Chronology of the Earliest Pottery in East Asia: Progress and Pitfalls

Article excerpt


The origin of pottery manufacture is one of the most important subjects in Old World prehistory. Since the mid-1960s, the Jomon of Japan was considered as the archaeological complex with the earliest pottery in the world dated to the Final Pleistocene, c. 12 70012 200 radiocarbon years ago (hereafter BP) (e.g. Morlan 1967). From the 1970s to the 1990s, the increased pace of archaeological and chronological studies in East Asia has brought to light new evidence of the Final Pleistocene pottery in other regions neighbouring Japan, such as China and the Russian Far East. In East Asia, the presence of pottery is very often associated with the Neolithic stage in prehistory (e.g. Barnes 1999: 17) although in the earliest sites important indicators of the Neolithic in its classical definition, agriculture and sedentism, are missing. Thus, the meaning of the term 'Neolithic' in East Asia is different from that in Europe and the Near East.

The main aim of this paper is to present an updated review of the chronological aspects of pottery origins in East Asia, with a critical evaluation of the latest summaries. The 'chronometric hygiene' approach (sensu Spriggs 1989) is applied to the archaeological complexes with the earliest pottery in East Asia, meaning that the radiocarbon dates are critically assessed and unreliable ones are rejected.

Materials and Methods

Pottery in this review is defined as 'clay that has been fashioned into a desired shape and then dried to reduce its water content before being fired or baked to fix its form' (Darvill 2002: 337-8). For our purpose, those sites with radiocarbon (hereafter 14C) dates directly associated with the earliest pottery were chosen. The sites are located in China, the Japanese islands, the Russian Far East, and Korea (Figure 1). All the sites meet the criteria for establishing the presence of pottery in archaeological contexts (Vandiver 1999). In order to evaluate the reliability of the [sup.14]C dates associated with pottery, critical assessment of different chronological aspects were addressed. These included the materials dated, the methods for direct [sup.14]C dating of pottery, the degree of association between [sup.14]C dates and potsherds from particular cultural layers, and the correspondence of the earliest pottery [sup.14]C dates with general cultural chronologies. Dates that do not fulfil the 'chronometric hygiene' requirements, such as consistency in stratigraphy, secure association of [sup.14]C-dated material and pottery, adequate reporting of original data, and correspondence to the general chronological outline of prehistoric cultural complexes, were rejected after explaining why they were not considered to be reliable.


In this paper, comprehensive summaries with lists of [sup.14]C dates were used (Wu & Zhao 2003; Keally et al. 2003; Kuzmin & Shewkomud 2003). Figure 2 shows the calibrated ages of the most reliable earliest [sup.14]C values associated with pottery. Chinese dates, originally reported with 5730 yrs [sup.14]C half-life, were re-calculated for the 'Libby value' of 5568 yrs, to be compatible with other dates produced elsewhere. The CALIB rev. 4.4.2 software was used for date calibration.


For the [sup.14]C data corpus used in this review, the materials dated include wood, charcoal, food residues (adhesions), human and animal bones, freshwater shells, and potsherds. Wood, charcoal, bone, shell, and humic acid pretreatment procedures are quite standardised now (cf. Taylor 1987:44-61). For the direct dating of food residues and pottery, several pretreatment protocols were used. For the extraction of carbon from the charred food remains attached to the potsherds, Nakamura et al. (2001) used acid-alkali-acid pretreatment, with a weaker concentration of alkaline solution to prevent the loss of carbon

Direct dating of pottery is a more difficult task (see review in Bonsall et al. …