Home | Index Japaneses Archaeology last revised:
May 11, 1999

Middle Jomon Pottery Chronology in Kanto


Kanto was among the most densely settled regions in Japan during the Jomon Period. And Middle Jomon there had by far the highest density of population of any subperiod within the Jomon Period. Today, among the regions of Japan, Kanto has the highest density of archaeologists specializing in the Jomon culture, and also the largest number of excavations of Jomon sites annually. And most of the Jomon specialists emphasize pottery analysis in their research. The result is a uniquely (in the world?) detailed pottery sequence for Middle Jomon -- 31 temporal types (Kuroo 1995; Kuroo, Kobayashi & Nakayama 1995) defined for a period that lasted only about 800 uncalibrated radiocarbon years, from about 4800 to 4000 B.P. (Keally & Muto 1982). That calculates to about 26 years for the average pottery type, about one generation each. [In calibrated radiocarbon years, the dates are roughly 5600-4400 B.P., about 1,200 years, giving each of the 31 pottery types about 39 years, or just a bit more than one generation.]

Pottery Types in Jomon Archaeology

Jomon pottery types are temporal markers, developed on the bases of large clusters of formal and stylistic attributes and disrgarding all functional attributes. A single type includes a very wide range of formal types and an even wider range of stylistic attributes and motifs. Jomon pottery, especially Middle Jomon pottery, is extremely elaborate and varied, making it ideal for refining types (Kidder 1968; Kobayashi & Ogawa 1988a, 1988b). The Middle Jomon potters took full advantage of the plasicity of the medium they worked in -- and of the human imagination.

In pottery typology studies, Japanese archaeologists begin with "stratigraphy" -- they attempt to isolate at the time of excavation all the pots and potsherds that are the debris of a single face-to-face community. Through minute analysis of vertical and horizontal stratigraphy, they then set about determining the sequence of these point-in-time pottery assemblages. The next step is to describe each assemblage in detail. The final step is to look for any unique traits, motifs, and clusters of traits and/or motifs that define this point-in-time pottery assemblage. The archaeologists give these types within-site names, such as "III-a-1", meaning type III, subtype a, sub-subtype 1 (which generally is assumed to temporally precede Type III-a-2 in the same system).

Good reports will then compare these within-site types to the types in the standard sequences, using statements like: "Type III is Katsuzaka pottery, subtype III-a is Katsuzaka I, and sub-subtype III-a-1 is early Katsuzaka I; Type III-a-2 is middle Katsuzaka I," and so on. Reports that do not give the relationships to the standard names are effectively useless to most other archaeologists.

There is a problem with Japanese pottery typology, however -- disagreement on the definition of the types. What one archaeologist calls Kasori E-x another calls Kasori E-y. Kuroo, Kobayashi and Nakayama (1995) have provided a rare chart showing the relationships among different views of the Kanto Middle Jomon pottery types.

Chronologies in Japanese Archaeology

It is impossible to say anything about chronologies in Japanese archaeology without being extremely critical of Japanese archaeology itself. The vast majority of Japanese archaeologists do not seem to understand absolute dating techniques very well, or even the real need for absolute dates in archaeology.
  1. The last roughly thorough summary of Jomon dates was published in 1982 (Keally & Muto 1982). Even at that time, the number of dates available for the Jomon period was small, and I see no indication that the number has increased nearly as much as the 17 years since then would suggest it should have.
  2. There seems to be no awareness of the difference in the half-life 5730 and 5563 years (although the existing dates are so few and so rough that the difference is hardly significant at this point).
  3. The use of calibrated dates is effectively unheard of in Japan, and most archaeologists do not seem to know what "calibrated date" means.
  4. Dates published in excavation reports generally are simply appended. Some reports give no hint of the cultural association of the dates; others bury the information in the fine print of the lengthy text. I have never seen a report that truly integrated the dates into the interpretation of the site.
  5. Some archaeologists seem to think AMS dates give older ages than standard techniques; others appear to think AMS gives younger ages. No archaeologists seem to understand clearly that AMS simply gives more accurate dates with less material.
  6. Most archaeologists seem to have trouble understanding the meaning of the plus-minus standard error given with most dates.
  7. Lab numbers and citations are frequently not given for dates, making it difficult to confirm them.
  8. Most archaeologists compare radiocarbon dates to obsidian hydration dates, fission-track dates, thermoluminescence dates, and calibrated radiocarbon dates, as if they were not aware that these are all different systems with different sources of error.

Table 1: Middle Jomon Pottery Chronology in the Kanto District.
Phase /
end Early Jomon Moroiso a, b 5,290±138 (N-38b)
5,260±110 (Gak-5368)
5,100±400 (M-240)
4,970±80 (TK-1)
4,730±90 (Gak-379a)
4,760±90 (Gak-379b)
6,200±150 (SKJ-4)
Moroiso c
5,800±200 (II-11)
5,900±300 (II-17)
Goryogadai 1a Goryogadai I Kagohata 5,650±100 (Suzuki 1977)
5,500±300 (Suzuki 1977)
5,600±350 (Suzuki 1977)
5,400 (II-12)
4,850±340 (FT-557-MS-1)
4,950±180 (FT-557-MS-1)
4,800±300 (FT-558-MS-1)
4,950±160 (FT-558-MS-2)
2 Goryogadai IIa Kyubeione
3a Goryogadai IIb
4a Goryogadai IIc
Katsuzaka 5a Katsuzaka I Mujinazawa 5,240±170 (Gak-8672)
6a Aramichi .
6b .
7a Katsuzaka II Tonai I 4,660±90 (N-3863)*
4,510±110 (N-3864)*
4,800±180 (Gak-8044)
4,760±140 (Gak-8045)
4,700±140 (Gak-8049)
4,410±120 (Gak-8047)
7b .
8a Tonai II .
8b .
9a Katsuzaka III Idojiri I .
9b Idojiri II .
9c Idojiri III 5,070±220 (Suzuki 1975)
4,900±400 (Suzuki 1975)
4,950±220 (Suzuki 1975)
Kasori E 10a Kasori E I Sori I old/new 4,790±80 (Gak-1068)
4,590±65 (N-3815)*
4,550±90 (N-3817)*
4,520±100 (N-3865)*
4,590±90 (N-3866)*
4,450±100 (N-3814)*
3,910±95 (N-3816)*
10b Sori I new
10c Sori II old
11a Kasori E II
Sori II new 4,170±105 (N-1431)
3,770±110 (Gak-7307)
5,200±150 (NKJ-7)
5,000 (NKJ-8)
4,600 (SKJ-14)
4,700 (SKJ-15)
4,800 (SKJ-16)
4,800±150 (SKJ-17)
4,700 (SKJ-21)
4,700±300 (FT-359-MS-1)
11c Sori III old
12a Kasori E III Sori III new 5,110±90 (Gak-8396)
4,340±110 (Gak-8395)
4,170±140 (Gak 8397)
12b Sori III new/
Sori IV
12c Sori V old
13a Kasori E IV . .
13b . .
Late Jomon /
Shomyoji I
14a Kasori E IV cont'd
Kasori E V
Shomyoji 1
. 4,090±110 (Gak-9453)
14b Shomyoji 2 .
14c Shomyoji 3 .
15a Shomyoji 4 .
15b Shomyoji 5 .
Shomyoji II . Shomyoji 6 . .
. Shomyoji 7 . .
Horinouchi . Horinouchi I, II . 5,150±160 (Gak-8440)
4,010±145 (N-3484)*
3,940±105 (N-1429)
3,880±150 (N-59)
3,840±190 (N-1430)
3,790±160 (Gak-8441)
3,580±130 (Gak-8443)
3,250±137 (N-128)
2,190±110 (Gak-8442)
3,700 (SKJ-25)
3,600±400 (II-19)
(after Kuroo, Kobayashi & Nakayama 1995; dates from Keally & Muto 1982).
* dates calculated with half-life 5,730 years.
RED dates are OB-FT (Suzuki 1974, unless cited otherwise).
BLUE dates are FT dates by Suzuki.

Suzuki (1974) uses the symbols NKJ (North Kanto Jomon), SKJ (South Kanto Jomon), and II (Izu Islands) like lab nos. for the dates. Keally and Muto (1982) give the full information on these dates.