Home | Index Bulletin of the International Jomon Culture Conference
Vol. 1, 2004, pp. 45-50

"Fakery" at the Beginning, the ending and the Middle of the Jomon Period

Charles T. Keally


Several times since 1999, archaeologists have published calibrated AMS radiocarbon dates for the Jomon period. These dates have caused "shock and confusion" among Japanese archaeologists, because they show the beginning, middle and end of the Jomon period to be older than was thought. The calibrated dates for the end of Jomon and the beginning of Yayoi are particularly shocking to Japanese archaeologists, because acceptance of these calibrated dates will mean that for 30 years, since calibration first became possible, the archaeologists have been writing a false history of the transition from the Jomon hunting-fishing-gathering culture to the Yayoi wet-rice farming culture. And they have been writing this false history simply because they did not know or understand a fundamental aspect of a fundamental archaeological dating method.


On April 17, 1999, the newspapers in Japan headlined the astonishing fact that the Jomon Period actually began (the oldest pottery in Japan was) 3,000 years earlier than had been thought -- 16,000 years ago, not 13,000 years ago (Jomon no kigen 1999). On May 20, 2003, the newspapers in Japan headlined the astonishing fact that the Yayoi Period actually began (the Jomon Period ended) 500 years earlier than had been thought -- in the 9th or 10th century B.C., not the 4th or 5th century B.C. (Gakkai 2003; Yayoi jidai 2003; Yayoi no hajimari 2003; "Yayoi" no kigen 2003). And on May 25, 2003, the newspapers in Japan reported that the Middle Jomon Period was 500 years older than had been thought -- that it ranged from 4,500 to 5,500 years ago, not from 4,000 to 5,000 years ago (Jomon Chuki 2003).

These "astonishing" new dates are both real and fake, both astonishing and not at all surprising. These new dates are real in the sense that they do in fact contradict the school textbooks and the popular knowledge, and apparently also the ideas of most archaeologists. But these new dates are fake in the sense that they are nothing more than the dates all archaeologists should have known for most of the past 20 or 30 years. They are also fake in the sense that the archaeologists reporting them did not report them as "astonishing" -- for example, "they do not contradict the dates we could have obtained with the poorer quality dates available since the 1950s" (Harunari et al. 2003:65).

More specifically, these dates are true in the sense that they do reflect the simple radiocarbon age measurements obtained since the first Jomon dates (and first Japanese dates) were published in 1951 (Libby 1951). But these dates are false because simple radiocarbon age measurements are not calendar dates, or true ages, and they can differ from true ages by several centuries or more, quite significant amounts of time for some archaeological hypotheses. These Jomon falsehoods differ from the Japanese Early and Middle Paleolithic hoax only in the fact that no one is purposefully planting faked information. But, like the Early and Middle Paleolithic hoax, these Jomon falsehoods are serious and they exist because archaeologists in Japan almost universally do not understand radiometric dating methods.

Since the mid 1960s, archaeologists have known that radiocarbon dates were not true calendar dates (Libby 1963; Rainey and Ralph 1966; Stuiver and Suess 1966; see Watanabe 1966:166). And it has been possible to adjust (calibrate) simple radiocarbon age measurements roughly to true calendar dates since the early 1970s (Clark 1975; Suess 1970; Tite 1972). Consequently, all archaeologists long ago should have known the calibrated (or historic) dates for the Jomon Period, and they should have been using those calibrated dates in their hypotheses, when necessary. Thus, these new dates should have been astonishing to no one.

Dating finds is fundamental to archaeological research, and every archaeologist should understand at least the essentials for using dates (Suzuki 1976; Kigoshi 1978; Bowman 1990; Nakamura 1999; Imamura 2000). For the most part, these essentials are (1) the sources of error in the dates, and (2) how dates obtained with one method compare to dates obtained with another method. Radiocarbon dating is the most common method used in archaeology, and it is the method providing these "astonishing" new dates on the Jomon Period. The main sources of error in radiocarbon (14C) dates are (1) the dated material and the finds being dated are not actually associated, and (2) the dated material is contaminated with younger or older carbon. Comparing radiocarbon dates to dates obtained with other methods requires that all dates be calibrated to the same standard, usually taken to be true calendar years. Variations in the amount of 14C in the atmosphere cause radiocarbon age measurements to be sometimes younger and sometimes older than calendar years. But radiocarbon ages can be calibrated roughly to true ages (calendar dates) using charts derived from saturation dating of tree rings of known calendar date (Kigoshi 1978; Suess 1970; Tite 1972; Stuiver et al. 1998). This is something archaeologists must do whenever (1) their hypotheses involve comparison to cultures with historic dates (the Jomon-Yayoi transition compared to China), or (2) their hypotheses require accurate knowledge of the lengths of time involved (how long it took the Jomon people to move from first making pottery to fully integrating pottery into their daily lives).

The "astonishing" new dates reported in the past few years were obtained by the AMS (accelerator mass spectrometry) radiocarbon dating method (Imamura 1991). AMS dating is a relatively new technique. It allows more precise dating and dating of much smaller samples, making it possible to directly date potsherds by dating charred adhesions on the potsherds or charred organic tempering in the potsherds. Thus, AMS dates are generally more valid than dates obtained by earlier radiocarbon dating techniques. However, AMS dating has made no significant change in the radiocarbon dates for the Jomon Period (Table 1). What has made a significant difference is that archaeologists finally, over 30 years late, have begun to publish calibrated radiocarbon dates (given as cal BP, or as cal BC and cal AD), in addition to simple radiocarbon dates (given as years BP, years before 1950).

The first compilation of Japanese radiocarbon dates was published in 1966 (Table 1, col. 1; Watanabe 1966). There still were very few dates (188) available for the Jomon (and Yayoi) Period in 1966;

Table:1 Jomon dates estimated from multiple dates listed in the cited publications

therefore, the estimated dates in Table 1 are rough. But, as can be seen, those rough estimates do not differ significantly from the estimated dates based on the more precise AMS dates and much larger number of measurements published recently (Table 1, col. 3). If these rough 1966 estimated dates are calibrated to approximate calendar dates, using a calibration chart published in 1972 (Tite 1972:89) (the oldest calibration chart in my hands, but not the oldest one published [Suess 1967, 1970]), we get the calendar dates for the Jomon Period (Table 1, col. 1, dates in parentheses) that should have been known and used at least as early as 1972, the date this calibration chart was published. But these calibrated (calendar, or historic) dates have never been used, to my knowledge. The next relatively complete compilation of Jomon radiocarbon dates was published in 1982 (Keally and Muto 1982). This compilation increased the number of radiocarbon dates from Watanabe's (1966) 188 Jomon and Yayoi dates to 439 Jomon-only dates. But it made no significant change in the estimated radiocarbon ages of the Jomon subperiods (Table 1, col. 2; Keally and Muto 1982). And the true ages (Table 1, col. 2, dates in parentheses), calibrated with a chart published in 1972 (Tite 1972:89), also do not differ significantly from the calibrated dates published recently (Table 1, col. 3).

Keally and Muto (1982) did not deal with the dates for the Plain pottery that precedes Linear-relief pottery (see Oda and Keally 1979), because this oldest pottery lacked directly associated dates at that time. But the age of this Plain pottery could already be estimated at ca. 13,000 BP in the early 1980s. At the Odai Yamamoto I site in Aomori Prefecture, Plain pottery had been found with a typical Chojakubo-Mikoshiba lithic assemblage (Miyake and Iwamoto 1979). This lithic assemblage was already known to be stratified under the Towada-Hachinohe Pumice (HP) (Yamanouchi and Sato 1967). And by 1982, this pumice was known to date ca. 12,500 BP (see Machida and Arai 1992: 235-236), indicating that the Chojakubo-Mikoshiba Culture probably dated to about 11,000 B.C. (Ikawa 1964: 98-99, quoting personal communication from Serizawa Chosuke). And there were several other sites in 1982 that were thought to belong to this same lithic culture and which also yielded Plain pottery or pottery that appeared to pre-date Linear-relief pottery: Ushirono Loc. A in Ibaragi Prefecture (Ushirono Iseki Chosadan 1976), Terao in Kanagawa Prefecture (Shiraishi 1980), and Sempukuji Cave in Nagasaki Prefecture (Aso 1984; see Oda and Keally 1979). Also, about that time, the Maeda Kochi site in Tokyo had a 13,000 BP radiocarbon date for Plain pottery (Keally and Miyazaki 1986; see Akigawa-shi 1983), but this date was not then, nor never has been, officially published. However, unlike the dates for Middle Jomon and the Jomon-Yayoi transition, the very old Linear-relief and Plain pottery dates could not be calibrated until quite recently (see Pilcher et al. 1984; Stuiver and Reimer 1993; Stuiver et al. 1998; Kitagawa and van der Plicht 2000). Nevertheless, no one should have been surprised when the true age of the oldest pottery in Japan was reported to be ca. 16,000 cal BP, rather than 13,000 BP (Nakamura and Tsuji 1999; Jomon no kigen 1999). And certainly no one should have thought this changed the relationship of the oldest Jomon pottery with the oldest pottery elsewhere in the world, because simple radiocarbon dates can be compared only to simple radiocarbon dates, and calibrated dates to calibrated dates, and these relationships remained unchanged (Taniguchi 1999, 2001, 2002; Taniguchi and Kawaguchi 2001).

For 30 years, archaeologists have been writing a false history of the Jomon period and of the Jomon-Yayoi transition, simply because they lacked the readily available knowledge and understanding of radiocarbon dates that they needed in order to write a reasonably accurate history of those times. This error makes quite clear the need for archaeologists to learn and understand the methods they borrow from the natural sciences.

The examples discussed in this paper are radiocarbon dating examples. But there are many other dating methods and much other information brought into archaeology from the natural sciences, and archaeologists are using these wrongly, too, in many cases. Hence, they are writing falsehoods about other aspects of the Jomon and other prehistoric Japanese cultures. Japan's Early and Middle Palaeolithic hoax has led to much introspection and to calls for Japanese archaeology to open up and to become interdisciplinary. Heeding these calls is urgent.