SOPHIA INTERNATIONAL REVIEW |
February 4, 2004 last revised: May 14, 2004 |
Home | Index |
by Charles T. Keally |
This is a revised version of a paper published in Sophia International Review no. 26, 2004. This journal can be obtained at:
Fax: +81-3-3238-4076
The oldest pottery in Japan has just been dated at 16,500 years, pushing back the date by 4,500 years and making Jomon pottery by far the oldest pottery in the world. (April 1999)Both of these news headlines are trueThe beginning of the Yayoi Period, and of iron use and rice farming in Japan, has been pushed back 500 years to about 900 BC, necessitating a major rewriting of early Japanese history. (May 2003)
Jomon Origins 4,500 Years Older: 16,500-Year-Old Potsherds in Aomori Prefecture (Asahi Shimbun [Yukan], April 17, 1999, p. 1)Leading archaeologists expressed amazement and confusion (Jomon no kigen 1999; Jomon furusa 1999; Kataoka 1999). Suddenly the oldest pottery in Japan was 4,500 years older than the oldest known pottery elsewhere in eastern Asia, and not 4,000 years older than the oldest pottery in Southwest Asia but 8,500 years older. (NOTE 1)Jomon Age Radically Older: 16,500-Year-Old Pottery (Asahi Shimbun [Yukan], April 17, 1999, p. 15)
Age Measurement Still Not Reliable (Yukan Yomiuri Shinbun, May 13, 1999, p. 15)
But, in fact, these ideas were gross misunderstandings of what actually happened; our fundamental knowledge of the beginnings of pottery technology in eastern Asia, and in the world, remained unchanged (Kataoka 1999; see Keally, Taniguchi & Kuzmin 2003; Fagan 1998). The simple radiocarbon ages (dates) for all regions of early pottery, including Japan, were still the same as they had been. The "new" date
Transmission of Rice Farming Began 500 Years Earlier: Also a Major Revision of Ancient History (Asahi Shimbun, May 20, 2003, p. 1)Again, leading archaeologists expressed amazement and confusion, and strong disagreement (Gakkai 2003; Hayamaru 2003; Ikki 2003; Rekishikan 2003). Such a change in the dating, if true, would radically rewrite the story of the beginning of the Yayoi culture and of rice farming in Japan, and it would also change our ideas about some major cultural events on the continent.Yayoi Beginning 500 Years Earlier: Kyushu Pottery Measured with the Carbon Method (Mainichi Shinbun, May 20, 2003, p. 1)
Yayoi Period Moved Earlier by 500 Years: Rice Farming Transmitted in the 10th Century BC (Hokkaido Shinbun, May 20, 2003, p. 1)
"Yayoi" Origin 500 Years Older: Need to Reexamine the Beginnings of Rice Farming (Yomiuri Shinbun, May 20, 2003, p. 1)
Five Centuries Older in One Leap, Overturning Common Knowledge: Scholars Express "Disbelief" (Asahi Shimbun, May 20, 2003, p. 35)The introduction of rice farming into Japan would no longer correlate with events late in the Chinese Chou (Zhou) Dynasty as thought, and as published in the textbooks, but rather with events at the beginning of the Chou Dynasty, or even the end of the Shang (Yin) Dynasty. The oldest iron tools in Japan would be older than the oldest iron tools in North China. And the length of time it took for rice farming to diffuse across western Japan would be lengthened by many centuries, negating the accepted view that it had diffused rapidly from Kyushu to the central mountains. These would truly be radical changes in Japan's history, if these new dates were true. No wonder archaeologists were upset by the announcement.Shock and Confusion in the Discipline: Also a Need to Rewrite the Textbooks (Mainichi Shinbun, May 20, 2003, p. 27)
The "Yayoi Shock" Overturns Our View of History: New Puzzles Such as the Background to Immigration (Nihon Keizai Shinbun [Yukan], May 23, 2003, p. 5)
Aging Yayoi is Spreading Disruption: Origin of Iron is the Greatest Cause of Doubt (Sankei Shinbun, May 31, 2003)
Radiocarbon dating is the most common dating method used in archaeology. (For a technical discussion of the method see Radiocarbon 2004.) Radiocarbon dates are considered to be absolute dates, but they are NOT calendar dates, only estimates of calendar dates, or true ages, and often only very rough estimates.
Like dates from all archaeology dating methods, radiocarbon dates are subject to a number of sources of potential error. One source of error is randomness; another is short- and long-term fluctuations in atmospheric radioactive carbon. Randomness means, in a crude way, that about 50% of radiocarbon dates are significantly different from the true radiocarbon age of the dated material, and that the other 50% will span a range of ages but with an obvious clustering tendency. Fluctuations in atmospheric radioactive carbon cause radiocarbon dates to differ irregularly from true dates.
The problem of randomness can be lessened by saturation dating
Stated a bit more simply: (1) radiocarbon age measurements always are only approximations of the actual radiocarbon age of a sample; the randomness in the system causes all radiocarbon age measurements to vary more or less from the actual radiocarbon age. And (2) radiocarbon dates are not calendar dates; they must be calibrated in order to bring them closer to true dates, or calendar dates. This should make it obvious that simple radiocarbon dates and calibrated radiocarbon dates cannot be used together
A note on technical terminology: Radiocarbon age measurements, or conventional radiocarbon dates, are given as years BP (before physics, or before 1950); calibrated radiocarbon dates are given as years cal BP, or as years cal BC or cal AD.
These dates are also given as RCYBP (radiocarbon years BP) and CALYBP (calibrated years BP). Radiocarbon ages, or dates, often are given in thousands of years, years ka, such as 12,000 years written as 12 ka. The word radiocarbon is often replaced with 14C |
In order to meet these goals in both North America and Japan, accurate dates are of utmost importance. Explanation requires accurate knowledge of cause and effect; accurate knowledge of cause and effect requires accurate dating of what came before and what came after. Accurate history, as well as accurate knowledge of origins, also requires accurate knowledge of what is older and what is younger
But archaeologists do not find calendars in prehistoric sites. They must use other methods to date their finds. However, none of these methods
For 25 years, Japanese archaeologists have been unwittingly publishing a false story of Japan's prehistory because they do not understand the dating methods that are providing the dates for that story of the human past in Japan, and in the wider world. It is this lack of knowledge and understanding of a methodology essential to the discipline that caused the amazed and confused reactions among archaeologists and journalists when the calibrated radiocarbon dates for the oldest pottery in Japan and for the beginning of the Yayoi Period were published in 1999 and 2003.
The discovery of Jomon rice paddies in 1978 caused considerable turmoil initially (Jomon-jin no ashi ato 1978; Yureru 1978; Kondo wa kuwa 1978; "Inasaku sekki" 1978). Now it causes confusion: there are two opinions about where to draw the Jomon-Yayoi boundary in the pottery sequence.
Some archaeologists have chosen to define "Yayoi" by pottery types instead of rice farming, and they have left the boundary where it was before 1978. They divide the Yayoi Period into Early, Middle, and Late subperiods.
Other archaeologists have chosen to keep the definition of "Yayoi" as Japan's first rice farming culture, and these archaeologists have moved the boundary back in time to include some of the final Jomon pottery types in an "Earliest" Yayoi period, reclassifying these Jomon pottery types as Yayoi pottery types. They divide the Yayoi Period into Earliest, Early, Middle, and Late subperiods.
Rice in Japan before Yayoi: There are rice grains and grain impressions in the clay of pottery from several Late Jomon sites dated a few centuries before Earliest Yayoi begins. This has been known since at least 1992 (Kigenzen 1992). (NOTE 5) And in 1999, evidence of rice dating back as far as 6,000 years ago (uncalibrated radiocarbon years) was reported from a number of sites in western Honshu and Kyushu (6000 nen 1999; Takahashi 2001). However, these finds are not considered evidence of "significant rice farming" in Japan and hence are not included within the definition of Yayoi culture.
Dating the Beginning of the Yayoi Period: Using the radiocarbon dates available in 1966 (Watanabe 1966), the boundary between the Jomon and the Yayoi periods can be estimated very roughly at about 400-500 BC for the first view (for Early Yayoi), and about 600 or 700 BC for the second view (for Earliest Yayoi).
Using the radiocarbon dates available in 1982 (Keally & Muto 1982; Nihon Senshi Jidai 2000, p. 86, citing Karatsu-shi 1982), the boundary between the Jomon and the Yayoi periods can be estimated roughly at 500 BC for the first view, and about 700 BC for the second view.
And using the new radiocarbon dates announced in May 2003 (Harunari et al. 2003), the boundary between the Jomon and the Yayoi periods can be estimated at about 600 BC for the first view, and 700 BC for the second view.
The simple radiocarbon dating announced in May 2003 gives essentially the same estimated dates for the Jomon-Yayoi boundary as did the meager data available in 1982 and 1966. But these new dates also were calibrated
In fact, I have just surveyed a number of recent publications on the Yayoi Period and have not found any of them using radiocarbon dates to determine the age of the period (for example, see Kanaseki and Osaka-fu 1995; Otsuka et al. 1998; Morioka 2001). They all use comparative artifact typology and the assumption that the dates from Korea and China are all accurate and infallible. A few of the works noted that there are some tree-ring dates available for the Yayoi Period, but they did not use these dates.
Most books and articles on the Yayoi Period give its beginning as 4th or 5th century BC (between 300 and 500 BC), and some give simply 300 BC. (NOTE 6) These also are the dates given in the newspapers in May 2003, and they are referred to as the "accepted view" (the teisetsu).
Calibration: By the late 1960s, archaeologists were aware that their radiocarbon dates were not giving them true historical dates, or calendar dates (Libby 1963; Rainey & Ralph 1966; Watanabe 1966; Kigoshi 1967a, 1967b, 1968). At that time it also was possible to estimate true dates back to about 4,000 years ago (Kigoshi 1968, p. 6). And by the late 1970s, archaeologists could "calibrate" these dates as far back as about 8,000 years (Tite 1972, p. 89; Clark 1975). In fact, in a 1978 paper on the radiocarbon dating of Jomon sites in Hokkaido, William M. Hurley and his colleagues did just that with their Japanese radiocarbon dates (Hurley et al. 1978), using the MASCA correction factor (see Michael & Ralph 1971, p. 28; Hole & Heizer 1977, p. 209). (NOTE 11)
My course lecture notes (for Prehistoric Japanese Cultures) "revised completely Nov. 1977", with some minor corrections in the "Fall 1979", have the dates thus:
Yayoi 14C dates are: Early (500-300 BC), Middle (300-0 BC), and Late (AD 0-250/300). Calibrated, the Early Yayoi dates would be about 900-400 BC. The Jomon-Yayoi boundary is 400±30 BC (Sakada Kunihiro 1975, p. 6). The Ukikunden site in Kyushu yielded Yuusu (Jomon) and Itazuke I (Yayoi) pottery together, suggesting a date of ca. 400-500 BC 14C (Chard 1974). (NOTE 7)
I began my web site on Japanese archaeology in 1997. The site gives these dates for the beginning of the Yayoi Period:
Uncalibrated radiocarbon ages...suggest the beginning of [Early] Yayoi dates to about 400 or 500 BC. Calibrated ages take the beginning back to 800 or 900 BC. (http://www.t-net.ne.jp/~keally/yayoi.html)According to the "new" calibrated radiocarbon dates published in 2003, the beginning of Early Yayoi is about 800 cal BC, and of Earliest Yayoi is about 900-1000 cal BC (Harunari et al. 2003). This is about the same as the calibrated dates I was teaching my classes 25 years ago. And these are the dates that would have been in general use for most of that time, too, if archaeologists had been reading the many articles and books on radiocarbon dating that were readily available to them (Watanabe 1966; Kigoshi 1967a, 1967b, 1968, 1978; Suzuki 1976; Nakamura & Nakai 1983; Imamura 1991, 2000; Nakamura 1999).
Japan's Oldest Iron: The date for the oldest iron in Japan is one reason so many archaeologists disagree with the new Yayoi radiocarbon dates (Hayamaru Yayoi 2003; Yayoi-ki "shinsetsu" 2003; Yayoi no hajimari 2003). They say that the oldest iron in North China dates to the 8th century BC, while the oldest iron presently known in Japan is dated to the beginning of the Yayoi Period in the 4th or 5th century BC. But the new Yayoi dates would make the oldest iron in Japan older than the oldest iron in North China, which, they say, is impossible (Ito 2003b; Iida 2004). (NOTE 8) Their thinking is not valid.
First, is our knowledge of the development of iron use and production in China, and Korea, complete and accurate? Probably not; this is archaeology. And, second, does iron working have to be older in China than in some other region of eastern Asia from which iron artifacts could have gotten to Japan? I see no reason why. But the archaeologists who oppose these new Yayoi dates clearly do see reason here.
There is a widespread assumption in Japanese archaeology
In fact, the early development of iron technology in China is still a matter of disagreement (Wagner 1999). And the Semipiatnaya Gorge site near Vladivostok in Primorye, in Russia, is said to have iron associated with dates given variously as 2980±30 BP (Derevianko 1965, p. 138) and 3010±80 (RUL-165; Chard 1962, p. 85, for a site called Pad' Semipiatniaia), or as the 11th to 12th century BC (Derevianko 1965, p. 138; Chard 1974, p. 96), although these dates are not confirmed (Wagner 1999). Nevertheless, Kuzmin (1995, p. 80) agrees with these dates, giving the "Early Iron Age sites [as dating] from the end of the 2nd to the 1st millenia BC", with the Neolithic/Early Iron Age boundry...ca. 3100-3300 BP (1400-1600 cal BC)" (Kuzmin et al. 1998).
In Japan, the oldest generally accepted evidence of iron use comes from the Magarita site in Fukuoka Prefecture (Hayamaru Yayoi 2003). This iron was found with Yuusu I pottery of the Earliest Yayoi period, giving it a simple radiocarbon age of about 700 BC or older (see Harunari et al. 2003). This date is 200-300 years older than the 4th-5th century BC that archaeologists give for the beginning of Yayoi. Even the dates available before 2003 (Watanabe 1966; Keally & Muto 1982; Imamura 2001) suggest that this site is older than 500 BC and most likely 600 BC. These simple radiocarbon dates become about 750-800 cal BC in calibrated years (see Stuiver et al. 1998).
There are several other sites yielding iron artifacts that date from the beginning of Early Yayoi, for example, the Saitoyama Shellmound in Kumamoto Prefecture (Wajima 1967, pp. 435-436), the Imagawa site in Fukuoka Prefecture (Saiko no tetsu 1980), and the Okamoto Yonchome site in Fukuoka Prefecture (Nihon saiko no tekken 1980).
There even are two Latest Jomon sites claimed to have evidence of metal use in Japan before the Yayoi Period began. Cutting and stabbing marks on human bones from the Itoku site in Kochi Prefecture were identified as those made by a metal implement (Jomon-jin no hone 2002; Kizu ato jinkotsu 2003). These bones were associated with middle to late Latest Jomon pottery, thought to date to around 2800-2500 BC [uncalibrated radiocarbon date]. The newpapers, however, reported the radiocarbon date [calibrated?] of this site as about 3200 BP, or about 1260-1130 BC [cal BC?] (Kizu ato jinkotsu 2003).
There also is the completely overlooked terminal Latest Jomon site in Kushiro City, Hokkaido, that yielded a fragment of an iron artifact in a burial (Kono 1973). Available dates (Watanabe 1966; Keally & Muto 1982) suggest this iron dates to at least 500 BC (ca. 750 cal BC) and possibly 600 BC (ca. 800 cal BC) (see Stuiver et al. 1998).
The age of 12,000 years for Japan's oldest pottery was derived in the 1960s from the radiocarbon dates for the Linear-relief Pottery found at the Fukui Cave site in Kyushu and the Kamikuroiwa Rockshelter site in Shikoku (Ikawa 1964, p. 100; Watanabe 1966; Esaka, Okamoto & Nishida 1967, p. 236; Kamaki and Serizawa 1967, p. 265). These radiocarbon dates were 12,700±500 BP (Gak-950) and 12,165±600 BP (I-944) respectively. But they were rounded to 12,000 BP for general use, which was valid, given the wide range and large standard deviation (plus-minus, or sigma) values.
But as early as 1964, one archaeologist felt he had reason to estimate the age of Japan's oldest pottery to be 13,000 years, or about 11,000 BC (Serizawa Chosuke, personal communication quoted by Ikawa 1964, pp. 98-99). And by the late 1970s, it was widely known that a number of sites had yielded small amounts of Plain Pottery that was clearly older than the Linear-relief Pottery at Fukui Cave and Kamikuroiwa Rockshelter (Oda & Keally 1979). These sites were Odai Yamamoto I in Aomori Prefecture (Miyake & Iwamoto 1979), Ushirono in Ibaragi Prefecture (Ushirono 1976) and Senpukuji in Nagasaki Prefecture (Aso 1984). Odai Yamamoto I was stratigraphically older than the Hachinohe Pumice, which then had a radiocarbon date of 12,700±260 BP (Gak-205) (Ikawa 1964, p. 98).
By 1980, archaeologists generally should have been giving the estimated age of Japan's oldest pottery, the Plain Pottery, as about 13,000 years. Instead, the mainstream of Japanese archaeology held to a date that was based on evidence from the 1960s, already more than 30 years old when these "surprising" new dates were announced in 1999.
Calibration Again: Calibrating radiocarbon dates for this oldest pottery was not possible until quite recently (see Bard et al. 1990; Stuiver & Reimer 1993; Stuiver et al. 1998). But by 1999, calibrating archaeological dates older than 12,000 years was rapidly becoming the global standard. Scientists investigating past climate changes, in particular, were studying the 18O/16O oxygen isotope ratios in the annual ice layers in the Greenland ice sheet (Stuiver, Grootes & Braziunas 1995) and the changing ratios of climate-sensitive micro-organisms in the annual sediments of lakes, effectively using calendar dates for the climate changes. Consequently, matching our knowledge of past human cultures to the natural environment they were adapting to now, by 1999, required archaeologists to calibrate their dates beyond 12,000 BP. This is what Taniguchi and his colleagues did with the finds from the 1998 excavation of the Odai Yamamoto I site (Odai Yamamoto I 1999).
In fact, at a major symposium in early December, 1998, five months before the Odai Yamamoto I dates were announced, one archaeologist predicted that the true age of the oldest pottery in Japan was about 15,000 years (Tsutsumi 1998, pp. 35-36, citing Kitagawa 1994 and Kitagawa & van der Plicht. 1998). Thus, archaeologists should not have been "surprised and confused" when Taniguchi Yasuhiro announced the calibrated dates for Odai Yamamoto I.
There are two academic organizations in the natural sciences that have made great efforts to work closely with archaeologists. These are the Japanese Society for Scientific Studies on Cultural Property and the Japan Association for Quarternary Research (geology). Both have roots going back at least into the 1960s. Yet fewer than 5% of Japanese archaeologists belong to these organizations.
Still, over the past 30 years, archaeological excavation reports with chapters on related natural science studies have gradually increased in number and are now rather common. But the natural science results usually are simply appended to the report and not integrated into the interpretations. And, when they are integrated, they usually are integrated poorly.
Japan's "Early Palaeolithic Hoax" is a world-class archaeological hoax involving 33 excavated sites and 153 surface-collected sites (Keally 2002). This hoax was in part a result of archaeologists ignoring, or openly contradicting, geological opinion for 20 years. Now the need to rewrite the school textbooks about the origins of rice farming and iron in Japan is the result of archaeologists willfully ignoring the natural science specialists in radiometric dating
And, despite the experience of the "Early Palaeolithic Hoax," some leading archaeologists still seem intent on ignoring or contradicting the opinions of highly experienced specialists in the related natural sciences. (NOTE 10)
Odai Yamamoto I Plain pottery is dated about 15,300-16,520 cal BP, or about 12,680-13,780 BP in uncalibrated radiocarbon years (Odai Yamamoto I Site 1999; Taniguchi 2001, 2002; Taniguchi & Kawaguchi 2001). The Plain Pottery from the Kitahara site in Kanagawa Prefecture has an uncalibrated radiocarbon age of 13,060-13,020 years, based on 5 samples (Taniguchi 2001, 2002; Taniguchi & Kawaguchi 2001). The calibrated age of this pottery would be about 15,650 years. The several other known Plain Pottery sites almost certainly date in the same range (Oda & Keally 1979). Linear-relief Pottery follows Plain Pottery in time. In uncalibrated radiocarbon years, the Kannoki Linear-relief Pottery would be about 13,000 years old. The Fukui Cave Linear-relief Pottery has an uncalibrated radiocarbon age of 12,700±500 years (Gak-950) (Kigoshi 1967c, p. 53; Kamaki & Serizawa 1967). Most other Linear-relief Pottery sites date a few centuries younger (Taniguchi 2002). Given the margin of error in all radiocarbon age measurements, it is stretching the information to say that the Kannoki Linear-relief Pottery is "the second oldest pottery" in Japan.
For the Kazahari site, D'Andrea (1995, pp. 207-208) gives two radiocarbon dates for rice grains associated with the Tokoshinai IV occupation of the site: 2540±240 BP (TO-2202) and 2810±270 BP (TO-4086), or 787 and 925 cal BC respectively. Tokoshinai IV pottery belongs near the end of Late Jomon and has an estimated date of about 3000-3100 BP, but the dates from the Kazahari site fit better with the known dates for early and middle Latest Jomon. However, there was no Latest Jomon occupation of the Kazahari site, and the dates are much too old for the next occupation there.
Ishikawa Hideji, a professor at Meiji University, studied the continental bronzes and compared them to the oldest Japanese find, a bronze arrowhead from the Imagawa site in Fukuoka Prefecture and dated to the beginning of Early Yayoi (Iida 2004). He claims this arrowhead matches the tip of the Liaoning type bronze spears, which are dated to the 8th century BC. He says this bronze would arrive in Japan "much later" and, therefore, the oldest bronze in Japan cannot date to the 8th century BC. Ishikawa feels it more probably dates to the 7th-5th century BC, perhaps a few centuries older than generally thought.
I see no reason why anything, object or idea, needs to appear in Japan "much later" than it appears on the continent. Even in the stone age it took only a few days to get from the Korean peninsula to the western part of Japan
To these doubts and criticisms, Imamura Mineo responded (Ito 2003a): "This dating method is used abroad for precise dating and it is published in [the highly respected journals] Science and Nature. If archaeologists want to question these Japanese dates, then they are taking on the whole world."
And other archaeologists point out that the new (older) dates make explainable some things that were not explainable with the old (younger) dates. Harunari Hideji (Ito 2003b) notes that a style of tooth extraction common in China until the Spring-Autumn Period (722-481 BC), but which disappeared after that, appeared in Japan in Early Yayoi (old dating