyagi Prefecture, the droppings of sika deer (Cervus nippon) were discovered – direct evidence for the presence of medium sized mammals in the area during the last glacial maximum.
All of the dates provided in this survey, however, are uncalibrated. Because of fluctuations in the radiocarbon reservoir, all radiocarbon dates need to be calibrated: the calibration of dates back into the Neolithic is now well-established, but the calibration of Palaeolithic dates is relatively new, it appears that radiocarbon dates around this period need to be pushed back by about 2,000 years. It is the effects of calibration that are causing such interest in the finds from Odai Yamamoto. A total of 8 AMS dates were obtained from the carbonised adhesions on five of the pottery sherds and three pieces of charred wood found in association with the pot sherds. These gave a range of some 13,500-13,800 uncalibrated radiocarbon years ago, currently the earliest dates for pottery in East Asia, and indeed anywhere else. The dates were produced by Nakamura Toshio of the Nagoya University Dating and Materials Research Center and Tsuji Sei-ichiro of the National Museum of Japanese History. The only AMS dating facility in Japan is located at Nagoya University, and this is where the carbonised accretions were dated. The three wood samples were dated by the Beta Laboratory in the USA. Nakamura proceded to calibrate his dates using the MacCALIB 3.0 program, and this pushed the dates back to 16,000-16,500 calibrated radiocarbon years ago (i.e. around 14,400 to 14,400 B.C.).
This is of great significance as it places the earliest known pottery in East Asia before the beginning of the first warm stages at the end of the Pleistocene, the Bølling interstadial (dated to 14,760 calibrated radiocarbon years ago) and the subsequent fluctuations of the Younger Dryas (dated to 12,890 calibrated years ago). This is before any of the Holocene environmental conditions appeared in what is now the Japanese archipelago and places early pottery usage firmly in the Late Palaeolithic. The results of more dates from this critical period are eagerly awaited.
The discoveries of early pottery and arrowheads at Odai Yamamoto, in conjunction with the evidence for the well-organised use of the landscape during the Late Palaeolithic and suggestions of a relatively settled lifestyle during the Incipient Jomon period at sites such as Oushikakubo, suggest that we need to reassess our view of what life was like at the Pleistocene-Holocene transition. Advances in dating techniques offer the possibility to understand these critical developments at a much finer scale of resolution. The May 2003 issue of the journal Kikan Kokogaku (Archaeology Quarterly) was devoted to the theme of researching the origins of Jomon culture that emphasises the regional diversity at this time, even within the seemingly relatively narrow confines of what was to become the Japanese archipelago. A conference in Cambridge in autumn 2001, to be published in the Journal of East Asian Archaeology in 2004, summarised findings of early pottery elsewhere in East Asia. Particular attention was given to the Amur River valley in eastern Siberia and China, where increasing numbers of sites with pottery dated to before 10,000 years ago are being discovered. There is little doubt that this part of the world was central to the development of pottery technology, only a few thousand years after it had been on the route for the first peopling of the American continent.
This article is an extract from the full article published in World Archaeology Issue 1. Click here to subscribe