Professional Documents
Culture Documents
Prehistoric Maritime Cultures and Seafaring in East Asia: Chunming Wu Barry Vladimir Rolett Editors
Prehistoric Maritime Cultures and Seafaring in East Asia: Chunming Wu Barry Vladimir Rolett Editors
Prehistoric Maritime Cultures and Seafaring in East Asia: Chunming Wu Barry Vladimir Rolett Editors
Chunming Wu
Barry Vladimir Rolett Editors
Prehistoric
Maritime Cultures
and Seafaring
in East Asia
The Archaeology of Asia-Pacific Navigation
Volume 1
Series Editor
Chunming Wu, The Center for Maritime Archaeology, Xiamen University,
Xiamen, Fujian, China
This series will publish the most important, current archaeological research on ancient navigation and sea routes
in the Asia-Pacific region, which were key, dynamic factors in the development of human civilizations spanning
the last several thousand years. Restoring an international and multidisciplinary academic dialogue through
cross cultural perspectives, these publications underscore the significance of diverse lines of evidence, including
sea routes, ship cargo, shipwreck, seaports landscape, maritime heritage, nautical technology and the role of
indigenous peoples. They explore a broad range of outstanding work to highlight various aspects of the
historical Four Oceans sailing routes in Asia-Pacific navigation, as well as their prehistoric antecedents, offering
a challenging but highly distinctive contribution to a better understanding of global maritime history.
The series is intended for scholars and students in the fields of archaeology, history, anthropology,
ethnology, economics, sociology, and political science, as well as nautical technicians and oceanic scientists
who are interested in the prehistoric and historical seascape and marine livelihood, navigation and nautical
techniques, the maritime silk road and overseas trade, maritime cultural dissemination and oceanic immigration
in eastern and southeastern Asia and the Pacific region.
The Archaeology of Asia-Pacific Navigation book series is published in conjunction with Springer under the
auspices of the Center for Maritime Archaeology of Xiamen University (CMAXMU) in China. The first series
editor is Dr. Chunming Wu, who is a chief researcher and was a Professor at the institute. The advisory and
editorial committee consists of more than 20 distinguished scholars and leaders in the field of maritime
archaeology of the Asia-Pacific region.
Advisory Board:
Editorial Board:
Editorial in Chief:
Editors
123
Editors
Chunming Wu Barry Vladimir Rolett
The Center for Maritime Archaeology Department of Anthropology
Xiamen University University of Hawaii at Manoa
Xiamen, Fujian, China Honolulu, HI, USA
This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Nephrite lingling-o ear ornament excavated from the Go Ma Voi archaeological site, Vietnam
(Institute of Archaeology, Hanoi, Vietnam. Drawing by Yidian Gao.)
Preface
vii
viii Preface
collaboration in Jiangxi, laughed aloud while exclaiming: “If I can move from
Mexico to China, you can certainly expand from Polynesia to China!”
Finally, an opportunity arose in 2001, when as Visiting Professor at Harvard
University, Rolett began a project on the Austronesian homeland in collaboration
with Gongwu Lin and Xuechun Fan of the Fujian Provincial Museum. This was the
Museum’s first international collaboration. Meeting in 2003 and discovering their
shared interests, Wu invited Rolett to attend the 2004 Annual Meeting of the
Association for the Study of Baiyue Ethnicities of southern China, held in
Wuyishan, a national park area in western Fujian. From here on, everyone involved
became excited to expand the China–US collaboration.
In 2006, Wu applied for and was awarded a Henry Luce Foundation/American
Council of Learned Societies Fellowship to spend the 2006/2007 academic year at
the University of Hawaii at Manoa, hosted by Rolett. Wu’s work in Hawaii
investigated the relationship between indigenous peoples of southeast China and the
Austronesian maritime cultures of Southeast Asia and Oceania. As stated in his
fellowship application, two of his project goals were: (1) to “overcome the aca-
demic barrier between Chinese and American archaeologists which seriously
restricts the understanding of the aboriginal history of Asia and Pacific” and (2) to
“enhance the exchange of different ideas, theory, and methodology of the two
countries’ archaeologists.” Spending a year together at the University of Hawaii
laid the foundation for future exchanges and collaboration, even as Rolett began
working with paleoenvironmental scientist Zhuo Zheng (Sun Yat-Sen University)
and Wu branched out to study China’s role in the trans-Pacific Manila galleon trade.
During the 2012/2013 academic year, Wu came to the USA again, this time with
a fellowship from the Harvard-Yenching Institute to undertake 10 months of
independent research at Harvard University as Visiting Scholar in the Department
of Anthropology, where he was hosted by Archaeologist Rowan Flad. The goal of
Wu’s fellowship project was to develop an alternative, indigenous-focused history
of the origin of maritime cultures in the seas surrounding China, to challenge the
Han-centric bias held by traditional Chinese historians. That research, and discus-
sions with an ever-expanding group of colleagues with overlapping interests and
areas of expertise, led directly to the 2017 conference on which this book is based.
In conclusion, we wish to express our gratitude to the scholars who attended the
conference and who contributed, so eagerly and generously, their diverse per-
spectives. Their expertise and collegiality created an ideal environment for
discussion and feedback. We also gratefully acknowledge the Xiamen University
Center for Maritime Archaeology and the University of Hawaii at Manoa,
Department of Anthropology, for supporting us, and the group as a whole, in
organizing and convening the conference. As the conference was held in Xiamen,
most of this responsibility was fulfilled by Xiamen University. A number of dis-
tinguished colleagues, not represented among the authors of this book, generously
agreed to moderate various panels, to lead discussions and to offer their professional
comments during the meeting. For this, we are grateful to: Hua Sun (Peking
University, School of Archaeology and Museology), Zhijun Zhao (Archaeology
Institute of the Chinese Academy of Social Sciences), Junlei Zheng (Sun Yat-Sen
Preface ix
xi
xii Introduction
In the opening Chap. 1, Chunming Wu discusses how coastal East Asian Neolithic
cultures served as an incubator for the development of the historic era Maritime Silk
Road sea route. He examines in detail, and refutes, the widely held notion that the
Maritime Silk Road emerged primarily as an extension of the inland Silk Road,
which reached its peak during China’s Tang Dynasty (AD 618–907). Wu
demonstrates how seafaring technology and the core trade routes of the Maritime
Silk Road, the “Four Seas” and “Four Oceans” navigation system recorded in
ancient China, are in fact deeply rooted in the Neolithic archaeological cultures of
coastal East and Southeast Asia. At the heart of these ancient maritime traditions are
indigenous cultures such as the Yi and Yue, as they are known to Han Chinese. As
Wu shows, the documented maritime traditions of these indigenous cultures once
dominated the Four Seas (the North, East, South, and West Oceans) surrounding
China.
Introduction xiii
In Chap. 7, Zhen Li explores Neolithic shell middens as well as sand dune and
cave sites along the Beibu Gulf coast, extending south of Guangdong. This tropical
coastline includes Guangxi and northern Vietnam on the mainland, in addition to
much of the coast of Hainan Island. Throughout the Beibu Gulf coast area, stratified
archaeological sites have revealed the existence of Pleistocene hunter–gatherer
populations overlaid by later settlements distinguished by Neolithic artifacts such as
pottery and ground stone adzes. Remarkably, however, the Neolithic occupations
bear little to no evidence of agriculture. Instead, the existing subsistence evidence
shows striking continuity between the Pleistocene and the Neolithic periods, with
an emphasis throughout on the exploitation of marine resources and wild plants,
especially sago-type palms (Yang et al. 2013).
Following these surveys of the China coast, Chap. 8, by Kim Dung Nguyen,
offers an overview of known archaeological cultures from the coast of northern
Vietnam. This area, also discussed by Higham in Chap. 2, features significantly in
the two-layer hypothesis. Nguyen, who has directed excavations at some of the
most important coastal sites in northern Vietnam, describes the area as a cultural
crossroads. Indeed, in addition to evidence for migrating populations from the
north, this area appears to have been the center of a Neolithic interaction sphere for
the manufacture and distribution of decorative jade ornaments. Taken together, this
evidence suggests the existence of interaction and exchange linking the Neolithic
population of northern Vietnam with coastal populations to both the north and the
south, and possibly the Philippines as well.
With Chap. 9, by Ling Qin and Dorian Fuller, we return to the topic of Neolithic
rice agriculture. In particular, Qin and Fuller question the notion that irrigated rice
agriculture was a driver of demographic expansion, including the crossing of the
Taiwan Strait during the Neolithic. While not denying that irrigated rice agriculture
produces much higher yields than shifting rice cultivation strategies, they argue that
the archaeological record does not support the notion that wet rice fueled outbound
and overseas migration. Instead, they propose that migrating Neolithic farmers
practiced mainly low-yield rice production based on shifting cultivation, which may
explain the scarcity of rice in the early Neolithic sites of southern China and
northern Vietnam.
Part I of this book addresses central questions relating to the “who,” “what,”
“where,” and “when” of early trade, seafaring and maritime-oriented cultures in our
study area, extending from China to Southeast Asia and the neighboring seas.
Questions concerning the “how” are equally important, but often much more
challenging. For example, did migrating farmers travel by water, as well as by land?
And if so, what kind of watercraft did they use? This question comes to the
Introduction xv
Neolithic through the Spanish colonial period. Events in this maritime history
include the initial crossing of the Balintang Channel separating Taiwan from the
northern Philippines, the establishment of regular contact between the Philippines
and central Vietnam during the Metal Age, and the trans-Pacific Manila galleon
trade, which foreshadowed the modern era of globalization.
The Ryukyus form an alignment of stepping stone islands reaching from
southern Japan to Taiwan. In the Chap. 17, Naoko Kinoshita examines archaeo-
logical evidence for seafaring and inter-island exchange in the Ryukyus. She argues
that Neolithic era voyaging networks in this region emerged largely in relation to
the intervisibility among islands. In general, connectivity among the northern
Ryukyus and Kyushu (Japan’s southernmost main island) was highly developed,
while far less contact occurred between the southern Ryukyus and Taiwan.
Kinoshita’s study helps to explain why Austronesians did not migrate from Taiwan
to Japan. It also offers a framework for developing and testing seafaring hypotheses
based on model simulations and experimental voyaging, a strategy already in use
for studies of Pacific voyaging (Finney 1994; Irwin 1992).
References
Bellwood, P. (2018). The search for ancient DNA heads east. Science, 361(6397), 31–32.
Chang, C. S., Liu, H. L., Moncada, X., Seelenfreund, A., Seelenfreund, D., Chung, K. F. (2015).
A holistic picture of Austronesian migrations revealed by phylogeography of Pacific paper
mulberry. Proceedings of the National Academy of Sciences, 112(44), 13537–13542.
Finney, B. (1994). Voyage of Rediscovery. Berkeley: University of California Press.
Hung, H. C., Carson, M. T., Bellwood, P., Campos, F. Z., Piper, P. J., Dizon, E. et al. (2011). The
first settlement of Remote Oceania: the Philippines to the Marianas. Antiquity, 85(329),
909–926.
Irwin, G. (1992). The prehistoric exploration and colonisation of the pacific. Cambridge:
Cambridge University Press.
Jiang, L., & Liu, L. (2005). The discovery of an 8000-year-old dugout canoe at Kuahuqiao in the
Lower Yangzi River, China. Antiquity, 79(305), 1–6.
Lipson, M., Cheronet, O., Mallick, S., Rohland, N., Oxenham, M., Pietrusewsky et al. (2018).
Ancient genomes document multiple waves of migration in Southeast Asian prehistory.
Science, 361, 92–95.
Matsumura, H., Hung, H. C., Higham, C., Zhang, C., Yamagata, M., Nguyen L. C. et al. (2019).
Craniometrics Reveal “Two Layers” of Prehistoric Human Dispersal in Eastern Eurasia.
Scientific Reports, 9(1), 1451.
McColl, H., Racimo, F., Vinner, L., Demeter, F., Gakuhari, T., Moreno-Mayar, J. V. et al. (2018).
The prehistoric peopling of Southeast Asia. Science, 361(6397), 88–92.
Rolett, B. V. (2007). Southeast China and the emergence of Austronesian seafaring. In T. Jiao
(Ed.), Lost Maritime Cultures: China and the Pacific (pp. 54–61). Honolulu: Bishop Museum
Press.
Yang, X., Barton, H. J., Wan, Z., Li, Q., Ma, Z., Li, M. et al. (2013). Sago-type palms were an
important plant food prior to rice in southern subtropical China. PLoS One, 8(5):e63148.
Contents
xix
xx Contents
Chun-Yu Chen Institute of History and Philology, Academia Sinica, Taipei City,
Taiwan
Eusebio Z. Dizon National Museum of the Philippines, Manila, Philippines
Lin Fu Department of History, Xiamen University, Xiamen, China
Dorian Q. Fuller Institute of Archaeology, University of London, London, UK;
School of Cultural Heritage, Northwest University, Xi’an, China
Mana Hayashi Tang Department of Anthropology, Washington University in
St. Louis, St. Louis, MO, USA
Maya Hayashi Tang Department of History, The Chinese University of Hong
Kong, Hong Kong, China
Charles F. W. Higham Department of Anthropology and Archaeology,
University of Otago, Otago, New Zealand
Hsiao-chun Hung Australian National University, Canberra, Australia
Tuukka Kaikkonen Canberra, Australia
Naoko Kinoshita Faculty of Humanities and Social Science, Kumamoto
University, Kumamoto, Japan
Su-Chiu Kuo Institute of History and Philology, Academia Sinica, Taipei City,
Taiwan
Yan Li Guangdong Institute of Cultural Relics and Archaeology, Guangzhou,
Guangdong, China
Zhen Li Institute of Cultural Relics Conservation and Archaeology of Guangxi,
Guangxi, China
Ting Ma School of Geography and Planning, Sun Yat-Sen University, Guangzhou,
China
xxi
xxii Contributors
xxiii
xxiv List of Figures
Fig. 8.3 Stone tools from the Ha Long Culture (left: main types
of stone tools; right: pointed tools from the Bai Cat
Don Site) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 146
Fig. 8.4 Map showing sites associated with the Ha Long Culture . . . . .. 147
Fig. 8.5 Stone and bone tools from the Ha Long Culture found
at the Hon Hai Co Tien Site in 2004 . . . . . . . . . . . . . . . . . . . .. 147
Fig. 8.6 Ha Long pottery from the Bai Ben Site . . . . . . . . . . . . . . . . . .. 148
Fig. 8.7 Excavation at the Trang Kenh Site, 1996 (photo courtesy
of Tang Chung) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Fig. 8.8 Dau Ram Site (photo courtesy of the author) . . . . . . . . . . . . . . 150
Fig. 8.9 Nephrite core from Trang Kenh . . . . . . . . . . . . . . . . . . . . . . . . . 151
Fig. 8.10 Nephrite discoid cores from Trang Kenh . . . . . . . . . . . . . . . . . . 151
Fig. 8.11 Evidence of regional cultural exchange, including: nephrite
T-section ring from the Phung Nguyen Culture (1 and 3);
nephrite T-section ring from the Man Bac Site (2); and similar
decorative motifs on pottery from the Man Bac (4), Phung
Nguyen (5), and Dau Ram (6) Sites . . . . . . . . . . . . . . . . . . . .. 153
Fig. 8.12 Nephrite ornaments found at the Trang Kenh Workshop
Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 154
Fig. 8.13 Nephrite objects and grinding saws from the Dau Ram
Workshop Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 155
Fig. 8.14 Bronze and Early Iron Age jade slit rings from the coast
of Vietnam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 155
Fig. 9.1 Map of Lower Yangtze River. 1. Kuahuqiao (跨湖桥), 2.
Hemudu (河姆渡), 3. Tianluoshan (田螺山), 4. Majiabang
(马家浜), 5. Caoxieshan (草鞋山), 6. Chuodun (绰墩), 7.
Chenghu (澄湖), 8. Jiangli (姜里), 9. Liangzhu ancient city
(良渚古城), 10. Maoshan (茅山) . . . . . . . . . . . . . . . . . . . . . . .. 163
Fig. 9.2 Paddy fields and agricultural tools of the Lower Yangtze
River a paddy fields of the Chuodun site (Fuller et al. 2009)
b paddy fields of the lower layer of the Maoshan site
(Illustration by L. Qin) c paddy fields of the top layer of the
Maoshan site (Zhuang et al. 2014) d agricultural tools of the
Lower Yangzte (Illustration by L. Qin) . . . . . . . . . . . . . . . . . .. 165
Fig. 9.3 Material culture reflects wetland management of the Lower
Yangtze River a animal images from Liangzhu jades and
pottery decoration (from exhibition at the Liangzhu Museum)
b canoe from the Kuahuqiao site (Zhejiang Provincial Institute
of Archaeology and Culture Relics et al. 2004) c canoe from
the Maoshan site (lower layer) (photograph by L. Qin) . . . . . .. 168
Fig. 9.4 Different forms of landscape engagement are reflected in
dietary stable isotopes. 5 different types can be recognized.
Type 1 (lower left) is the Lower Yangtze type characterized
by C3 wild plants, freshwater wetland resources and terrestrial
List of Figures xxxiii
Table 5.1 Time table showing shell middens in four different regions
during the pre-Qin period . . . . . . . . . . . . . . . . . . . . . . . . . . .. 91
Table 5.2 Domesticated animal and plant remains found in shell
middens of China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 94
Table 9.1 Estimated rice consumption, land requirements and carrying
capacity for Yangtze River Valley . . . . . . . . . . . . . . . . . . . .. 175
Table 9.2 Estimated millet consumption, land requirements and
carrying capacity for Yellow River Valley . . . . . . . . . . . . . .. 177
Table 12.1 Neolithic culture phases of central and northern coastal
Fujian (based on Lin 2005) . . . . . . . . . . . . . . . . . . . . . . . . . .. 224
Table 12.2 Neolithic culture phases of Taiwan and the Straits islands
(after Hung and Carson 2014) . . . . . . . . . . . . . . . . . . . . . . . .. 227
Table 14.1 Spatiotemporal structure of prehistoric cultures in Taiwan . .. 253
xxxix
Part I
Neolithic and Metal Age Maritime Cultures
of South China and Southeast Asia
Chapter 1
A Synthetic Analysis of the Neolithic
Origins of Eastern and Southeastern
Asia’s Maritime Silk Road
Chunming Wu
Abstract The “Silk Road” is a focus of historical and archaeological research that
explores the ancient routes of transportation between the East and West. Most
academic scholars in the last few decades have assumed that the Maritime Silk
Road emerged much later than the one on land and was part of the Silk Road’s
broader geographic shift from land to sea. However, according to recent investi-
gations of southeast China’s cultural heritage in prehistoric archaeology and
ethno-archaeology, the maritime transportation known as the “Four Oceans”
Navigation system has in fact existed since the Neolithic Age, when the prehistoric
ancestors of indigenous Yi (夷) and Yue (越) peoples of coastal China began
navigating between mainland and offshore islands. Thus, the development of the
maritime Silk Road in the southeast of China and Asia occurred no later than the
land Silk Road of the inland northwest areas of China and inner Asia. The pre-
historic Yue ethnicity and proto-Austronesian peoples in mainland southeastern
China initiated first nautical techniques with simple sailing devices, compound
canoes and celestial navigation. These developments thousands years ago became
the fundamental building blocks for the development of the Maritime Silk Road
during the late historical period. As a result, the Maritime Silk Road was not, as
previously assumed, the product of cultural change in the land Silk Road or the shift
of ancient China’s economic center from north to south.
1.1 Introduction
Various interpretations of both the “Silk Road” and the “Maritime Silk Road” have
been proposed by historians and geographers investigating ancient transport
between China and other countries. Historians and historical archaeologists have
generally viewed both terms as symbolic concepts representing the transportation
routes between East and West that crossed land and sea, respectively (Chavannes
C. Wu (&)
The Center for Maritime Archaeology, Xiamen University, Xiamen, China
e-mail: wu_chunming@hotmail.com
1903; Richthofen 1877). They also insisted that the Maritime Silk Road had
emerged much later than the land Silk Road and had been the result of cultural
change and the geographical shifting of ancient China’s social and economic center
from the northern plains to the southeastern coastal region. They hypothesized that
this change was driven by the decline of the inland Silk Road, combined with road
blockages from war and other turmoil that raged in northwest China and inner Asia
from the Six dynasty to the Tang dynasty.
Historical accounts record how ancient Chinese maritime transportation had
included a series of sub-regional navigation practices based on what was known as
the “Four Seas” and “Four Oceans.” The “Four Oceans” system began in a series of
different seaports in the coastal region of southeast China. Yet prehistoric archae-
ology has revealed that the historical “Four Oceans” framework actually originated
in the Neolithic seafaring routes of the native Yi (夷) and Yue (越) in southeast
China. These prehistoric navigators created an early maritime economy,
marine-based culture and primitive sailing techniques. They also completed the first
nautical voyages across the seas around China, which took place no later than the
height of the Han dynasty’s inland Silk Road. As a result, both the Silk Road and
the Maritime Silk Road developed independently in two areas, northwest inland
China and southeast coastal China, and the emergence of a seafaring route had little
to do with changes or challenges to the Silk Road on land.
Starting in the early twentieth century, historians and archaeologists proposed the
concept of a “Maritime Silk Road” and promptly fell into a debate over what this
actually meant. Most historians focused on historical trade between East and West and
defined the “Maritime Silk Road” as nautical transportation between China and Europe
via the South China Sea and the Indian Ocean. This hypothesis was first proposed by
the European explorers and sinologists who began traveling through the inner regions
of Asia in the late nineteenth century. They used the phrase “Silk Road” to encompass
all the ancient, inland trade that took place between the East and the West, while the
phrase “Maritime Silk Road” came to symbolize the ancient sea route between China
and Europe (Chavannes 1903: 208; Richthofen 1877: 477, 496, 500, 507).
The first European explorer to coin the term “Maritime Silk Road” was the
French sinologist, Edouard Chavannes. His book, Documents Sur les Tou-kiue
occidentaux (Historical Documentation of the Western Tou-Kiue), published in
1903, described the historical routes of the Chinese silk trade between the East and
the West. He listed both land and sea routes, including the “older trade route” across
inner Asia and a “new sea route” that crossed the Indian Ocean and passed through
a series of Indian seaports. Thus, the so-called “Maritime Silk Road” was imme-
diately identified as a later development than the inland Silk Road (Chavannes
1903: 208).
1 A Synthetic Analysis of the Neolithic Origins … 5
At Peking University, Professor Meicun Lin (林梅村) voiced the same opinion:
Since the Tang dynasty, the main transportation route between East and West gradually
shifted toward the sea. In time, this Maritime Silk Road completely replaced the traditional
inland Silk Road as the main route between ancient China and foreign countries (Lin 2006: 4).
From 1990 to 1991, the United Nations Educational, Scientific and Cultural
Organization (UNESCO) organized an international inspection of the surviving
cultural heritage “along” the historical Maritime Silk Road. Starting in Italy, the
inspection team sailed through the Mediterranean Sea, the Suez Canal, the Red Sea,
the Indian Ocean, Malacca Straits and the South China Sea to arrive in eastern Asia.
By “following” the ancient path of the “Maritime Silk Road” across Europe and
into Asia, this inspection affirmed the existing historical and archaeological research
indicating the crucial geographic role of the Maritime Silk Road in historically
linking the East and the West (Su 1991, pp. 1–40; OCQZ.UNESCO 1991, pp. 1–3).
6 C. Wu
Historians and archaeologists generally recognize the Silk Road and the Maritime
Silk Road as an integrated transportation system between East and West. Scholars
have also assumed that the Maritime Silk Road across the South China Sea and the
Indian Ocean had been the result of cultural changes to Euro-Asian trade after the
inland Silk Road across inner Asia declined in the Tang dynasty. According to
historical documents, however, ancient Chinese maritime transportation had already
developed systematically, not only in the South China Sea and the Indian Ocean,
but also in the North and East China Seas. In a series of historical documents,
contemporary observers describe the Maritime Silk Road as an integrated naviga-
tion system of “Four Seas” and “Four Oceans.”
During the Zhou and Han dynasties, the “Four Seas” referred to a remote region
far away from the center of the empire, illustrating the diverse regions utilized by
early seafaring and maritime cultures in the seas around China.
The chapter of “Admonition of King Yi (伊训)” of Shangshu (Book of Early
History 《尚书》) recorded:
The love and respect should exist in all family of our mother land from central state to the
territorial boundary at the four seas (Ruan, ed. 2009, pp. 344).
The chapter of “Writing of the Master Gao (告子)” of Mengzi (Book of the
Master Mencius 《孟子》) stated:
The four seas had been the gullied boundary 141 of the Xia (夏) kingdom during the Yu
(禹) period (Ruan, ed. 2009, pp. 6008).
Han Shu (History of the Han Dynasty 《汉书》) recorded in the Vol.
96 “Biography of the West Territory (西域传)”:
The water stream in the west of Yutian (于阗, now Hetian [和田] county of Xinjiang [新
疆]) flow westward to the West Sea, while the water stream in the east of Yutian flow
eastward to the saltlake and to be the source of Yellow River….The Tiaozhi (or ntiochia, 条
支, Now Iran and Iraq) state is located next to the West Sea (Ban, G. 1962, pp. 3881, 3888).
Houhan Shu (History of the Later Han Dynasty 《后汉书》) also said in the
chapter of “Biography of the Southwest Barbarians” (南蛮西南夷列传) that the
great empire of Daqin (大秦, ancient Rome) had been located next to the West Sea
(Fan 1965, pp. 2851).
The earliest hydrogeographical book Shuijing Zhu (Commentary on the
Waterways Classic 《水经注》) also recorded:
Anxi (安息, or Parthian, now Iran) empires is located next to the West Sea….Chongling
(葱岭, now Pamirs [帕米尔] plateau) is located at the watershed of West Sea and head-
water of Yellow River (Li and Wang 1984, pp. 31–33).
During the subsequent Tang and Song dynasties, the historical “Four Seas”
navigation system developed into a more extensive network of sea routes that
historical documents define as “Four Oceans Navigation” (Chen 1992). “Four
Oceans Navigation” refers to a series of historical sea routes and sub-regional
navigational systems focused around the southeastern coast of China (mainly Fujian
and Guangdong regions), illustrating the existence of a trans-border maritime
community in China’s seas before European contact. Historical references to this
system include the following.
The geographical work of Song dynasty Lingaai Daida (Interlocution on the
History of South Coast of China, 《岭外代答》) recorded in the chapter of
“Foreign Oceanic Nations (海外诸蕃国)”:
Samboja (三佛齐, now Sumatra in Indonesia) is located to the big South Ocean in which
there are thousands of islands. The big East Ocean is located to the east of Yavadvipa (阇
婆, now Java in Indonesia) and next to the women state (Zhou 1996, pp. 37–42).
Nanhai Zhi (Biography of the South China Sea 《南海志》) included the
chapter of the “Foreign Oceanic Nations (诸番国)”. It recorded:
The Foni (佛坭, now Brunei) state rules the small East Ocean region including the
Philippines and north of Kalimantan Island. The Saltanah Sulu (苏禄, now Sulu) and
Yavadvipa rule the big East Ocean including Java, south of Kalimantan, Sulawesi, Timor,
Maluku islands (Chen 1986, pp. 37–38).
Zhang Xie said in Dongxiyang Kao (General Survey on the East and West
Oceans 《东西洋考》) of late Ming dynasty:
“The South China Sea covered two navigation regions resulting different maritime cultures,
the West Ocean and East Ocean. Brunei state is located in the watershed region where the
West Ocean separates from the East Ocean.” The East Ocean includes Taiwan, Philippines,
and the area of northeast Kalimantan. “Taiwan is included in the small East Ocean” (Zhang
1981a, b, pp. 102, 184, 185).
Because the West Ocean route connected with the South Ocean route in the
South China Sea, the South Ocean Navigation system was generally included as
part of the West Ocean Navigation network. Thus a series of seaports such as
Cochin (交趾, the coast of north Vietnam), Champa (占城, the coast of south
Vietnam), Malacca and Timor were listed as being located in the West Ocean
(Zhang 1981a, b, pp. 1, 21, 66, 87).
Thus, the historical documents of ancient China illustrate how the maritime nav-
igation of the “Four Seas” and the “Four Oceans” was a system for pluralistic cultural
integration in the seas around China. “Four Ocean Navigation” refers to a series of
sub-regional sea routes that branch out from the southeast coast of China. One of these
was the “East Ocean” to Taiwan, Philippines and eastern Indonesia, which had been
the main route for the movement of proto-Austronesians out into the Pacific since the
Neolithic Age. The “South and West Ocean” stretched across the South China Sea
and the Indian Ocean and was known as the “South China Sea Route via Xuwen and
Hepu” (Xuwen Hepu Nanhaidao, 徐闻、合浦南海道), or the “Canton Sea Route”
(Guangzhou Tonghai Yidao, 广州通海夷道). The “North Ocean” crossed the Yellow
Sea and Bohai Sea and was known as the “Nautical Route to Koryo and Balhae from
Tengchow” (Dengzhou Haixing Ru Gaoli Bohaidao, 登州海行入高丽、渤海道).
This navigation system and the resulting maritime cultural interaction effectively
created an extensive maritime community in the seas around China before European
contact and their maritime globalization. As a result, the Maritime Silk Road became
not just a linear sea route between the West and the East, but a diverse and com-
plicated maritime transportation system that included a series of sub-regional sea
routes through oceans located to the north, east, south and west. The previously
accepted hypothes is regarding how the Maritime Silk Road between East and West
emerged from the decline and subsequent geographical shift of the inland Silk Road
thus deeply misunderstood the history of navigation systems in the coastal China and
the ancient “Four Seas” and “Four Oceans” (Fig. 1.1).
Though some cultural interaction between the inland Silk Road and the Maritime
Silk Road might have occurred in the course of transportation between East and
West, it seems clear that the Maritime Silk Road was not a cultural continuation or
1 A Synthetic Analysis of the Neolithic Origins … 9
Fig. 1.1 A historical reconstruction of the Four Oceans navigation system in the seas around
China
replacement for the inland Silk Road following the Tang dynasty. Prehistoric and
early historic archaeology reveals that the northern plains region of China (or
Zhongyuan, 中原) had been the cultural center for advanced inland agricultural
development since the Neolithic Age, while the prehistoric and historical inhabi-
tants of China and Southeast Asia’s coastal regions had been more focused on
maritime culture. Both maritime culture and inland agriculture later became key
components in the pluralistic cultural integration that resulted in the emergence of
ancient Chinese civilization.
Before the 1970s, when the meanings of the “Silk Road” and the “Maritime Silk
Road” were first introduced to the Chinese academy, Chinese archaeologists and
ethnologists had already begun a series of research projects focusing on the pre-
historic cultural diversity of eastern Asia—specifically the differences between the
maritime cultures of southeastern China and Asia and the inland farming culture of
central and northwestern China.
For example Professor Huixiang Lin (林惠祥), from Amoy (Xiamen) University
of southeast China, explored the Neolithic cultural characteristics of the “Maritime
Region of Southeastern Asia” (亚洲东南海洋地带). He described how the pre-
historic culture of southeast coastal China had differed from that of northern China
through its connections and mutual interaction with the maritime cultures of the
neighboring peninsulas and islands to the south, including Malaysia and even parts
of Polynesia (Lin 1937).
10 C. Wu
Professor Chunshen Lin (凌纯声) from Academia Sinica, went a step further by
reconstructing the regional maritime cultural community, or “Asian Mediterranean
Cultural Circle” (亚洲地中海文化圈) that covered the south and east coasts of
China, parts of the mainland and the archipelagoes of Southeast Asia. He identified
the “Asian Mediterranean Culture” as belonging to the Yi (夷) and Yue (越), two
prehistoric and early historic indigenous groups that had been characterized as
maritime culture through their patterns of marine shell utilization, canoe techniques
and tattoo. Lin then compared these maritime cultural systems of southeastern
China and Asia with the inland farming cultures of northwestern China and central
Asia that anchored the traditional imperial civilization of ancient China (Lin 1954).
The leading authority on Chinese archaeology, Professor Binqi Su (苏秉琦),
from the CASS Archaeology Institute, argued that the Neolithic culture of China
should be divided into two different regional systems: the maritime culture and
marine subsistence practices observed along China’s southeast coast from
Shandong (山东) Peninsula to southern Guangdong (广东), and the agricultural
culture focused on the inland areas of the northwest, including the central plains,
that had given rise to China’s earliest recognized civilization (Su 1978).
Building upon this argument, the inland trade route across inner Asia that is
known as the “Silk Road” and the corresponding sea route in the Asia-Pacific region
known as the “Maritime Silk Road” were not necessary an integrated transportation
system, but instead probably developed separately over several thousand years. Thus
the Maritime Silk Road was not a continuation of the inland Silk Road of inner Asia,
but rather its own independent trade network. Simply put, the historical hypothesis
regarding maritime navigation “replacing” land-based routes overlooks a rich and
deep history of maritime transportation off the coasts of China.
In fact, this Maritime Silk Road developed no later than its inland
counterpart. Archaeological investigations and ethnohistorical analysis reveal that
prehistoric maritime transportation and early navigation were developed in the seas
around China by several non-Han indigenous groups, including the Yi and Yue,
since the early Neolithic Age. Indeed the Yi and Yue, who lived in the coastal
regions of eastern and southern China, carried out Neolithic seafaring and
marine-focused cultural practices at least 7000 to 8000 years ago. These native
maritime cultures and their Neolithic seafaring activities essentially built the early
foundation for the historical Maritime Silk Road. Along the coasts of southeast
China, native peoples created diverse, complex and multi-center maritime cultures
with seascape settlements along beaches, sea bays, estuaries and islands. They also
practiced marine subsistence including fishing and shell gathering, as indicated by a
series of Neolithic shell midden sites (Yuan 1995). Archaeological excavations
have revealed some important regional areas with intensively distributed shell
mound sites, including Jiaodong (胶东) Peninsula and the Miaodao (庙岛)
archipelagoes, Taihu (太湖) Basin, Qiantangjiang (钱塘江) Estuary and the
Zhoushan (舟山) archipelagoes, the coastal region of Taiwan strait, the Pearl River
Delta, and the north coast of the South China Sea. These areas reflect the prehistoric
and early historic development of marine adaptations and maritime cultural prac-
tices among the indigenous Yi and Yue (Fig. 1.2). Their sub-regional seascape use
1 A Synthetic Analysis of the Neolithic Origins … 11
patterns and seafaring practices of the Neolithic and early historic periods estab-
lished an important maritime cultural basis for the historical development of “Four
Oceans” navigation (Jiao 2012; Wu 2012).
The North Ocean Navigation network, or the sea route between the coastal regions
of the Bohai Sea (渤海), the Yellow Sea (黄海) of China, Korea (Koryo, 高丽) and
the Sea of Japan (or Balhae State, 渤海国), represented an important sub-regional
maritime culture of ancient China as well as a section of the historical Maritime Silk
Road. The archaeological excavation of prehistoric sites has revealed how the North
Ocean Navigation system originated in the Neolithic seafaring of indigenous
eastern Yi (东夷) ancestors who crossed Bohai Strait between Shandong (山东)
Peninsula and Liaodong (辽东) Peninsula.
12 C. Wu
According to historical documents from the Zhou (周) and Han (汉) dynasties,
the earliest maritime culture emerged among the ancient ancestors of the indigenous
eastern Yi living in the coastal region of Jiaodong (胶东) which were comprised
of 9 sub-regional branches including the Quan-Yi (畎夷), Lai-Yi (莱夷), Huai-Yi
(淮夷), Dao-Yi (岛夷), and etc.
The chapter of “Tribute of the Yu Period (禹贡)” in Shangshu (Book of Early
History 《尚书》) recorded:
The Dao-Yi (island barbarian) dress in furs and live in Jizhou (冀州), next to Jieshi (碣石)
and the estuary of the Huanghe (黄河) river…the Huai-Yi live in Xuzhou (徐州), including
the Haidai (海岱) and Huaihe (淮河) river basins, and dress with pearl shell ornaments, eat
marine fish, and make bamboo weavings and fine textiles. They sail along the Huaihe and
Sishui (泗水) rivers to reach the Huanghe river (Ruan, ed. 2009, pp. 311–312).
The chapter of the “Zi Han (子罕 Rare Words of Confucius)” in Lunyu (Analects
of the Master of Confucius 《论语》) recorded:
Confucius would like to live in the Jiu-Yi (九夷, or nine barbarians) region (Ruan, ed.
2009, pp. 5409).
During the Han and Jin (晋) dynasties, historic accounts recorded the maritime
transportation networks between the coastal region of northern China, Korea and
Japan:
In the chapter of “Annals of Economy and Finance (货殖列传)” of Shiji
(Records of the Historian 《史记》), Q. Sima (司马迁) said:
Yan (燕, now Beijing) had been a flourishing capital with rich marine resources in the
coastal region between Bo (渤) and Jie (碣, now Liaoning Province). Being associated with
Wuhuan (乌桓) and Fuyu (夫余) states, Yan traded easily with Huimo (秽貉), Korea and
Zhenfan (真番) states along the coast of the Korean Peninsula (Sima 1959, pp. 3265).
The early transportation of Wo people to ancient China took place along the route of
Liaodong (辽东) Peninsula. After the Six and Song dynasties, they came to pay tribute by
the southern sea route (Ma 2000, pp. 2554).
The most influential sea route of the North Ocean Navigation network was the
“Nautical Route to Koryo and Balhae State from Tengchow” (Dengzhou Haixing
Ru Gaoli Bohai Dao, 登州海行入高丽、渤海道) recorded by Jiadan (贾耽) of the
Tang Dynasty and collected in chapter of the “Annals of Geography (地理志)” of
Xin Tangshu (New History of the Tang Dynasty 《新唐书》). He wrote:
The navigation starts from Tengchow (登州) at the east cape of Shangdong Peninsula,
along Daxie (大谢) Island, Guixin (龟歆) Island, Mo (末) Island, and Wuhu (乌湖) Island
for 150 kilometers, then across the Wuhu Sea to Duli (都里) town, to the east of Mashi (马
石) Mountain, for 100 kilometers, and then reaching Wugu (乌骨) city at the estuary of
Yalujiang (鸭绿江) River through the Qingni (青泥) Sea, Shirenwan (石人汪) Sea, and the
Tuotuo (橐驼) Gulf for 400 more kilometers (Ouyang and Song 1975, pp. 1147).
After that the route continues southeast, along the west coast of the Korean
Peninsula and across the Tsushima Strait and the Sea of Japan to reach the “Balhae
(渤海) State territory” (Ouyang and Song 1975, pp. 1147).
The chapter of “Annals of the Koryo (高丽传)” of Songshi (History of the Song
Dynasty 《宋史》) recorded that Chenjin (陈靖) as the Song imperial court’s
ambassador to Korea had taken an identical sea route, sailing east from Tengchow
to Korea:
The boat sails from Zhigang (芝冈) Island (now Zhifu, 芝罘) to the Wengjing (瓮津)
seaport of Korea across Huanghai Sea for 80 kilometers to the Haizhou (海州) of Koryo,
another 50 kilometers to Yanzhou (阎州), 20 kilometers to Baizhou (白州), and 20 kilo-
meters to the capital of Koryo (Tuotuo 1977, pp. 14046–14052).
These historical accounts described routine North Ocean Navigation across the
Bohai Strait to the Korean Peninsula during an era from the pre-Qin to Han, Tang
and Song dynasties. Archaeological investigations of the Neolithic maritime cul-
tures distributed along this historical sea route reflect the prehistoric origins of such
voyages across the Bohai Strait. Hundreds of Neolithic sites along the coast of
Jiaodong (胶东) Peninsula and the Miaodao (庙岛) archipelagoes have been
investigated in the last half century, and they illustrate the early development of
prehistoric maritime cultures and seascape use patterns which originally emerged
among the indigenous eastern Yi (Fig. 1.3). These prehistoric maritime cultures
were identified as both a part of the “Haidai (海岱) cultural tradition” over the
central territory of Shandong provincial region and a unique sub-regional maritime
cultural circle in its east coast. These hybridizing cultures were dated to 7000BP to
4000BP with a chronology of Period I of Baishicun (白石村), Period I of
Qiujiazhuang (邱家庄), Period I of Zijingshan (紫荆山), and then a part of
Longshan (龙山) Culture and Yueshi (岳石) Culture (Han 1986; Yan 1986). More
than hundreds of Neolithic shell midden sites were investigated, both a great
amount of various marine shells as oyster, clam, meretrix, spiral shell, marine
fishes, some of territorial animal bones and a few species of plants seeds and
pollens were collected and identified, showing the mixture content of maritime
14 C. Wu
cultural settlement and the compound subsistence of both marine fishing and ter-
ritorial hunting & gathering (IACASS 1999; Wang 2008; Yuan 1998).
Comparative research on these prehistoric maritime cultures on either side of the
Bohai Strait reveals the dissemination of Neolithic cultural practices across the strait
via seafaring. Likewise, the cultural deposits at dozens of Neolithic sites on the
islands of the Miaodao archipelagoes have been identified as similar to the types
found on both the Jiaodong and Liaodong Peninsulas. Thus, Period I of Baishicun
(白石村一期) Culture in the Jiaodong region, which dates to 7000 BP, shares similar
content with findings from the Lower Layer Type of Xiaozhushan (小珠山下层) in
the Liaodong region. Meanwhile the Lower Layer Type of Qiujiazhuang (邱家庄下
层) (dating to 5700-5400BP), the Lower Layer Type of Zijingshan 紫荆山下层)
(dating to 5500-5100BP) and Period II of Beizhuang (北庄二期) Culture (dating to
5100-4400 BP) in the Jiaodong region show a striking similarity to the cultural
content from the Lower Layer Type of Guojiazhuang (郭家村下层) in the Liaodong
region (Fig. 1.4). Period I of Tuojidakou (砣叽大口一期), of Longshanian age, from
the Jiaodong region also shares similar content with the Upper Layer Type of
Guojiazhuang (郭家村上层) in the Liaodong region (Figs. 1.5 and 1.6).
1 A Synthetic Analysis of the Neolithic Origins … 15
JFig. 1.3 Distribution of Neolithic sites across the Bohai Strait, including : 1, Santangdi (三堂地);
2, Hazadi (蛤渣地); 3, Hapidi (蛤皮地); 4, Xiangshuiwan (响水湾); 5, Wangtun (王屯); 6,
Shichengshan (石城山); 7, Zhenxing (振兴); 8, Tasitun (塔寺屯); 9, Dongnankou (东南口); 10,
Wujiacun (吴家村); 11, 12, 13, Xiaozhushan (小珠山); 14, Tangwa (塘洼); 15, Liutiao (柳条);
16, Nanyao (南窑); 17, Shangmashi (上马石); 18, Liqiangzi (李墙子); 19, Shabao (沙包); 20,
Liangzigou (亮子沟); 21, Shigou (石沟); 22, Wenjia (文家); 23, Suangtuozi (双坨子); 24, Dayan
(大盐); 25, Dazuizi (大嘴子); 26, Donggang (东岗); 27, Lizatai (蛎渣台); 28, Datuozi (大坨子);
29, Xiankedi (蚬壳地); 30, 31, 32, Guojiacun (郭家村); 33, Xiaoheishi (小黑石); 34, 35,
Shanqian (山前); 36, 37, Beicun (北村); 38, Dongcun (东村); 39, 40, Dakou (大口); 41, Houkou
(后口); 42, Chenghou (城后); 43, Xidajing (西大井); 44, 45, 46, Beizhuang (北庄); 47, Nanhe
(南河); 48, Wanggou (王沟); 49, 50, Zijingshan (紫荆山); 51, Hanjiatuan (韩家疃); 52, Xidongjia
(西董家); 53, 54, Liujiagou (刘家沟); 55, Xiyingzi (西营子); 56, Dachijia (大迟家); 57,
Nanwangchu (南王储); 58, Tangjia (唐家); 59, 60, Shaojia (邵家); 61, Fanjia (范家); 62,
Dazhongjia (大仲家); 63, Laodian (老店); 64, Loujiazhuang (楼子庄); 65, Houdali (后大里); 66,
Nanjie (南截); 67, Shengjiazhuang (盛家庄); 68, Beilizhuang (北里庄); 69, Lujiagou (鲁家沟);
70, Yandi (燕地); 71, Qiaozhi (桥芝); 72, 73, Beichengzi (北城子); 74, 75, Guzhendu (古镇都);
76, 77, Yangjiaquan (杨家圈); 78, Qiujiazhuang (邱家庄); 79, Zhishui (芝水); 80, 81, Baishicun
(白石村); 82, Dongbozi (东泊子); 83, Wuhou (午后); 84, Sunjiatuan (孙家疃); 85, Xixishan (西
系山); 86, Haduiding (蛤堆顶); 87, Houlufang (后炉房); 88, Jingzibu (荆子埠); 89, Nantai (南
台); 90, Laoyingding (老莹顶); 91, Zhaogezhuang (照格庄); 92, Lizha (蛎渣); 93, Diancun (店
村); 94, Rengliuzhuang (仁柳庄); 95, Shendaokou (神道口); 96, Beidianzi (北店子); 97, Yihe (义
和); 98, Gounanzhuang (沟南庄); 99, Majiatanghou (马家汤后); 100, Xidoushan (西豆山); 101,
Liujia (柳家); 102, Beilangge (北兰格); 103, 104, Suanyuanzi (蒜园子); 105, Zhongyang (中扬);
106, Guanli (观里); 107, 108, Yujiadian (于家店); 109, Yangjiatuan (杨家疃); 110, Xixiandu (西
贤都); 111, Shaogezhuang (韶格庄); 112, Changqing (长清); 113, Damengge (大孟格); 114,
Shangniantou (上碾头); 115, Luojizhuang (落鸡庄); 116, Henan (河南); 117, Luji (鲁济); 118,
Qiujiawa (仇家洼); 119, Shalidian (沙里店); 120, Lincun (林村); 121, Shiyang (石羊); 122,
Maitian (脉田); 123, 124, Panjiazhuang (潘家庄); 125, Wenjiabu (温家埠); 126, Gongjia (宫家);
127, Machang (马场); 128, Dasongjia (大宋家); 129, Hekou (河口); 130, Renhe (人和); 131,
Quanshuitou (泉水头); 132, 133, Beiqian (北阡); 134, Yangjiaoyuan (羊角园); 135, Bunan (埠
南); 136, Xiaohugezhuang (小胡各庄); 137, Pizitou (陂子头); 138, Xing’an (辛安); 139,
Chengdingzi (城顶子); 140, 141, Xiaoguan (小管); 142, Pangguzhuang (盘古庄); 143, Geziling
(葛子岭)
The similarities between these cultural deposits must have been the result of
prehistoric cultural interaction and cultural exchange across the Bohai Strait,
illustrating the crucial role of the Liaodong Peninsula as the first stop in maritime
cultural dissemination from mainland Jiaodong to coastal east Asia (Tong 1989;
Wang and Li 1990). Also worth noting, the distance between Jiaodong and
Liaodong Peninsulas is less than 100 kilometers and there are more than 30 islands
scattered across the strait. Most of these islands are separated from each other by a
few miles or less, with the exception of the widest one, known as the Laotieshan (老
铁山) waterway, which spans 22.8 miles between North Huangcheng (北隍城)
Island and Liaodong. These densely distributed islands acted like stepping stones
for the diffusion of prehistoric maritime cultural practices. The discovery on these
islands of dozens of Neolithic sites, all of which contain similar cultural deposits to
those found in Jiaodong and Liaodong, illustrates the movement of early maritime
culture and primitive seafaring eastward across the strait.
16
Fig. 1.4 Evidence of cultural interaction between Jiaodong and Liaodong regions from 5500-4500 BP (revised from Tong 1989)
C. Wu
1 A Synthetic Analysis of the Neolithic Origins …
Fig. 1.5 Evidence of cultural interaction between Jiaodong and Liaodong from 4500-4000 BP (revised from Tong 1989)
17
18 C. Wu
The East Ocean, which included the East China Sea and Taiwan Strait, was a busy
and flourishing region for historical maritime navigation. Archaeological excava-
tion of prehistoric sites has revealed that the indigenous ancestors of the Bai-Yue
(百越) and proto-Austronesians carried out the first seafaring voyages along the
coastal regions of the East China Sea, across the Taiwan Strait and the Bashi
Channel of the Philippines, and into the archipelagoes of Southeast Asia and the
Pacific, beginning in the Neolithic Age.
Historical accounts of the indigenous East-Yue (东越, or Yu-Yue, 于越) living
in Zhejiang and Min-Yue (闽越) in Fujian illustrate how they initiated the earliest
maritime cultures and prehistoric seascapes in the East Ocean by creating maritime
settlement patterns, fishing and practicing marine subsistence, and developing early
nautical traditions.
The chapter of the “Principle of Serving King (恃君览)” in the Lvshi Chunqiu
(Analects of the Political Opinion of Master Lv 《吕氏春秋》) recorded:
Yangzhou (扬州) region is located in southeast China, where the aboriginal Yue lives…
Various branches of Bai-Yue people are distributed in the region to the south of the
Yanghan (扬汉, Yangtse) Rivers (Lv and Chen 2002, pp. 1331).
The chapter of “South Area of Inland Sea (海内南)” in the Shanhaijing (Classics
of the Mounts and Seas 《山海经》) recorded:
Both Ou (瓯) and Min (闽), as branches of the aboriginal Bai-Yue, live in the coastal region
next to the northwestern mountain (Yuan, ed. 2014, pp. 237).
Yuejueshu (The History of the Lost Yue Nation 《越绝书》) was an important
historical literature of Yue people. The chapter of “Biography of the Yue Territory
(越绝外传记地传)” recorded:
The aboriginal Yue live in the coastal region and specialize in bird farming…The character
of aboriginal Yue is frank and rough. They live in mountainous coast and travel by water,
taking the boat with oar as their main transportation tool. They skilledly sail the boat as fast
as the howling wind (Yuan and Wu 1985, pp. 57, 58).
Even the great and wise king Tang (汤) of the Shang (商) dynasty drives a boat poorly,
while the aboriginal Yue are skilled at driving a boat well (Liu and Chen 2010, pp. 131).
Bowu Zhi (Biography of the Things and Cultures 《博物志》) distinguished the
different livelihood of coast and inland people in the chapter of “Variants of
Ethnicity (五方人民)”. It recorded:
The native people living on the southeast coast enjoy aquatic food, while people living in
the northwest like to eat the products of livestock. The southeast aquatic foods include
tortoise, clam, snail, and mussels which are taken to be delicious by the local people (Zhang
2012, pp. 10).
During the Han, Tang, Song and Yuan dynasties, historical accounts described
the route of the East Ocean Navigation system. Voyagers departed from Fujian and
Zhejiang in the southeast coastal region and headed eastward to Yizhou (夷州),
Penghu (or Pescadores, 澎湖), Dongfan (东番, or Taiwan), Mait (麻逸) and the
Sandao (三岛) or Sanyu (三屿) Islands, carrying out maritime trade along the way.
The chapter of “Biography of Sunquan (孙权传)” of Sanguozhi (History of the
Wu of Three Kingdoms 《三国志(吴书)》) recorded:
“In 2nd year (AD.230) of Huanglong (黄龙) reign of Wu (吴) kingdom, the king ordered
generals Weiwen (卫温) and Zhuge Zhi (诸葛直) leading a troop with 10,000 soldiers to
search for Yizhou and Chanzhou (澶洲) which were located in the open sea.” (Chen and
Pei 2006, pp. 674).
The chapter of “Annal of the Liuqiu Nation (流求传)” of Suishu (History of the
Sui Dynasty 《隋书》) recorded:
“Liuqiu (流求) is an island state within five days’ eastern bound sailing from Jian’an county
in Fujian” (Wei and Linghu 1982, pp. 1823).
land, instead anchoring their boat in the river and calling for local aboriginals to come
trade” (Zhao 1985, pp. 25–26).
During the Ming and Qing dynasties, the sea routes of East Ocean Navigation
developed into a more explicit, concrete and stable network. There were even a
series of historical navigation guide books printed in this period, such as
Dongxiyang Kao (General Survey on the East and West Ocean 《东西洋考》),
Shunfeng Xiangsong (Sea Routes with Successful Sailing 《顺风相送》), and
Zhinan Zhengfa (Guide to the Right Sea Routes 《指南正法》) (Zhang 1981a, b,
pp. 170–191; Xiang 1961, pp. 13–99, 101–195). These described dozens of sea
routes connecting mainland Southeast China, the Philippines and eastern Indonesia.
22 C. Wu
JFig. 1.7 Distribution of Neolithic sites in the lower reaches of the Yangtse and Qiantangjiang
Rivers, including: 1, Kaizhuang (开庄); 2, Qingdun (青墩); 3, Fenghuangshan (凤凰山); 4,
Sanxingcun (三星村); 5, 6, Shendun (神墩); 7, Qintoushan (祁头山); 8, 9, Panjiatang (潘家塘);
10, Luhuadang (芦花荡); 11, 12, Xi’xi (西溪); 13, Gaochengdun (高城墩); 14, Dongshancun (东
山村); 15, 16, Sidun (寺墩); 17, Xindumiao (新渎庙); 18, Xudun (圩墩); 19, Anjidun (庵基墩);
20, Hongkoudun (洪口墩); 21, Cimadzui (赤马嘴); 22, Luotuodun (骆驼墩); 23, Qiucheng (邱
城); 24, Jiangjiashan (江家山); 25, Shizishan (狮子山); 26, Kunshan (昆山); 27, Xiaoshancun (小
山村); 28, 29, Tadi (塔地); 30, Huiguanshan (汇观山); 31, Lucun (芦村); 32, 33, Wujiabu (吴家
埠); 34, Yaoshan (瑶山); 35, Fanshan (反山); 36, Mojiaoshan (莫角山); 37, 38, Dengjiashan (邓
家山); 39, Hengshan (横山); 40, 41, Miaoqian (庙前); 42, Xuzhuang (许庄); 43, 44, Luodun (罗
墩); 45, Qiandixiang (钱底巷); 46, Xiangtatou (象塔头); 47, Pengzhudun (彭祖墩); 48,
Jialingdan (嘉菱荡); 49, Qiuchengdun (丘城墩); 50, 51, Yuecheng (越城); 52, Yujiadu (俞家渡);
53, Xuxiang (徐巷); 54, Yujiadun (俞家墩); 55, Taipingqiao Cun (太平桥村); 56, 57, Longnan
(龙南); 58, Gangcheng (港城); 59, Luojiajiao (罗家角); 60, Xinqiao (新桥); 61, Xindili (新地里);
62, Xubuqiao (徐步桥); 63, 64, Caoxieshan (草鞋山); 65, 66, Zhaolingshan (赵陵山); 67, 68,
Zhanglingshan (张陵山); 69, Siqian (寺前); 70, Shaoqinshan (少卿山); 71, 72, Fuquanshan (福泉
山); 73, 74, Songze (崧泽); 75, Tongli (同里); 76, Dawang (大往); 77, Guangfucun (广福村); 78,
Meiyan (梅堰); 79, Duxing (独行); 80, 81, Guangfulin (广富林); 82, Tanjiawan (谭家湾); 83,
Wujiaban (吴家浜); 84, Dafengtang (大坟塘); 85, Shuangqiao (双桥); 86, Quemuqiao (雀慕桥);
87, Pingqiudun (平邱墩); 88, Chashan (查山); 89, Tinglin (亭林); 90, Maqiao (马桥); 91,
Zhongjiagang (钟家港); 92, Wujiaqiang (吴家墙); 93, Guojia Shiqiao (郭家石桥); 94,
Fengqiaogang (坟桥港); 95, Majiabang (马家浜); 96, Pengcheng (彭城); 97, Zhuangfenqiao
(庄桥坟); 98, Shushan (蜀山); 99, Maocaoshan (茅草山); 100, Xiasun (下孙); 101, Jinshan (金
山); 102, Kuhuqiao (跨湖桥); 103, Mianquanshan (眠犬山); 104, 105, Loujiaqiao (楼家桥); 106,
Shangshan (上山); 107, Shangdi (上地); 108, Shanbei (山背); 109, Xiaohuangshan (小黄山);
110, Jinjishan (金鸡山); 111, Wuguishan (乌龟山); 112, Shunhuli (舜湖里); 113, Taojia (陶家);
114, Dakeng (大坑); 115, Zhushan (猪山); 116, Zhuqiao (祝桥); 117, Jiangjia (蒋家); 118,
Qinxian (琴弦); 119, Dushan (独山); 120, Jiangdoushan (豇豆山); 121, Shuikoushan (水口山);
122, Ma’an (马鞍); 123, Xianrenshan (仙人山); 124, Mamanqiao (马慢桥); 125, Niutoushan (牛
头山); 126, Yangqi’ao (杨岐岙); 127, Wengjiashan (翁家山); 128, Qiancihu (前溪湖); 129,
Tianwu (田屋); 130, Huangjiashan (黄家山); 131, Maohu (茅湖); 132, Pengqiao (彭桥); 133,
Zishan (鲻山); 134, Xinzhoujia (新周家); 135, Zhangshu (樟树); 136, Xiangshan Fotang (相山佛
堂); 137, Wangjia (王家); 138, Kengshanlong (坑山陇); 139, Tianluoshan (田螺山); 140,
Xiazhuang (下庄); 141, Hemudu (河姆渡); 142, 143, Xiangjiashan (鲞架山); 144, 145, Cihu (慈
湖); 146, 147, Baziqiao (八字桥); 148, Majiadun (马家墩); 149, Yangdun (洋墩); 150,
Wangjiadun (王家墩); 151, 152, Xiaodongmen (小东门); 153, Shenjiao (蜃蛟); 154, Qian’ao (钱
岙); 155, Dongjiatiao (董家跳); 156, 157, Mingshanhou (名山后); 158, 159, Tashan (塔山); 160,
Hongmiaoshan (红庙山); 161, Wangjiatai (王家台); 162, Sunjiashan (孙家山); 163, Hamoshan
(蛤蟆山); 164, Peiyincun (培荫村); 165, Tangjiadun (唐家墩); 166, Yangtandun (洋坦墩); 167,
Liangmaopeng Dun (凉帽蓬墩); 168, Hebangdun (河蚌墩); 169, Baiquan (白泉)
There were nine East Ocean Navigation sea routes recorded in the Dongxiyang
Kao, of which one was the route between mainland southeast China and the
Philippines archipelagoes known as “Navigation from Taiwushan (太武山) of
Zhangzhou (漳州) to Miyan (密雁) harbor of Luzon via Penghu.” Others were sea
routes linking different islands of the East Ocean archipelagoes, in the Philippines
and east of Indonesia. Most of East Ocean Navigation routes recorded in the
Shunfeng Xiangsong at the seaports of Fujian and Guangdong, such as Meizhou (湄
洲), Quanzhou, Zhangzhou, and Nan’ao (南澳), made their way to Penghu, Luzon,
Sulu, Brunei in the East Ocean archipelagoes. The section “Topography of the East
Ocean” (东洋山形水势) in the Zhinan Zhengfa described the landscape features
1 A Synthetic Analysis of the Neolithic Origins … 23
along the sea route from Dadan (大担), in Zhangzhou, to Malaoyang (麻老央)
harbor in the south of Luzon via Penghu and Taiwan. More East Ocean sea routes
between Fujian, Guangdong and Luzon via Penghu and Taiwan were recorded in
Zhinan Zhengfa, and together these paths formed a complicated network of East
Ocean Navigation. Moreover, since the late Ming dynasty the East Ocean sea route
had linked up with the Spanish Manila Galleon, extending eastern Asian navigation
into the early stages of globalization.
These historical routes of the East Ocean fit closely with the Neolithic seafaring
patterns of the native Bai-Yue and proto-Austronesian ancestors, as reconstructed
by archaeologists. Archaeological investigations of prehistoric sites have revealed
how the Neolithic ancestors of Bai-Yue in mainland southeast China once sailed
across the Taiwan Strait and the Bashi Channel of the Philippines, then emigrated
into the archipelagoes of Southeast Asia and the Pacific to become the first
proto-Austronesians.
The existence of a thriving Neolithic seascape was discovered in the coastal
regions of mainland southeast China. From the Yangtse Delta of the East China Sea
to the coastal areas along the Taiwan Strait, the Neolithic coastal settlements had
spread widely from the mainland to adjacent islands, illustrating the strength of this
nautical culture. The south coast of Hangzhou Bay (杭州湾, an estuary of the
Qiantangjiang River, 钱塘江) is just one example. Typological and chronological
research of this region revealed that a single Neolithic culture had come out of the
mountainous inland and extended, first to the shore and then out into the coastal
islands between 10000 and 4000BP. In the Puyangjiang (浦阳江) watershed
located in the western sub-region of the southern Hangzhou Bay, this Neolithic
culture spread into the lower reaches of the basin and the river estuary. The earliest
Shangshan (上山) Culture (dating to between 10000 and 8500BP) emerged at sites
in Shangshan and Xiaohuangshan (小黄山), in the upper reaches of the
Qiantangjian River. By the time the middle period Kuahuqiao (跨湖桥) and
Loujiaqiao (楼家桥) Cultures (dating to between 8500 and 5000 BP) and the late
period Liangzhu (良渚) Culture (dating to between 5000 and 4000BP) emerged, it
had extended into the lower basin and river estuary, where more than 25 settlement
sites have been discovered.
On the opposite side of southern Hangzhou Bay, Yaojiang (姚江) River Basin
occupies the eastern sub-region and here, too, a similar Neolithic culture spread
through the prehistoric coastal areas and islands. The early period of Hemudu (河姆
渡) Culture, dating to between 7000 and 6000BP, appeared at the prehistoric off-
shore islands of Hemudu, Zishan (鲻山) and Tianluoshan (田螺山), land forma-
tions that have now become separate mountains in the coastal plain. The late period
Hemudu Culture (dating to between 6000 and 5000BP) flourished in the prehistoric
marine accumulation plain along the coast, where more than 20 settlements sprang
up in the wake of the dropping sea level. During this period, the Hemudu Culture
extended across the strait to the Zhoushan (舟山) archipelagoes, as evidenced by
the Baiquan (白泉) site; also the earliest known settlement on the island. Soon the
Liangzhu Culture (dating to between 5000 and 4000BP) was thriving on the east
coast of southern Hangzhou Bay. More than 30 sites have been discovered, 10 of
24 C. Wu
Fig. 1.8 Distribution of Neolithic sites across the Taiwan Strait, including: 1, Niutougang (牛头
岗); 2, Shizigang (狮子岗); 3, Shanqianshan (山前山); 4, Sanmenbao (三门宝); 5, Xitang (溪潭);
6, Muyang (穆阳); 7, Wuqu (武曲); 8, 9, Huangguashan (黄瓜山); 10, Yacheng (牙城); 11,
Zhongfang (中房); 12, 13, Huangqiyu (黄岐屿); 14, Houmenluan (后门峦); 15, 16, Liangdao (亮
岛); 17, 18, Zhipinglong (炽坪陇); 19, Yushan (玉山); 20, Baofengshan (宝峰山); 21, 22,
Zhanghuban (樟湖板); 23, 24, Niutoushan (牛头山); 25, Zhailishan (寨里山); 26, Meixian (梅
仙); 27, Xinqiao (新桥); 28, 29, Xitou (溪头); 30, Qiapushan (洽浦山); 31, 32, 33,
Zhuangbianshan (庄边山); 34, 35, 36, Tanshishan (昙石山); 37, Fucun (浮村); 38, Panshishan
(磐石山); 39, 40, Dongzhang (东张); 41, 42, Pingtang Guishan (龟山); 43, Keqiutou (壳丘头);
44, Hupuqian (湖埔乾); 45, Xiying (西营); 46, Chitanghou (祠堂后); 47, Nancuochang (南厝场);
48, Yinloushan (音楼山); 49, 50, Yishan (蚁山); 51, Shizishan (狮子山); 52, An’shan (庵山); 53,
Jinmen Guishan (龟山); 54, 55, Fuguodun (富国墩); 56, Zhaizishan (寨子山); 57, Guankou (灌
口); 58, Songbaishan (松柏山); 59, Xiangshan (香山); 60, Huotian (火田); 61, 62, Mulinshan (墓
林山); 63, Kengbei (坑北); 64, Damaoshan (大帽山); 65, 66, Lazhoushan (腊洲山); 67, 68,
Chenqiao (陈桥); 69, Shiweishan (石尾山); 70, 71, Guoye (菓叶); 72, Liyushan (鲤鱼山); 73,
Zhishanyan (芝山岩); 74, 75, Xuntangpu (讯塘埔); 76, Dalongdong (大龙峒); 77, 78, Dabenkeng
(大坌坑); 79, Yuanshan (圆山); 80, Zhiwuyuan (植物园); 81, 82, Anhe (安和); 83, Xidadun (西
大墩); 84, Zhongleng (中冷); 85, 86, Huilai (惠来); 87, 88, Chongguang (重光); 89, 90, Bajiacun
(八甲村); 91, 92, Nanguanli (East) (南关里东); 93, 94, 95, Fengbiou (凤鼻头); 97, 98, Kengding
(垦丁); 99, 100, Beinan (卑南); 101, Qihedong (奇和洞); 102, Nanshanta Dong (南山塔洞)
1 A Synthetic Analysis of the Neolithic Origins … 25
which are in the Zhoushan archipelagoes. This spread illustrates the successful
development of Neolithic seascapes and the extent of cultural and maritime
exploration by the native Yu-Yue’s prehistoric ancestors (Fig. 1.7; Cao 2012).
The coastal region of Fujian, which centers around the lower Minjiang (闽江)
River Basin, is another example. Here, hundreds of densely packed prehistoric shell
midden sites were discovered along the estuary and the coast, reflecting the rise and
subsequent prosperity of the local maritime culture from 8000 to 3000BP. The
earliest maritime culture was investigated at the Liangdao (亮岛) I Site in the Mazu
(马祖) archipelagoes, located in the estuary of the Minjiang River and dating to
between 8320 and 8160 BP (Chen 2012). The Keqiutou (壳丘头) Culture and
Lower Layer Type of Tanshishan (昙石山) Culture (dating to between 6000 and
5000BP) were discovered not only in the lower Minjiang Estuary as Tanshishan,
Xitou (溪头) and the Lower Layer of Zhuangbianshan (庄边山), but also at off-
shore islands such as Keqiutou at Pingtan (平潭), Fuguodun (富国墩) at Jinmen
(金门), and Lazhoushan (腊洲山) at Zhao’an. Meanwhile the Middle Layer Type
of Tanshishan Culture, dating to between 5000 and 4000BP, was discovered at
more than 20 prehistoric settlement sites, while the Upper Layer Type of
Tanshishan Culture and Huangtulun (黄土仑) Culture (dating to between 4000 and
3000BP) were discovered at more than 40 prehistoric sites in the coastal region of
Fujian. This breadth and longevity illustrates the growth of maritime culture and
Neolithic seascapes in the west coast region of the Taiwan Strait, where the early
seafaring practices of the indigenous Min-Yue and maritime emigration of the
proto-Austronesians began (Fig. 1.8; Wu 1995; Rolett et al. 2011).
The typological comparison of research conducted on the Neolithic artifacts of
the coastal mainland and near shore islands also revealed the early seafaring and
primitive navigation carried out by the ancient ancestors of the East-Yue and
Min-Yue. Meanwhile, the Neolithic seafaring of the Bai-Yue ancestors among the
Eastern Ocean archipelagoes resulted in the early diffusion of proto-Austronesians
into the islands of the southwest Pacific. For example, in the estuary of the
Qiantangjiang River and Zhoushan archipelagoes, evidence of the Hemudu and
Liangzhu Cultures (dating to between 7000 and 4000BP) have been discovered at a
series of offshore Zhoushan Islands. Reflecting the diffusion of Neolithic marine
culture from the mainland to the near shore islands of the East Ocean, these have
come from the Tangjiadun (塘家墩), Shizilu (十字路), Dazhi (大支), and
Chaomian (潮面) Siteson Dinghai (定海) Island; the Mantoushan (馒头山) Site on
Daishan (岱山) Island; the Sunjiashan (孙家山) and Hamashan (蛤蟆山) Sites on
Daqu (大衢) Island; and the Caiyuanzhen (菜园镇) Site on Shengsi (嵊泗) Island
(Wu 1983). In the coastal region of Fujian, the cultural deposits of the Keqiutou on
Pingtang Island, the Fuguodun, Jinguishan (金龟山) and Pubian (浦边) Sites on
Jinmen Island, and the Damaoshan (大帽山) Site of Dongshan (东山) Island are
each identified respectively as the same culture types of the Lower and Middle
Layers of Tanshishan, thus illustrating how Neolithic near shore navigation and
cultural diffusion actually occurred.
Other Neolithic cultural similarities have been discovered between the Taiwan
Strait, the Bashi Channel of the Philippines, and other straits in the Southeast Asia
26 C. Wu
Fig. 1.9 A comparison of Neolithic painted pottery from Fujian and Taiwan, including: (Above)
Fujian Neolithic painted pottery of the Upper Level Type of Tanshishan (昙石山), 1-9, and the
Middle Level Type of Tanshishan from the Tanshishan Site (昙石山), 10-11 and 14-17, and the
Xitou Site (溪头), 12-13; and (Below) Taiwan Neolithic painted pottery from the Fengbitou Site
(凤鼻头), 1-8, 10-12, and 14-15; the Shejiao Site (社脚), 9 and 13; and the Zhishanyan Site
(芝山岩), 16-20
Fig. 1.10 Comparison of the Neolithic and indigenous cultures of Fujian and Taiwan (Wu 1994),
including: indigenous artifacts from the modern era, 1-8; and from Kending (垦丁), 9; Niaosong
(鸟崧), 10; Nantou (南投), 11 and 13; Hualian (花莲), 12; Fengbitou (凤鼻头), 14; Huangtulun
(黄土仑), 16 and 24; Keqiutou (壳丘头) 17-18; Tanshishan (昙石山), 19; Zhuangbianshan
(庄边山), 20; Xitou (溪头), 21 and 23; Baizhuduan (白主段), 22
typological series as those of China’s eastern coast including the Corded Pattern
Pottery Type, Red Fine Pottery Type, Gray and Black Stoneware Type, and Check
Pattern Stamped Pottery Type. This research detailed the early emergence and
growth of Neolithic cultures in the west coastal plain of Taiwan, which may have
been due to the continuing emigration of indigenous peoples from mainland China
to Taiwan (Figs. 1.9 and 1.10; Wu 1994; Wu and Li 1992). The same combination
of prehistoric maritime cultural dissemination and indigenous emigration had also
occurred across the Bashi Channel of the Philippines and a series of sea straits in the
southeastern Asia and Pacific archipelagoes. Both the typical stone adze of
Polynesia and the Lapita cultural complex of the Pacific Islands have been inves-
tigated and identified as the result of this cultural diffusion from mainland southeast
China. Thus, these discoveries support the theory that Pacific Islanders originated
from mainland southeastern China. Gradually, and with the help of archaeologists,
the sea route of proto-Austronesian emigration from the coast of Fujian and
Guangdong and across the Taiwan Strait has been reconstructed, illustrating how
these seafaring peoples made their way to Taiwan, the Philippines, Indonesia and
eventually the archipelagoes of the Pacific (Bellwood 1997, 2005; Chang 1987; Lin
1958; Rolett et al. 2002; Wu 2008; Wu and Chen 2003; Wu and Cao 2005).
28 C. Wu
The South Ocean Navigation network, next to the west coast of South China Sea,
and the West Ocean Navigation network in the Indian Ocean were some of the
busiest and most important routes of China’s historical Maritime Silk Road.
Historical accounts and prehistoric archaeological discoveries have demonstrated
how both the coastal sea route of “South China Sea Route via Xuwen and Hepu”
(Xuwen Hepu Nanhaidao 徐闻合浦南海道), during the Qin and Han dynasties,
and the off-shore sea route of “Canton Sea Route” (Guangzhou Tonghai Yidao,
广州通海夷道), during the Tang and Song dynasties, could be traced back to the
Neolithic diffusion of the indigenous South-Yue (南越), West-Ou (西瓯), Luo-Yue
(骆越) and Dan’er (儋耳) maritime cultural practices across the coastal region of
Lingnan (岭南).
Historical accounts from the Zhou and Han dynasties documented how
indigenous peoples such as the South-Yue, West-Ou, Luo-Yue and Dan’er had
active maritime cultures which they practiced in the coastal region of southern
China.
The chapter of “Annals of the South Yue Kingdom (南越列传)” of Shiji
(Records of the Historian 《史记》) recorded:
The southern barbarians live in the humid and low land of south China, with the Min-Yue
kingdom located to the east and West-Ou, Luo-Yue and Luoguo (裸国) kingdoms to the
west (Sima 1959, pp. 2970).
The chapter of “Record the Kings Meeting (王会解)” of Yi Zhoushu (The Lost
Historical Literature of Zhou Dynasty 《逸周书》) recorded:
“During the Shang (商) dynasty, the aboriginal states of Oudeng (瓯邓), Guiguo (桂国),
Sunzi (损子), Chanli (产里), Baipu (百濮), and Jiujun (九菌) on the southern coast pre-
sented marine pearls, tortoise shells, ivory, rhinoceros horns, peacock feathers, and etc. as
their tribute to the northern Central Empire” (Huang et al. 2007, pp. 908–915).
1 A Synthetic Analysis of the Neolithic Origins … 29
During the Han and Tang dynasties, the regularly used coastal and off-shore sea
routes of the South China Sea were recorded in detail.
The chapter of the “Annals of Geography (地理志)” of Hanshu (《汉书》)
discribled the sailing route from coast of south China to Indian ocean:
“Setting sail from Zhangsai (障塞) of Rinan (日南) prefecture, Xuwen and Hepu, the boat
traveled for 5 months to Duyuan (都元) State, for 4 months to Yilumo (邑卢没) State, 20
days to Chenli (谌离) State, and more than 10 days to Fugandulu (夫甘都卢) State. Setting
sail again from Fugandulu State, the boat traveled for 2 more months to Huangzhi (黄支)
State…8 months more to Pizong (皮宗) State, and 2 months to Rinan and Xianglin (象林)
Territories. The Yichengbu (已程不) State is located to the south of Huangzhi State.”
(Ban, G. 1962, pp. 1671).
Historians have studied these narratives and identified Huangzhi State as the area
on the east coast of India, while Yichengbu State is now Sri Lanka. These accounts
reveal the earliest coastal sea routes from Panyu (番禺, now Canton or Guangzhou)
and Dongye (now Fuzhou) heading into the Indian Ocean along the Beibu Gulf
(Bac Bo, 北部湾) and to the east of coastal Vietnam, by the stops of coastal
seaports such as Rinan, Xuwen and Hepu, but later declined (Han 1958; Wang
1992). During the Jin (晋) dynasty, a new off-shore sea route across the South
China Sea were described in the travel notes of the famous Chinese monk Faxian
(法显). Faxian wrote:
The boat sailed eastbound, returning to China from Shizi (师子国, Ceylon, now Sri Lanka)
State, by stopping at Yepoti (Yāva—dvīpa, 耶婆提) State…The boat set sail again from
Yepoti, heading northeast for Guangzhou, across the boundless open sea, and navigating by
the stars. (Zhang 1985, pp. 167–171).
During the Tang dynasty, the famous off-shore navigation route known as the
“Canton Sea Route” emerged, crossing the South China Sea and Malacca Strait to
reach the Persian Gulf and eastern Africa. Descriptions of the Canton Sea Route
were the most systematic narration recording the South and West Ocean Navigation
systems of the Tang dynasty. The domestic section of this route set sail from
Guangzhou (Canton) and headed southward across South China Sea, passing
Hainan (海南) Island to the northeast.
“Canton Sea Route” (Guangzhou Tonghai Yidao) originally recorded by Jiadan
(贾耽) of Tang dynasty and collected in chapter of the “Annals of Geography (地理
志)” of Xin Tangshu (New History of the Tang Dynasty 《新唐书》). The boat set
sail from Canton and headed southeast for 100 kilometers to Tunmenshan (屯门山,
now Hong Kong) Mountain, then sailed westward for 2 days to Jiuzhoushi
(九州石, now northeast of Hainan Island) and southward for another 2 days to
Xiangshi (象石, now southeast of Hainan Island)…The boat sailed eastward from
30 C. Wu
1 A Synthetic Analysis of the Neolithic Origins … 31
JFig. 1.11 Distribution of Neolithic sites along the coast of the South China Sea, including: 1, 2,
Chenqiao (陈桥); 3, Shiweishan (石尾山); 4, Sanduo (三舵); 5, Yinzhou (银州); 6, Xiankezhou
(蚬壳洲); 7, Maichuangang (船埋岗); 8, Tongxingang (通心岗); 9, 10, Jinlansi (金兰寺); 11,
Wanfu’ang (万福庵); 12, 13, Haogang (蠔岗); 14, Guye (古椰); 15, Youyugang (鱿鱼岗); 16,
Hedan (河宕); 17, Yuanzhou (圆洲); 18, Cuntou (村头); 19, 20, 21, Xiantouling (咸头岭); 22, 23,
24, Dahuangsha (大黄沙); 25, 26, Xiaomeisha (小梅沙); 27, Shatou Xincun (沙头角新村); 28,
Zhaoxiacun (灶下村); 29, Chiwancun (赤湾村); 30, Zhaogang (灶岗); 31, Luoshandi (罗山地);
32, Maodi (猫地); 33, 34, Baishuijing (白水井); 35, Waisha (外沙); 36, 37, Houshanwan (后沙
湾); 38, Dong’aowan (东澳湾); 39, Yapowan (亚婆湾); 40, Lengjiaozui (棱角嘴); 41, Nansha (南
沙); 42, 43, Longxue (龙穴); 44, Tangxiahuan (棠下环); 45, Xiguapu (西瓜铺); 46, Xiaza (下栅);
47, Shuijingkou (水井口); 48, Shuiyong (水涌); 49, Xiedijiao (蟹地角); 50, Nanshawan (南沙
湾); 51, Heisha (Hac sa 黑沙); 52, Baojingwan (宝镜湾); 53, Suochiwan (锁匙湾); 54, 55,
Caotangwan (草堂湾); 56, Chishwan (赤沙湾); 57, 58, Yonglang (涌浪); 59, 60, Longgutang (龙
鼓滩); 61, Shijiaozui (石角嘴); 62, Shazhou (沙洲); 63, Longguzhou (龙鼓洲); 64, Baimang (白
芒); 65, Shaluowan (沙螺湾); 66, Batougu (扒头鼓); 67, 68, Xiediwan (蟹地湾); 69,
Dongwanzhai Bei (东湾仔北); 70, 71, Dongwan (东湾); 72, Xiaoyazhou (小鸭洲); 73,
Fuyuwan (鯆鱼湾); 74, Daguiwan (大鬼湾); 75, 76, Dawan (大湾); 77, Luxucheng (芦须城);
78, 79, Shenwan (深湾); 80, Chunkangwan (春坎湾); 81, Shaxia (沙下); 82, Haoyong (蠔涌); 83,
Liyudun (鲤鱼墩); 84, 85, Xinjie (新街); 86, 87, Yingdun (英墩); 88, Shigong (石贡); 89, 90,
Qiaoshan (桥山); 91, Yi’nian (移辇); 92, 93, Gaogaodun (高高墩); 94, Baijaodun (芭蕉墩); 95,
Dadundao (大墩岛); 96, 97, Yapushan (亚菩山); 98, Sheshan (社山); 99, Malanzui Shan (马兰嘴
山); 100, Beijiaoshan (杯较山); 101, Ha Lung; 102, 103, Phung Nguyen; 104, Cai Beo; 105, Son
Vi; 106, 107, Trang Kenh; 108, 109, Da But; 110, Hoa Loc; 111, 112, Quynh Van; 113, 114,
BauTro
Sri Vijaya (佛逝, Samboja 三佛齐 in Song dynasty, now southeast of Sumatra)
State for 4 or 5 days to Heling (诃陵, or Yavadvipa [阇婆] in the Song dynasty,
now Java) State. (Ouyang and Song 1975, pp. 1153–1154).
Archaeological investigations have revealed that the historical sea routes of the
South and West Ocean Navigation systems, including early coastal routes such as
the “South China Sea Route via Xuwen and Hepu” of the Han dynasty and the later,
off-shore route known as the “Canton Sea Route” of the Tang dynasty, originated in
the Neolithic maritime cultures of the indigenous South-Yue and Luo-Yue living in
the coastal regions of south China and the Indochina Peninsula (Wu 2010, 2011).
More than 100 shell midden and sand dune sites excavated along the north coast of
the South China Sea illustrated the successful development of Neolithic seascapes
and maritime settlement patterns. Most of the shell midden sites were distributed in
the Pearl River Delta and its estuary region, while more sand dune sites were
present on the coast and near shore islands, both of which included evidence of rich
ecological resources such as marine shells and fish bones (Zhu 1994). These pre-
historic settlements vary in geographical ecology and subsistence patterns. The sites
along sea bays were filled with the remains of marine oyster and clam shells; river
estuary sites were most heavily characterized by evidence of fresh water clams,
marine clams and oysters; and the riparian sites had a high density of fresh water
clam shells, thus illustrating different livelihoods involving fishing and aquaculture
(Yuan 1999). In terms of their chronology, more than 10 shell midden and 30 sand
dune sites of Middle Neolithic age (dating to between 6000 and 5000 BP) were
discovered in the Pearl River Delta and the coastal bay, while more than 40 shell
32 C. Wu
midden and 140 sand dune sites of Late Neolithic age (dating to between 5000 and
3000BP) were found. This increase shows the continuing development of prehis-
toric maritime culture and Neolithic seascapes in the region (Fig. 1.11).
Rock art investigated in Zhuhai (珠海), Macao and Hong Kong also depicts
fishing boats and sacrifices at sea, strengthening the evidence for a maritime
seascape (Xu and Liang 1991). The success of the Pearl River Delta’s prehistoric
maritime settlements created a solid foundation for the development of early Panyu,
an historical seaport, into the capital of the South-Yue kingdom (Cao 2003). In
addition, dozens of Neolithic sand dune and shell midden sites (dating to between
5000 and 2500 BP) in the coastal region around Hainan Island reflect the devel-
oping prehistoric maritime practices of local indigenous peoples (He 2012).
The typological comparison of the artifacts from these maritime settlement sites
has revealed that prehistoric maritime cultural exchange and diffusion had been
active both on the southern coast of Lingnan and the northern coast of the Indochina
Peninsula, thus laying the foundation for subsequent navigation of the South China
Sea. The earliest maritime cultural interaction and primitive seafaring was observed
between the mainland coast of the Pearl River Delta and the near shore islands such
as the Qiao (淇澳), Sanzao (三灶), Hengqin (横琴), Dong’ao (东澳), Gaolan
(高栏), and Hebao (荷包) islands of Zhuhai (ZHMM 1991, 1999); the Dachan
(大铲) and Neilingdin (内伶仃) islands of Shenzhen (深圳) (Huang and Wen
1990); the Dayu (大屿) and Chilajiao (赤邋角) islands of Hong Kong; and Jiu’ao
(九澳) Island of Macao (Meacham 1994; Ou and Tang 1988; Tang and Zheng
1996). Archaeological evidence of cultural practices in these near shore islands
revealed typological similarities with those in the Pearl River Delta that date to
between 7000 and 4000BP, further indicating the diffusion Neolithic maritime
culture to the shoreline islands (Shang and Chen 1990).
The maritime cultural exchange between the coastal regions of Lingnan and the
Indochina Peninsula was also reflected in the prehistoric cultural similarities
observed across the Beibu Gulf. Archaeologists discovered that the Cai Beo and
Quynh Van Cultures on the northeast coast of Vietnam (dating to between 6500 and
4500BP) had been highly similar to the cultural deposits found at more than 10
shell midden sites in Dongxing (东兴), Fangcheng (防城), and Qinzhou (钦州)
counties in the Beibu Gulf region of Guangxi, as well as more than 20 shell midden
and sandy dune sites at Xiantouling (咸头岭) and Houshawan (后沙湾) from the
Middle Neolithic age, found in the estuary region of the Pearl River and charac-
terized by chipped point stone choppers (haolizhuo, 蠔蛎琢, also known as oyster
pickers) and coarse corded pottery. The Ha Long and Bau Tro Cultures in the
coastal region of Vietnam (dating to between 4500 and 3500BP) were also similar
to the Neolithic cultures of the Longshan (龙山) period and the Xia (夏) dynasty
and found in the coastal regions of Guangdong (广东), Guangxi (广西) and Hainan
(海南) of southern China. These deposits are characterized by stone bark-cloth
beaters, shouldered stone axes, and LingLing-O type stone earrings (Fig. 1.12; Gan
2008; Tang 2000; Wu 2008, 2011).
During the Bronze Age, the Dong Dau, Go Mun and Dong San Cultures on the
coast of Vietnam were likewise identified as similar to that of the indigenous
1 A Synthetic Analysis of the Neolithic Origins … 33
Fig. 1.12 Distribution of the stone stepped adze, shouldered axe, bark-cloth beater and
Ling-Ling-O earring illustrates cultural exchange across the South China Sea. Artifacts shown
come from: 1, Lower Layer of Hemudu, Yuyao of Zhejiang, China (7000-6000 BP, 河姆渡); 2,
Qianxihu Site, Yuyao of Zhejiang, China (5000-4000 BP, 前溪湖); 3, Huangguashan, Xiapu of
Fujian, China (4000-3500 BP, 黄瓜山); 4, Upper Layer of Xitou Site, Minhou of Fujian, China
(3500-3000 BP, 溪头); 5, 6, 11, 12, Yuanshan Culture, Taipei of Taiwan, China (3500-2000 BP, 圆
山); 7, 8, 13, 14, Luzon, Philippine. (2200-1000 BP); 9, 16, Lower layer of Shixia, Qujiang of
Guangdong, China (4500-4000 BP, 石峡); 10, 15, Haifeng, Guangdong, China (4000-3000 BP, 海
丰); 17, 18, Nongshan, Wuming of Guangxi, China (4000-3000 BP, 武鸣); 19, 20, Cai Beo Culture,
Vietnam (5000-4000 BP); 21, Bianhe, Vietnam (4000-3000 BP); 22, Nan River Basin, Thailand
(4000-3000 BP); 23, Thailand (4000-3000 BP); 24, Baishuixi, Taipei of Taiwan, China (5000-4000
BP, 台北); 25, Tapengken, Taipei of Taiwan, China (5000-4000 BP, 大坌坑); 26, 27, Luzon,
Philippine (3000-2000 BP); 28, Xiantouling, Shenzhen of Guangdong, China (7000-5000 BP, 咸头
岭); 29, Longxue, Zhongshan of Guangdong, China (7000-5000 BP, 龙穴); 30, 31, Phung Nguyen
Culture, Vietnam (5000-3500 BP); 32, Lo Grach, Vietnam (5000-3500 BP); 33, Badong, Vietnam
(4000-3000 BP); 34, Sulitani, Thailand (4000-3000 BP); 35, Nakongzutangmali, Thailand
(4000-3000 BP); 36, 37, 38, Beinan Culture, Taitung of Taiwan, China (3000-2000 BP, 卑南);
39, 40, 41, Luzon, Philippine (2000-1000 BP); 42, 43, Middle layer of Shixia, Qujiang of Guandong,
China (4000-3500 BP, 石峡); 44, Nanya, Hongkong, China (4000-3000 BP, 南Y岛); 45, Yangshan,
Wuming of Guangxi, China (3500-3000 BP, 武鸣); 46, Guogailing, Tiandong of Gguangxi, China
(3500-3000 BP, 锅盖岭); 47, Phung Nguyen Culture, Vietnam (5000-3500 BP); 48, 49, Go Mun
Culture, Vietnam (3000-2500 BP); 50, 51, 52, Sa Huynh Culture, Vietnam (2500-2000 BP)
34 C. Wu
Min-Yue (闽越), South-Yue (南越) and Xi-Ou (西瓯) of southern China, demon-
strating the cultural interaction across the trans-border coast of the South China Sea as
well as the cultural unity of the indigenous Bai-Yue (百越) system (Wu 2010).
Archaeologists investigating the typology and distribution of stone stepped adzes and
stone shouldered axes, which were the most common Neolithic artifacts found in
southern China and Southeast Asia, discovered that the stone stepped adze originated
in the lower reaches of the Yangtze River before diffusing southward, heading mainly
across the East Ocean from Fujian to Taiwan, the Philippines and the Pacific
archipelagoes. The stone shouldered axe originated on the Pearl River Delta and also
spread southward, mainly traveling across the South Ocean and along the coast of the
Indochina Peninsula and mainland Southeast Asia (Fu 1988).
1.6 Conclusion
Previous hypotheses regarding the Maritime Silk Road that connected East and
West via the South China Sea and the Indian Ocean do not accurately reflect the
complicated and diverse circumstances of ancient navigation routes in the seas
around China dating back to the Neolithic Age. This Maritime Silk Road did not
necessary emerge later than the inland Silk Road, nor did it likely serve as a cultural
continuation or replacement for the land route. Instead, historical accounts and
archaeological investigations have revealed that the Maritime Silk Road was based
on “Four Seas” and “Four Oceans” sea routes that had originated in the seafaring
Neolithic ancestors of indigenous Yi and Yue peoples in the Asia-Pacific region.
These groups created not only flourishing maritime cultures and Neolithic seascapes
in the coastal regions of eastern Asia, but they also charted the earliest navigation
routes between mainland Asia, coastal islands, and the off-shore archipelagoes of
East Asia, Southeast Asia and the Pacific. As a result, prehistoric maritime cultural
practices and Neolithic seafaring effectively represent the foundations of the his-
torical Maritime Silk Road and “Four Oceans” Navigation. The prehistoric
Maritime Silk Road was thus one of the most important cultural factors from the
indigenous Yi and Yue to subsequently contribute to what became a pluralistic
ancient Chinese civilization.
References
Anders, H. S. (1936). Sidenvaigen (The silk road) (Hong Jiang 江红, & Peijun Li 李佩娟, Trans.
《丝绸之路》). Urumqi: Xinjiang Renmin Press (新疆人民出版社).
Ban, G. 班固. (1962). Han Shu (History of the Han Dynasty, 《汉书》, originally published in
East Han dynasty), New edition. Beijing: Zhongua Shuju (中华书局, Zhonghua Book
Company).
Bellwood, P. (1997). Prehistory of the Indo-Malaysian Archipelago. Honolulu: University of
Hawaii Press.
1 A Synthetic Analysis of the Neolithic Origins … 35
Bellwood, P. (2005). First farmers: The origins of agricultural societies. Oxford: Blackwell.
Boulnoi, L. (1963). La route de la soie (The silk road) (Sheng Geng 耿升, Trans. 1982 《丝绸之
路》). Urumqi: Xinjiang Renmin Press (新疆人民出版社).
Cao, J. 曹峻. (2003). A preliminary study on maritime essentials in the capitals of the Bai-Yue
Kingdoms. In C. Tang 邓聪, & C. Wu 吴春明 (Ed.), Archaeological study of Southeast China
(Vol. 3, pp. 224–242). Xiamen: Xiamen University Press (Baiyue Ducheng Haiyangxing Chutan,
《百越都城海洋性初论》, in Dongnan Kaogu Yanjiu, 《东南考古研究》, 厦门大学出版社).
Cao, J. 曹峻. (2012). An investigation into the change and maritime adaptation of prehistoric
settlements on the southern coast of Hangzhou Bay. In C. Wu 吴春明 (Ed.), Maritime cultural
heritage and archaeology in seas around China (pp. 28–38). Beijing: Science Press (Kexue
Chubanshe) (Hangzhou Nan’an Shiqian Juluo de Bianqian yu Haiyang Shiying, 《杭州湾南
岸史前聚落的变迁与海洋适应》, in Haiyang Yichan yu Kaogu 《海洋遗产与考古》, 科
学出版社).
Chang, K. C. 张光直. (1987). Archaeology in southeastern coastal China and the origin of the
Austronesians. In The Museum of Sichuan University, Research Society of Bronze Drum of
Ancient China (Ed.), Southern ethnology and archaeology (Vol. 1, pp. 1–14). Chengdu:
Sichuan University Press (Zhongguo Dongnan Haian Kaogu yu Nandao Yuzu Qiyuan Wenti,
《中国东南海岸考古与南岛语族的起源问题》, in Nanfang Minzu Kaogu 《南方民族考
古》, 四川大学出版社).
Chavannes, E. (1903). Historical documentation of the western Tou-Kiue (Documents sur les
Tou-kiue occidentaux) (Chengjun Feng 冯承钧, Trans. 2004, Xi Tujue Shiliao 《西突厥史
料》). Beijing: Zhonghua Book Company (中华书局).
Chen, C. 陈仲玉. (2012). The Neolithic maritime settlement sites of the Mazhu Islands. In C. Wu
吴春明 (Ed.), Maritime cultural heritage and archaeology in the seas around China (pp. 39–
47). Beijing: Science Press (Kexue Chubanshe) (Mazu Liedao Xinshiqi Shidai de Haiyang
Juluo, 《马祖列岛新石器时代的海洋聚落》, in Haiyang Yichan yu Kaogu《海洋遗产与考
古》, 科学出版社).
Chen, D. 陈大震. (1986). Dade Nanhai Zhi Canben (Biography of the South China Sea
[incomplete copy], 《大德南海志残本》, originally published in 8th of Dade [大德] reign
[1304] of Yuan dynasty). New edition. Guangzhou: Guangzhoushi Difangzhi Yanjiusuo (广州
市地方志研究所, Guangzhou Municipal Institute of Chorography).
Chen, G. 陈高华, Wu, T. 吴泰, & Guo, S. 郭松义. (1991). The maritime silk road. Beijing: China
Ocean Press (Haiyang Chubanshe) (Haishang Sichou Zhilu 《海上丝绸之路》, 海洋出版社).
Chen, J. 陈佳荣. (1992). On the East, West, South and North Ocean in the Song, Yuan, Ming and
Qing Dynasties. Journal of Maritime History (Haijiaoshi Yanjiu), 1, 9–15 (Song, Yuan, Ming,
Qing Zhi Dongxii Nanbei Yang, 《宋元明清之东西南北洋》, 《海交史研究》).
Chen, S. 陈寿, & Pei, S. 裴松之. (Ed.). (2006). Sanguozhi (History of the Three Kingdoms,
《三国志》, originally published in Jin dynasty), New edition. Beijing: Zhongua Shuju (中华
书局, Zhonghua Book Company).
Chen, Y. 陈炎. (1982). A preliminary study of the Maritime Silk Road. Historical Research
(Lishi Yanjiu), 3, 161–177 (Luelun Haishang Sichou Zhilu 《略论海上丝绸之路》, 《历史
研究》).
Fan, Y. 范晔. (1965). Houhan Shu (History of the Later Han Dynasty 《后汉书》, originally
published in Liu-Song dynasty), New edition. Beijing: Zhongua Shuju (中华书局, Zhonghua
Book Company).
Fu, X. 傅宪国. (1988). On stepped stone adzes and shouldered stone implements. ACTA
Archaeological Sinica (Kaogu Xuebao), 1, 1–36 (Lun Youduan Shiben yu Youjian Shiqi, 《论
有段石锛和有肩石器》, 《考古学报》).
Gan, X. 干小莉. (2008). A study of jade earrings of the projection type and prehistoric cultural
diffusion across South China Sea areas. Cultural Relics in Southern China (Nanfang Wenwu),
3, 109–112, 93 (Cong Tu-niu-xing Jue Kan Huannanhai Shiqian Wenhua de Jiaoliu, 《从凸纽
形玦看环南海史前文化的交流》, 《南方文物》).
Han, R. 韩榕. (1986). A preliminary study of the prehistoric cultures in Jiaodong Peninsula. In
Shandong Provincial Qilu Archaeology Series Editorial Office (Ed.), Symposium on the
36 C. Wu
Prehistoric Culture of Shandong (pp. 96–119). Jinan: Qilu Press (Jiaodong Shiqian Wenhua
Chutan, 《胶东史前文化初探》, in Shandong Shiqian Wenhua Lunwenji, 《山东史前文化
论文集》, 齐鲁书社).
Han, Z. 韩振华. (1958). The maritime transportation between China, India and Southeast Asia
during the second century BC to the first century BC. Journal of Xiamen University (Xiamen
Daxue Xuebao), 2, 195–227 (Gongyuanqian Ershiji zhi Yishiji Jian Zhongguo yu Yindu
Dongnanya de Haishang Jiaotong——Hanshu Dilizhi Yueditiao Moduan Kaoshi, 《公元前二
世纪至一世纪间中国与印度东南亚的海上交通——汉书地理志粤地条末段考释》, 《厦
门大学学报》).
He, G. 何国俊. (2012). Prehistoric settlement and maritime cultures on the coast of Hainan Island.
In C. Wu 吴春明 (Ed.), Maritime cultural heritage and archaeology in the seas around China
(pp. 48–56). Beijing: Science Press (Kexue Chubanshe) (Huan Hainandao de Shiqian Juluo yu
Haiyang Wenhua, 《环海南岛的史前聚落与海洋文化》, in Haiyang Yichan yu Kaogu,
《海洋遗产与考古》, 科学出版社).
Huang, C. 黄崇岳, & Wen, B. 文本亨. (1990). Review of 10 years’ work on archaeological and
cultural relics in Shenzhen. Cultural Relics (Wenwu), 11, 53–56, 43 (Shenzhen Wenwu Kaogu
Gongzuo Shinian, 《深圳文物考古工作十年》, 《文物》).
Huang, H. 黃怀信, Zhang, M. 張懋镕, & Tian, X. 田旭東. (2007). Yi Zhoushu Huijiao Jizhu
(Annotations and Commentaries of the Lost Historical Literature of Zhou Dynasty, 《逸周书
汇校集注》), Shanghai: Shanghai Guji Chubanshe (上海古籍出版社, Shanghai Classic
Publishing House).
IACASS (Institute of Archaeology, Chinese Academy of Social Science). (1999). Shell midden
sites in the Jiaodong Peninsula studies in environmental archaeology. Beijing: Social Science
Academic Press (Sheke Wenxian Chubansh) (Jiaodong Bandao Beiqiu Yizhi Huanjing Kaogu,
《胶东半岛贝丘遗址环境考古》, 社科文献出版社).
Jiao, T. 焦天龙. (2012). Issues of prehistoric maritime settlement archaeology of China. In C. Wu
吴春明 (Ed.), Maritime cultural heritage and archaeology in the seas around China (pp. 11–
18). Beijing: Science Press (Kexue Chubanshe) (Shiqian Zhongguo Haiyang Juluo Kaogu de
Ruogan Wenti, 《史前中国海洋聚落考古的若干问题》, in Haiyang Yichan yu Kaogu,
《海洋遗产与考古》, 科学出版社).
Li, D. 郦道元, & Wang, G. 王国维. (Ed.). (1984). Shuijingzhu Jiao (Commentary on the
Waterways Classic, 《水经注校》, originally published in North Wei dynasty and annotated
in Qing dynasty). New edition. Shanghai: Shanghai Renmin Chubanshe (上海人民出版社,
Shanghai People’s Publishing House).
Lin, C. 凌纯声. (1954). The ancient chinese maritime culture and the “Asian Mediterranean.”
Journal of Overseas Culture (Haiwai Zazhi), 3, 7–10. Zhongguo Gudai Haiyang Wenhua yu
Yazhou Dizhonghai, 《中国古代海洋文化与亚洲地中海》, 《海外杂志》).
Lin, H. 林惠祥. (1937). A Neolithic Site in Wuping, Fukien. Paper for “The Third Congress of
Prehistorians of the Far East” (Singapore), Xiamen University library collection.
Lin, H. 林惠祥. (1955). A study on the stone age artifacts in Taiwan. Journal of Xiamen
University (Xiamen Daxue Xuebao), 4, 135–155 (Taiwan Shiqi Shidai Yiwu de Yanjiu 《台湾
石器时代遗物的研究》, 《厦门大学学报》).
Lin, H. 林惠祥. (1958). The stepped adze: One of the characteristics of neolithic culture in the
southeastern region of China. ACTA Archaeological Sinica (Kaogu Xuebao), 3, 1–23
(Zhongguo Dongnan Diqu Xinshiqi Shidai Tezheng Zhiyi: Youduan Shiben, 《中国东南地区
新石器文化特征之一:有段石锛》, 《考古学报》).
Lin, M. 林梅村. (2006). Lectures on the archaeological issues of the Silk Road. Beijing: Peking
University Press (Beijing Daxue Chubanshe) (Sichou Zhilu Kaogu Shuwu Jiang, 《丝绸之路
考古十五讲》, 北京大学出版社).
Liu, A. 刘安, & Chen, J. 陈静. (Ed.). (2010). Huai’nanzi (The Book of the Prince of Huainan,
《淮南子》, originally published in Han dynasty). New edition. Zhengzhou: Zhongzhou Guji
Chubanshe (中州古籍出版社 Zhongzhou Ancient Book Publishing House).
Luan, F. 栾丰实. (2008). A analysis on the prehistoric emergence, development and dispersal of
rice agriculture in Haidai region. In F. Luan, & K. Miyamoto (Ed.), Research works of
1 A Synthetic Analysis of the Neolithic Origins … 37
anthropology and early agriculture in the Haidai region (pp. 41–55). Beijing: Science Press
(Kexue Chunbanshe) (Haidai Diqu Shiqian Shiqi Daozuo Nongye de Chansheng, Fazhan yu
Kuosan 《海岱地区史前时期稻作农业的产生、发展与扩散》, in Haidai Diqu Zaoqi
nongye he Renleixue Yanjiu 《海岱地区早期农业和人类学研究》, 科学出版社).
Lv, B. 吕不韦, & Chen, Q. 陈奇猷. (Ed.). (2002). Lvshi Chunqi Xinjiaoshi (Analects of the
Political Opinion of Master Lv with new annotation. 《吕氏春秋新校释》, originally
published in Warring States period). New edition. Shanghai: Shanghai Guji Chubanshe (上海
古籍出版社, Shanghai Classic Publishing House).
Ma, D. 马端临. (2000). Wenxian Tongkao (Textual Research on Historical Narration, 《文献通
考》, originally published 11th of Dade [大德] reign [1307] of Yuan dynasty). New edition.
Hangzhou: Zhejiang Guji Chubanshe (浙江古籍出版社, Ancient Book Publishing House).
Meacham, W. (1994). Archaeological investigations on Chek Lap Kok Island. Hong Kong: Hong
Kong Archaeological Society.
OCQZ.UNESCO (Organizing Committee of the Quanzhou International Seminar of the
Comprehensive Investigation of Maritime Silk Route of UNESCO). (1991). China and
maritime silk route. Fuzhou: Fujian People’s Publishing House (Fujian Renmin Chubanshe)
(Zhongguo yu Haishang Sichou Zhilu 《中国与海上丝绸之路》, 福建人民出版社).
Ou, J. 区家发, & Tang, C. 邓聪. (1988). The archaeological excavation of the Neolithic Dongwan
Sand Dune Site at Dayushan Island in Hong Kong. In Guangdong Provincial Museum, Qujiang
County Museum (Ed.), Symposium Commemorating the 30th Anniversary of the Discovery of
the Maba Homo Sapiens Fossil (pp. 208–217). Beijing: Cultural Relics Press (Xianggang
Dayushan Dongwan Xinshiqi Shidai Shaqiu Yizhi de Fajue 《香港大屿山东湾新石器时代
沙丘遗址的发掘》, in Ji’nian Mabaren Huashi Faxian Sanshi Zhounian Wenji, 《纪念马坝
人化石发现三十周年文集》, Wenwu Chubanshe, 文物出版社).
Ouyang, X. 欧阳修, & Song, Q. 宋祁. (1975). Xin Tangshu (New History of the Tang Dynasty,
《新唐书》, originally published in Song dynasty). New edition. Beijing: Zhongua Shuju (中
华书局, Zhonghua Book Company).
von Richthofen, F. (1877). China, ergebnisse eigener reisen und derauf gegrondeter studien 1.
Berlin: Verlag von Dietrich Reimer.
Rolett, B. V., Jiao, T., & Lin, G. (2002). Early seafaring in the Taiwan Strait and the search for
Austronesian origins. Journal of East Asian Archaeology, 4, 307–319.
Rolett, B. V., Zheng, Z., & Yue, Y. (2011). Holocene sea-level change and the emergence of Neolithic
seafaring in the Fuzhou Basin (Fujian, China). Quaternary Science Reviews, 30, 788–797.
Ruan, Y. 阮元. (Ed.). (2009). Shisanjing Zhushu (The Thirteen Classics, with annotations and
commentaries, 《十三经注疏》, originally published in Song dynasty and re-edited & printed
by Ruan Yuan in 20th year of Jiqqing 嘉庆 reign of Qing dynasty). New edition. Beijing:
Zhongua Shuju (中华书局, Zhonghua Book Company).
Shang, Z. 商志香覃, & Chen, S. 谌世龙. (1990). A study of the cultural characteristics and related
issues of the prehistoric sand dune sites in the Pearl River Estuary. Cultural Relics (Wenwu),
11, 48–52, 67 (Huan Zhujiangkou Shiqian Shaqiu Yizhi de Tedian ji Youguan Wenti, 《环珠
江口史前沙丘遗址的特点及有关问题》, 《文物》).
Sima, Q. 司马迁. (1959). Shiji (Records of the Historian, 《史记》, originally published in Han
dynasty), New edition. Beijing: Zhongua Shuju (中华书局, Zhonghua Book Company).
Su, B. 苏秉琦. (1978). A preliminary study of the Neolithic archaeology of the southeastern coast
of China. Cultural Relics (Wenwu), 3, 40–42 (Luetan Woguo Dongnan Yanhai Diqu de
Xinshiqi Shidai Kaogu 《略谈我国东南沿海地区的新石器时代考古》, 《文物》).
Su, W. 苏文福. (Ed.). (1991). Comprehensive investigation of “Maritime Silk Route” sponsored
by UNESCO, Quanzhou: The Organizing Committee for UNNESCO Investigation of
Maritime Silk Route (Lianheguo Jiaokewen Zuzhi Haishang Sichou Zhilu Zonghe Kaocha
《联合国教科文组织“海上丝绸之路”综合考察》, Quanzhou Haishang Sichou Zhilu
Kaocha Huodong Chouweihui Bangongshi 泉州“海上丝绸之路”考察活动筹委会办公室).
Tang, C. 邓聪, & Zheng, W. 郑炜明. (1996). Archaeology at the Heisha Site in Macao. Hong
Kong: Chinese University Press of Hong Kong (Ao’men Heisha, 《澳门黑沙》, Xianggang
Zongwen Daxue Chubanshe, 香港中文大学出版社).
38 C. Wu
Tang, C. 邓聪. (2000). The oceanic emigration of prehistoric Mongolian ethnicity: A case study
on the discovery of Neolithic bark-cloth beaterson the South China Coast. Southeast Culture
(Dongnan Wenhua), 1, 6–22 (Shiqian Menggu Renzhong haiyang Kuosan Yanjiu——Lingnan
Shupibu Wenhua Faxian Jiqi Yiyi, 《史前蒙古人种海洋扩散研究——岭南树皮布文化发
现及其意义》, 《东南文化》).
Tong, W. 佟伟华. (1989). On the primitive cultural exchange between Jiaodong and Liaodong
Peninsulas. In B. Su (Ed.), Symposium of Archaeological Culture (Vol. 2, pp. 78–95). Beijing:
Cultural Relics Press (Wenwu Chubanshe) (Jiaodong Bandao yu Liaodong Bandao Yuanshi
Wenhua de Jiaoliu 《胶东半岛与辽东半岛原始文化的交流》, in Kaoguxue Wenhua Lunji
《考古学文化论集(二)》, 文物出版社).
Tuotuo 脱脱. (1977). Songshi (History of the Song Dynasty, 《宋史》, originally published in
Yuan dynasty). New edition. Beijing: Zhongua Shuju (中华书局, Zhonghua Book Company).
Wang, D. 汪大渊. (1981). Daoyi Zhilue (Biography of the Foreign Island Barbarians, 《岛夷志
略》, originally published in Yuan dynasty), new edition annotated by Su, J. 苏继庼. Beijing:
Zhongua Shuju (中华书局, Zhonghua Book Company).
Wang, F. 王富强. (2008). An archaeological perspective on the subsistence economy of the
Jiaodong region before the Zhou Dynasty. In F. Luan 栾丰实, & K. Miyamoto 宫本一夫
(Ed.), Research works of anthropology and early agriculture in the Haidai region (pp. 59–71).
Beijing: Science Press (Kexue Chunbanshe) (Zhoudai Yiqian Jiaodong Diqu Jingji Xingtai de
Kaoguxue Guancha 《周代以前胶东地区经济形态的考古学观察》, in Haidai Diqu Zaoqi
nongye he Renleixue Yanjiu 《海岱地区早期农业和人类学研究》, 科学出版社).
Wang, X. 王锡平, & Li, B. 李步青. (1990). Study of the prehistoric cultural interaction between
Jiaodong and Liaodong Peninsulas. In The Society of Chinese Archaeology (Ed.), The
Proceeding of the Sixth Annual Conference of Chinese Archaeology Society (pp. 89–95). Beijing:
Cultural Relics Press (Wenwu Chubanshe) (Lun Jiaodong Bandao yu Liaodong Bandao Shiqian
Wenhua Guanxi 《论胶东半岛与辽东半岛史前文化关系》, in Zhongguo Kaogu Xuehui
Diliuci Nianhui Linwenji 《中国考古学会第六次年会论文集》, 文物出版社).
Wang, Z. 王子今. (1992). On transportation in the East and South Ocean of the Qin and Han
Dynasties. Journal of Maritime History (Haijiaoshi Yanjiu), 1, 1–8 (Qinhan Shiqi de
Dongyang yu Nanyang Hangyun, 《秦汉时期的东洋与南洋航运》, 《海交史研究》).
Wei, Z. 魏征, & Linghu, D. 令狐德. (1982). Suishu (History of the Sui Dynasty, 《隋书》,
originally published in Tang dynasty). New edition. Beijing: Zhongua Shuju (中华书局,
Zhonghua Book Company).
Wu, C. 吴春明. (1994). On the prehistoric foundation of the Gaoshan Nationality’s Culture as
exhibited in primitive ceramics. Archaeology (Kaogu), 11, 1022–1027, 1046. (Cong Yuanshi
Xhitao Tantao Gaoshanzu Wenhua de Shiqian Jichu, 《从原始制陶探讨高山族文化的史前
基础》, 《考古》).
Wu, C. 吴春明. (1995). A preliminary study of the prehistoric and early cultures of the pre-Qin
and Han Dynasties in the Minjiang River Basin. ACTA Archaeological Sinica (Kaogu
Xuebao), 3, 147–172 (Minjiang Liuyu Xianqian Lianghan Wenhua de Chubu Yanjiu 《闽江
流域先秦两汉文化的初步研究》, 《考古学报》).
Wu, C. 吴春明, & Chen, W. 陈文. (2003). An ethno-archaeological review of the “Fujian-Taiwan
Hypothesis” on the origin of Austronesians. Ethno-National Studies (Minzu Yanjiu), 4, 75–83,
109 (Nandao Yuzu Qiyuan Yanjiuzhong Mintaishuo de Shangque, 《”南岛语族”起源研究
中”闽台说”的商榷》, 《民族研究》).
Wu, C. 吴春明, & Cao, J. 曹峻. (2005). The wrong research methodology for the origin of
Austronesians. Journal of Xiamen University (Xiamen Daxue Xuebao), 3, 85–93 (Nandao
Yuzu Qiyuan Yanjiuzhong de Sige Wuqu, 《南岛语族起源研究中的四个误区》, 《厦门大
学学报》).
Wu, C. 吴春明. (2008). On the prehistoric cultural interaction between the Philippines and
Southern China. Archaeology (Kaogu), 9, 36–50. (Feilvbin Shiqian Wenhua yu Huanan de
Guanxi, 《菲律宾史前文化与华南的关系》, 《考古》).
Wu, C. 吴春明. (2010). A study on the issues relating to the Ou-Luo Kingdom and Dongsan
Culture discovered in Vietnam. In C. Tang 邓聪, & C. Wu 吴春明 (Ed.), Archaeological study
1 A Synthetic Analysis of the Neolithic Origins … 39
of Southeast China (Vol. 4, pp. 389–400). Xiamen: Xiamen University Press (Dongshan
Wenhua yu Ouluo Guo Wenti, 《东山文化与瓯雒国问题》, in Dongnan Kaogu Yanjiu,
《东南考古研究》, 厦门大学出版社).
Wu, C. 吴春明. (2011). The Neolithic cultural interaction between the lower basin of the Red
River of Vietnam and the coast of South China. In Research Society of Bai-Yue Ethnicities
(Ed.), Study of Bai-Yue Ethnicities (Vol. 2, pp. 44–56). Hefei: Anhui University Press (Anhui
Daxue Chubanshe)(Honghe Xiayou Xinshiqi Shidai Wenhua yu Huanan de Guanxi, 《红河下
游新石器时代文化与华南的关系》, in Baiyue Yanjiu, 《百越研究》, 安徽大学出版社).
Wu, C. 吴春明. (2012). Maritime culture and maritime archaeology in the seas around China. In C.
Wu 吴春明 (Ed.), Maritime cultural heritage and archaeology in the seas around China (pp. 1–
10). Beijing: Science Press (Kexue Chubanshe) (Haiyang Wenhua yu Haiyang Kaogu, 《海洋文
化与海洋考古》, in Haiyang Yichan yu Kaogu《海洋遗产与考古》, 科学出版社).
Wu, C. 吴春明, & Li, J. 李家添. (1992). A preliminary chronology for the prehistoric culture of
the west coast of Taiwan. Cultural Relics in Southern China (Nanfang Wenwu), 3, 39–46
(Taiwan Xihaian Shiqian Wenhua Biannian Chulun, 《台湾西海岸史前文化编年初论》,
《南方文物》).
Wu, Y. 吴玉贤. (1983). An archaeological perspective on primitive maritime transportation in the
shoreline region of Ningpo. Prehistoric Studies (Shiqian Yanjiu), 1, 156–162 (Cong Kaogu
Faxian Tan Ningpo Yanhai Diqu Yuanshi Jumin de Haishang Jiaotong, 《从考古发现谈宁波
沿海地区原始居民的海上交通》, 《史前研究》).
Xiang, D. 向达. (Ed.). (1961). Liangzhong Haidao Zhenjing (Two Ancient Nautical Guides, with
anotation and commentaries, 《两种海道针经》, originally made in Ming and Qing dynasties
with lost name of authors). New edition. Beijing: Zhongua Shuju (中华书局, Zhonghua Book
Company).
Xu, H. 徐恒彬, & Liang, Z. 梁振兴. (1991). Investigations of the rock art and settlement site at
Baojingwan, Gaolang Island, Zhuhai. In Zhuhai Municipal Museum, Guangdong Provincial
Institute of Archaeology and Cultural Relic, Guangdong Provincial Museum (Ed.),
Archaeological discovery and research in Zhuhai (pp. 282–294). Guangzhou: Guangdong
Renmin Press (Gaolangdao Baojingwan Shike Yanhua yu Guyizhi de Faxian yu Yanjiu, 《高
栏岛宝镜湾石刻岩画与古遗址的发现与研究》, in Zhuhai Kaogu Faxian yu Yanjiu, 《珠
海考古发现与研究》, 广东人民出版社).
Yan, W. 严文明. (1986). A preliminary study of the prehistoric cultures on Jiaodong Peninsula. In
Shandong Provincial Qilu Archaeology Series Editorial Office (Ed.), Symposium on the
Prehistoric Culture of Shandong (pp. 63–95). Jinan: Qilu Press (Jiaodong Yuanshi Wenhua
Chulun, 《胶东原始文化初论》, in Shandong Shiqian Wenhua Lunwenji, 《山东史前文化
论文集》, 齐鲁书社).
Yan, W. 严文明. (2000a). Eastward transmission of ancient Chinese agriculture and its impact on
the archaic Japan society. In W. Yan 严文明 (Ed.), The Origins of agriculture and Rise of
civilization (pp. 44–46). Beijing: Science Press (Kexue Chubanshe) (Zhongguo Fudai Nongye
Wenhua de Dongchuan Dui Riben Zaoqi Shehui Fazhan de Yingxiang 《中国古代农业文化
的东传对日本早期社会发展的影响》, in Nongye Qiyuan yu Wenming Fasheng 《农业发
生与文明起源》, 科学出版社).
Yan, W. 严文明. (2000b). Notes on Archaeology in Shandong peninsula. In W. Yan 严文明
(Ed.), The Origins of agriculture and Rise of civilization (pp. 302–308). Beijing: Science Press
(Kexue Chubanshe) (Jiaodong Kaogu Ji 《胶东考古记》, in Nongye Qiyuan yu Wenming
Fasheng 《农业发生与文明起源》, 科学出版社).
Yuan, J. 袁靖. (1995). Some issues regarding research on the shell midden sites along the
southeast coast of China. Archaeology (Kaogu), 12, 1100–1109 (Guanyu Zhongguo Dalu
Dongnan Yanhai Beiqiu Yizhi Yanjiu de Jige Wenti 《关于中国大陆东南沿海贝丘遗址研
究的几个问题》, 《考古》).
Yuan, J. 袁靖. (1998). Review of a few issues relating to environmental archaeology at the shell
midden sites of Jiaodong Peninsula. Southeast Culture (Dongnan Wenhua), 2, 36–39 (Guanyu
Jidong Bandao Beiqiu Yizhi Huanjing Kaogu Yanjiu de Jidian Sikao, 《关于胶东半岛贝丘遗
址环境考古研究的几点思考》, 《东南文化》).
40 C. Wu
Yuan, J. 袁靖. (1999). Environmental archaeological concerns regarding the shell midden sites of
the Pearl River Delta. In C. Tang 邓聪, & C. Wu 吴春明 (Ed.), Archaeological study of
Southeast China (Vol. 2, pp. 147–149). Xiamen: Xiamen University Press (Zhujiang
Sanjiaozhou Beiqiu Yizhi de Huanjing Kaoguxue Wenti, 《珠江三角洲贝丘遗址的环境考
古学问题》, in Dongnan Kaogu Yanjiu, 《东南考古研究》, 厦门大学出版社).
Yuan, K. 袁康, & Wu, P. 吴平. (Ed.). (1985). Yuejueshu (The History of the Lost Yue Nation,
《越绝书》, originally made in Warring States period and re-edited in Eastern Han dynasty).
New edition. Shanghai: Shanghai Guji Chubanshe (上海古籍出版社, Shanghai Classic
Publishing House).
Yuan, K. 袁珂. (Ed.). (2014). Shanhaijing Jiaozhu (Commentaries on the Classics of the Mounts
and Seas, 山海经校注). Beijing: Beijing Lianhe Chubanshe (北京联合出版社, Beijing United
Publishing House).
Zhang, C. 张崇根. (1981a). Linhai Shuitu Yiwuzhi Jijiao (Annotations and Commentaries of
Biography of the Environment and Product of the Linhai Prefecture, 《临海水土异物志辑
校》, originally made and published by Shenying [沈莹] in Three Kingdoms period). Beijing:
Nongye Chubanshe (China Agriculture Press, 农业出版社).
Zhang, H. 张华. (2012). Bowuzhi Waiqizhong (Biography of the Things and Cultures, attached 7
other works, 《博物志外七种》, originally published in Jin [晋] dynasty). New edition.
Shanghai: Shanghai Guji Chubanshe (上海古籍出版社, Shanghai Classic Publishing House).
Zhang X. 张燮. (1981b). Dongxiyang Kao (General Survey on the East and West Oceans, 《东西
洋考》, originally published 45th year of Wanli [万历] reign of Ming dynasty [1617]), new
edition of Xie, F. 谢方 ed., Beijing: Zhongua Shuju (中华书局, Zhonghua Book Company).
Zhang, X. 章巽. (1985). Faxian Zhuan Jiaozhu (Annotation of the Biography of Monk Faxian,
《法显传校注》), Shanghai: Shanghai Guji Chubanshe (上海古籍出版社, Shanghai Classic
Publishing House).
Zhao, R. 赵汝适. (1985). Zhufan Zhi (Biography of the Foreign Nations, 《诸番志》, originally
published in Song dynasty). New edition. Beijing: Zhongua Shuju (中华书局, Zhonghua Book
Company).
Zhou, Q. 周去非. (1996). Lingaai Daida (Interlocution on the History of South Coast of China,
《岭外代答》, originally published in 5th of Chunxi [淳熙] reign [1178] of Song dynasty).
New edition. Shanghai: Shanghai Yuandong Chubanshe (上海远东出版社, Shanghai Far East
Publisher).
Zhu, F. 朱非素. (1994). The Shell Mound and Sand Dune Sites and their Settlement in the Pearl
River Delta, in Center for Chinese Archaeology and Art, ICS, The Chinese University of Hong
Kong (Ed.), Ancient Cultures of South China and Neighbouring Regions: Essays in Honor of
Professor Cheng Te-K’un on the Occasion of the Sixtieth Anniversity of His Academic Career
(pp. 219–228). Hong Kong: The Chinese University Press (Zhujiang Sanjiaozhou Beiqiu
Shaqiu Yizhi he Juluo Xingtai, 《珠江三角洲贝丘、沙丘遗址和聚落形态》, in Nan
Zhongguohai ji Linjin Diqu Fuwenhua Yanjiu ——Qinzhu Zhengdekun Jiaoshou Congshi
Xueshu Huodong Liushi Zhounian Lunwenji, 《南中国及邻近地区古文化研究——庆祝郑
德坤教授从事学术活动六十周年论文集》, 中文大学出版社1994年).
ZHMM (Zhuhai Municipal Museum). (1991). Archaeological discovery and research in Zhuhai.
Guangzhou: Guangdong Renmin Press (Zhuhai Kaogu Faxian yu Yanjiu, 《珠海考古发现与
研究》, 广东人民出版社).
ZHMM (Zhuhai Municipal Museum). (1999). Archaeological investigations of the Suochiwan site
at Hebao Island, Zhuhai. In C. Tang 邓聪, & C. Wu 吴春明 (Ed.), Archaeological study of
Southeast China (Vol. 2, pp. 71–82). Xiamen: Xiamen University Press (Guangdong Zhuhai
Hebaodao Suochiwan Yizhi Diacha, 《广东珠海荷包岛锁匙湾遗址调查》, in Dongnan
Kaogu Yanjiu, 《东南考古研究》, 厦门大学出版社).
Chapter 2
A Maritime Route Brought First
Farmers to Mainland Southeast Asia
Charles F. W. Higham
Abstract The domestication of rice first took place in the Yangtze River Valley. It
is argued that the expansion of farming communities to the south reached mainland
Southeast Asia starting in the late third millennium BC. The conjunction of new
archaeological and bioanthropogical information, and the re-examination of older
reports, is beginning to shed light on the southward expansion of Neolithic rice
farmers. The existing evidence suggests that a maritime expansion took place,
originating in the lower Yangtze and spreading south along the coast of Fujian to
Lingnan and then into Southeast Asia. This shift brought farmers into a wide range
of new habitats long densely inhabited by indigenous hunter-gatherers. Three key
sites document this maritime expansion in Southeast Asia. Man Bac is located in
Bac Bo, the Red River area of Northern Vietnam; An Son is one of several sites in
the Dong Nai Valley of Southern Vietnam; and Khok Phanom Di is located on the
former estuary of the Bang Pakong River in Central Thailand, where a new analysis
of cranial and dental variables has linked the inhabitants to the migrating farmers.
Yet the population’s adaptation to a marine estuarine habitat made rice cultivation
marginal at best, and the new settlers turned instead to hunting and gathering, even
as they continued to maintain a fully Neolithic material culture.
2.1 Introduction
The two models for identifying when and how domestic rice came to be cultivated
in mainland Southeast Asia (MSEA) are diametrically opposed. The first advocates
local continuity, a scenario in which “the spread of farming is seen mainly as the
result of the adoption and/or diffusion of agricultural technology by the descendants
of in situ hunter–gatherers without necessarily the spread of new languages or
genes” (Pietrusewsky 2010). This has found support in two distinct patterns of tooth
morphology known as Sundadont and Sinodont (Turner 1990). The former is
C. F. W. Higham (&)
Department of Anthropology and Archaeology, University of Otago, Otago, New Zealand
e-mail: charles.higham@otago.ac.nz
Fig. 2.1 Map of East and Southeast Asia showing the sites and locations mentioned in the text,
including: 1. Nanzhuangtou, 2. Jiahu, 3. Peiligang, 4. Cishan, 5. Baligang, 6. Bashidang, 7.
Shangshan, 8. Kuahuqiao, 9. Tianluoshan, 10. Hemudu, 11. Dingshishan, 12. Da But, 13. Shixia,
14. Baiyangcun, 15. Man Bac, 16. Sham Wan, 17. An Son, 18. Nong Nor, 19. Khok Phanom Di,
20. Ban Kao, 21. Non Ratchabat, 22. Ban Non Wat, 23. Non Nok Tha, 24. Ban Chiang, 25. Non
Pa Wai, 26. Weidun and Songze, 27. Huxi; and the following regions: A. Yellow River, B.
Yangtze River, C. Sichuan, D. Yunnan, E. Guangxi, F. Guangdong, G. Red River Valley, H. Dong
Nai Valley, I. Khorat Plateau, J. Central Thailand
2 A Maritime Route Brought First Farmers … 43
prevalent in Southeast Asia, while the latter is concentrated further north. The
second “two layer” hypothesis argues for the migratory expansion of rice and millet
farmers into MSEA from centers of domestication located to the north. A parallel
theory featuring the spread of farming via Taiwan into Island Southeast Asia was
originally proposed by Chang and Goodenough (1985) and subsequently champi-
oned by Bellwood’s studies of coastal and inland prehistoric settlements.
This review of the expansion of farming communities via a coastal route into
MSEA was inspired by the recent reassessment of the settlement of Europe by early
farmers, in which isotopic, genetic, human biological and archaeological analyses
were combined to trace the establishment of the European Neolithic (Whittle and
Bickle 2014). The first essential point is to establish where and when rice was
domesticated.
In the Yangtze River Valley the path to domestication has been identified at
Baligang (Deng et al. 2015), Shangshan (10,000-8000 BC), Huxi (7000-6400 BC),
and Tianluoshan (Fuller et al. 2009; Zheng et al. 2016). The presence of domes-
ticated rice stimulated the outward spread of farming communities into the Sichuan
Basin and south into Yunnan, Guangxi and Guangdong. Southern China was then
occupied by hunter-gatherers, best seen in the Dingshishan Culture sites and the Da
But Culture in northern Vietnam. Across Southeast Asia, in general, there are many
upland rock shelter sites, and doubtless many more such settlements exist in
drowned Sundaland. These hunter-gatherers are often described in the Chinese and
Vietnamese literature as Neolithic due to their pottery and polished stone adzes.
Oxenham and Matsumura (2010: 129) have advocated for the term “Pre-Neolithic
Pottery using Cultures” to distinguish them from farmers. Almost universally, the
former interred the dead in a flexed position with very few, if any, mortuary
offerings. Whatever system of nomenclature is preferred, the contrast between these
complexes is the basis for the “two layer hypothesis” which featured the intrusion
of farmers into a hunter-gatherer ecumene and the corresponding clash of human
biology, language, gene flows and cultural configurations. In order to shed light on
this process, I will now examine new information from a series of important
archaeological sites (Fig. 2.1).
Recent research in the Fuzhou Basin has made a crucial contribution to tracing a
possible maritime route of expansion taken by rice farmers originating in the lower
reaches of the Yangtze (Ma et al. 2013). The recovery of rice phytoliths from
Zhuangbianshan has placed initial rice cultivation in the Tanshishan Phase of the
local sequence, dating to 3000-2300 BC, when the sites were situated on small
islands. It is particularly relevant that the red slipped ceramics of this phase are
strikingly similar to those of early farmer settlements on Taiwan (Rolett et al. 2011;
Yue et al. 2015), suggesting that movement along the coast was accompanied by
the crossing of the Taiwan Strait. Rice became increasingly dominant in the
44 C. F. W. Higham
subsistence of this region during the second millennium BC, as expressed in the
Hangguashan Period. In tracing the spread of rice farming from Fujian south to
Vietnam and the Gulf of Siam, it must be remembered that this plant is not adaptive
to salinity. Thus in the early Neolithic Fuzhou sites, which are dominated by shell
middens, rice was probably cultivated in rain-fed fields as a relatively minor
component of primarily marine-oriented patterns of subsistence.
The initial farmer settlement of the strategic Red River plains is associated with the
Phung Nguyen Culture. Excavations at Man Bac have revealed occupation
beginning in about 2000 BC and a cemetery containing typically Neolithic inter-
ments of the dead in a supine position. Mortuary offerings were sparse, often
comprising a single pottery vessel and never more than five (Huffer and Hiep 2010).
The pottery vessels recall those found at Sham Wan, Hong Kong, where they have
been dated to the late third millennium BC. The inhabitants of Man Bac cultivated
rice and raised pigs and dogs. They also hunted deer, wild cattle and rhinoceros
(Sawada et al. 2010) and fished in estuaries, brackish lagoons and along a mangrove
shore.
Prima facie a typically intrusive farmer settlement, bioarchaeological research
has portrayed a more subtle situation. The form and non-metric variables of the
crania reveal two groups of individuals, one closely akin to the Neolithic site of
Weidun in the Yangtze Valley and the other to the local hunter-gatherers (Dodo
2010; Matsumura 2010a). Metric and non-metric dental traits support the same
conclusion. The latter indicate genetic input from the indigenous hunter-gatherers,
while the former suggest immigration from the north (Matsumura 2010b).
Mitochondrial DNA has been employed to assess biological affinities (Shinoda
2010). Haplogroups D and G are highly represented in East Asian farmer popu-
lations, while F and B are more likely to be found in Southeast Asian
hunter-gatherers. Man Bac burials 1, 5, 9, 10 and 31 have Neolithic East Asian
crania, and their haplogroups are DG, DG, F, ND and ND. Burials 27, 30 and 32
have Hoabinhian/Australo-Melanesian crania and haplogroups F, F1b and F. As
Bellwood (2007) has stressed, recovering skeletal remains opens a direct window
onto the actual humans, and at this site they reveal the existence of two separate
groups.
The Dong Nai River in southern Vietnam drains extensive flood plains before
entering the South China Sea. The cultural sequence at An Son began with an
occupation phase dated between 2300-2000 BC and lacking any evidence of rice
2 A Maritime Route Brought First Farmers … 45
cultivation or the presence of domestic animals (Bellwood et al. 2013). The earliest
pottery was sand tempered while during the second cultural phase, pottery vessels
were decorated with incised patterns and tempered with rice chaff temper (Sarjeant
2012). Neolithic material culture included shouldered and rectangular stone adzes
and one-piece barbed fish hooks. The cemetery, dating to between 1500 and 1000
BC, contained the graves of adults and infants laid out in a supine position.
Mortuary offerings were more abundant than at Man Bac, and one adult was
interred with nine pottery vessels and three stone adzes. Ornaments included shell
beads.
Subsistence after the initial occupation period included rice identified through
DNA as Oryza sativa japonica, the variety domesticated in the Yangtze River
Valley (Castillo et al. 2016). These remains were associated with the raising of
domestic dogs and pigs, the former of which was more common and used for food.
Freshwater fish and aquatic turtles were also strongly represented. A superficial
glance at the material culture and subsistence at An Son during its second phase
suggests intrusive Neolithic settlement by rice farmers. Again, however, the
bioarchaeology suggests a more nuanced interpretation. Whereas cranial mor-
phology reveals an intrusive population with northern affinities, the teeth belonging
to these individuals are more akin to those of the indigenous hunter-gatherers.
Between 5000 and 4000 years ago, the period when rice farmers were expanding
southward from the Yangtze River Valley into southern China, the sea level was
higher than at present. The coast, particularly where estuaries and embayments
formed, would have provided rich and predictable marine resources. Nong Nor and
related sites were located along the shore of a sheltered marine embayment of the
Gulf of Siam (Higham and Thosarat 1998; Fig. 2.2). Dating to the 24th century BC,
the site comprises a one-phase occupation layer densely packed with marine
shellfish that indicate a low energy, sandy marine shore with access to mudflats
(Mason 1998). Crabs of the mangal and mudflats were abundant. The inhabitants
had easy access to the open sea, and brought bull and tiger sharks as well as eagle
rays back to the settlement. Mammalian bones are few, and most were worked or
modified. Porpoises were also hunted. No pig or dog bones were identified, and
flotation failed to recover a single fragment of rice.
The inhabitants of Nong Nor were proficient potters who created sand-tempered
forms decorated by smoothing, cord marking and incised patterns. Stone was a
valued import. Just four polished adzes were found, and they had been regularly
sharpened with whetstones. Good quality bone was also rare and was converted into
barbed fishhooks and awls. Pottery vessels accompanied the single burial that was
found: that of an adult female interred in the crouched, seated position typical of
hunter-gatherers. O’Reilly (1998) has concluded that Nong Nor was a seasonal base
46 C. F. W. Higham
Fig. 2.2 The location of Nong Nor and contemporary hunter-gatherer sites relative to the former
coastline, as it appeared in 2300 BC
from Nong Nor. Polished stone adzes, whetstones, awls and barbed bone fishhooks
also parallel those from the earlier site. In her analysis of the seeds and charcoal,
Thompson (1996) has reconstructed the surrounding landscape during this initial
settlement phase: a mangrove estuarine habitat backed by salt flats and punctuated
by streams flanked by fresh to brackish water swamps. The date for this phase,
about 2000 BC, is also significant. This was the precise period when we find the
first evidence for the arrival of rice farmers along the coast of Vietnam and an
admixture between the newcomers and hunter-gatherers at Man Bac and An Son.
There are no burials from this initial occupation phase of Khok Phanom Di, but rice
husks were present in a cultural context otherwise dominated by maritime and
estuarine gathering and fishing.
The changing environment over the next five centuries has been illustrated by
the recovery of minute shellfish, ostracodes and forams (Mason 1991; McKenzie
1991). Meanwhile the mortuary sequence, prolific artifactual remains and the
bioarchaeology of the inhabitants present a rare opportunity to integrate cultural and
environmental changes. Burials were superimposed, in discrete clusters, over about
17 generations. There are seven mortuary phases (MP). An estuarine mangrove
habitat dominated MP1-3A, while the sea level fell and fresh water swamps formed
during MP3B-4, then reverted to marine conditions in MP5-7.
The fishing and gathering of marine resources, particularly shellfish and crabs,
dominated except for during the brief period marked by lower sea levels. The few
mammal bones included pigs and macaques, both found wild in mangroves. While
some pigs might well have been domestic, there is no doubt about the exotic status
of the few dogs identified. There were, however, marked changes in the transition
from Ceramic Period 1 of the initial settlement to the next ceramic phase, when
grog replaced sand temper and new forms and decorative patterns were introduced
(Fig. 2.3); a shift that Vincent (2004) interprets as evidence of new human arrivals.
Indeed, this is reflected in changing mortuary traditions. The first graves were cut
down into the initial occupation layer, and the dead laid out in a supine position
save for one child who was found flexed. There were three adults and the same
number of infants, and between them mortuary offerings totalled just 12 shell beads.
Strontium isotopes in the teeth of all three adults identify them as immigrants
(Bentley et al. 2007). The ensuing phase saw the establishment of six discrete
clusters of inhumation graves that were to accumulate until the end of MP4. Each
contained the graves of men, women and infants. Mortuary offerings included
pottery vessels that shone with burnishing and came in a variety of forms decorated
with complex, incised designs typical of the Southeast Asian Neolithic. Shell beads
were now common, with one man wearing over 39,000.
During MP3B the sea level fell, and fresh water habitats formed. The isotopes in
the teeth from this period reveal the arrival of women raised in a different habitat.
New pottery forms were placed with the dead, and granite hoes were used, along
with shell harvesting knives. Domestic rice was found in the stomach contents of a
woman whose carbon isotopes showed that marine food was only a minor part of
her diet (Fig. 2.3). Domestic rice remains were also found in the faeces recovered
from a male grave (Thompson 1996), and together with the remains of a beetle and
48 C. F. W. Higham
(a) (c)
(b)
(d)
Fig. 2.3 Evidence for rice and Neolithic ceramics from Khok Phanom Di, including: A. rice
impressions on clay found on the surface of a potsherd; B. the stomach contents of a woman, burial
56, that contained rice husks and freshwater fish bones and scales; C. looking into a pottery vessel
from burial 11, mortuary phase 6, whose incised design is widely paralleled in Southeast Asian
Neolithic sites; and D. Pottery vessels with distinct forms but characteristic incised and impressed
motifs
mouse hairs these findings suggest the presence of rice stores. Khok Phanom Di
was, at about 1700 BC, a rice farming community that traded widely by sea and
river.
Then the sea levels rose and marine conditions returned. Shell knives and granite
hoes were no longer found, but during MP5 the dead were interred with remarkable
wealth. A woman potter wore over 120,000 shell beads, shell discs on her chest, ear
ornaments and a bangle, and was buried with her anvil, burnishing stones and
2 A Maritime Route Brought First Farmers … 49
Fig. 2.4 An unrooted neighbour-joining tree analysis applied to the Smith’s mean measure of
divergence matrix based on a battery of 21 non-metric dental traits. Samples in red are mentioned
in the text (reproduced with the permission of Dr. Hirofumi Matsumura and Dr. Marc Oxenham)
superb ceramic vessels. An infant in the adjacent grave wore 12,600 beads, while a
man of the same period was associated with 57,000 beads. In the subsequent phase,
two wealthy women and a child were buried within a raised, clay-walled building.
Several attempts to recover DNA from these human bones have failed, and the
biological affinities of the Khok Phanom Di population remained enigmatic until a
recent analysis of the non-metric dental and metric cranial variables (Matsumura &
Oxenham 2014; Fig. 2.4). These studies relate the site closely to Weidun and
Songze in the Yangtze Delta region. Matsumura has concluded that there was
probably a rapid coastal movement by rice farmers to the Gulf of Siam, where they
met and mixed with indigenous hunter-gatherers.
Archaeological research has identified cultural changes that took place starting
around 2,000 BC and involving people biologically akin to those of the Yangtze
River Valley. These represent the fifth of eleven demographic thrusts from the
centers of rice domestication proposed by Fuller et al. (2010). The archaeological
signature expressed in burial practices, the decorations on ceramic vessels, and the
presence of domestic rice and millet, pigs, and dogs, is so clear that it suggests the
opening of one chapter in the prehistory of Southeast Asia just as its predecessor
closed.
Still, key new evidence comes from human genetics. Hunter-gatherers have been
identified as the principal contributors to mtDNA in modern populations, with a
leavening from the intrusive Neolithic farmers. Until the appearance of recent
bioanthopological research, I was uncertain how best to interpret the overall
sequence of Khok Phanom Di. Now the study of cranial morphometrics and
non-metric dental variations of the inhabitants have shown, beyond a reasonable
doubt, that incoming rice farmers interacted with the resident hunter-gatherers; and
aside from a brief window of opportunity when the sea level fell back, this pop-
ulation of farmers largely became hunter-gatherers themselves. Meanwhile, bioar-
chaeological research has likewise identified mixed populations of indigenous and
intrusive groups at An Son and Man Bac.
Much emphasis has been placed on gaining insight into the expansion of farmers
into Taiwan and beyond. That they could cross the Taiwan Strait and proceed south
to the Philippines makes it clear that these early farmers were adept deep sea sailors.
Yet, far less attention has been given to the possibility of a parallel, coastal
movement along the shores of Southern China to Southeast Asia. New fieldwork in
the Fuzhou Basin has now revealed rice farmer settlements that link the seminal
Yangtze River sites with those in the lower reaches of the Pearl River and the
Phung Nguyen complex of Bac Bo. Further on, and again by a coastal route, we
find early rice farmers also settling the Dong Nai and Bang Pakong River Valleys.
References
Deng, Z., Qin, L., Gao, Y., Weisskopf, A. R., Zhang, C., & Fuller, D. Q. (2015). From early
domesticated rice of the Middle Yangtze Basin to millet, rice and wheat agriculture:
Archaeobotanical macro-remains from Baligang, Nanyang Basin, Central China (6700-500
BC). PLoS ONE, 10(10), e0139885. https://doi.org/10.1371/journal.pone.0139885.
Dodo, Y. (2010). Qualitative cranio-morphology at Man Bac. In M. Oxenham, H. Matsumura, &
D. K. Nguyen (Eds.), Man Bac: The excavation of a Neolithic Site in Northern Vietnam.
Canberra: Australian National University. Terra Australis 33:33–42.
Fuller, D. Q., Qin, L., Zheng, Y., Zhao, Z., Chen, X., Hosoya, L. A., et al. (2009). The
domestication process and domestication rate in rice: Spikelet bases from the Lower Yangtze.
Science, 323, 1607–1610.
Fuller, D. Q., Sato, I., Castillo, C. C., Qin, L., Weisskopf, A. R., Kingwell-Banham, E. J., et al.
(2010). Consilience of genetics and archaeobotany in the entangled history of rice.
Archaeological and Anthropological Science, 2, 115–131.
Gamble, C. (2007). No Neolithic revolution. Cambridge Archaeological Journal, 17, 91–93.
Higham, C. F. W., & Thosarat, R. (Eds.). (1998). The excavation of Nong Nor, a prehistoric site in
Central Thailand. Oxford: Oxbow Books and University of Otago Studies in Prehistoric
Anthropology No. 18.
Higham, C. F. W., & Thosarat, R. (2004). The excavation of Khok Phanom Di, a prehistoric site in
Central Thailand, volume VII: summary and conclusions, Research Report No. XLVIII.
London: The Society of Antiquaries of London.
Huffer, D. G., & Hiep, T. H. (2010). Man Bac burial descriptions. In M. Oxenham, H. Matsumura,
& D. K. Nguyen (Eds.), Man Bac: The excavation of a Neolithic Site in Northern Vietnam.
Canberra: Australian National University. Terra Australis 33:135–168.
Mason, G. M. (1991). The molluscan remains. In C. F. W. Higham & R. Bannanurag (Eds.), The
excavation of Khok Phanom Di, a prehistoric site in Central Thailand, volume II: the
biological remains (Part I), Research Report No. XLVIII (pp. 259–319). London: The Society
of Antiquaries of London.
Mason, G. M. (1998). The shellfish, crab and fish remains. In C. F. W. Higham & R. Thosarat
(Eds.), The excavation of Nong Nor, a prehistoric site in Central Thailand (pp. 173–211).
Oxford: Oxbow Books and University of Otago Studies in Prehistoric Anthropology No. 18.
Matsumura, H. (2010a). Quantitative cranio-morphology at Man Bac. In M. Oxenham, H.
Matsumura, & D. K. Nguyen (Eds.), Man Bac: The excavation of a Neolithic Site in Northern
Vietnam, Terra Australis 33 (pp. 21–32).
Ma, T., Zheng, Z., Wang, Q. C., Rolett, V. B., & Lin, G. W. (2013). Study on the phytolith and
sporopollen of Zhuangbianshan site-New evidence to the rice farming activities of Tanshishan
Culture, Fujian. Lingnan Journal of Archaeological Research, 13, 32–41. (in Chinese with
English abstract).
Matsumura, H. (2010b). Quantitative and qualitative dental morphology at Man Bac. In M.
Oxenham, H. Matsumura, & D. K. Nguyen (Eds.), Man Bac: The excavation of a Neolithic Site
in Northern Vietnam, Terra Australis 33 (pp. 43–63).
Matsumura, H., & Oxenham, M. (2014). Demographic transitions and migration in prehistoric
East/Southeast Asia through the lens of nonmetric dental traits. American Journal of Physical
Anthropology, 155, 45–65.
McKenzie, K. G. (1991). The Ostracodes and Forams. In C. F. W. Higham & R. Bannanurag
(Eds.), The excavation of Khok Phanom Di, a prehistoric site in Central Thailand, volume II:
The biological remains (Part I), Research Report No. XLVIII (pp. 139–46). London: The
Society of Antiquaries of London.
O’Reilly, D. J. W. (1998). Nong Nor phase one in a regional context. In C. F. W. Higham & R.
Thosarat (Eds.), The excavation of Nong Nor, a prehistoric site in Central Thailand (pp. 509–
522). Oxford: Oxbow Books, Oxford and University of Otago Studies in Prehistoric
Anthropology No. 18.
Oxenham, M., & Matsumura, H. (2010). Man Bac: Regional, cultural and temporal context. In M.
Oxenham, H. Matsumura, & D. K. Nguyen (Eds.), Man Bac: The excavation of a Neolithic Site
in Northern Vietnam, Terra Australis 33 (pp. 127–33).
52 C. F. W. Higham
Abstract This study aims to clarify the characteristics of early coastal peoples in
Southeast China and their relationship to the emergence of a Neolithic transition in
this region. In prior decades, the region’s pottery-bearing Neolithic sites were
thought to reflect rice farming societies (which were identified as an “Early
Neolithic Culture”) linked to ancient Austronesian language groups. However,
these beliefs may need to be revised now that archaeological findings have revealed
an ongoing reliance on coastal and maritime resources, rather than rice agriculture,
in this region since about 5000 BC, or even much earlier. More specifically, the
evidence suggests that mixed-origin complex foragers successfully occupied the
coastal zones of Fujian (福建), Guangdong (广东), Guangxi (广西), Hainan (海南),
and perhaps Taiwan (台湾) around 5000 to 3000 BC. Only later did these groups
experience variable degrees of cultural transformation and move towards rice
agriculture as a result of their contact with migrating farmers from the middle and
lower Yangtze Valley.
3.1 Introduction
Coastal Southeast China has long been regarded as one of the key homelands for
the people that first settled Southeast Asia and the Pacific Islands. For example,
archaeologists have interpreted the “Earliest Neolithic” sites (as they are called in
Chinese archaeology) dating to 5000-3000 BC and found in the coastal region of
Fujian-Guangdong (福建-广東) as the result of proto-Austronesian groups’ earliest
dispersals in this region (e.g., Chang 1994; Chang and Goodenough 1996;
Bellwood 1997). Within this broad region, the Pearl River (珠江) Delta in
H. Hung (&)
Australian National University, Canberra, Australia
e-mail: hsiao-chun.hung@anu.edu.au
C. Zhang (&)
Peking University, Beijing, China
e-mail: cscazc@pku.edu.cn
Guangdong in particular has been suggested as the ancestral homeland of the early
Neolithic groups of Taiwan (台湾), seen in sites of the Taiwanese Dabenkeng
Culture (大坌坑文化) (Tsang 2005: 69) starting about 3500 BC or earlier. The
“farming-language dispersal model” similarly suggests that ancient rice farming in
coastal Southeast China pushed early farmers to migrate into Taiwan and beyond
(Bellwood 2005: 126–127).
A recent study of coastal site chronology has revealed that the ongoing use of
marine resources can be traced back to at least 5000 BC in the ancient assemblages
of both northern and southern China (Zhang and Hung 2016). The northern sites, on
the Jiaodong (胶东) and Liaodong (辽东) Peninsulas, contain evidence of millet
farming and generally large-scale settlements in addition to large shell middens,
illustrating a long-term reliance on coastal and maritime habitats as sources of fish
and shellfish. Southern sites along the coastlines of Fujian (福建), Guangdong (广
东), Guangxi (广西), Hainan (海南) and Taiwan (台湾), contain evidence of
smaller settlements than the north in terms of the size of sand dune and shell midden
sites.
Thus far more than 200 midden sites have been identified in coastal China,
including 150 in the Guangdong (广东) and Guangxi (广西) Provinces (Yuan
1995; Jiang 1997; Huang 2008). Within this southern cluster, the shell midden sites
may have supported more stable year-round, multi-year settlement than what
occurred at the sand dune sites. The regional distribution of these sites extended
outside the modern political boundary of China to include, for example, some
similar coastal sites in northern Vietnam (Nguyen et al. 2004) (Fig. 3.1).
This chapter reconsiders the ancient lifeways of coastal groups in southern China
based on their subsistence economies, patterns of using land and resources, rela-
tions with neighboring groups, and overseas contacts. The coastal traditions
apparently changed around 3000-2500 BC, when rice farming had spread into
Liangnan (岭南) (south of the Nanling 南岭 Mountains) and Southeast China. In
this context, the coastal-maritime groups of southern China and northern Vietnam
witnessed a profound transformation in their subsistence economies and their
general manner of living.
From 5000 BC onwards, numerous coastal sites in southeast China have been
recorded and categorized as either sand dune or shell midden sites. The earlier sites
in Fujian belong to the Keqiutou (壳丘头) shell midden group. In Guangdong,
Hong Kong and Macau, the most representative archaeological assemblage is
Xiantouling (咸头岭) (also called Dawan, 大湾), which occurs mostly at sand dune
sites but also a few shell midden sites. The shell midden sites in Chao’an (潮安), of
Guangdong, and in Fangcheng (防城) and Qinzhou (钦州), of Guangxi, are
believed to date to before 3000 BC (Wei 2012). Some archaeologists have sug-
gested that these sites represent the Chenqiao Culture (陈桥文化) (Wei 2012: 153).
3 The Origins, Expansion and Decline … 55
Fig. 3.1 Representative coastal sites along the South China Coast from 5000 to 3000 BC,
including: 1. Dabenkeng (大坌坑); 2. Changguang (长光); 3. Xinyuan (新园); 4. Fengbitou (凤鼻
头); 5. Daowei I (岛尾I), Daowei II (岛尾II); 6. Chipinglong (炽坪垅); 7. Keqiutou (壳丘头); 8.
Jinguishan (金龟山) and Fuguodun (富国墩); 9. Chenqiaocun (陈桥村) and Shiweishan (石尾
山); 10. Xiantouling (咸头岭); 11. Dahuangsha (大黃沙); 12. Dameisha (大梅沙); 13. Haogang
(蚝岗); 14. Dawan (大湾) and Shenwan (深湾); 15. Houshawan (后沙湾); 16. Caotangwan (草堂
湾); 17. Xiankezhou (蚬壳洲); 18. Gaomiao (高庙); 19. Liyudun (鲤魚墩); 20. Yapushan (亚菩
山), Malanzuishan (马兰嘴山) and Beixiaoshan (杯校山); 21. Cai Beo; 22. Da But, Con Co Ngua,
Go Trung; 23. Quynh Van; 24. Bau Du; 25. Yingdun (英墩); 26. Lianziwan (莲子湾). Gaomiao
(高庙) (Site 18) is regarded as the possible origin for the Xiantouling-associated sites 10–17
(around the Pearl River Delta)
Meanwhile, several sand dune and shell midden sites have also been documented in
the islands of Hainan (海南) and Taiwan (see below) (Fig. 3.2).
Shell midden sites are abundant along the eastern coast of Fujian and on the
surrounding islands. The most productive excavations have occurred at Keqiutou
(壳丘头) on Pingtan Island (平潭岛) (Fujian Museum 1991; Jiao et al. 2009a;
2009b); at Fuguodun (富国墩) (Lin 1973; Huang 1982) and Jinguishan (金龟山)
(C. Y. Chen 1997, 1999; W. C. Chen 1999, 2004, 2007; Chen et al. 1997, 2001) on
Jinmen Island (金門岛); at Chipinglong (炽坪垅) on Dongju Island (东莒岛)
(Chen et al. 2004); and at Daowei I (岛尾I) and Daowei II (岛尾II) on Liang Island
56 H. Hung and C. Zhang
Fig. 3.2 Representative archaeological assemblages for coastal hunter-gatherers along the South
China Coast from 5000 to 3000 BC
(Liangdao) (亮岛) in Mazu (马祖) (Chen 2013; Chen and Chiu 2013; Chen et al.
2012) (Figs. 3.3, 3.4).
Some of these sites have been described as part of a single broad-based material
assemblage from the Keqiutou Culture (壳丘头文化), even though their respective
dates have been variable. The Keqiutou Site itself dates to 4000-3500 BC, while the
Fuguodun Site of Jinmen dates to approximately 5500-3000 BC and the Jinguishan
Site dates to about 5800-3600 BC. Shells found at the Chipinglong Site on Dongju
Island, Mazu, have been dated to as early as 4000 BC, although four samples have
been dated later, around 2000 BC. Recent excavations of the Daowei I and
Daowei II Sites on Liang Island (Liangdao), Mazu, have been dated to 6300-5500
BC based on 11 samples from marine shells, carbonized charcoal and human
remains (Chen 2013; Chen and Chiu 2013). Thus apart from Chipinglong, whose
dates remain uncertain, the other sites have all been confirmed as dating prior to
3000 BC.
Regarding the cultural characteristics of these sites, Keqiutou, Fuguodun,
Jinguishan, Chipinglong, Daowei I and Daowei II have all been categorized as shell
3 The Origins, Expansion and Decline … 57
Fig. 3.3 Standing at Daowei I (岛尾 I), a Shell Midden Site on Liangdao Island (亮岛) in Mazu
(马祖) (this photo and all subsequent photos in this chapter taken by Hsiao-chun Hung unless
otherwise noted)
Fig. 3.4 Daowei I (岛尾 I), a Shell Midden Site on Liangdao Island (亮岛) in Mazu (马祖)
The generally high quality of preservation at Keqiutou allowed for a fair rep-
resentation of faunal remains as well. Terrestrial animal remains from the first
excavation were mostly from Sika deer (Cervus nippon), Red deer (Cervus ela-
phus), Sambar deer (Rusa unicolor), and wild pig. Marine remains included those
of wrasses (Labridae Family), sea turtle, baleen whale, common orient clam, oyster,
cockle, and small sea snails. The second season of excavation revealed 19 types of
shellfish, most prominently the common orient clam (Meretrix lusoria), followed
by the Chinese domestic oyster (Ostrea plicatula) and the blood cockle (Arca
granosa). Substantial numbers of marine fish bones were also recovered. However,
no plant remains were discovered in the floatation samples.
Excavations at Fuguodun of Jinmen revealed more than 22 types of shellfish and a
small number of animal bones (Lin 1973; Chen 2007). The total shellfish remains
from two excavation squares (1 sq m each to depths of 80–110 cm) weighed 203.8 kg
and 154.2 kg, while the total weight of animal bones amounted to only 166.23 grams
over the course of the 2004 excavation (Chen 2007). Meanwhile, at Daowei I on
Liangdao Island in Mazu, abundant shellfish and animal bone remains were recov-
ered. At least 24 taxa of shellfish were found, including members of the Families
Mytilidae, Patellidae, Muricidae, Trochidae, Balanidae, Ostreidae, Fruticicolidae,
Vermetidae, Veneridae, Arcidae, Turbinidae, Conidae, Columbellidae, and
Chamidae. The 7.4 kilograms of bone remains found here included specimens from
the Families Sparidae, Ariidae, Trichiuridae (cutlassfish), Carcharhinidae, Dasyatidae
(whiptail sting ray), Labridae (wrasses), Sciaenidae, Suidae, Cervidae, Delphinidae,
Muridae and the Order of Testudines (turtle). Fish bones comprised the bulk of the
vertebrate remains (Chen and Chiu 2013).
Xiantouling (咸头岭) Culture sites are located at the mouth of the Pearl River, in
the modern-day Shenzhen (深圳), Zhuhai (珠海), Hong Kong and Macau areas.
This cultural group relied chiefly on marine resources for survival, and thus far
more than 20 Xiantouling sites have been surveyed or excavated in the offshore
islands and the Pearl River Delta (Tang 1999). Most of these sites have been
categorized as sand dunes, but 3 shell middens were also identified along the river
terrace.
Among the most important sites to have been excavated are Xiantouling
(Shenzhen Museum and Department of Anthropology, Sun Yat-Sen University
1990a), Dahuangsha (大黃沙) (Shenzhen Museum and Department of
Anthropology, Sun Yat-Sen University 1990b), and Dameisha (大梅沙) (Shenzhen
Museum 1993), in Shenzhen (深圳); Caotangwan (草堂湾) (Liang 1991) and
Houshawan (后沙湾) (Li 1991) in Zhuhai (珠海); Dawan (大湾) (Qu et al. 1994:
237–240) and Sham Wan (深湾) (Hong Kong Archaeological Society 1978) in
Hong Kong; and Xiankezhou (蚬壳洲) in Gaoyao (高要) (Guangdong Provincial
3 The Origins, Expansion and Decline … 59
Cytherea and Arca. The river snails belong primarily to the Genera Viviparus and
Semisulcospira, and vertebrate animal remains include specimens of fish and turtles
as well as deer, pig, bovines, elephant, rabbit and bird.
3.3.1 Hainan
The Hainan sites consist primarily of shell middens dating to between 5000 and
3000 BC, with a few sand dune sites appearing toward the end of this period.
According to the Southern China Team One, Institute of Archaeology, Chinese
Academy of Social Sciences and Hainan Provincial Museum (2016), the chrono-
logical sequence of southeast Hainan is represented by the cultural remains of three
sites: Yingdun (英墩) (ca. 4000-3500 BC?) in Sanya (三亚), and Lianziwan (莲子
湾) (ca. 3000 BC) (Fig. 3.5) and Qiaoshan (桥山) (ca. 1500-1000 BC) in Lingshui
(陵水). The Yingdun Site was a shell midden while Lianziwan and Qiaoshan were
both sand dune sites. Still, in our view, the dating may yet be refined as a
cross-regional comparison of pottery forms and styles has suggested findings at
Yingdun perhaps as early as 5500-4500 BC.
The findings at Yingdun, and in Hainan more generally, have been interpreted as
the result of early migrations of maritime foragers who were adapting to their new
island environment after leaving the continent. The pottery and other materials are
notably different from the preceding, older hunter-gatherer sites of Hainan, sug-
gesting that they likely represent an influence from the nearby continental mainland.
The association with marine foragers is further consistent with the large numbers of
marine shellfish and fish remains at these sites.
3.3.2 Taiwan
The Dabenkeng Culture (大坌坑文化) (also called the TPK Culture) is regarded as
Taiwan’s oldest “Neolithic” assemblage, dating from around 3500 BC through
2200 BC. It also represents the oldest pottery-bearing horizon. Early Dabenkeng
findings can be distinguished around 3500-2800 BC, although they may go back
even further based on a date of 4500 BC from Xinyuan (新园) (Yan 2013). After
2800 BC, a later Dabenkeng component is quite well represented, and mostly it has
been dated in the rage of 2800-2200 BC (Hung and Carson 2014).
More than 40 sites have revealed evidence of the early or late Dabenkeng
assemblages. The most informative excavations have been at Dabenkeng (大坌坑)
in Taipei; the lower layer of Fengbitou (凤鼻头) in Kaohsiung (高雄) (Chang
1969); Nanguanli (南关里), Nanguanlidong (南关里東) (Tsang et al. 2004), and
3 The Origins, Expansion and Decline … 61
Fig. 3.5 Lianziwan (莲子湾) Sand Dune Site on Hainan (海南) Island
Fig. 3.6 The 2015 season of archaeological excavation at the Changguang (长光) Sand Dune,
Taidong (台东), in eastern Taiwan
Extended burials and evidence of the practice of tooth extraction have been found.
These people raised domesticated dogs, but it is unclear whether they had
domesticated pigs at this point (Tsang et al. 2004).
Based on the excavated animal remains, the inhabitants of Nanguanli and
Nanguanlidong relied primarily on marine resources, with some contribution from
hunter-gatherer activities on land. Among the 48 kilograms of vertebrate remains
studied at Nanguanli (sample count: 123,878), the taxonomies of 44 kilograms were
identifiable. This included 25 kilograms of fish, or 56.8% of the total. Much of the
remaining dietary protein came from wild deer and wild pigs, which accounted for
21 kilograms, while notably fewer remains were attributed to birds and reptiles (Li
2013: 626–628). Overall, the abundant fish remains, along with the large amounts
of shellfish pieces in the massive midden, indicate that the site’s ancient occupants
obtained the majority of their food from coastal and marine habitats.
Previously, many of these coastal sites were accepted as “Neolithic” due to the
presence of pottery and other similar evidence in their material assemblages; yet the
indications of typically Neolithic subsistence economies, residential patterns, or
other social aspects usually associated with the period have been lacking.
3 The Origins, Expansion and Decline … 63
Moreover, studies of the ancient human remains have not been integrated with other
lines of evidence until now. In order to clarify the cultural contexts of coastal sites
in southern China from 5000 to 3000 BC, new efforts have focused on coordinating
the lines of evidence regarding subsistence, burial practices, and physical affiliation.
3.4.1 Subsistence
at Cai Beo (on Cat Ba Island, northern Hai Phong off the northeast coast of
Vietnam) and Bau Du (in Quang Nam, central Vietnam).
Overall, the non-pottery tool assemblages reflect fishing and shell fishing, and
little evidence points to plant-food resources or other aspects of the ancient diet.
Together with the various fishing net sinkers, the most distinctive oyster pick tools
illustrate a primary dietary focus on coastal and marine foods. Other tools included
a shaped chopping tool and a hand axe-like device that might possibly be related to
forest clearing but could also serve a broad range of general-utility tasks.
These clues reveal that the people living at coastal sand dune and shell midden
sites may have shared similar subsistence economies, despite their differences in
environmental setting and locally-specific forms of cultural expression. They cer-
tainly relied on coastal and marine zones for their primary sources of protein, and
these areas also probably shaped their lifestyles, more generally. They also obtained
limited amounts of animal foods from forest and land-based hunting. While the
plant foods in their diet do not appear to have been based on rice or millet farming,
there is relatively little evidence of the other kinds of plants they may have been
eating.
Human burial features offer the most direct evidence for investigating ancient
cultural and biological affiliations. Due to the limited number of burials that have
thus far been examined in the study region from 5000-3000 BC, the conclusions
remain tentative.
From the Late Pleistocene through the Middle Holocene in southern China, two
major population groups can be distinguished (Matsumura et al. 2017a, b, 2019;
Hung et al. 2017). These are consistent with the “Two Layer Model” proposed for
Southeast Asia. The earlier group has been described as hunter-gatherers who
practiced flexed-body burial traditions and are generally identified as
Australo-Papuan. The later group showed evidence of farming, extended body
position burial traditions, and East Asian (Mongoloid) affinity. Naturally, the dating
and the specific details varied somewhat from place to place, but the overall,
large-scale pattern depicts a two-part relative chronology.
Some of the known coastal sites from 5000-3000 BC in southern China, have
been difficult to interpret as belonging to either the older Australo-Papuan
hunter-gatherer context or the younger Mongoloid farming context. It is possible
that elements from both associations may have blended at these sites over a short
period of time.
In Guangdong and Guangxi, ancient burials have been reported with extended,
flexed and crouched body positions at different sites. In Guangdong, the
Xiankezhou (蚬壳洲) Shell Midden Site (5130 ± 100 years uncal BP) contains the
burials of 28 individuals in flexed positions (Guangdong Provincial Museum and
Gaoyao Cultural Affairs Bureau 1990; Guangdong Provincial Museum et al. 1991)
3 The Origins, Expansion and Decline … 65
(Fig. 3.7). In Guangxi, 8 human burials with flexed or crouched body positions
were unearthed from the Liyudun (鲤鱼墩) Shell Midden Site (5050 ± 100 years
uncal BP) (Li et al. 2013; Guangdong Provincial Institute of Cultural Relics and
Archaeology et al. 2015). Flexed-position burials also appear to have been the
general trend in Hainan and coastal northern Vietnam before 3000 BC. In Hainan, a
new discovery at the Yingdun (英墩) Shell Midden features a human skeleton in
flexed position, apparently older than 4000 BC. In Vietnam, several coastal sites
dating to 5000-3000 BC have revealed human skeletons in flexed positions,
including Cai Beo, Da But, Quynh Van and Bau Du (e.g. Nguyen et al. 2004).
The flexed or crouched position burial traditions represent a continuation of the
older burial customs practiced by the indigenous hunter-gatherers in earlier cave
sites or inland shell midden sites (Hung et al. 2017). Many of these burials have
also shown evidence of spreading ochre or hematite over the deceased individuals,
as seen among remains found at Chenqiaocun (陈桥村) in Guangdong (Cultural
Relic Management Committee of Guangdong Province 1961). This custom
has been observed at older indigenous hunter-gatherer sites including Zengpiyan
66 H. Hung and C. Zhang
(甑皮岩), Changtan (长塘) and Xijin (西津) (Yang and He 2011); in the inland
areas of Guangxi; at the Hoabinhian cave sites of Vietnam (Nguyen et al. 2004:
179); and at several coastal sites such as Bau Du, in central Vietnam.
At least two sites with unusual burial formats along the South China Coast also
deserve consideration. First, a jar burial was documented at Luoshanzui (罗山嘴)
(Mo 2003); and second, two extended-position burials were reported at Haogang
(蚝岗), in Guangdong (Guandgong Provincial Institute of Cultural Relics and
Archaeology et al. 2008). The associated artifacts have been generally interpreted as
coming from 5000-3000 BC, but have not yet been exposed to radiocarbon dating.
On the island of Liangdao (亮島) in the Taiwan Strait, near Fujian, two human
skeletons have been discovered from two different layers in the shell midden of
Daowei I (岛尾I). In the lower layer, Liangdao Man No. 1 was in a flexed position
and dated to about 6300 BC. In the upper layer, Liangdao Man No. 2 was in an
extended position and dated to about 5500 BC. The cranial and dental metrics
conducted on the Liangdao burials by a recent study concluded that the older,
flexed-position skeleton was Australo-Papuan while the younger, extended-position
skeleton was East Asian or Mongoloid (Matsumura et al. 2019). Curiously, both of
those burial features appear to pre-date the earliest recognized pottery-bearing
Neolithic horizon of the region, including the Keqiutou Culture of Fujian (dating to
4500-3500 BC), the Xiantouling Culture of Guangdong (dating to 5000-3500 or
3000 BC), and the early Dabenkeng Culture of Taiwan (dating to 3500-2800 BC).
As previously noted, the ancient people who inhabited the inland caves and
riverside shell middens of Guangxi, as well as the inland Hoabinhian sites of
Vietnam, have been confirmed as belonging to Australo-Papuan groups and were
buried in the typical flexed position of ancient hunter-gatherers. Their later
descendants at the coastal sites of southern China so far have not been identified
through morphometric measurements in physical features of 5000–3000 BC, but
they had been interred in the older flexed-position tradition.
The emergence and expansion of coastal foragers in Southeast China and northern
Vietnam relates to two key factors. First, the more stable and reliable Holocene
weather conditions, firmly established by 7000-5000 BC, provided overall warmth
and a predictable seasonal rainfall that proved propitious for hunting and gathering.
Second, the rapid development of rice farming in the middle-lower Yangtze River
region must have affected the coastal groups, albeit indirectly, with an awareness of
the spreading population of rice farmers. Perhaps this led to the relocation of some
inland hunter-gatherers to the southeastern coastal areas where they could continue
a lifestyle that did not include rice farming.
From 5000 to 3000 BC, the ancient coastal groups in southern China may have
shared some aspects of their subsistence practices and lifestyle, but they appear to
have developed from several different origins simultaneously. These groups most
3 The Origins, Expansion and Decline … 67
likely descended from the various hunter-gatherer groups that inhabited this large
and diverse region for many thousands of years previously, and as they grew they
combined various cultural practices that had become popularity by 5000 BC.
Meanwhile, rice and millet farming was spreading rapidly, transforming many
social groups in the middle-lower Yangtze River region; yet this influence mani-
fested differently in the coastal zones where people instead sustained their
non-farming traditions through 3000 BC. Thus, based on the current evidence, the
early maritime hunter-gatherers in coastal southern China around 5000 BC can be
traced to at least three major sources prior to 5000 BC: (1) native hunter-gatherers
who lived near the ancient coast prior to 5000 BC (but so far their sites have not yet
been found), (2) the inland areas of Fujian-Guangdong-Guangxi; and (3) the middle
and lower Yangtze River region (Fig. 3.8).
Most of the inhabitants of early shell midden sites in coastal southern China were
probably the descendants of Late Pleistocene or Early Holocene hunter-gatherers
who had lived in caves or riverside shell middens further inland. This interpretation
is based primarily on the observed continuity of stone tool technologies (Yang et al.
2015), similar burial practices (e.g., the flexed position), and non-rice-farming
subsistence patterns. In Guangxi, the chronology of shell middens has revealed an
overall age gradient ranging from the oldest sites inland to the youngest sites dating
to 5000-3000 BC on the southeastern coast. Finding themselves in need of food,
perhaps some of the descendants of the originally inland-focused foragers gradually
expanded toward the southeastern coastal areas of Guangxi and western
Guangdong, eventually reaching Hainan and the other coastal islands.
Similar inland-to-coastal movements appeared in northern Vietnam after 5000
BC. Here the older pre-5000 BC assemblages have been well defined at inland sites
inhabited by hunter-gatherers who used stone tools of the Hoabinhian and
Bacsonian traditions. After 5000 BC those stone tool traditions continued, with
some modifications, and coincided with the emergence of new pottery-bearing
coastal settlements.
The middle Yangtze River was one of the major sources for the early coastal
foragers who came to inhabit Guangdong. Most Chinese archaeologists agree that
the white and painted pottery styles of the Xiantouling (咸头岭) Culture, in the
68 H. Hung and C. Zhang
Fig. 3.8 In addition to the indigenous hunter-gathers along the ancient coast prior to 5000 BC, a
few possible external origins can be considered for the early maritime hunter-gatherers in coastal
southern China
Pearl River Estuary, derived from Gaomiao (高庙), Tangjiagang (汤家岗) and Daxi
(大溪) sources in Hunan (湖南) (e.g., He 1994: 71–78; Pei 1999: 117–131; Bu
2007) (Figs. 3.9, 3.10, 3.11, 3.12). Among these sources, the ancient people buried
at Gaomiao (高庙), in Hunan (Fig. 3.13), around 5000-4000 BC have been con-
firmed as having an Australo-Papuan physical affiliation (Matsumura et al. 2017b).
Comparative studies of pottery from ancient hunter-gatherer sites in southwest
Hunan, northern Guangxi, and southwest Hubei illustrate the importance of these
areas in understanding the early links between Hunan and Guangdong. In view of
this context, Xiantouling (咸头岭) may represent a mixed-origin assemblage from
5000 BC or earlier.
As mentioned above, one of the Liangdao (亮島) burials (Liangdao Man No. 2)
suggests an unexpectedly early East Asian (or Mongoloid) presence in 5500 BC,
older than any other such finding in this region. Meanwhile, evidence of typical rice
and millet farming by East Asian (or Mongoloid) groups does not appear in Fujian
3 The Origins, Expansion and Decline … 69
Fig. 3.9 Examples of white pottery from the Xiantouling (咸头岭) Site in Guandgong (广东)
(Courtesy: Shenzhen Museum)
Fig. 3.10 Examples of white pottery from the Xiantouling (咸头岭) Site (Shenzhen Municipal
Institute of Archaeology 2013)
until 3000 BC. As a result, this ancient East Asian buried at Liangdao most likely
came from the Lower Yangtze. This time range also suggests relations with the
Kuahuqiao- Hemudu (跨湖桥-河姆渡) cultural assemblages in the lower Yangtze
70 H. Hung and C. Zhang
Fig. 3.11 Example of painted pottery from the Xiantouling (咸头岭) Site (Shenzhen Municipal
Institute of Archaeology 2013)
Fig. 3.12 Example of white pottery from the Gaomiao (高庙) Site in Hunan (湖南) (Courtesy: He
Hang 贺刚)
River, or perhaps with another unknown group in southern Zhejiang. Either way,
these populations evidently ventured to explore their overseas neighbors to the
south, including Liangdao (亮島).
From 5000 to 3000 BC, the cultural groups in coastal southeastern China
engaged in frequent interaction with each other. For example, the early Dabenkeng
(大坌坑) Culture in Taiwan may have developed from the Keqiutou (壳丘头)
3 The Origins, Expansion and Decline … 71
Culture of the Fujian coast, and also may have been related to the Xiantouling (咸
头岭) Culture in the Pearl River Delta. These cultural exchanges must have
involved overseas travel. Likewise, the later Xiantouling (咸头岭) forager popu-
lations of coastal Guangdong maintained long-term cultural interactions with the
Gaomiao (高庙) and Daxi (大溪) Cultures, in the middle Yangtze region; with the
Dingsishan (顶蛳山) Culture in southern Guangxi; and perhaps even with the
inhabitants of Zengpiyan (甑皮岩) during Zengpiyan phase 5 (甑皮岩第5 期), in
northern Guangxi.
These complex coastal societies began to decline between 3000 and 2500 BC.
The numbers of Xiantouling (咸头岭) sand dune sites and Keqiutou-Fuguodun (壳
丘头-富国墩) shell middens fell steeply. The climate at that time had become
cooler and perhaps more arid than it was previously, and the average annual
temperature in southern China was about 2 degrees lower than at present. These
changes may have reduced the reliability of natural food resources. At the same
time, farming groups from the middle and lower Yangtze River Valley began to
migrate directly into Lingnan (岭南) and the neighboring areas, and the first
agricultural settlements appeared at Shixia (石峡), in northern Guangdong;
Tanshishan (昙石山), in Fujian; and Nanguanli (南关里) (Late Dabenkeng phase),
in Taiwan. The later Dingsishan phase 4 (顶蛳山第4期) cultural assemblage in
Guangxi also appears to have been associated with rice agriculture.
72 H. Hung and C. Zhang
3.6 Conclusion
At present, no trace of rice or millet farming has been found prior to 3000 BC along
the south coast of China, although the groups inhabiting this area since at least 5000
BC left behind diverse pottery and artifact assemblages, different scales of settle-
ments, stone tool workshops, and cemeteries. Varied forms and decorations of
pottery included cord marking, paddle impression, incision, punctate, red slip,
paint, and specially-made white ware. The pebble tools found here are apparently
derived from older Hoabinhian traditions. Other stone tools included partly polished
or finely polished shouldered or stepped axes and adzes. The world’s earliest bark
cloth beaters have also been found in this region. With the exception of a few
extended-position burials like that of Liangdao Man No. 2, most of the human
burials from these maritime forager sites illustrate the same flexed body position as
those of the earlier indigenous foragers that lived further inland.
According to the “Two Layer Model,” the transition from the older to the
younger “layer” probably occurred after 7000 BC in northern and eastern China,
but this shift was delayed until about 3000-2500 BC in the coastal areas of southern
China, and occurred still later in northern Vietnam. The older layer refers to
hunter-gatherer groups that practiced flexed-position burial and have been identified
as Australo-Papuan, while the younger layer refers to farming groups who practiced
Fig. 3.14 The formation of two main populations in Southern China from the Late Paleolithic
through Early Neolithic contexts
3 The Origins, Expansion and Decline … 73
extended-position burial and have generally been identified as East Asians (or
Mongoloids). The majority of early coastal inhabitants between 5000 and 3000 BC
were therefore probably Australo-Papuan, rather than East Asian (Mongoloid); but
some interaction and movement likely occurred between these groups.
Subsequently, we see strong East Asian affinities in the human remains found at the
Tanshishan (昙石山) cultural sites, which are associated with the earliest farming
groups of the Fujian coastal region (Fig. 3.14).
The Neolithic Transition in coastal southeast China was complicated. At this
stage of research we can confirm that a critical transition occurred around 3000 BC,
but this transition followed the sustained appearance of coherent cultural groups
and deposits from 5000 to 3000 BC. These assemblages illustrate the prominence of
coastal and marine foods and lifestyles independent of rice or millet farming, yet
exactly which non-agricultural plant foods were being consumed here during this
period has yet to be determined. With the rising spread of rice farming by 3000 BC,
some of the coastal groups likely travelled overseas to Taiwan, south to Vietnam, or
elsewhere, even as some others remained in their homelands and underwent pro-
found societal change.
References
Chu, C. Y., Yang, F.P., Chong, Y. X., Chang, Y. S., & Song, Y. J. (朱正宜、杨凤屏、钟亦兴、
张益生、宋昱洁). (2015). The salvage excavation of the Dachang Bridge Site lying on the
Yanshui stream bed. Tainan: Suguo Company (Yanshuixi Dachangqiao Yizhi Qiangjiu Fajue
《盐水溪大昌桥遗址抢救发掘》庶古文创事业股份有限公司, 台南, 2015年).
Cultural Relic Management Committee of Guangdong Province (广东省文物管理委员会).
(1961). Shell midden in Chaoan, Guangdong. Archaeology, 11, 577–584 (Guangdong Chaoan
De Beiqiu Yizhi, Kaogu < 广东潮安的贝丘遺址 > 《考古》11:577-584, 1961年).
Deng, Z. H., Hung, H. C., Fan, X. C., Huang, Y. M., & Lu, H. Y. (2018). The ancient dispersal of
millets in southern China: new archaeological evidence. The Holocene, 28(1), 34–43. (2018).
https://doi.org/10.1177/0959683617714603.
Fujian Museum (福建省博物館). (1991). A brief report on the excavation at Keqiutou in Fujian.
Archaeology, 7, 581–600 (Fujian Pingtan Keqiutou Yizhi Fajue Jianbao, Kaogu < 福建平潭
壳丘头遗址发掘简报 > 《考古》7:581–600, 1991年).
Guangdong Provincial Institute of Cultural Relics and Archaeology, Dongguang Municipal
Museum and Hoagang Site Museum of Dongguan (广东省文物考古研究所、东莞市博物
館、东莞蚝岗遗址博物馆). (2008). A brief report on the preliminary excavation of the
Haogang Site in the south city district of Dongguan City. Huanan Archaeology, 2, 151–179
(Dongwanshi Nanchengqu Haogang Yizhi Chubu Fajue Jianbao, Huanan Kaogu. < 东莞市南
城区蚝岗遗址初步发掘简报 > 《华南考古》2:151–179, 2008年).
Guangdong Provincial Institute of Cultural Relics and Archaeology, Zhanjiang City Museum and
Suixi County Museum (广东省文物考古研究所、湛江市博物館、遂溪县博物館). (2015).
Brief report on the excavation at the Liyudun Neolithic Shell Midden Site in Suixi,
Guangdong. Cultural Relics, 7, 4–18 (Guangdong Suixi Liyudun Xinshiqi Shidai Beiqiu Yizhi
Fajue Jianbao. Wenwu. < 广东遂溪鲤鱼墩新石器时代贝丘遗址发掘简报 > ,《文物》
Relics 7:4–18, 2015年).
Guangdong Provincial Museum (广东省博物館). (1961). The Neolithic shell midden sites in
Dongxing, Guangdong. Archaeology, 12, 644–649 (Guangdong Dongxing Xinshiqi Shidai
Beiqiu Yizhi, Kaogu. < 广东东兴新石器时代贝丘遗址 > 《考古》12:644–649, 1961年).
Guangdong Provincial Museum and Gyaoyao Cultural Affairs Bureau (广东省博物館、高要县
文化局). (1990). Neolithic shell midden site discovered at Xiankezhou, Gaoyao County,
Guangdong. Archaeology (Kaogu), 6, 565–568 (Guangdong Gaoyaoxian Xiankezhou Faxian
Xinshiqi Shidai Beiqiu Yizhi, Kaogu. < 广东高要县蚬壳洲发现新石器时代贝丘遗
址 > 《考古》6:565–568, 1990 年).
Guangdong Provincial Museum, Zhaoqing Cultural Affairs Bureau and Gyaoyao County Museum
(广东省博物館、肇庆地区文化局、高要县博物館). (1991). Xiankezhou shell midden site
in Longyi Town, Gaoyao County. Cultural Relics, 11, 8–13 (Gaoyaoxian Longyixiang
Xiankezhou Beiqiu Yizhi, Wenwu. < 高要县龙一乡蚬壳洲贝丘遗址 > 《文物》11:8–13,
1991年).
He, J. J. (何介鈞). (1994). Prehistoric painted pottery and the Daxi Culture of the Zhujiang River
mouth. In C. Tang (Ed.), Ancient cultures of South China and neighbouring regions (pp. 71–
8). Hong Kong: Chinese University of Hong Kong (< 环珠江口的史前彩陶与大溪文
化 > 《南中国及邻近地区古文化研究》,香港中文大学, 香港,1994年).
Higham, C. (1996). The Bronze Age of Southeast Asia. Cambridge: Cambridge University Press.
Hong Kong Archaeological Society (香港考古学会). (1978). Report on investigations of the
Sham Wan Site on Lamma Island. Bulletin of the Hong Kong Archaeological Society, 3, 1–293
(Nanyadao Shenwan Kaogu Yizhi Diaocha Baogao, Xianggang Kaogu Xuehui
Zhuankan. < 南丫岛深湾考古遗址调查报告 > 《香港考古学会专刊》3:1–293, 1978年).
Hu, Y. W., Li, F. J., Wang, C. S., & Richards, M. P. (胡耀武、李法軍、王昌燧). (2010). Analysis
of the C- and N- stable isotopes in human remains found at the Liyudun Site in Zhanjiang,
Guangdong: a preliminary exploration of prehistoric life among Neolithic peoples in Southern
China. Acta Anthropologica Sinica, 29(3), 264–269 (Guangdong Zhanjiang Liyudun Yizhi
Rengu De C N Wending Tongweisu Fenxi:Huanan Xinshiqi Shidai Xianmin Shenghuo Fangshi
Chutan, Renleixue Xuebao < 广东湛江鲤鱼墩遗址人骨的C、N稳定同位素分析:华南新石
器时代先民生活方式初探 > 《人类学学报》29 (3): 264–269, 2010年).
76 H. Hung and C. Zhang
Huang, Q. S. (黃启善). (2008). Study of the prehistoric shell midden sites in Guangxi. Lingnan
Archaeology Research, 7, 34–36 (Guangxi Shiqian Beiqiu Wenhua Yizhi De Yanjiu, Lingnan
Kaogu Yanjiu. < 广西史前贝丘文化遗址的研究 > 《岭南考古研究》7: 34–36. 2008年).
Huang, S. C. (黃士強). (1974). Report on archaeological investigations at Bajiacun, Guirenxiang,
Tainan, Taiwan. Bulletin of the Department of Archaeology and Anthropology, 35/36, 62–68
(Tainanxian Guirenxiang Bajiacun Yizhi Diaocha, Guoli Taiwan Daxue Kaogu Renlei
Xuekan< 台南县归仁乡八甲村遗址调查 > 《国立台湾大学考古人类学刊》35/36:62-68.
1974年).
Huang, S. C. (黃士強). (1982). Jinmen archaeological survey. In J. Z. Yin. (Ed.), Chinese
traditional folk arts field survey report (pp. 22–27).Taipei: Department of Anthropology,
National Taiwan University (Jinmen Kaogu Diaocha, Zhongguo Minjian Chuantong Jiyi
Fangcha Baogao. < 金门考古調查 > 《中国民间传统技艺访查报告》,教育部社会教育司
委託国立台湾大学人类学系之研究计划报告, 台北》, 1982年).
Hung, H. C., & Carson, M. T. (2014). Foragers, fishers and farmers: origins of the Taiwanese
Neolithic. Antiquity, 88, 1115–1131.
Hung, H. C., Zhang, C., Matsumuraand, H., & Li, Z. (2017). Neolithic transition in Guangxi: the
long development of hunter-gatherer society in southern China. In H. Matsumura, H. C. Hung,
Z. Li, & K. Shinoda (Eds.), Bio-anthropological studies of Early Holocene hunter-gatherer
sites at Huiyaotain and Liyupo in Guangxi, China (pp. 205–228). Tokyo: National Museum of
Nature and Science No. 47.
Institute of Cultural Relics and Archaeology of Hunan Province (湖南省文物考古研究所).
(2000). Preliminary report on the excavations at Gaomiao in Qianyang, Hunan. Cultural Relics,
4, 4–23 (Hunan Qianyang Gaomiao Yizhi Fajue Jianbao. Wenwu< 湖南黔阳高庙遗址发掘
简报 > 《文物》4:4–23, 2000年).
Jiang, T. Y. (蒋廷瑜). (1997). Study on shell midden sites in Guangxi. Guangxi Ethnic Studies
(Guangxi Minzu Yanjiu), 4, 130–144 (Guangxi Beiqiu Yizhi De Kaocha Yanjiu, Guangxi
Minzu Yanjiu < 广西贝丘遗址的考察研究 > 《广西民族研究》4: 130–144, 1997年).
Jiao, T. L., Fan, X. C., & Rolett, B. (焦天龙、范雪春、罗莱). (2009b). The Keqiutou Site and
Early Neolithic Culture in the Taiwan Straits. Fujian culture heritage and museums (Fujian
Wenbo), 2, 8–12 (Keqiutou Yizhi Yu Taiwan Haixia Xinshiqi Shidai Zaoqi Wenhua, Fujian
Wenbo < 壳丘头遗址与台湾海峡新石器时代早期文化 > 《福建文博》2:8–12, 2009年).
Jiao, T. L., Fan, X. C., Rolett, B., & Lin, G. W. (焦天龙、范雪春、罗莱、林公務). (2009a).
Excavation report for the Keqiutou Site in Pingtan in 2004. Fujian culture heritage and
museums (Fujian Wenbo), 1, 1–15 (2004 Nian Pingtan Keqiutou Yizhi Fajue Baogao, Fujian
Wenbo < 2004年平潭壳丘头遗址发掘报告 > 《福建文博》1:1–15, 2009年).
Li, K. T. (2013). First Farmers and their Coastal Adaptation in Prehistoric Taiwan. In A.
P. Underhill (Ed.), A companion to chinese archaeology (pp. 612–633). Hoboken, NJ: John
Wiley & Sons Inc.
Li, F. J., Wang, M. H., Feng, M. Q., Chen, C., & Zhu, H. (李法軍、王明辉、冯孟钦、陈诚、朱
泓). (2013). Osteometric analysis of human skulls from the Liyudun Neolithic Site. Acta
Anthropologica Sinica, 32(3), 302–318 (Liyudun Xinshiqi Shidai Jumin Tougu De Xingtaixue
Fenxi, Renleixue Xuebao < 鲤鱼墩新石器时代居民头骨的形态学分析 > 《人类学学报》
32(3): 302–318, 2013年).
Li, Z. W. (李子文). (1991). The excavation of the Houshawan Site on Qiao Island. In Zhuhai
Museum et al. (Eds.), Archaeological studies and discoveries in Zhuhai (pp. 3–21).
Guangzhou: Guangzhou People’s Press (Qi’aodao Houshawan Yizhi Fajue, Zhuhai Kaogu
Faxian Yu Yanjiu < 淇澳岛后沙湾遗址发掘 > 《珠海考古发现与研究》,广东人民出版
社, 广州, 1991年).
Liang, Z. X. (梁振兴). (1991). The excavation of the Caotongwan Site on Sanzao Island. In
Zhuhai Museum et al. (Eds.), Archaeological studies and discoveries in Zhuhai (pp. 22–33).
Guangzhou: Guangzhou People’s Press (Sanzaodao Caotangwan Yizhi Fajue, Zhuhai Kaogu
3 The Origins, Expansion and Decline … 77
Yuan, J. (袁靖). (1995). Some questions relating to coastal shell middens in mainland China.
Archaeology, 12, 1100–109 (Guanyu Zhongguo Dalu Yanhai Diqu Beiqiu Yizhi Yanjiu De Jige
Wenti, Kaoguo 〈关于中国大陆沿海地区贝丘遗址研究的几个问题〉《考古》12:1100–
1109, 1995年).
Zhang, C., & Hung, H. C. (張弛、洪晓纯). (2010). The emergence of agriculture in southern
China. Antiquity, 84(323), 11–25.
Zhang, C. & Hung, H. C. (張弛、洪晓纯). (2016). Early maritime adaptation in coastal China.
Relics from the South, 3, 1–13 (Zhongguo Yanhai De Zaoqi Haiyang Shiyingxing Wenhua,
Nanfang Wenwu 〈中国沿海的早期海洋适应性文化〉《南方文物》3:1–13, 2016年).
Chapter 4
Pollen Evidence for Human-Induced
Landscape Change Reveals the History
of Agricultural Development
in Southeastern China
T. Ma (&)
School of Geography and Planning, Sun Yat-Sen University, Guangzhou, China
e-mail: 229372051@qq.com
Z. Zheng (&)
School of Earth Science and Geological Engineering, Sun Yat-Sen University,
Guangzhou, China
e-mail: eeszzhuo@mail.sysu.edu.cn
4.1 Introduction
Rice (Oryza sativa) was first domesticated around 8,000 years ago in the coastal
wetland environments of China’s Lower Yangtze region (including the Hangzhou
Bay area and the Yangtze Delta), and this area quickly became a center for early
rice agriculture (e.g., Fuller 2011; Fuller et al. 2009; Zong et al. 2007). The sub-
sequent expansion of freshwater wetlands in this region during the Middle
Holocene encouraged the development and expansion of rice-based agriculture over
thousands of years (Zong et al. 2011). Agriculture was highly developed in the
Liangzhu (良渚) Culture by about 6000 BP, and the cultivation of domesticated rice
became a major subsistence activity (Atahan et al. 2008; Cao et al. 2006; Qin et al.
2011).
By contrast, Neolithic rice agriculture in southeastern China lagged significantly
behind that of the Yangtze region. At present, it appears that rice cultivation
emerged in southeastern China around 5000–4000 years ago (Chi and Hung 2010;
Ma et al. 2016a); however, much of the evidence in this region suggests that rice
cultivation initially formed only one part of a broad-spectrum Neolithic subsistence
economy centered around fishing, hunting and gathering (Ma et al. 2016a, b; Yue
et al. 2015; Zong et al. 2013). In fact, when compared to the Yangtze area the
history and spread of rice agriculture in southeastern China is far from systemati-
cally understood.
The development of agriculture is always associated with anthropogenic changes
to the landscape due to activities such as intensified land reclamation and
slash-and-burn agriculture (Boyle et al. 2011; Kaplan et al. 2011). Therefore,
palynological records may contribute important information about the timing and
extent of anthropogenic modifications to the landscape that might further elucidate
the history of agricultural practices in Southeast China during the Holocene.
Although Neolithic rice cultivation emerged about 5000 years ago, pollen data from
the Pearl River Delta and the Fuzhou Basin show that the shift to economic
dependency on rice agriculture occurred after 2500 BP (Yue et al. 2015; Zong et al.
2013). In order to better understand the history of agricultural practices in the study
region, this chapter examines anthropogenic landscape changes in southeastern
China based on a series of important palynological records.
Fig. 4.1 Physiographic maps showing the pollen records mentioned in this article
84 T. Ma and Z. Zheng
Fig. 4.2 Variation of pollen percentages during the Late Holocene in delta areas of southeastern
China, including: a Core GZ-2 (Wang et al. 2009); b Core ZK-2 (Zheng et al. 2004); c Core HP-1
(Zheng et al. 2004); and d Core FZ4 (Yue et al. 2015)
4 Pollen Evidence for Human-Induced Landscape Change Reveals … 85
anthropogenic impact on the regional landscape at this time was rather limited. This
interpretation also correlates with evidence that the regional Neolithic subsistence
economy centered on fishing and hunting (Yang et al. 2013; Ma et al. 2016a),
indicating rice cultivation was limited in South China at that time.
In contrast to earlier periods, the pollen record from the last three thousand years
depicts a solidly anthropogenic environment. Landscape changes are evidenced by
abrupt declines in Castanopsis and Quercus arboreal pollen, representing a decrease
in natural forest lands. Meanwhile, the relative percentages of Dicranopteris spores
show a sharp increase in addition to an expansion of other pioneer taxa such as
Pinus and Artemisia. Most notably, this period is characterized by an overwhelming
rise in the frequency of Poaceae caused by the spread of cultivated rice. Together,
these shifts in pollen assemblages suggest intense human disturbances and reveal
the onset of large-scale agricultural development in southeastern China (Fig. 4.2).
The pollen records from hinterland and mountainous areas (Fig. 4.3) are rep-
resented by the brief pollen assemblages from cores SZY (1007 m a.s.l.) and GT-2
(1667 m a.s.l.). Both of these cores show obvious declines in evergreen broad-
leaved trees after 1000 BP, coupled with increases in Pinus and Poaceae pollen as
Fig. 4.3 Brief pollen assemblages from cores SZY and GT-2, representing hinterland and
mountainous areas in southeastern China (adapted from Ma et al. 2017)
86 T. Ma and Z. Zheng
Acknowledgements This work was supported by the National Natural Science Foundation of
China (Grants 41472143 and 41701222).
References
Atahan, P., Itzstein-Davey, F., Taylor, D., Dodson, J., Qin, J., Zheng, H., et al. (2008).
Holocene-aged sedimentary records of environmental changes and early agriculture in the
lower Yangtze, China. Quaternary Science Reviews, 27, 556–570.
Boyle, J. F., Gaillard, M. J., Kaplan, J. O., & Dearing, J. A. (2011). Modelling prehistoric land use
and carbon budgets: A critical review. Holocene, 21, 715–722.
Cao, Z. H., Ding, J. L., Hu, Z. Y., Knicker, H., Kögel-Knabner, I., & Yang, L. Z. (2006). Ancient
paddy soils from the Neolithic age in China’s Yangtze River Delta. Die Naturwissenschaften,
93, 232–236.
Chen, W. M. (1989). A review of agricultural production in the Nanyue Kingdom during the early
Han dynasty. Study of Ethnics in Guangxi 3, 76–79 (汉初南越国农业生产述评. 广西民族研
究, 1989).
Chi, Z., & Hung, H. (2010). The emergence of agriculture in southern China. Antiquity, 84, 11–25.
Fuller, D. Q. (2011). Pathways to Asian Civilizations: tracing the origins and spread of rice and
rice cultures. Rice, 4, 78–92.
Fuller, D. Q., Qin, L., Zheng, Y. F., Zhao, Z. J., Chen, X. G., Hosoya, L. A., et al. (2009). The
domestication process and domestication rate in rice: Spikelet bases from the Lower Yangtze.
Science, 323, 1607–1610.
Kaplan, J. O., Krumhardt, K. M., Ellis, E. C., Ruddiman, W. F., Lemmen, C., & Goldewijk, K. K.
(2011). Holocene carbon emissions as a result of anthropogenic land cover change. Holocene,
21, 775–791.
Ma, T., Zheng, Z., Rolett, B. V., Lin, G., Zhang, G., & Yue, Y. (2016a). New evidence for
Neolithic rice cultivation and Holocene environmental change in the Fuzhou Basin, Southeast
China. Vegetation History & Archaeobotany, 25, 375–386.
Ma, T., Tarasov, P. E., Zheng, Z., Han, A., & Huang, K. (2016b). Pollen- and charcoal-based
evidence for climatic and human impact on vegetation in the northern edge of Wuyi mountains,
China, during the last 8200 years. Holocene, 26, 1616–1626.
Ma, T., Zheng, Z., Man, M., Dong, Y., Li, J., & Huang, K. (2017). Holocene fire and forest
histories in relation to climate change and agriculture development in southeastern China.
Quaternary International, 488, 30–40. https://doi.org/10.1016/j.quaint.2017.07.035.
Qin, J., Taylor, D., Atahan, P., Zhang, X., Wu, G., & Dodson, J. (2011). Neolithic agriculture,
freshwater resources and rapid environmental changes on the lower Yangtze, China.
Quaternary Research, 75, 55–65.
88 T. Ma and Z. Zheng
Rolett, B. V. (2012). Late Holocene evolution of the Fuzhou Basin (Fujian, China) and the spread
of rice farming. In L. Giusan, D. Q. Fuller, K. Nichol, R. K. Flad, & P. D. Clift (Eds.),
Climates, landscapes and civilizations, geophysical monograph series 198 (pp. 137–143).
Washington, DC: American Geophysical Union.
Rolett, B. V., Zheng, Z., & Yue, Y. F. (2011). Holocene sea-level change and the emergence of
Neolithic seafaring in the Fuzhou Basin (Fujian, China). Quaternary Science Reviews, 30,
788–797.
Wang, J., Wang, X., Cao, L., Zheng, Z., Yang, X., & Yang, J. (2009). The Holocene pollen
characteristics and their paleoenvironmental significance from Core GZ-02 in the Pearl River
Delta. Journal of Palaeogeography, 11, 661–669 (珠江三角洲GZ-2孔全新统孢粉特征及古
环境意义. 古地理学报, 2009).
Xie, C. (2004). A new pattern of ethnic relations in South China during the Tang and Song
Dynasties. Zhejiang Academic Journal, 5, 87–94 (唐宋时期南方民族关系的新格局. 浙江学
刊, 2004).
Yang, S. X., Zheng, Z., Huang, K. Y., Zong, Y., Wang, J., Xu, Q., et al. (2012). Modern pollen
assemblages from cultivated rice fields and rice pollen morphology: application to a study of
ancient land use and agriculture in the Pearl River Delta, China. Holocene, 22, 1393–1404.
Yang, X., Barton, H. J., Wan, Z., Li, Q., Ma, Z., Li, M., et al. (2013). Sago-type palms were an
important plant food prior to rice in southern subtropical China. PLoS ONE, 8, e63148.
Yue, Y., Zheng, Z., Rolett, B. V., Ma, T., Chen, C., Huang, K., et al. (2015). Holocene vegetation,
environment and anthropogenic in- fluence in the Fuzhou basin, southeast China. Journal of
Asian Earth Sciences, 99, 85–94.
Zheng, Z. (1998). Study of human disturbance to the vegetation in the most recent few thousand
years in coastal areas of Southeast China. Ecological Science, 17, 29–36 (近几千年华南沿海
地区植被的人为干扰. 生态科学, 1998).
Zheng, Z., Deng, Y., Zhang, H., Yu, R., & Chen, Z. (2004). Holocene environmental changes in
the tropical and subtropical areas of South China and the relation to human activities.
Quaternary Science, 24, 387–393 (华南沿海热带-亚热带地区全新世环境变化与人类活动
的关系. 第四纪研究, 2004).
Zong, Y. (1992). Post-glacial stratigraphy and sea-level changes in the Han River Delta, China.
Journal of Coastal Research, 8, 1–28.
Zong, Y., Chen, Z., Innes, J. B., Chen, C., Wang, Z., & Wang, H. (2007). Fire and flood
management of coastal swamp enabled first rice paddy cultivation in East China. Nature, 449,
459–463.
Zong, Y., Innes, J. B., Wang, Z., & Chen, Z. (2011). Mid-Holocene coastal hydrology and salinity
changes in the east Taihu area of the lower Yangtze wetlands, China. Quaternary Research, 76,
69–82.
Zong, Y., Zheng, Z., Huang, K. Y., Sun, Y. Y., Wang, N., Tang, M., et al. (2013). Changes in sea
level, water salinity and wetland habitat linked to late agricultural development in the Pearl
River Delta plain of China. Quaternary Science Reviews, 70, 145–157.
Chapter 5
Subsistence Patterns Associated
with Shell Middens from the Pre-Qin
Period in the Coastal Region of China
Luo Zhao
Abstract During the pre-Qin period, the subsistence patterns from nearly 500 shell
middens in the coastal region of China fall into the following three categories:
fishing-hunting-foraging, mixed, and agricultural. Before about 7,000 years ago the
fishing-hunting-foraging pattern was predominant. However, around that date
increasing food production, animal domestication and farming of crops becomes
more visible in the coastal shell middens, indicating a shift in subsistence patterns
toward the mixed type. During this period the level of food production in the
northern region was also significantly higher than in the south. Only the Beiqian
Site, in the Jimo County of Shandong, contained evidence of an agricultural pattern
of subsistence. The emergence of greater shellfish consumption signalled local
adaptations to the environment during this period. The subsequent transformation
and disappearance of subsistence patterns from shell middens thus relates closely to
local levels of agriculture and food production, as well as the cultural context and
social complexity of the time.
5.1 Introduction
L. Zhao (&)
Shanghai Cultural Heritage Conservation and Research Centre, Shanghai, China
e-mail: pseudozhao@163.com
and other subjects of inquiry (Cai 1998, 2004; Wu 2004; Pan and Chen 2011; Zhao
2014, 2018).
The shell midden is a kind of archaeological site which contains large amounts
of shell remains abandoned by ancient people, in addition to other artifacts and
animal bones (Zhao 2014). The subsistence patterns typically seen in shell middens
are most often representative of fishing and hunting economies, due to their location
near water or off shore. However, archaeological investigations of shell middens in
China have increasingly discovered evidence of other food production activities as
well, including plant cultivation and animal domestication.
The prehistoric and historic subsistence patterns typical of hunter-gatherer,
agricultural and domesticated food activities have been studied by scholars all over
the world. In particular, Bruce Smith and Zhongpei Zhang have advanced argu-
ments about the development of subsistence patterns based on the analysis of shell
middens in the coastal region of China during the pre-Qin period. Smith’s theory of
low-level food production proposed a picture of the general subsistence landscape
for ancient peoples, and divided the entire history of human economic development
into three stages: hunting-fishing-foraging, agriculture, and the middle ground
(Smith 2001). Zhang suggested that a “farming and animal husbandry” stage pre-
ceded the agricultural economy in northern China, stressing that the time span
between hunter-gatherer subsistence and agriculture were longer than a few thou-
sand years (Zhang 2011). Building on these theories, the subsistence patterns
associated with Chinese coastal shell middens can be divided into three stages:
fishing-foraging-hunting, mixed, and agricultural.
Nearly 500 shell midden sites have been reported in mainland China, not including
Taiwan. Most of these are located in the coastal provinces, though a few have also
been discovered in the inland provinces of Yunnan (云南) and Hunan (湖南). The
highest concentration of shell middens occurs in the southern coastal region of
China, especially in Guangdong (广东) Province (Fig. 5.1).
As shown in Fig. 5.1, these shell middens fall into four regions, Zone I—IV,
according to their geographical locations and cultural content. Zone I is located on
the Liaodong (辽东) and Jiaodong (胶东) Peninsulas; Zone II in the Taihu Lake (太
湖) Basin; Zone III in the coastal region of Fujian (福建) Province; and Zone IV in
southern China, including the coastal areas of Guangdong, Guangxi (广西) and
Hainan (海南) Provinces. Of these four distribution zones, Zones I and II represent
the north regional type, while Zones III and IV represent the south regional type.
Shell middens along the coastline contained marine mollusk species in both north
and south regions, while the great majority of freshwater mollusk species appeared
in shell middens along river banks and inside some caves but not by the seashore in
Guangxi and Guangdong in the southern region (Zone IV).
5 Subsistence Patterns Associated with Shell Middens … 91
Fig. 5.1 Shell middens of the pre-Qin period discovered in costal mainland China
Table 5.1 Time table showing shell middens in four different regions during the pre-Qin period
Region Cultural types Dating (BP)
Zone I Lower Xiaozhushan to Shuangtuozi Stage III 6500-3000
Baishicun Stage I to Zhaogezhuang type
Zone II Majiabang Culture to Songze Culture 6500-5000
Zone III Keqiutou Culture to Huangtulun Culture 6500-2700
Zone IV Bailiandong Cave to Gantuoyan Site 26000-2500
The dates associated with these shell middens vary widely (Table 5.1). More
than half of the shell midden sites on the Liaodong Peninsula of Zone I are
Neolithic sites from before 4000 BP, while most of the 27 shell midden sites on the
Jiaodong Peninsula are dated to 6500-5500 BP. In Zone II, most of the shell
middens sites are dated to 6500-5500 BP, with a few sites dated specifically to 5000
BP. More than half of the sites along the Fujian coast in Zone III are dated to the
Neolithic Age, ranging from the Keqiutou (壳丘头) to the Huangguashan (黄瓜山)
Culture. In Zone IV archaeologists have discovered the earliest shell midden at
Bailian (白莲洞) Cave in Liuzhou (柳州), dating to 26,000 BC, and the latest one
at the Gantuoyan (感驮岩) Site in Napo (那坡), from 2,500 BP.
92 L. Zhao
Domesticated plant and animal remains in the archaeological record provide the
most direct evidence for studying the subsistence of the prehistoric and early his-
toric inhabitants of coastal China. By identifying and assessing the species found in
excavated materials, the animal and plant resources utilized by ancient inhabitants
can be deduced, allowing us to analyze the methods of food exploitation, pro-
duction and subsistence.
Among the coastal shell middens, the excavated animal remains from 55 sites have
been studied and the resulting analyses published, concluding that animal species
were likely exploited by the people who settled this area and also practiced hunting,
fishing, foraging and the domestication of animals. Assessment of these unearthed
faunal remains has revealed evidence of the following: mammals including dogs,
sambas, sika, small muntjac, red muntjac, pigs, cattle, bears, tigers, foxes, hogs, and
whales; birds including pheasant, wild pigeon, sand chicken, goose, duck, and
swan; and fish including black carp, carp, grouper, Oriental catfish, sharks, black
turtle, soft-shelled turtle, and a series of mollusks and arthropods. Among the
mammals, deer was the most common type of remains, followed by pigs and dogs.
A number of marine and fresh water species of mollusks were clearly exploited.
The coastal shell middens in Zone I, II and III contained various marine species
such as Ostreagigas Thunberg, Meretrixmeretrix (Linnaeus), Rapanavenosa
(Valenciennes), Chlorostomarustica Gmelin, and Chlorostomarustica Gmelin
(DASDU et al. 2011; Fu 1984; IA CASS 2007). Fresh water mollusks were more
common in inland shell middens along river banks in Zone IV and included
Cipangopalidian hainaensis (Kobelt), Bellanya quadratus (Benson), Rivularia
porcellanea (Kobelt), Limnoperna sp., Unio douglasiae (Gray), and Lamprotula
leai (Gray) (Lv 2011). Still, wild animals appear to have been the most important
source of meat; confirming that hunting and fishing were crucial to obtaining animal
food resources among most of the populations that deposited shell middens in the
coastal region of China.
Research on the plant remains from these shell middens has revealed new
insights on the paleo-environment, ancient economies and subsistence strategies of
local inhabitants by identifying large specimens such as charred seeds, fruits, and
pollen and phytolith remains. The plant remains from such shell middens have been
discussed by scholars of both archaeobotany and environmental archaeology.
Species such as pine, spruce and fir help to illustrate the historical environment of
sites, while edible plants species such as acorns, hazelnuts, lotus seeds, olives,
ginkgo, neem, melon, rice, millet, wheat, millet, soybeans and barley may have
served the dietary needs of ancient peoples. According to the existing
5 Subsistence Patterns Associated with Shell Middens … 93
Pig and dog have been the most common prehistoric domesticated animals iden-
tified among the remains from the shell middens of coastal China (Table 5.2).
Yunbing Luo has systematically researched the domesticated pigs remains of pre-
historic and early historical China and argues that they were used for both meat and
sacrificial rituals in ancient China (Luo 2012). According to the quantitative
analysis of mammalian bones unearthed from different shell middens, three levels
of domesticated pig utilization were identified. The first level includes the sites in
which more than half of mammalian bones are domesticated pig, such as the
Guojiacun (郭家村) Site in Dalian (大连) City, the Beiqian (北阡) Site in Jimo (即
墨) County, and the Wanbei (万北) Site in Shuyang (沭阳) County. Among these,
the domesticated pig accounted for 86% of the total mammalian remains in the
early stage of the Wanbei Site (Li 1991). The second level includes sites in which
domesticated pigs were the most common species of mammal but amounted to less
than 50% of the total mammalian remains. The Beiwutun (北吴屯) Site in Dalian
City, the Tanshishan (昙石山) Site in Minhou County, and the Huangguashan (黄
瓜山) Site in Xiapu (霞浦) County are all examples of this level. The domesticated
pig ratio in the Tanshishan Culture reached more than 40% (Luo 2012: 214), while
the domesticated pig remains at Huangguashan Site accounted for more than 25%
of the total mammalian remains (Jiao 2009). The third level includes the sites in
which the domesticated pig was not a primary species for human utilization, such as
the Hedang (河宕) Site in Foshan (佛山) City, the Qiujiang (秋江) Site in
Hengxian (横县) County, and the Baishawan (白沙湾) Site in Xiangzhou (象州)
County.
Shell middens also illustrate how domesticated dogs were another common
animal raised by the prehistoric peoples of China’s coastal region. In ancient
Chinese society dogs played a special role in social and economic life, helping
humans with hunting and guarding in addition to being used for sacrifices and meat
(Wu 2014). Except for the Xuecheng (薛城) Site in Gaochun (高淳) County and
the Dingsishan (顶蛳山) Site in Yongning (邕宁) County, the bones of domesti-
cated dog were generally found together with domesticated pig. Thus, since
domesticated pigs would have provided much more abundant meat than dogs and
wild deer and shellfish were also identified as supplementary sources of meat at
these sites, the shell midden findings indicate that dogs are less likely to have been a
primary source of meat for ancient peoples.
94 L. Zhao
Table 5.2 Domesticated animal and plant remains found in shell middens of China
Sites Animals Plants Dating (BP)
Dog Pig Broomcorn Foxtail Rice
millet millet
Dalian ✓ ✓ ✓ ✓ ✓ 5500-4000
Guojiacun
Dalian Beiwutun ✓ ✓ 6500-5000
Dalian Dapanjia ✓ 4500-4000
Beihai ✓ ✓ ✓ 6000-5000
Donggang
Changhai ✓ ✓ 6500-4000
Xiaozhushan
Yantai ✓ ✓ ✓ 6500-6100
Baishicun
Yantai ✓ Beixin Culture to early
Qiujiazhuang Dawenkou Culture
Rushan ✓ early Dawenkou Culture
Wengjiabu
Penglai ✓ ✓ early Dawenkou Culture
Dazhongjia
Yantai ✓ early Dawenkou Culture
Hadingdui
Jimo Beiqian ✓ ✓ ✓ ✓ ✓ 6100-5500
Shuyang ✓ ✓ 6540 ± 90
Wanbei
Gaochun ✓ middle and late Majiabang
Xuecheng Culture
Jintan ✓ 6500-5500
Sanxingcun
Minhou ✓ ✓ ✓ 5500-3500
Tanshishan
Minhou Xitou ✓ ✓ 5000-3500
Xiapu ✓ ✓ 4300-3500
Huangguashan
Xiapu ? ✓ ✓ ✓ 3700-3400
Pingfenghshan
Foshan Hedang ✓ 4300-4000
Gaoyao ✓ Late Neolithic to Shang-Zhou
Maogang periods
Dongguan ✓ ca. 4000
Cuntou
Hengxian ? 8000-7000
Qiujiang
(continued)
5 Subsistence Patterns Associated with Shell Middens … 95
Jiaodong and Liaodong Peninsulas, though the scale of rice agriculture here is still
unknown. Typical dry crops such as millet, wheat and barley were also found from
this period in the shell middens of this region. The use of flotation techniques on
excavated seeds revealed that cultivated broomcorn millet was the dominant veg-
etable food produced by the early Dawenkou Culture at the Beiqian Site in Jimo
County, while cultivated rice agriculture was relatively underdeveloped (Nie 2013;
Wang et al. 2012). Broomcorn millet and foxtail millet remains were also unearthed
in some shell middens in the Jiaodong Peninsula of Zone I, including the Guojiacun
Site in Dalian City, the Donggang (东岗) Site in Behai (北海) County, and the
Xiaozhushan (小珠山) Site in Changhai (长海) County. However, the details
surrounding corn agriculture remain vague. In southern China, domesticated rice
remains from after 5,000 BP were collected from the Tanshishan Site in Minhou
County, the Huangguashan and Pingfengshan (屏风山) Sites in Xiapu County, and
the Gantuoyan (感驮岩) Site in Napo (那坡) County. Meanwhile, the identification
of barley and wheat at the Huangguashan and Pingfengshan Sites suggests the
spread of these species southward along the coast (Zhao 2017).
The settlement patterns evident in the shell middens of China’s coastal region lasted
for 7,000 years during the pre-Qin period, from 10,000-3,000 BP. The shell mid-
dens themselves, along with the settlements and subsistence strategies they repre-
sent, passed through 3 stages: appearance, change and decline.
5 Subsistence Patterns Associated with Shell Middens … 97
Along the coastal region of China, the earliest shell middens have been found in the
cave sites of the northern mountainous region of Zone IV. Shell deposits in a strata
dated to 26,000 BP were discovered in Bailiandong (白莲洞) Cave of Liuzhou (柳
州) City, Guangxi Province, but most of the early shell middens found in this region
date to between 10,000 and 8,000 BP; for examples, the Dushizi (独石仔) Site in
Yangchun (阳春) County, Huangyandong (黄岩洞) Cave in Fengkai (封开)
County, and Niulandong (牛栏洞) Cave in Yingde (英德) County.
The appearance of shell middens at this moment, in the Late Pleistocene to the
Early Holocene, was similarly observed all over the world. K.V. Flannery has
suggested that the dramatic environmental changes in the postglacial period led to
the demise of many large ungulates, disrupting the food chain and the resources
available to prehistoric peoples. Faced with the dual pressures of population growth
and the dramatic decrease of large animals, prehistoric humans were forced to
exploit new, smaller food resources such as small animals, plants, fish, shellfish,
nuts, roots and grass seeds (Flannery 1969). The presence of bones from extinct
mammal species including rhinoceros, tapir and stegodon in the Huangyandong,
Dushizi and Niulandong Caves supports Flannery’s hypothesis. The appearance of
shell middens in the mountainous caves of northern Guangdong during the tran-
sition period between Paleolithic and Neolithic might therefore have been the result
of the extinction of large animals. Ultimately, the subsistence patterns evident in
most of the shell middens from Zone IV illustrate the practice of
fishing-hunting-foraging activities until 4,000 BP.
Starting around 6,500 BP, the subsistence patterns evident in China’s coastal shell
middens changed to a mixture of both food production and hunting and gathering.
The domestication of animals and crops may have been the main reason for the
subsequent changes in the social and cultural indicators found in shell middens.
The first change was a shift from the traditional nomadic lifestyle of fishing and
hunting to residential settlements dedicated to developing local food production. At
the Beiwutun (北吴屯) Site of Dalian (大连) City, strata dating to between 6,500
and 5,500 BP contained numerous, complicated house remnants of different sizes as
well as domesticated pig remains (LPICRA et al. 1994). Likewise, at the Beiqian
(北阡) Site, of the late Dawenkou (大汶口) Culture in Jimo (即墨) County of
Shandong, a large sedentary settlement with sub-region zoned housing and burials
was discovered along with a large quantity of domesticated plant and animal
remains.
The second change occurred in social stratification; a shift evident in the burial
remains found in shell middens like the ones at Sanxingcun (三星村) Site, in Jintan
98 L. Zhao
(金坛) County, where 1,001 burials were found. The funerary objects contained in
these burials ranged from 0–20 pieces. The most common number of burial objects
was 5–6, but a few had 10–20, and some had none. The richest burial is M636,
whose funerary objects are much more numerous than the others and include such
complicated luxury goods as carvings, jade earrings, Jue (玦), ivory artifacts and
plate-shaped carving implements; testament to the higher social standing of the
tomb occupant (JATS 2004). Such patterns of social stratification are absent from
the shell middens of the southern region.
The decline of shell middens in the 4 zones of coastal China happened at different
times according to local environmental, cultural and economic conditions.
After 7,000 BP the shell middens on the river banks of the Yongjiang (邕江)
River began to decrease, and few other shell middens have been found. Among
these are the Nanshawan (南沙湾) Site in Xiangzhou (象州) County, the Hecun (河
村), Jiangbian (江边) and Chongtang (冲塘) Sites in Chongzuo (崇左) County, and
the Gantuoyan (感驮岩) Site in Napo (那坡). According to zooarchaeological
studies of the shellfish remains from the Yongjiang River shell middens, the pre-
historic residents who deposited these shell middens did not overexploit the
shellfish resources, suggesting that a shortage of shellfish resources was not the
reason shell middens began to decline during this period (Lv 2010). Naihan He has
argued that shell middens in Guangxi decreased during the Late Neolithic due to the
rapid development of agriculture (He 1985). The high concentration of rice phy-
toliths in the fourth stage of the Dingdingshan Site, dating to 6,000 BP, indicate the
emergence of rice cultivation in that area.
By 5,000 BP shell middens had almost disappeared from the Jiaodong
Peninsula. As noted in earlier chapters, high levels of food production based on
plant and animal domestication are evident in the costal shell middens of the
Jiaodong Peninsula and the Taihu Lake Basin of Zone I and Zone II. In the Haidia
(海岱) area, food production advanced rapidly from the late Dawenkou (大汶口)
Culture to the Longshan (龙山) Culture. In particular, during the Longshan Culture
period, the proportion of millet and rice farming relative to other crops increased
significantly on the Jiaodong Peninsula (Luan 2005). A similar process unfolded in
the development of rice farming and livestock husbandry in the Taihu (太湖) Lake
Basin during the Songze (崧泽) and Liangzhu (良渚) Cultures, as rice and pigs
became the most important food sources. In Zone III, rice farming also underwent
significant developments in the Huangguashan (黄瓜山) Culture, although con-
temporaneous food production in the southern areas of this zone illustrated only
limited development of domesticated foods, as seen at the Yishan (蚁山) and
Yinloushan (音楼山) Sites of Jingjiang (晋江) County and the Baitang’ao (白塘澳)
Site in Dongshan (东山) County. The ultimate decline of shell middens occurred
late in the Huangtulun (黄土仑) Culture.
5 Subsistence Patterns Associated with Shell Middens … 99
5.5 Conclusion
The shell middens in the coastal region of pre-Qin China represent an important
type of archaeological site for better understanding the maritime characteristics of
early Chinese subsistence strategies. The patterns of subsistence developed gen-
erally from hunting-fishing-gathering to a mixture of gathering and domesticating
before transitioning to agriculture. The latter two subsistence patterns are charac-
terized by the intentional production and utilization of domesticated animals and
plants as food resources. The cultural changes observed in the subsistence patterns
associated with shell middens reflect inhabitants’ changing lifestyles, from nomadic
to sedentary, as well as social stratification. Regionally, the level of food production
taking place in the north was higher than the south. Thus, the Beiqian Site had
already transitioned into agriculture at the same time that shell middens continued
to appear in the south, where food production remained limited for a relatively long
time.
The shell middens of the north and south coasts of China illustrate the devel-
opment of a variety of maritime cultures during the pre-Qin period. The maritime
tradition of the southern region lasted for a longer period, and agricultural activity
there grew more slowly. By contrast, the maritime tradition in the northern region
lasted for a relatively shorter period, declining more rapidly as it was incorporated
into the broader framework of the continental agricultural system of the Late
Neolithic.
Acknowledgements I wish to thank Dr. Wu Chunming for inviting me to join the international
conference, “The prehistoric maritime silk road: new research on Neolithic seascapes of East Asia”
at Xiamen (October 29–November 2, 2017). I am also indebted to Mr. Li Yan and Dr. Fu Lin for
their reviews and comments on this paper.
References
Cai, B. (蔡保全). (1998). The living environment and resource exploitation of coastal resources in
the shell middens of Fujian. Journal of Xiamen University, Arts & Social Sciences, 3, 106–111
(Cong BeiqiuYizhi Kan Fujian Yanhai Xianmin de Juzhu Huanjingyu Ziyuan Kaifa, 《从贝丘
遗址看福建沿海先民的居住环境与资源开发》, Xiamen Daxue Xuebao, 《厦门大学学
报》).
Cai, B. (蔡保全). (2004, May 7). Cultural exchange and changes in dietary habits or depletion of
shellfish resources: An answer to Mr. Wu Xiaoping on “the causes of the disappearance of shell
middens in Fujian Province.” Chinese Cultural Relics News, p. 7 (Shi Wenhua Jiaoliu Yu
Yinshi Xiguan Bianqian Haishi Beilei Ziyuan de Kujie——jiu “Fujian Beiqiu Yizhi Xiaowang
Yuanyin” da Wu Xiaoping Xiansheng, 《是文化交流与饮食习惯变迁还是贝类资源的枯竭
——就“福建贝丘遗址消亡原因”答吴小平先生》,Zhongguo Wenwu Bao,《中国文物
报》).
Claassen, C. (1998). Shells. Cambridge: Cambridge University Press.
DASDU (Department of Archaeology of Shandong University), Institute of Cultural Relics
Conservation and Archaeology of Qingdao (ICRCAQD), & Jimo Musuem (JMM). (2011).
100 L. Zhao
Brief Report on the Excavation of the Beiqian Site, Jimo, Shandong Province, in 2007.
Archaeology, 11, 3–23 (Shandong Jimoshi Beiqian Yizhi 2007 Nian Fajue Jianbao,《山东即
墨市北阡遗址2007年发掘简报》, Kaogu, 《考古》).
Flannery, K. V. (1969). Origins and Ecological Effects of Early Domestication in Iran and the Near
East. In P. J. Ucko & G. W. Dimbley (Eds.), The domestication and exploitation of plants and
animals (pp. 73–100). London: Gerald Duckworth & Co., Ltd.
Fu, R. (傅仁义). (1984). Study of the Animal Bones Unearthed at Guojiacun. Acta Archaeologica
Sinica, 3, 331–334 (Dalian Guojiacun Yizhi de Dongwu Yigu, 《大连郭家村遗址的动物遗
骨》, Kaogu Xuebao 《考古学报》).
He, N. (何乃汉). (1985). The emergence and development of prehistoric agriculture in Guangxi.
Agricultural Archaeology, 2, 90–95 (Guangxi ShiqianShiqiNongye de Chansheng he
FazhanChutan,《广西史前时期农业的产生和发展初探》, NongyeKaogu,《农业考古》).
Hu, Y. (胡耀武), Wang, G.(王根富), & Cui, Y(崔亚平). (2007). Study on the Inhabitants’
Recipes of the Sanxingcui Site, Jintan, Jiangsu Province. Chinese Science Bulletin, 52(1), 85–
88 (Jiangsu Jintan Sanxingcun Yizhi Xianmin de ShipuYanjiu, 《江苏金坛三星村遗址先民
的食谱研究》, Kexue Tongbao《科学通报》).
IA CASS (Institute of Archaeology CASS). (2007). Shell Midden Sites in the Jiaodong Peninsula:
Studies in environmental archaeology. Beijing: Social Sciences Academic Press (Jiaodong
Bandao Beiqiu Yizhi Huanjing Kaogu《胶东半岛贝丘遗址环境考古》).
JATS (The Joint Archaeological Team at Sanxingcun). (2004). Excavation of a Neolithic site at
Sanxingcun, Jintan, Jiangsu. Cultural Relics, 2, 4–26 (Jiangsu Jintan Sanxingcun Xinshiqi
Shidai Yizhi, 《江苏金坛三星村新石器时代遗址》, Wenwu, 《文物》).
Jiao, T. (焦天龙). (2009). The significance of economic changes in the coastal region of Fujian
Province during the Neolithic Age. Fujian Cultural Relics and Museology, Z, 47–54 (Fujian
Yanhai Xinshiqi Shidai Jingji Xingtai de Bianqian ji Yiyi,《福建沿海新石器时代经济形态
的变迁及意义》, Fujian Wenbo,《福建文博》).
Li, M. (李民昌). (1991). The assessment of Unearthed Neolithic Animal Remains from the
Wanbei Site, Shuyang, Jiangsu Province. Southeast Culture, Z, 183–189 (Jiangsu Shuyang
Wanbei Xinshiqi Shidai Yizhi Dongwu Guge Jianding Baogao,《江苏沭阳万北新石器时代
遗址动物骨骼鉴定报告》, Dongnan Wenhua,《东南文化》).
LPICRA (Liaoning Provincial Institute of Cultural Relics and Archaeology), Dalian
Municipal CPAM (DMCPAM), & Zhuanghe Municipal Office for the Preservation of
Ancient Monuments (ZMOPAM). (1994). The Neolithic Site at Beiwutun, Dalian city. Acta
Archaeologica Sinica, 3, 343–380 (Dalian Beiwutun Xinshiqi Shidai Yizhi,《大连市北吴屯
新石器时代遗址》, KaoguXuebao《考古学报》).
Luan, F. (栾丰实). (2005). The Appearance, Development and Pervasion of Prehistoric Rice
Cultivation Agriculture in the Haidai Area. Journal of Literature, History and Philosophy, 6,
41–47 (Haidai Diqu Shiqian Daozuo Nongye de Chansheng, Fazhan he Kuosan,《海岱地区
史前稻作农业的产生、发展和扩散》, Wen Shi Zhe《文史哲》).
Luo, Y. (罗运兵). (2012). The domestication, raising and ritual use of pigs in ancient China.
Beijing: Science Press (Zhongguo Gudai Zhulei Xunhua, Siyangyu Yishixing Shiyong,《中国
古代猪类驯化,饲养与仪式性使用》, Kexue Chunanshe, 科学出版社).
Lv, P. (吕鹏). (2010). Zooarchaeological study on the shell middens along the Yong River,
Guangxi. Unpublished Ph.D. Thesis, Institute of Archaeology, CASS, Beijing (Guangxi
Yongjiang Liuyu de Dongwu Kaoguxue Yanjiu, 《广西邕江流域贝丘遗址的动物考古学研
究》).
Lv, P. (吕鹏). (2011). Study of Fauna from Shell Midden Sites along the Yong River, Guangxi.
Quaternary Sciences, 31(4), 715–722, (Guangxi Yongjiang LiuyuBeiqiu Yizhi Dongwuqun
Yanjiu,《广西邕江流域贝丘遗址动物群研究》, DisijiYanjiu《第四纪研究》).
Nie, Z. (聂政) (2013). Settlement and Subsistence on the Jiaodong Peninsula during Early
Dawenkou Culture. Unpublished Ph.D. Thesis, Shandong University, Jinan (Jiaodong Bandao
Dawenkou Wenhua Zaoqi de Juluo yu Shengye,《胶东半岛大汶口文化早期的聚落与生
业》).
5 Subsistence Patterns Associated with Shell Middens … 101
Pan, Y. (潘艳), Chen, C. (陈淳). (2011). On the origin of agriculture and theoretical change of the
broad spectrum revolution. Southeast Culture, 4, 26–34 (Nongye Qiyuan yu “Guangpu
Geming” Lilun de Bianqian,《农业起源与“广谱革命”理论的变迁》, Dongnan Wenhua,
《东南文化》).
Smith, B. D. (2001). Low-level food production. Journal of Archaeological Research, 9(1), 1–43
(KaoguxueYanjiu,《考古学研究》).
Tsang, C. H. (臧振华). (2006). The Archaeology of Taiwan. Taipei: Artist Press (Taiwan Kaogu,
《台湾考古》, Yishujia Chubanshe, 艺术家出版社).
Wang, F. (王芬), Fan, R. (樊榕), & Kang, H. (康海涛). (2012). Stable Isotope Analysis on Human
Bones Unearthed from the Beiqian Site, Jimo: the Food Recipe of Coastal Ancestors. Chinese
Science Bulletin, 57(12), 1037–1044 (Jimo Beiqian Yizhi Rengu Tongweisu Fenxi: Yanhai
Xianmin de Shiwu Jiegou,《即墨北阡遗址人骨稳定同位素分析:沿海先民的食物结构》,
KexueTongbao,《科学通报》).
Wu, X. (吴小平). (2004). The causes for the disappearance of shell middens in Fujian Province.
Agricultural Archaeology, 1, 27–28 (Yetan Fujian BeiqiuYizhi Xiaowang de Yuanyin,《也谈
福建贝丘遗址消亡的原因》, Nongye Kaogu,《农业考古》).
Wu, Z. (武庄). (2014). A current study and research prospects of the domesticated dog during the
pre-Qin period. Cultural Relics in Southern China, 1, 65–73 (Xianqin Shiqi Jiaquan Yanjiu de
Xianzhuang yu Zhanwang, Nanfang Wenwu,《南方文物》).
Zhang, Z. (张忠培). (2011). Closing Speech at Seminar on “Past, Present and Future of
Archeology in the Northeast and Inner Mongolia Area. Northern Cultural Relics, 1, 107–110
(Zai “Dongbei ji Neimenggu Dongbu Diqu Kaogu de Guoqu, Xianzaiyu Weilai” Xueshu
Yantaohui Bimushishang de Jianghua, 《在“东北及内蒙古东部地区考古的过去、现在与
未来”学术研讨会闭幕式上的讲话》, BeifangWenwu,《北方文物》).
Zhao, L. (赵荦). (2014). Study of shell middens in pre-Qin Coastal China. Unpublished Ph.D.
Thesis, Fudan University, Shanghai (ZhongguoYanhai Xianqin Beiqiu Yizhi Yanjiu, 《中国
沿海先秦贝丘遗址研究》).
Zhao, L. (赵荦). (2018). Reviews of shell middens studies in China. Fujian Cultural Relics and
Museology, 1, 22–28 (Woguo Beiqiu Yizhide Yanjiu Pingshu,《我国贝丘遗址研究述评》,
Fujian Wenbo,《福建文博》).
Zhao, Z.(赵志军). (2009). Archaeobotanical research on the origins of agriculture and Chinese
Civilization. In Ministry of Science and Technology, State Administration of Cultural Heritage
(Ed.), Collections of Chinese Civilization Project, Technology and Economy, Volume I
(pp. 79–91). Beijing: Science Press (Youguan Nongye Qiyuan he Wenming Qiyuan de Zhiwu
Kaoguxue Yanjiu,《有关农业起源和文明起源的植物考古学研究》, Zhonghua Wenming
Tanyuan Gongcheng Wenji, Jishu yu Jingji Juan I, 《中华文明探源工程文集•技术与经济卷
•I》).
Zhao, Z. (赵志军). (2017). Discussion of an Ancient Ocean Channel Based on Flotation Results
from Plant Remains at the Nanshan Site. Speeches at the International Conference of
“Prehistoric Archaeology of Southeast China and the Pacific Rim,” Mingxi, Fujian (Cong
Nanshan Yizhi Fuxuan Jieguo Tan Gudai Haiyang Tongdao,《从南山遗址浮选结果谈古代
海洋通道》).
Chapter 6
A Preliminary Analysis
of the Development of Neolithic Culture
in Coastal Region of Guangdong
Yan Li
Geologists have reconstructed historic sea level changes in the Pearl River Delta
region from the Late Pleistocene to the Holocene (Li et al. 1991), and archaeolo-
gists have confirmed these changes based on evidence from cultural artifacts found
in Hong Kong (Shang and Wu 2010). Building on this research, our focus centers
around sea level change and environmental variations in this region starting around
6000 BP.
According to geographic research, the primary coastal topographies present in
Guangdong fall into four categories also relevant to prehistoric settlements: sand
dam-lagoon, drowned valley coast, estuarine delta, and eroded terrace (Fig. 6.1).
Y. Li (&)
Guangdong Institute of Cultural Relics and Archaeology, Guangzhou, Guangdong, China
e-mail: 13602842657@189.cn
Fig. 6.1 Distribution of coastal geomorphology in Guangdong (Made by Zhaoxuan Zeng and
Weifeng Huang)
(1) Sand dam-lagoon. Sand dam-lagoons appear on both the east and west coasts of
Guangdong. They are most heavily concentrated on the east coast from
Huidong (惠东) County to Shantou (汕头) City. To the west, they stretch from
Yangdong (阳东) to Wuchuan (吴川).
(2) Drowned valley coast. Drowned valleys were originally river or mountain
valleys that were subsequently flooded and entirely submerged by rising sea
levels. These appear mainly in the central coast area from Taishan (台山) to
Huidong (惠东). Drowned valley coasts are generally associated with ancient,
underwater riverbeds, zigzag coastlines or sea bays, and islands and peninsulas.
(3) Estuarine delta. The largest estuarine delta is located in the lower reaches of the
Pearl River Delta. Others include the estuaries of the Hanjiang River (韩江), the
Tangjiang River (潭江), the Moyangjiang River (莫阳江), and the Jianjiang
River (鉴江).
(4) Eroded terrace. Eroded terraces are primarily located around the Leizhou (雷
州) Peninsula.
(5) Within single, relatively small regions these varied topographies can overlap;
thus the Pearl River Delta contains both sand dam-lagoons and drowned
valleys.
Another important coastal topographic landscape relating to the development of
maritime culture is islands. Most of the coastal sea islands in Guangdong are
offshore, and are located mainly in two areas: the ocean along the Taishan (台山)
and Dayawan (大亚湾) Coast, and the ocean off the Zhanjiang (湛江) Coast;
6 A Preliminary Analysis of the Development … 105
although a few are also located off the coast of Shantou (汕头). Most of these
islands are within 30 nautical miles of the mainland, and are associated with sea
bays and rich marine resources well suited to human subsistence.
Following investigations at the Neolithic Baojingwan (宝镜湾) Site in Zhuhai
(GDICRA et al. 2004), archaeologists proposed the idea of a Prehistoric Culture of
Sea Islands based on the theory that Neolithic peoples moved out into the islands
due to the rich marine resources available there. The Investigation of Prehistoric
Sites in The Pearl River Delta (2000), by professor Hui Zhao of Peking University
(TPSIPRD 2000), synthesized and evaluated the previous research before proposing
a set of methodological principles for continued archaeological investigation and
research on the Neolithic cultures of the coastal region.
“Firstly, we should inspect the deposit strata and collect any cultural artifacts in order to
identify the date and size of the site. Secondly, we should collect different kinds of animal
remains to analyze the livelihoods and ecological context for the prehistoric population.
Thirdly, we should investigate the geographical landscapes of the sites” (pp. 355).
Zhao’s analysis illustrates the changes in different types of shell middens during
the Neolithic Age, using the comparison of mollusk remains from different shell
mounds to divide the sites into 3 types: near estuarine, estuarine, and beach.
During the early stage, there were not many shell mounds. Though people in this stage
tended to collect shellfish resources in the estuarine region, this activity does not appear to
have correlated with specific settlements. However, during the late stage shell mound
settlements rapidly multiplied along the estuarine region, illustrating the obvious tendency
of subsistence region (pp. 399).
Zhao also divided sites into 3 types according to topography: hillock, terrace,
and the beach sand dune. The terrace type settlements were the most suitable for
subsistence and transportation. Meanwhile, other archaeologists and geologists
have also discussed the prehistoric environment, resources and their relationship
with human life in works such as Evolution of the Geographical Environment of
Zhuhai (珠海) and its Interaction with Ancient Cultures in the last 6000 years, by
professor Pingri Li (Li, P. 1991), and A Research on the Sand Dam Sites of Zhuhai,
by professor Shande Zhao (1991).
According to systematic archaeological investigations, the Neolithic and Early
Bronze Age cultures along the coastal region of Guangdong are most concentrated
in several regions. On the east coast, these include: the Meilong (梅陇), Chenqiao
(陈桥) and Nanao (南澳) microlith sites in the region surrounding the Hangjiang
River Estuary; a series of sand dune sites in Lianjiang (练江) River Estuary; and the
Bazaiyuan (菝仔园) and Shakeng (沙坑) Sites in Jieshi Bay (碣石湾), Honghai
Bay (红海湾) and Shanwei (汕尾) City (GDPM 1961b). In the central coast of the
Pearl River Estuary they include the Xiaomeisha (小梅沙), Xiantouling (咸头岭) in
Dapengwan (大鹏湾) Bays of Shenzhen; a series of sites in Hong Kong, Zhuhai,
and Macau such as Hac Sa (黑沙) Site, Longxue (龙穴) Site in Zhongshan, Guye
(古椰) Site in Gaoming, Cuntou (村头) Site in Dongguan, Lujingcun (鹿颈村) Site
106 Y. Li
Among the various prehistoric sites discovered in the coastal region of Guangdong,
the richest deposits have been found on the Pearl River Delta and its estuary. Based
primarily on the information found in this Pearl River Delta region, the chrono-
logical sequence of Guangdong’s prehistoric coast includes the following seven
stages.
6 A Preliminary Analysis of the Development … 107
Fig. 6.2 Shorelines of the Pearl River Delta since 6000 BP (from Li et al. 1991: 78)
The Xiantouling Site is located next to the Diefu Lagoon and River, on the sand
dunes of Diefuwan (跌福湾), of northeast Dapengwan (大鹏湾), in Shenzhen City
(SZICRAA 2013: 3–5). The site articulates with both the macro-environment of the
drowned valley coast and the micro-environment of a sand dam-lagoon landscape.
The combined environments of the lagoon’s fresh water/plain, and the shoal’s
seawater/beach provide plenty of flora and fauna ideal for supporting human life.
Artifacts were discovered in large quantities at the site, including a painted ring foot
pottery Dou (豆) plate, a painted ring foot pottery pot and cup, a kettle, Fu (釜)
cooking ware, and their associated pedestals. The decorative patterns found on
pottery include painting and fine corded patterns, stamped patterns, carving pat-
terns, stamped shell patterns, and embossed line patterns characterized by painted
108 Y. Li
Fig. 6.3 Shorelines of the Hangjiang River Delta since 6000 BP (from Li et al. 1991: 153)
patterns (Figs. 6.7, 6.8). These findings are chronologically divided into 3 stages: a
first stage from 7000-6400, a second from 6400-6200 BP, and a third from
6200-6000 BP (SZICRAA 2013: 27–41). Most of the researchers acknowledge that
Xiantouling Culture emerged through the cultural dissemination and influence of
certain Neolithic cultures from the Hunan Province, specifically the Gaomiao (高
庙), Tangjiagang (汤家岗), and Daxi (大溪) Cultures.
The Guye Shell Mound Site is located on the Guye Hillock, along the Gaoming (高
明) River in Foshan (佛山) City (Li and Cui 2017), across the Xijiang (西江) River
6 A Preliminary Analysis of the Development … 109
Fig. 6.4 Distribution of Neolithic and Early Bronze Age sites in East Guangdong
Fig. 6.5 Distribution of Neolithic and Early Bronze Age sites in Pearl River Delta
110 Y. Li
Fig. 6.6 Distribution of Neolithic and Early Bronze Age sites west of Guangdong
from the Xiqiaoshan (西樵山) Site (Fig. 6.9). Most of the mollusk shell remains are
of the freshwater type, while a small number of intertidal zone oyster remains were
also collected. This suggests that this paleo-estuarine topography was located far
from the beach. Similar findings associated with the Guye Culture were also
identified in sites along the Pearl River Delta, including Phase I and II of the
Baojingwan Site (GDICRA et al. 2004), Phase I of the Caotangwan (草堂湾) Site
in Zhuhai (Liang and Li 1991), the Hudi (虎地) and Shaxia (沙下) Sites in Hong
Kong (AMOHK and HNPIA 2004; AMOHK and GZMICRA 2007), the Hac Sa
Site in Macao (Tang and Zheng 1996), the Dianchan Xicun Site in Taishan, and
layer 2 of the Dahuangsha (大黄沙) Site in Shenzhen (SZMM and ADSYSU
1990).
6 A Preliminary Analysis of the Development … 111
Fig. 6.8 Samples of pottery from the Xiankezhou and Haifeng Sites, 1–2; Haifeng (海丰), 3; and
Xiankezhou (蚬壳洲), 4–6
Discovered and excavated in 2006, the Guye Site was rated one of top ten most
important archaeological findings in China that year. Preliminary research on
artifacts from the site was carried out from 2011 to 2015, and indicated that the site
illustrated an offshoot of the Xiantouling Culture, dating to 5900-5000 BP (Li and
Cui 2017). Further investigations have shown that this Guye Culture extends from
Taishan, in the west, to Hangkong in the southeast and from the Dhilaodun (史老
墩) Site of Yinde (英德) County and the Shixia (石峡) Site of Qujiang (曲江)
County in the north down to the southern coast. The site was chronologically
divided into four phases. Phases I-III belong to the Early Stage of the culture and
date to 5900-5200 BP, while Phase IV belongs to a Late Stage dating to 5200-5000
BP. The culture associated with these four phases is characterized by sandy coarse
paste pottery, including Fu (釜) round bottom cooking ware and ware pedestals, as
well as fine paste pottery including straight neck round bottom pots, ring foot pots,
112 Y. Li
Fig. 6.9 Samples of pottery from the Guye (古椰) and Shaxia (沙下) Sites
ring foot plates and Dou (豆) ring foot plates. Stone artifacts from the site include
shouldered axes and other microliths of the later period.
Typologically, Guye Culture inherited some elements of Xiantouling Culture
such as Fu (釜) pottery cooking ware, ring foot plates, and carving pattern deco-
rations, illustrating a cultural continuity in this region. However, the cultural dif-
ferences between them are also obvious; for example, there is no painted pottery in
Guye Culture, even though this was characteristic of Xiantouling Culture, while the
Xiantouling Culture did not have the shouldered stone implements that became
highly typical of Guye Culture.
External cultural influences have also been identified in the successive phases of
Guye Culture. In Phase I of the Early Stage there is evidence of cultural commu-
nication from north to south. For example, the Duiziling (堆子岭) Culture from the
Xiangjiang River Basin appears to have played an important role in the formation of
Guye Culture. In Phase II of the Early Stage, cultural influences from the lower
Yangtze River northeast of Guye appear, including Dou (豆) high ring foot bowls,
decoration patterns from the Hudi Site, and spinning wheels from the Yonglang (涌
浪) Site that originated in the Songze (崧泽) Culture of the lower Yangtze River.
6 A Preliminary Analysis of the Development … 113
Finally, in Phase III of the Early Stage of Guye Culture, cultural influences from the
Liangzhu (良渚) Culture appear, including ring foot pots.
The Early Stage I of the Yuanzhou Site of Dongguan City suggests that the local
cultural characteristics of the coastal region of Guangdong arose and developed
subsequent to Guye Culture (GDICRA and DGMM 2000). Early Stage I Yuanzhou
Culture continued the cultural tradition of a stamped stripes line pattern and carving
on pottery characteristic of the Guye Culture but also developed a set of real
geometrical stamped patterns such as checks, rectangular checks, a trellis, zigzags,
and the compound pattern of checked parallel and diagonal lines. Some forms of
pottery from Yuanzhou I, such as the long neck pottery pot with short ring foot of
Type C Fu (釜) cooking ware, have been identified as variations on styles observed
among Guye artifacts. Meanwhile a newer form, the Ding (鼎) tripod, has been
identified as emerging from the cultural influence of Hutoupu (虎头埔) Culture,
from northeast of the Guangdong coast. Similar cultural forms can also be seen in
Phase III of the Baojingwan (宝境湾) Site in Zhuhai (Fig. 6.10).
The local cultural tradition of Early Stage I Yuanzhou Culture can be seen con-
tinuing on through Phase III of the Baojingwan Site in Zhuhai City (GDICRA et al.
2004) and the Early Stage II of the Yuanzhou Site, in such artifacts as ring foot pots
with plantain leaf decorative patterns (Fig. 6.11). New forms of the ring foot pot,
with a folded shoulder or belly and new geometric stamped patterns featuring
compounded parallel and diagonal lines, also appeared in Early Stage II of
Yuanzhou Culture. The same pottery styles have been observed in the Late Stage of
the Langyong (涌浪) Site in Hong Kong (AMOHK 1997) and the Chiwan (赤湾)
Sand Dune Site of Shenzhen, both dating to the same period as the Central Stage of
Shixia (石峡) Culture in northern, mountainous Guangdong (Yang 1994).
The archaeological deposit at the Houshawan Site in Zhuhai includes 2 layers and 2
phases representing successive cultures (Li, Z. 1991). The Phase II Houshawan
Culture presents new culture styles similar to Phase III of the Baojingwan Culture. The
same cultural styles have also been identified at the following sites: the Lujingcun (鹿
颈村) Shell Mound Site of Nansha, in Pangyu County (GZMICRA 2005: 279);
114 Y. Li
Fig. 6.10 Samples of pottery from Early Stage I of the Yuanzhou (圆洲) Site
Phase I of the Yinzhou (银洲) Site in Sanshui County; Phase I of Youyugang (鱿鱼岗)
Site in Nanhai County (GDICRA 2001); and the Hedan (河宕) Site in Foshan (佛山)
County (GDPM 2006) (Fig. 6.12).
Pottery remains from this period illustrate how certain forms from the previous
period, such as the ring foot pot with a folded shoulder or belly, continued while a
set of new forms also developed. This included the Houshawan Type I pot, char-
acterized by a wide shoulder, round belly and short ring foot and decorated with
stamped zigzag and embossed line patterns (Fig. 6.12). Typologically, this pottery
is similar to the artifacts from the Early Stage of Phase III at the Shixia Site
(GDICRA et al. 2014: 469).
6 A Preliminary Analysis of the Development … 115
Fig. 6.11 Samples of pottery form Early Stage II of the Yuanzhou (圆洲) Site and Phase III of the
Baojingwan (宝镜湾) Sites
Fig. 6.12 Samples of pottery from Phase I of the Youyugang (鱿鱼岗) Site and Phase I of the
Yinzhou (银洲) Site
116 Y. Li
The Early Stage of Phase I at the Cuntou Site presents a new type of culture in the
Pearl River Delta and estuary region. Similar cultural findings have also been
identified in a series of sites including the following: the Early Stage of Phase II at
the Youyugang (鱿鱼岗) Site in Nanhai County; M19 and M23 of the Hedan (河
宕) Site in Foshan (佛山) County; M040 and M042 of the Wubeiling (屋背岭)
Site; and M16 and M20 of the Huangzhuyuan (黄竹园) Site in Shenzhen
(Fig. 6.13). The typical pottery artifacts from this period are small Dou (豆) ring
foot dishes, handle and spout pots, and concave bottom pots, illustrating similarities
with the characteristic styles of the Guangfulin (广福林) Culture in the lower
Yangtze River (Chen 2014). This cultural form has been dated from the last stage of
the Longshan (龙山) Period to the Early Stage of the Xia (夏) Dynasty.
The Late Stage of Phase I and Phases II-III of the Cuntou Site contain archaeo-
logical evidence characterizing the culture of the Xia and early Shang (商)
Dynasties in the Pearl River Delta and Estuary. These findings are further divided
into two stages according to cultural change, the Early and the Late Stages.
The Early Stage includes cultural remains from the following: Late Stage of
Phase I at the Cuntou Site; Phase II of the Yinzhou Site; Late Stage of Phase II at
the Youyugang Site; the Zaogang (灶岗) Site in Nanhai County; layer 3 of the
Hedan Site in Foshan City; the Houhai (后海) Sand Dune Site in Shenzhen City;
the Lingjiaozui (棱角嘴) Sand Dune Site; Phase 1 of the Yapowan (亚婆湾) and
Dongaowan (东澳湾) Sites in Zhuhai City; the Lujingcun Site in Guangzhou; and
Phase III of the Yuanzhou Site in Dongguan City. The pottery vessel most char-
acteristic of this stage is the concave bottom pot and the folded shoulder ring foot
pot, with decorative zigzags combining cloud and thunder patterns that have been
identified as the result of influences from the Early Stage of the Maqiao (马桥)
Culture in the lower Yangtze River. The stone ring with a T-shape section and the
Go (戈) stone dagger are also characteristic of this stage.
The Late Stage includes cultural remains from the following: Phase II and III of
the Cuntou Site; M004 and M058 of the Wubeiling Site; M8 of the Huangzhuyuan
(黄竹园) Site; Phase III of the Yinzhou Site; the Maogang Site in Gaoyaomao
County; the Shang Dynasty remains from the Tangxiahuan (棠下环) Site in
Zhuhai; and Phase II of the Dongwanzai Site in Hong Kong. Most of the pottery
forms from this stage continued from the Early Stage, while some new forms like
the Gu (觚) cup, Dou (豆) low ring foot bowl, and Zun (尊) pot also appeared. The
6 A Preliminary Analysis of the Development … 117
Fig. 6.13 Samples of pottery from the Cuntou (村头) Site, 1–5; the Hedan (河宕) Site, 6–8; and
the Yapowan (亚婆湾) Site, 9
decorative zigzag combining cloud and thunder patterns declined during this stage,
while the number of checked and rhomboid patterns increased.
Meanwhile, the prehistoric cultures on the eastern coast of Guangdong devel-
oped in a different chronological sequence that featured three stages (Wei 2012).
Stage I is represented by Layer 3 of the Chenqiao (陈桥) Site in Chaoan (潮安)
County and the Shakengbei (沙坑北) Site in Haifeng (海丰) County. Pottery from
this stage features sandy grey paste pots and bowls with burnished and red slip or
carving line patterns as decoration. A few painted ring foot plates were also col-
lected from the Shakengbei Site. The stone tools include chopping tools, blade end
scrapers, awls, and needles. The culture of this stage coincides with the early period
of Xiantouling Culture in the Pearl River Estuary.
118 Y. Li
Stage II is represented by the Hutoupu (虎头铺) Site and Group I of the Guishan
(龟山) Site. The representative pottery includes fine gray paste stoneware short ring
foot pots, round belly pots and bowls, with stamped pattern decorations. This stage
is dated to the same period as early Yuanzhou Culture and Phase II of the
Houshawan Culture in the Pearl River Delta.
Stage III is represented by the Houshan (后山) Burial Site in Puning (普宁)
County and Group II of the Guishan Site. Typical artifacts from this stage are fine
grey stoneware round bottom pots, concave bottom pots, spout pots, and round
bottom bowls decorated with stamped checked and double line rhomboid patterns.
This stage is dated to the same period as the Cuntou and Wubeiling Cultures of the
Pearl River Delta.
The Neolithic cultures of Stages 3–6 in the Pearl River Delta and Estuary developed
their own local cultural traditions and spread a series of cultural influences
throughout the coastal region of the South China Sea.
The culture of Stage 3 developed the same kind of geometric stamped pottery
discovered in Stage I of the early period of Yuanzhou Culture in Dongguan City, a
style represented by the checked stamped pattern and a compound of checked
parallel and diagonal lines. The typical ring foot pot from the early period of the
Yuanzhou Culture was also discovered at the Wubeiling Site in Shenzhen
(SZCRCAC 2006: 54) and the Halong Culture in northern Vietnam, showing
Neolithic cultural diffusion along the coast. The typical round bottom pot of the
Hutoupu Culture, found at the Hengling (横岭) Site, was also identified in Phase III
of the Baojingwan Site, including the pot (T10③B:1) that illustrates a cultural
influence extending from the east to the central region along the coast of
Guangdong.
120 Y. Li
During Stage 4, represented by Phase III of the Baojingwan Site and the early
period II of the Yuanzhou Site, the Hutoupu Culture from east of Guangdong
spread to the east estuary of the Pearl River. Most of the evidence for this move-
ment, including stone and pottery from the Yonglang Site in Hong Kong, were
vessels largely typical of the Hutoupu Culture. Still, Phase III of the Baojingwan
Site retained the tradition of stamped pottery in the Pearl River Delta.
Stage 5, represented by the Lujingcun Site and Phase II of the Houshawan
Culture, illustrated cultural influences on Huntoupu Culture. The two Cong (琮)
ritual objects discovered at the Santuo (三舵) Shell Midden on the east coast of
Guangdong (Mao 1985; Yang et al. 1990), whose style matches that of late
Liangzhu Culture (Fang 2008), show the cultural diffusion of Liangzhu Culture
from the lower Yangtze River. Since the same style of artifact is entirely absent in
the Jiangxi and Fujian regions, these exotic Cong (琮) of Hutoupu Culture must
have been transported to the east coast of Guangdong by sea.
Stage 6 is represented by the Early Stage of Phase I at the Cuntou Site and
Huangzhuyuan (黄竹园) Sand Dune Site in the Yantian (盐田) District of
Shenzhen, and continued the same kinds of cultural interaction with surrounding
regions (SZMU et al. 2008). For example, the fine orange paste pottery bowl from
this stage, with a handle and round bottom and stamped with zigzag patterns, was
quite similar to pottery from the Jiaoshan (角山) Site in Yingtang (鹰潭) City,
Jiangxi Province. This illustrates the extension of cultural influences from north to
south during this period (JXPCRT et al. 1987, 2017: 413). Two jade semi-circular
Huang (璜) pendants (Site M16, 2-1/2) from the same site were likewise identified
as evidence of cultural diffusion from northern China (Huan 2016; Gao 2016).
Stage 7 presents a lot of new information about subsistence and the cross border
cultural interactions in the Pearl River Delta and Estuary. Domesticated pigs
remained, while T-cross section jade rings and jade Yazhang (牙璋) tablets from
the Cuntou Site illustrate the development of agriculture and broader regional
relationships.
The study of pig bones from the Cuntou Site revealed 89 individuals, some of
which were identified as belonging to domesticated species. Some domesticated
dog remains were also present at the Cuntou Site. More than 200 pieces of pig bone
remains were collected from the Hedan Site in Foshan City, most of which were
identified as domesticated. These findings reveal the development of agriculture in
the region, including livestock farming and grain cultivation.
6 A Preliminary Analysis of the Development … 121
The Jade Yazhang tablet is one of most typical ritual objects of ancient Chinese
civilization, first found at the Cuntou Site of Dongguan and the Dawan Site of Hong
Kong during this stage (Fig. 6.14). Similar object (BT08 ②: 2) have also been
discovered in the Late Stage of Phase II at the Gantuoyan (感驮岩) Site in Napo (那
坡) County, Guangxi (GXARCRT and NPCM 2003). The pottery from Phase II of
the Gantuoyan Culture is also similar to pottery from Stage 7 in the Pearl River
Delta. For example, the folded shoulder pots with corded pattern, the carved stripes
pattern and the painted decorations are also found at the Cuntou Site of Dongguan,
and even at the Phung Nguyen Site and in the Ha Long Culture of Vietnam. These
findings suggest that the jade Yazhang tablets of Gantouyan, Guangxi and Phung
Nguyen, Vietnam both originated in the Pearl River Estuary (Chen, et al. 2016).
A similar pattern has been observed with T-shaped section jade rings, which
have been discovered at the Huangzhuyuan Sand Dune in Shenzhen (Site M8, 1),
the Dawan Site in Hong Kong, and the Tangxiahuan Site in Zhuhai (GDICRA et al.
1998). Although it is thought to have originated in the Pearl River Estuary, this type
of artifact has also been found in Guangxi and northern Vietnam. The resulting
dispersal supports the argument for early maritime transportation along the north
coast of the South China Sea.
6.4 Conclusion
The Neolithic cultures along the north coast of the South China Sea developed
successively between 7000 and 3500 BP and illustrate a series of cultural inter-
actions with the regions to the north, east and west. The Pearl River Delta and its
estuary constitute the central region with the highest concentration of Neolithic
sites. Yet many other sites were also discovered along the eastern coast of
Guangdong, suggesting that cultural dissemination between the central and eastern
regions occurred mainly via coastal corridors and offshore islands. This theory is
confirmed by the fact that less Neolithic sites have been discovered along the
western coast of Guangdong, and therefore cultural interaction between the Pearl
River Estuary and the Red River Estuary in north Vietnam likely occurred via the
offshore sea route of the South China Sea. This early maritime road set an important
foundation for maritime transportation along the Beibu Gulf by Xuwen (徐闻) and
Hepu (合浦) during the Han Dynasty.
With regard to the prehistoric economy and early subsistence, these Neolithic
cultures developed different approaches to utilizing ocean and marine resources.
The Xiantouling Culture developed a hunting and gathering economy after
migrating from the northwest Hunan region to the coast and the Pearl River
Estuary, where the people became primarily dependent upon abundant marine
resources in the estuaries and coastal gulf. Guye Culture basically carried on the
same model, although the cultural influence of Songze and Liangzhu Cultures from
the lower Yangtze River also arrived during this period, accompanied by jade work
techniques and probably the use of a sea route, as well. Beginning with the Early
Stage of Phase I at the Yuanzhou Site to Phase II of the Houshawan Site, a new
period of heightened regional cultural interaction began. Cultural dispersals during
this period included the arrival of the Shixia Culture in the Pearl River Delta, the
influence of the Early Stage of Yuanzhou Culture on the Ha Long Culture of
Vietnam, and the spread of the Hutoupu Culture from the east coast to the
Yonglang Site of Hong Kong. At this point livestock farming and agriculture also
emerged, while the collection of fish and other marine resources declined by
comparison.
Acknowledgements I would like to thank Mr. Rutian Zhu for drawing the images for this paper.
References
AMOHK (Antiquities and Monuments Office of Hong Kong). (1997). A preliminary report on the
archaeological excavation of the Yonglang Neolithic Site in Hong Kong. Archaeology, 6, 5–53
(Xianggang Yonglang Xinshiqi Shidai Yizhi Fajue Jianbao, 《香港涌浪新石器时代遗址发
掘简报》, Kaogu, 《考古》).
AMOHK (Antiquities and Monuments Office of Hong Kong) & GZMICRA (Guangzhou
Municipal Institute of Cultural Relic and Archaeology). (2007). A preliminary report on the
6 A Preliminary Analysis of the Development … 123
archaeological excavation of the DI region of the Shaxia Site in Xigong, Hong Kong. Huaxia
Archaeology, 4, 3–34 (Xianggang Xigong Shaxia Yizhi DIQu Dajue Jianbao, 《香港西贡沙
下遗址DI区发掘简报》, Huaxia Kaogu,《华夏考古》).
AMOHK (Antiquities and Monuments Office of Hong Kong) & HNPIA (Henan Provincial
Institute of Archaeology). (2004). A preliminary report on the archaeological excavation of the
C02 and DII02 regions of the Shaxia Site in Xigong, Hong Kong, 2002. Huaxia Archaeology,
4, 3–47 (2002 Niandu Xianggang Xigong Shaxia Yizhi C02Qu he DII02Qu Kaogu Fajue
Jianbao,《2002年度香港西贡沙下遗址C02区和DII02区考古发掘简报》, Huaxia Kaogu,
《华夏考古》).
Chen, J. (陈杰) (2014). A preliminary study on the Guangfulin Culture. In Shanghai Museum
(Eds.), Collection of academic works on archaeological excavations and studies of the
Guangfulin Site (pp. 221–237). Shanghai: Shanghai Classics Publishing House (Guangfulin
Wenhua Chulun, 《广福林文化初论》, Guangfulin Kaogu Fajue yu Xueshu Yanjiu Lunji,
《广福林考古发掘与学术研究论集》).
Chen, X., Li, Y., & Wang, L. (2016). A study on the distribution and related issues at Yazhang in
the Lingnan Region. Huaxia Archaeology, 4, 100–107 (Guanyu Lingnan Suojian Yazhang de
Fenbu ji Xiangguan Renshi, 《关于岭南所见牙璋的分布及相关认识》, Huaxia Kaogu,
《华夏考古》).
Fang, X. (方向明). (2008). Further study of jades from the Shixia Culture. In Lingnan
archaeological research (Vol. 7, pp. 55–65). Hong Kong: China Academic Review Press
(Shixia Wenhua Xiangguan Yuqi Jiben Yanjiu zhi Buchong,《石峡文化相关玉器基本研究
之补充》, Lingnan Kaogu Yanjiu,《岭南考古研究》).
Gao, J. (高江涛). (2016). A preliminary analysis on connecting jade huang and bi from the Taosi
Site. South Culture Relics, 4, 89–97 (Taosi Yizhi Chutu Duolian Huanbi Chutan, 《陶寺遗址
出土多璜联璧初探》, Nanfang Wenwu, 《南方文物》).
GDCRCAC (Guangdong Cultural Relics Conservation Administration Committee). (1961a).
Primitive cultural sites in the southern area of Guangdong Province. Archaeology, 11, 595
(Guangdong Nanlu Diqu Yuanshi Wenhua Yizhi,《广东南路地区原始文化遗址》, Kaogu,
《考古》).
GDCRCAC (Guangdong Cultural Relics Conservation Administration Committee). (1961b). The
shell mound sites in Chao’an County, Fuangdong. Archaeology, 11, 77–584 (Guangdong
Chao’an de Beiqiu Yizhi, 《广东潮安的贝丘遗址》, Kaogu, 《考古》).
GDICRA (Guangdong Institute of Cultural Relics and Archaeology). (2001). The excavation
report for the Youyugang Shell Mound Site in Nanhai City. In GDICRA (Ed.), Memorial
Works for the 10th Anniversary of the Establishment of the Guangdong Provincial Institute of
Archaeology (pp. 282–328). Guangzhou: Lingnan Fine Arts Press (Nanhaishi Youyugang
Beiqiu Yizhi Fajue Baogao, 《南海市鱿鱼岗贝丘遗址发掘报告》, Guangdongsheng
Wenwu Kaogu Yanjiusuo Jiansuo Shizhounian Wenji, 《广东省文物考古研究所建所十周
年文集》).
GDICRA (Guangdong Institute of Cultural Relics and Archaeology) & DGMM (Dongguan
Municipal Museum). 2000. The excavation of the Yuanzhou Shell Mound Site in Dongguan,
Guangdong Province. Archaeology, 6, 11–23 (Guangdong Dongguan Yuanzhou Beiqiu Yizhi
de Dajue, 《广东东莞市圆洲贝丘遗址的发掘》, Kaogu, 《考古》).
GDICRA (Guangdong Institute of Cultural Relics and Archaeology), GDPM (Guangdong
Provincial Museum), & QJDMSGGD (Qujiang District Museum of Shaoguan City Guangdong
Province). (2014). Archaeological excavation report on the Shixia Sites, 1973–1978 (pp. 469).
Beijing: Cultural Relic Press (Shixia Yizhi 1973-1978Nian Kaogu Fajue Baogao,《石峡遗址
——1973-1978年考古发掘报告》).
GDICRA (Guangdong Institute of Cultural Relics and Archaeology) & PCBZH(Pingsha Cultural
Bureau of Zhuhai). (1998). A preliminary report on the excavation of the Tangxiahuan Site in
Pingsha, Zhuhai. Cultural Relics, 7, 4–16 (Zhuhai Pingsha Tangxiahuan Yizhi Fajue Jianbao,
《珠海平沙棠下环遗址发掘简报》, Wenwu, 《文物》).
GDICRA (Guangdong Institute of Cultural Relics and Archaeology) & ZHMM (Zhuhai Municipal
Museum). (2004). The Baojingwan Site in Zhuhai—Report on archaeological excavations of
124 Y. Li
the sea-island prehistoric cultural site (pp. 152–153). Beijing: Science Press (Zhuhai
Baojingwan——Haidaoxing Shiqian Wenhua Yizhi Fajue Baogao, 《珠海宝镜湾——海岛
型史前文化遗址发掘报告》).
GDPM (Guangdong Provincial Museum). (1961a). Neolithic shell mound site in Dongxing
County, Guangdong. Archaeology, 12, P644*649, 688 (Guangdong Dongxing Xinshiqi
Shidai Beiqiu Yizhi, 《广东东兴新石器时代贝丘遗址》, Kaogu, 《考古》).
GDPM (Guangdong Provincial Museum). (1961b). The Neolithic cultural remains of Eastern
Guangdong. Archaeology, 12, 650 (Guuangdong Dongbu Diqu Xinshiqi Shidai Yicun,《广东
东部地区新石器时代遗存》, Kaogu, 《考古》).
GDPM (Guangdong Provincial Museum). (2006). The excavation report from the Hedan Site,
Foshan, 1977–1978. Guangzhou: Guangdong Renmin Press (《佛山河宕遗址-1977年至
1978年夏发掘报告》).
GXARCRT (Guangxi Zhuang Autonomous Region Cultural Relic Team) & NPCM (Napo County
Museum). (2003). Excavation report on the Gantuoyan Sites in Napo County, Guangxi.
Archaeology, 10, 35–56 (Guangxi Napoxian Gantuoyan Yizhi Fajue Jianbao, 《广西那坡县
感驮岩遗址发掘简报》, Kaogu,《考古》).
GZMICRA (Guangzhou Municipal Institute of Cultural Relic and Archaeology) (Ed.). (2005).
Accumulated research works on Guangzhou Archaeology (pp. 279). Beijing: Cultural Relic
Press (Zhuci Cunlei,《铢积寸累》).
Huan, T. (黄翠梅). (2016). Development of Early Jade Cultures of Taiwan and Circum South
China Sea Region as Revealed from Jue Slit Earrings with Protrusions. National Taiwan
Museum Journal, 1, 65–90 (Cong Daitu Jueshi Lun Taiwan yu Huannanhai zaoqi Yuwenhua
de Fazhan, 《从带突玦饰论台湾与环南海早期玉文化的发展》,《国立台湾博物馆学
刊》).
JXPCRT (Jiangxi Provincial Culture Relic Team) & YTMP(Yingtang Municipal Museum).
(1987). A Preliminary Report on the Jiaoshan Kiln Site of the Shang Dynasty, Yingtang.
Historical Relics in Jiangxi, 2, 32–43 (Yingtang Jiashan Shangdai Yaozhi Shijue Jianbao, 《鹰
潭角山商代窑址试掘简报》, Jiangxi Lishi Wenwu, 《江西历史文物》).
JXPCRT (Jiangxi Provincial Culture Relic Team) & YTMP(Yingtang Municipal Museum).
(2017). Archaeological excavation of the Jiaoshan Kiln Site, 1983–2007 (pp. 413). Beijing:
Cultural Relic Press (Jiaoshan Yaozhi 1983–2007 Nian Kaogu Fajue Baogao, 《角山窑址-
1983-2007年考古发掘报告》).
Li, P. (李平日). (1991). Evolution of the geographical environment of Zhuhai and its interaction
with ancient cultures in the last 6000 years. In Guangdong Institute of Cultural Relics and
Archaeology (GDICRA) & Zhuhai Municipal Museum (ZHMM) (Eds.), Archaeological
Discoveries and Research at Zhuhai (pp. 265–272). Guangzhou: Guangdong People’s Press
(Liuqian Nian Lai Zhuhai Gudili Huanjing Yanbian yu Guwenhua Yicun, 六千年来珠海古地
理环境演变与古文化遗存, Zhuhai Kaogu Faxian yu Yanjiu, 《珠海考古发现与研究》).
Li, Z. (李子文). (1991). Excavation of the Houshawan Site on Qi’ao Island. In Guangdong
Institute of Cultural Relics and Archaeology (GDICRA) & Zhuhai Municipal Museum
(ZHMM) (Eds.), Archaeological Discoveries and Research in Zhuhai (pp. 3–21). Guangzhou:
Guangdong People’s Press (Qi’aodao Houshawan Yizhi Fajue,《淇澳岛后沙湾遗址发掘》,
Zhuhai Kaogu Faxian yu Yanjiu, 《珠海考古发现与研究》).
Li, Y. (李岩), & Cui, Y.(崔勇). (2017). A preliminary analysis of the Guye Shell Mound Site.
Journal of Hunan Archaeology, 13, 226–251 (Fuye Beiqiu Yizhi Chishi, 《古椰贝丘遗存初
识》, Hunan Kaogu Jikan,《湖南考古辑刊》).
Li, P. (李平日), Qiao, P.(乔彭年), Zheng, H.(郑洪汉), Fang, G.(方国祥), & Huang, G(黄光庆).
(1991). Environmental evolution of the Pearl River Delta in the last 10,000 years. Beijing: The
Ocean Press (Zhujiang Sanjiaozhou Yiwannian Lai Huanjing Yanbian,《珠江三角洲一万年
来环境演变》).
6 A Preliminary Analysis of the Development … 125
Liang, Z. (梁振兴), & Li, Z.(李子文). (1991). Excavation of the Caotangwan Site on Sanzao
Island. In Guangdong Institute of Cultural Relics and Archaeology (GDICRA) & Zhuhai
Municipal Museum (ZHMM) (Eds.), Archaeological Discoveries and Research in Zhuhai
(pp. 22–33). Guangzhou: Guangdong People’s Press (Zhuhai Kaogu Faxian yu Yanjiu, 《珠海
考古发现与研究》).
Mao, Y. (毛衣明). (1985). Discovery of a Neolithic pit containing jade at the Tianqianyu Site in
Haifeng County. In Chinese Archaeological Society (CAS) (Ed.), Chinese Archaeological
Yearbook (1985) (pp. 202). Beijing: Cultural Relic Press (Haifengxian Tianqianyu Faxian
Xinshiqi Shidai Yuqikeng,《海丰县田墘圩发现新石器时代玉器坑》, Zhongguo Kaoguxue
Nianjian,《中国考古学年鉴》).
Shang, Z. (商志香覃), & Wu, W.(吴伟鸿) (2010). Documentation of Hong Kong Archaeological
Research. Beijing: Cultural Relic Press (Xianggang Kaoguxue Xuyan,《香港考古学叙研》).
SZCRCAC (Shenzhen Cultural Relics Conservation Administration Committee). (2006). Record
of the last 7000 years in Shenzhen. Beijing: Cultural Relic Press (Shenzhen 7000 Nian,《深圳
7000年》).
SZICRAA (Shenzhen Institute of Cultural Relics and Archaeological Authentication). (2013).
Archaeological report on the Xiantouling Site excavation, 2006. Beijing: Cultural Relic Press
(Shenzhen Xiantouling 2006 Nian Fajue Baogao,《深圳咸头岭 2006年发掘报告》).
SZMM (Shenzhen Municipal Museum) & ADSYSU (Anthropology Department of Sun Yat-Sen
University). (1990). A preliminary report on archaeological excavation of the Dahuangsha
Sand Dunes Site in Shenzhen, Guangdong. Cultural Relics, 11, 2–22, 28 (Guandong
Shenzhenshi Dahuangsha Shaqiu Yizhi Fajue Jianbao,《广东深圳市大黄沙沙丘遗址发掘简
报》, Wenwu, 《文物》).
SZMU (Shenzhen Municipal Museum), SZMCRAO (Shenzhen Municipal Cultural Relic
Administration Office), YTDCRAOSZ (Yantian District Cultural Relic Administration Office
of Shenzhen), & SZICRAA (Shenzhen Institute of Cultural Relics and Archaeological
Authentication). (2008). A preliminary report on the excavation of the Huangzhuyuan Site in
Yantian, Shenzhen, Guangdong. Archaeology, 10, 17–32 (Guangdong Shenzhenshi Yantianqu
Huangzhuyuan Yizhi Fajue Jianbao, 《广东深圳市盐田区黄竹园遗址发掘简报》, Kaogu,
《考古》).
Tang, C. (邓聪), & Zheng, W. (郑炜明). (1996). Archaeology of the Heisha Site, Macao. Hong
Kong: Chinese University Press (Aomen Heisha,《澳门黑沙》).
TPSIPRD (Team for Prehistoric Sites Investigation in Pearl River Delta). (2000). Investigation of
Prehistoric Sites in the Pearl River Delta. In School of Archaeology and Museology of Peking
University (Ed.), A Collection of Studies on Archaeology (Vol. 4, pp. 355–403). Beijing:
Science Press (Zhujiang Sanjiaozhou Shiqian Yizhi Diaocha, 珠江三角洲史前遗址调查,
Kaoguxue Yanjiu,《考古学研究》).
Wei, J. (魏峻). (2012). The chronology of the archaeological culture of the pre-Qin period in
Eastern Guangdong and Southern Fujian. In School of Archaeology and Museology of Peking
University (Ed.), A Collection of Archaeological Studies (Vol. 9, pp. 140–165). Beijing:
Science Press (Yuedong Minnan Diqu Xianqin Kaoguxue Wenhua de Denqi he Puxi,《粤东
闽南地区先秦考古学文化的分期与谱系》, Kaoguxue Yanjiu,《考古学研究》).
Yang, Y. (杨耀林). (1994). Investigation and excavation of pre-Qin sites in Shenzhen. In
Shenzhen Municipal Museum (Ed.), Archaeological Discoveries and Research in Shenzhen
(pp. 60). Beijing: Cultural Relic Press (Shenzhenshi Xianqin Yizhi diaocha yu Shijue,《深圳
市先秦遗址调查与试掘》, Shenzhen Kaogu Faxian yu Yanjiu, 《深圳考古发现与研究》).
Yang, S. (杨少祥), & Zheng, Z. (郑政魁). (1990). Discovery of jade cong and bronze weapons in
Haifeng County, Guangdong. Archaeology, 8, 743–751 (Guangdong Haifengxian Faxian
Yucong he Qingtong Bingqi, 《广东海丰县发现玉踪和青铜兵器》, Kaogu,《考古》).
Zhang, C. (张弛), & Hung, H. (洪晓纯). (2012). Neolithic Fishing, Hunting and Gathering
Cultures in Southern China and the Adjacent Areas. In School of Archaeology and Museology
of Peking University (Ed.), A Collection of Archaeological Studies (Vol. 9, pp. 415–434).
Beijing: Science Press (《中国华南及其邻近地区的新石器时代采集渔猎文化》,
Kaoguxue Yanjiu,《考古学研究》).
126 Y. Li
Zhao, D. (赵德善). (1991). Research on Sand Dam Sites in Zhuhai. In Guangdong Institute of
Cultural Relics and Archaeology (GDICRA) & Zhuhai Municipal Museum (ZHMM) (Eds.),
Archaeological Discoveries and Research in Zhuhai (pp. 254–264). Guangzhou: Guangdong
People’s Press (Zhuhai Shati Yizhi Yanjiu, 《珠海沙堤遗址研究》, Zhuhai Kaogu Faxian yu
Yanjiu, 《珠海考古发现与研究》).
ZJICRA (Zhejiang Institute of Cultural Relics and Archaeology). (2012, January 6). A Neolithic
jade and stone implements workshop site at Fangjiazhou, Tonglu, Zhejiang. China Cultural
Relics News, p. 5 (Zhejiang Tonglu Fangjiazhou Xinshiqi Shidai Yushiqi Zhizaochang Yizhi,
《浙江桐庐方家洲:新石器时代玉石器制造场遗址》, Zhongguo Wenwubao, 《中国文物
报》).
Chapter 7
Early Maritime Subsistence
and Adaptive Ocean Cultures
Along the Beibu Gulf Coast
Zhen Li
Abstract The many Neolithic shell middens and dune sites along the coast of the
Beibu Gulf point to a specific lifestyle and subsistence pattern focused around
fishing, hunting and gathering based on marine resources. The central role of a
stable marine economy in local subsistence highlights a set of distinct maritime
cultural characteristics identified with the early culture of the Beibu Gulf coastal
region. This culture existed from about 7000 to 4000 years ago and was charac-
terized by adaptation and utilization rather than production of oceanic resources.
This early maritime culture grew out of early hunter-gatherer cultures in this region
dating to the Late Paleolithic and Early Neolithic, and emerged as a result of these
populations’ emigration to the coastal areas of Beibu Gulf.
7.1 Introduction
Beibu Gulf, which was originally called the Gulf of Tonkin, is a semi-enclosed gulf
in the northwest part of the South China Sea. It lies to the west of Leizhou (雷州)
Peninsula in Guangdong (广东) and Hainan (海南) Island; to the south of the
Guangxi (广西) Zhuang (壮) Ethnic Autonomous Region; and to the east of the
northeast and north-central parts of Vietnam. Surrounded by the mainland of China,
Vietnam and China’s Hainan Island, the gulf covers an area of nearly 130,000 square
kilometers and is linked to the Qiongzhou (琼州) Strait and the South China Sea.
Located in subtropical and tropical areas of the South China Sea, the Beibu Gulf
is rich in natural marine resources that offer a stable source of natural and wild food
for human survival. The coastline is comprised of tortuous sand and shallow beach
water, which would have made it convenient for the fishing and gathering of
prehistoric peoples. In search of more space, many primitive hunters and gatherers
of the Late Paleolithic and Early Neolithic were attracted by the rich marine
Z. Li (&)
Institute of Cultural Relics Conservation and Archaeology of Guangxi, Guangxi, China
e-mail: nnlizhen@163.com
resources of the Beibu Gulf, and ultimately settled this coastal region. As a result, a
large number of Neolithic sites are widely distributed along the coast of the Beibu
Gulf. According to investigations and preliminary analysis on the findings from
these sites, a large proportion of these Neolithic cultural sites contain distinct
oceanic characteristics.
According to the existing statistics, more than 100 Neolithic maritime cultural sites
have been discovered and investigated along the coast of the Beibu Gulf. These
sites are mainly distributed on the northwest and southeast coast of the Beibu Gulf,
including coastal Vietnam, Guangxi and China’s Hainan, and the highest concen-
tration were found on the northeast coast of Vietnam (Fig. 7.1). They are comprised
of several types including shell middens, sand dunes, and cave sites. As the most
common type, shell middens and sand dune sites represent the main source of
information about the region’s early maritime cultural heritage.
Fig. 7.1 Distribution of early maritime cultural sites along the Beibu Gulf
7 Early Maritime Subsistence and Adaptive Ocean Cultures … 129
7.2.1 Guangxi
The Beibu Gulf borders Beihai (北海), Qinzhou (钦州), Fangchenggang (防城港) and
other cities in south of Guangxi. At this point more than a dozen early maritime cultural
sites have been discovered along the coast. These include primarily coastal shell mid-
dens and dune sites, with coastal shell middens being the most commonly found type.
The main coastal shell middens include: Yapushan (亚菩山) Mountain;
Malanzuishan (马兰嘴山) Mountain; Beijiaoshan (杯较山) Mountain; Sheshan (社
山) Mountain; Xieling (蟹岭) Mountain; Fantaoping (番桃坪); Yingpan (营盘)
Village; Haotanjiao (蠔潭角) and Dadun (大墩) Islands in Fangchenggang City;
Bajiaodun (芭蕉墩) in Qinzhou City; and Gaogaodun (高高墩) in Baihutou
Village, Beihai City. All of these sites are located on hills near the beach, and some
are found on small islands or in front of mountains near the beach. They are
generally about 10 meters above sea level and are located next to a fresh water
stream flowing into the sea. The most well preserved and heavily excavated sites
include Yapushan, Beijiaoshan, and Sheshan in Fangchenggang City.
Several of the excavated sites illustrate how the stratified deposits of these shell
middens consist of large quantities of marine shells as well as aquatic and terrestrial
animal remains discarded after human consumption. The cultural artifacts from the
sites include substantial amounts of pottery, stone implements, bone and shell
artifacts. The aquatic and terrestrial animal remains have been identified as deer,
elephants, rabbits, birds, fish, turtles, clam (cytherea, sp.), big clam (aroa, sp.),
oyster (qstrea, sp.), river snail (viviparus), and snail (semiewecospira), among
others. Both chipped and polished stone implements have been found, but the
chipped variety is more common. Indeed, chipped stone implements are the most
common tools found at these sites overall. They include a wide range of complex
types, all of which are core artifacts made from river pebble. More than 80% of
these chipped stone implements are quartz sandstone. Made by hammering, most of
them have a flat oval shape with a thick and heavy pattern. The main types are the
Haolizhuo chopper (蠔蛎啄, a special heavy stone implement with a pointed tip for
breaking oysters), chopping tools, shaped hand axe tools, triangular stone tools, net
weights, concave shaped implements, and stone bola (of which the Haolizhuo is a
leading example). Some roughly polished stone tools have also been found
including axes, adzes, chisels, millstones, pestles, cake-shaped artifacts and stone
saddle querns, as well as shouldered shape axes, shouldered shape adzes and
trapezoid shape adzes.
All of the pottery is made of a coarse paste mixed with sand and clam powder.
Their firing temperature was low and their bodies are thin. The most common
exterior colors were red, grey black and brown, and some of the pottery bears a red
slip on the surface. The most common decoration is the fine corded pattern, though
the basket pattern and lineation pattern also appear. The main types of pottery found
are pots, Fu (釜) cooking ware and round bottom bowls. A few bone and shell
artifacts made by grinding have also been found. The bone artifacts include awls,
arrowheads and a large quantity of perforated ornaments made from fish vertebrae.
130 Z. Li
Shell artifacts included shovels, ring ornaments and net weights. Shell net weights,
used as fishing tools, were made by boring a round hole into the top of a clam shell
(GDPM 1961).
Sand dune sites have been discovered at Xishapo (西沙坡) and Niushihuantang
(牛屎环塘) in Beihai City, and at Shangyangjiao (上洋角) and Niyizui (妮义嘴) in
Qinzhou City. None of these sites have been excavated and their features and
cultural content are therefore unknown. Shangyangjiao is one of the best preserved
sites, located at the estuary of Sanniangwan (三娘湾) Bay in Qinzhou City. The site
links the Shangyanjiao Mountains in the north to the coastal terraces in the east,
west and south. To the east of the site, a small stream flows from the north, into the
sea to the south. The site covers an area about 52 meters square and stands 4 meters
above sea level. The stone artifacts collected from here include chipped pointed
implements, axes, and flakes as well as polished stone axes, adzes, chisels and
knives. Pottery shards from the site are made from a black paste mixed with coarse
sand (GDCRCAC 1961).
7.2.2 Guangdong
Guangdong Province connects with the Beibu Gulf via the west coast of Leizhou
(雷州) Peninsula. To date, the only site discovered in this region is the Liyudun
(鲤鱼墩) Shell Midden Site. Located in Dongbianjiao (东边角) Village, Jianghong
(江洪) Town, Suixi (遂溪) County, this site has a relative altitude of about 2 meters
and covers an area of about 1,200 square meters. It was excavated in 2002 and 2003
by the Institute of Cultural Relics and Archaeology of Guangdong. A stratigraphic
deposit containing a large amount of shell was discovered, in addition to a series of
cultural deposits including eight Neolithic tombs with flexed burials, two houses,
and a number of artifacts such as pottery, stone, bone and shell (GDICRA 2015).
Both chipped and polished stone implements were also found at this site. The
majority of the stone tools that were discovered are chipped stone implements,
including pointed implements, chopping tools, net weights, stone hammers, stone
anvils, and cake-shaped artifacts. Among these, the most common findings were
pointed implements and net weights. The polished stone implements included
adzes, chisels and stone beaters or stamps. Most of the pottery were made from a
reddish-brown coarse paste, while some others are made from a fine paste or a fine
paste mixed with fine sand. Among these is one type characterized by decoration
with a red ochre slip. The primary decorations on the pottery are straight or slant
lines scratched by clam shells and stamped corded patterns, basket patterns and line
patterns. Some of the plain pottery are also decorated with a serration pattern on the
rim. Pottery types included pots, Fu (釜) cooking ware, bowls, ring foot plates and
spinning wheels, among others. Excavated bone artifacts include shovel tools and
necklaces made from fish vertebrae, while shell artifacts include perforated tools
made from oyster shells. A large number of marine remains were collected from the
site, including more than 10 kinds of mollusk shells such as different types of
7 Early Maritime Subsistence and Adaptive Ocean Cultures … 131
clams, Meretrix meretrix, oysters, and others. Two kinds of fish remains were also
represented, mackerel and Cyprinidae. A few terrestrial animal bones were col-
lected from deer, buffalo, wild boar, and small chamois.
7.2.3 Hainan
At the point where the western coast of Hainan Island connects with the Beibu Gulf,
a series of Neolithic sites have been discovered. These sites are located in Lingao
(临高), Danzhou (儋州), Changjiang (昌江), Dongfang (东方), Ledong (乐东) and
other counties and include shell middens and sand dune sites. Some of the shell
middens are located several meters above sea level on the hills by the beaches of sea
bays, while the largest site measures more than ten thousand square meters. The
stratum deposit from these sites contains a large number of mollusk shell remains
such as whorl, mussels and oysters. Some other sites have been found on the dunes
near stream estuaries in Hainan, but these prehistoric sites were dated to 5000 BP or
later (Hao 1998; Hao and Sun 2010; He 2012).
The most representative Neolithic site is the Xinjie (新街) Site in Dongfang
County, the oldest and largest Neolithic shell midden discovered on Hainan Island.
The site is located on a sand dune in the south bank of the Beili (北黎) River, 2.5 km
away from the estuary, and measures about 16,000 square meters. About 0.5–1.1
meters thick, the cultural stratum deposit at this site includes a large number of both
mollusk shells (including whorl, clams and oysters) and terrestrial animal remains
(including cattle, wild boars, deer and porcupines). Artifacts found here include
stone implements and pottery. The stone implements are mostly chipped pebble and
include choppers and axe shape tools, though a few polished stone axes and adzes
were also found. A small amount of sandy coarse pottery consisted of simple types
such as round bottom Fu (釜) cooking ware and jars (HNCRCAC 1990).
7.2.4 Vietnam
A large number of Neolithic sites have been found along the coast of Vietnam, and
especially the Beibu Gulf Coast from Quang Ninh Province in the north to the north
central Quang Binh Province. A few of these are cave sites, but most are shell
midden and dune sites. Located on the beach and along the shore inland region,
these sites contain the characteristics typical of Vietnam’s Neolithic maritime
cultures such as GaiBeo, Ha Long, Da But, Quynh Van, BauTro and others.
Remains of the GaiBeo Culture have mainly been found on the Beibu Gulf coast
of Cat Ba Island in Haiphong Province. These sites are usually 2 to 6 meters above
sea level. GaiBeo Site is the only place where deposits from GaiBeo Culture have
been excavated. This site covers 18,000 square meters and stands 6 to 7 meters
above sea level. The strata deposits here contain a large number of bones from
132 Z. Li
marine fish, turtles, and the shells of largely marine oysters and scallops. Many
terrestrial animal remains were also found at the site such as masked civet, brown
bear, deer, elk, wild boar, sheep, elephant and monkey. Most of the stone artifacts
are chipped stone implements, including chopping tools, tray shaped implements,
pointed tools and hammers, among others. The polished stone implements include
oblong shaped axes and shouldered axes, all of which are sharpened only at the
blade. Only a small amount of pottery was found, made from a coarse paste and
shaped into thick, simple types. This included some pottery with a wide open
mouth, round body and round bottom and others with a straight mouth and round,
flat bottom. Most of pottery has plain surfaces, and some are decorated with simple
patterns such as the corded pattern, the shell rim printed pattern, the basket pattern
or the scratched line pattern. A chronological study of the GaiBeo Site shows that it
contains two different cultures, the GaiBeo Culture in the early period and the Ha
Long Culture in the late period (Tấn 1998).
Sites associated with the Ha Long Culture have been found mainly on the coastal
islands and low mountains of Quang Ninh and Haiphong Provinces on the northeast
coast of Vietnam. They include shell middens, sand dunes and cave sites, although
the most common type was the beach sand dune site. Excavated Ha Long Culture
sites include GaiBeo, Thoi Gieng and Hien Hao, and the associated artifacts consist
mainly of stone implements and pottery. The stone implements fall into two groups,
chipped and polished. The small number of chipped stone implements included
pointed tools, hand axe type implements and triangle artifacts. The polished stone
implements found here include different shapes of axe including trapezoidal, rect-
angle, triangle, shouldered, and stepped adzes as well as shouldered stepped adzes,
but also net weights, rings, slotted rings, concave hammers and burrs. The pottery
of Ha Long Culture is mainly made from a reddish paste mixed with crushed shell
and fine sand, either undecorated or with a corded pattern or a corded pattern mixed
with a wave or S-pattern. Pottery shapes include Fu (釜) cooking ware with an open
mouth and round bottom, long necked pots with ring feet, bowls with flat bottoms,
bowls with ring feet, and plates with pentagon shaped mouths. Some of the pottery
also has spouts and ears. A few tombs with mostly flexed burials and squatting
burials were unearthed here as well (Tấn 1998).
The sites associated with the Da But Culture are mainly located on the coastal
plains of Thanh Hoa and Ninh Binh Provinces in the lower reaches of Ma River
Valley. Seven shell middens and dune sites have been discovered in this region,
including Da But Site, Con Co Ngua Site, Lang Cong Site, Go Trung Site and
others. These sites are generally located on small mountains next to the water. The
stratigraphic deposit measured up to 5 meters thick and contained a large quantity
of clam shells, a few freshwater shells, and terrestrial animal bones that were
discarded following human consumption. Cultural artifacts were also abundant,
including pottery, stone implements, and objects of bone and shell, and a large
number of tombs have been found. The stone implements included plate-shaped
chipped tools, short axes and tortoise shell shaped tools. The polished stone
7 Early Maritime Subsistence and Adaptive Ocean Cultures … 133
implements included axes with polished blades, trapezoid shaped axes, perforated
round and flat artifacts, net weights, stone pestles, millstones and grinding rods.
Bone artifacts included fishing harpoons, arrowheads and awls. Most of the pottery
was made from coarse grey brown sandy paste, with basket and corded patterns.
The most common pot type was Fu (釜) cooking ware with varied rim shapes and
round bottoms. The tombs were characterized by flexed burials, and the subsistence
of the region’s inhabitants appears to have been focused on collecting and fishing
for aquatic products, especially shellfish including freshwater snails, mussels,
oysters, clams, in addition to marine scallops, clams, fish and tortoise (Ruan 2006:
341–346).
Deposits from the Quynh Van Culture have been found primarily in the long,
narrow coastal area of NgheAn and Ha Tinh Provinces, and were particularly
prominent on the gulf beach around QuynhLuu County in Nghe An Province.
These sites are mostly shell middens with large quantities of scallop remains,
although they also contain the remains of clams, snails, oysters, and large numbers
of terrestrial animals. The stratigraphic deposit varies between different sites,
generally measuring from two to five meters thick but sometimes as much as seven
meters thick. These sites are located primarily on terraces two to four meters above
the surrounding plains and can cover anywhere from a few thousand to 10,000
square meters. Cultural artifacts found here include stone and bone implements and
pottery. The stone implements are both chipped and polished, although the chipped
ones were most prominent. These chipped stone implements are made from beach
pebbles and include large chopping tools, plate shaped tools, elliptical shape tools,
triangle shape tools, tortoise shell shape tools, axe shaped tools, pointed tools,
scraping tools and stone flakes. A small number of rectangular or trapezoid
blade-polished axes, shouldered axes, grinders, hammers, anvils and burrs were
also found. Most of the pottery was made from coarse black, thick paste mixed with
sand using simple techniques. Their decoration patterns are minimal and most
frequently the vessels were plain, although some had corded patterns or line and
basket patterns. The shapes of the pottery include both pointed bottom and round
bottom, the former being the most common. Large pointed bottom pots with line
patterns are representative artifacts of the Quynh Van Culture. The bone artifacts
from here are mainly made from fish bones and include bone needles, conical tools,
concave chisels and knives. A small number of shell tools made from clam or oyster
shells were also found, including knives. Many tombs were also unearthed at the
Quynh Van Site. They were characterized by round pits, single flexed burials and
squatting flexed burials, and typically included a few funerary artifacts such as
stone tools and perforated shell ornaments (Tấn 1998).
Evidence of the BauTro Culture is mainly found in the coastal areas of NgheAn,
Ha Tinh, and Quang Binh Provinces, including shell middens and sand dune sites.
The shell middens of BauTro Culture are similar to those of the Quynh Van
Culture, although the BauTro middens are larger than those of Quynh Van. The
BauTro cultural layer is 3 to 5 meters thick and contains deposits of marine fish,
134 Z. Li
shellfish, and various kinds of terrestrial animal bones. The stone artifacts found
here include some chipped tools and large quantities of polished stone tools such as
various axes (round blade axe, square axe, stepped axe and shouldered axe) and
stone ornaments like rings, tubular beads and slotted rings. The pottery is mostly
made from fine sandy paste with scratching and stamped patterns, some of which
are burnished with red slip. The bone artifacts are mainly conical tools made of fish
bones, while the tombs from these sites contain flexed squatting burials (Tấn 1998).
The archaeological investigations and research offer some idea of the dating and
preliminary chronology of early maritime cultures along the coast of the Beibu
Gulf. The dating of the sites along the Vietnam coast is complete and fairly certain,
but the dating of those along the coast of southern China remain uncertain with the
exception of the C14 dates obtained for the Liyudun Site in Suixi, Guangdong.
Large numbers of early maritime sites have been excavated on the coast of
Vietnam, and most of these have been well-researched. Thus, the cultural conno-
tations and chronological sequences of these sites are relatively well understood.
Most of these sites have been dated using C14, and the earliest date comes from the
Da But Site, at 6540 ± 60 BP. The GaiBeo culture was dated to 6480 ± 40 BP,
while other cultures including Ha Long, Quynh Van and BauTro were dated to
about 5000 BP. The cultural content of Da But is similar to the Dingshishan (顶蛳
山) Culture in terms of burial customs, industries of stone tools and pottery, and the
types and decoration patterns of pottery. In particular the straight neck pots (or Fu
cooking ware) with basket patterns are highly similar to those of the Second and
Third Phase of the Dingshishan Culture. According to these dates, the earliest
period of Da But Culture began around 7000 BP. Further typological comparisons
of the cultural deposits in northern Vietnam and southern China also indicates that
the early maritime cultures along the coast of the Beibu Gulf date to between 7000
and 4000 years ago. Both GaiBeo and Da But Cultures coincided with this shift,
dating to about 7000–4000 years ago, while Ha Long Culture has been dated to
about 5000–4000 years ago; Quynh Van culture to about 5500-4000 years ago; and
BauTro Culture to about 4500–3500 years ago.
Liyudun Shell Midden is the only coastal site on the Beibu Gulf of China that has
been formally excavated. According to the analysis of stratigraphic deposits and
excavated material, the cultures associated with the Liyudun Site can be divided into
four different phases. Pottery from the First Phase is made from dark brown or red
sandy paste and has deeply impressed basket patterns and thick and fine corded
patterns. The coarse nature of the paste and low mode of firing for this pottery is
similar to pottery from the Second Phase of the Dingshishan Site (顶蛳山), in
Guangxi, which dates to 8000 BP (GXAT AICASS 1998). Pottery from the Second
7 Early Maritime Subsistence and Adaptive Ocean Cultures … 135
Phase is made from reddish brown coarse sandy paste and has burnished corded
patterns. The main type for this pottery is bowl shaped Fu (釜) cooking ware that
resembles the Type 1 Fu (釜) cooking ware found in the Third Phase of the
Dingshishan Site, whose C14 dates include 5160 ± 100 BP, 4820 ± 100 BP, and
5050 ± 100 BP according to charcoal and shell collected at the site. Pottery from the
Third Phase of the Liyudun Shell Midden is mainly comprised of fine paste bowls
with red slip, and some sandy paste wares which is similar to that of the Fourth
Phase. Pottery from the Fourth Phase are characterized by sandy coarse paste Fu (釜)
cooking ware and reddish brown fine paste ring foot plates. Generally, artifacts from
the Third and Fourth Phases are similar to cultural features from the Shang (商)
Dynasty in the Pearl River Delta, but the C14 date for these deposits is 4660 ± 100
BP (GDICRA 2015). Thus, the dates for the First and Second Phases of the Liyudun
Shell Midden are similar to that of the Third Phase of the Dingshishan Site in
Guangxi, while the date of the Fourth Phase is closer to 4500 BP.
Due to the relatively small areas excavated and limited number of artifacts
collected, the dates for Yapushan, Malanzuishan, Beijiaoshan, Sheshan and other
shell middens on the Guangxi Coast of Beibu Gulf remain uncertain. The earliest
excavations by the Guangdong Provincial Museum suggested that these cultures
belonged to the early Neolithic period of the Guangdong region and emerged later
than the Xiqiaoshan (西樵山) Culture (GDPM 1961: 644–688). Recently, some
scholars have proposed that the cultural features of these sites are actually similar to
those of the Chenqiaocun and Shiweishan Shell Middens on the eastern coast of
Guangdong, as well as other shell middens on the coast and in the islands of Fujian.
These sites share the same period as the Late Phase of the Dawan (大湾) Culture in
the Pearl River Delta (Zhang and Hung 2008: 415–434). The artifacts found at these
sites included chipped stone implements and polished stone implements that
resemble those from the Da But, GaiBeo, and Ha Long Cultures on the Vietnam
coast, as well as those from the Liyudun Site on the Leizhou Peninsula. The paste
and type of pottery is also similar to that of Da But Culture and the Second Phase of
the Liyudun Shell Midden, while the bone ornament made of fish vertebrae is the
same as those from the Second Phase of the Liyudun Shell Midden. Still, the stone
shouldered axe, adze, and some pottery pots with opening rims and narrow necks
are more similar to those of the Third Phase of the Liyudun Shell Midden and
GaiBeo, Quynh Van and Ha Long Cultures. As a result, the date for shell middens
along the Guangxi Coast of Beibu Gulf is about 6000 BP, an age that falls between
the GaiBeo Culture, the Third Phase of the Liyudun Shell Midden, and the Quynh
Van Culture.
The preliminary analysis and cultural comparison of artifacts from the sites
along the coast of Beibu Gulf therefore dates the early maritime culture in this
region to about 7000-4000 BP.
136 Z. Li
Due to its geographical characteristics as a semi-closed sea bay, the Beibu Gulf is a
relatively independent marine unit. Although early maritime cultural sites extend
throughout the vast coastal areas and developed continuously through a long his-
torical period, these cultures also indicate a series of discrete regional characteris-
tics. These include the following:
Maritime settlement pattern. Most of the sites are generally located on small hills
or terraces associated with the beach and sand dunes of the sea gulf, though some
are located on offshore islands and sand dune terraces along the river estuaries. This
represents geographic evidence for the oceanic environments in which these
Neolithic settlements were established.
Shell middens and sand dunes. Site types include coastal shell middens and dune
sites, though most are coastal shell middens. These contain large amounts of
mollusk shells, marine and terrestrial animal bones that indicate the specific com-
ponents of the prehistoric diet.
Chipped stone tools. Many findings such as pottery, stone implements, bone and
shell artifacts were collected, though the chipped stone implements were the most
common. The main types of chipped stone tools are Haolizhuo pointed choppers,
chopping tools, hand axes, stone hammers and net weights. Among these the
Haolizhuo, a pointed tool for cutting open oyster mollusks, is the tool most char-
acteristic of this oceanic region.
Maritime subsistence. The aquatic and terrestrial animal remains from these sites
reveal a mixed subsistence pattern of fishing, hunting, and gathering that relied
heavily on the use of stable marine fish and shellfish mollusk resources, the col-
lection of edible plants and fruits, and the hunting of terrestrial animals. No obvious
agricultural remains have been discovered at these sites. The shell middens, in
particular, illustrate a maritime subsistence strategy that depended primarily on
marine fish and shellfish resources, with sea water mollusk shells and marine fish
bones appearing in the heaviest concentrations. In addition, the C and N stabilized
isotopes from human bones collected at the Liyudun Site also revealed marine
species to be the primary dietary source, while terrestrial animals served as a
supplementary food source (Hu et al. 2010: 264–269).
The oceanic adaption to fishing and gathering rather than marine aquaculture.
The maritime subsistence of these early coastal settlements on the Beibu Gulf
illustrates an adaptation to the use of marine resources rather than the development
of aquaculture or marine production. Most of these sites are next to the shallow
water of the sea bay, which was convenient for collecting coastal plants and ani-
mals. The main marine resources included both offshore mollusk and marine fish.
7 Early Maritime Subsistence and Adaptive Ocean Cultures … 137
Judging from the prevalence of small fish bones and the discovery of a few of small
net weights, these Neolithic humans were carrying out offshore fishing and gath-
ering rather than engaging in the production of sea food or marine aquaculture.
The beach household and flexed burial. A single house remnant was found in the
First and Second Phase of the Liyudun Site, while a series of tomb remnants with
flexed burials were also discovered at the Liyudun Site and at sites associated with
the Da But, Ha Long, Quynh Van, and BauTro Cultures. These remains indicate the
location of households and the broader pattern of maritime settlement next to sea
environments.
Dating and sustained development. Maritime settlements along the coast of the
Beibu Gulf lasted for about 3000 years, beginning around 7000 BP and disap-
pearing around 4000 BP.
7.5 Conclusion
Early maritime cultural sites along the coast of the Beibu Gulf are mainly
distributed along the north and west sides of the Beibu Gulf, especially on the
northern coast of Vietnam and the southern coast Guangxi, China. The maritime
cultural heritage of this region resides primarily in coastal shell middens and sand
dune sites.
Early maritime culture along the coast of the Beibu Gulf arose and developed out
of the hunting and gathering cultures present in the surrounding areas, especially
the northern region of Vietnam and Guangxi, during the Late Paleolithic and Early
Neolithic Ages. As these early hunter-gatherers migrated from the inland regions to
the coastal areas of the Beibu Gulf, they adapted to their new environments. Thus,
the chipped stone tools from the Da But, GaiBeo, and Quynh Van Cultures and the
Fangcheng Shell Midden of Guangxi are all similar to the stone implements
technique observed in the Hoabinhian Culture. For example, this pattern applies to
the almond shaped chopper, pointed chopper, tortoise shell shape tools, oblong
shaped tools, short axe, plate shaped chopper, and chopping tool. Polished stone
implements from the Da But, GaiBeo, and Quynh Van Cultures and the Fangcheng
Shell Midden are also similar to those from the Dingshishan Culture in the southern
region of Guangxi, while their burial patterns are similar to those of the Hoabinhian
Culture and the Dingshishan Culture. The composition, shapes and patterns of
pottery from the Da But Culture, Liyudun Culture and Fangcheng Shell Midden are
also similar to that of the Dingshishan Culture. Thus, the Da But Culture appears to
have been a continuation of the Dingshishan Culture (Zhang and Hung 2008: 415–
434). The existing evidence makes it clear that the Early Stage of the coastal
maritime culture in the Beibu Gulf developed from the Dingshishan Culture. The
earliest inhabitants of these coastal settlements may therefore have originated
mainly from hunter-gatherers of the inland region, including the Dingshishan
Culture. Based on the timing of the emergence of stone shouldered axes and
138 Z. Li
stepped adzes in the Beibu Gulf, cultural influences and migration to this region
from the Pearl River Delta and southeastern coast of China appear to have occurred
during the Late Period, via coastal or sea routes.
Most of the Neolithic sites along the coast of the Beibu Gulf from 7000-4000 BP
show evidence of a maritime culture whose main subsistence strategies were marine
fishing, gathering and hunting. The early maritime economy in the Beibu Gulf
illustrates a dependence on the utilization of marine resources, rather than aqua-
culture or marine production. As agriculture subsequently spread to the south coast
of China and into southeast Asia, adaptive maritime subsistence patterns gradually
changed. Due to this shift from marine based subsistence to a more agricultural
culture that spread from north to south, the early maritime culture of the Beibu Gulf
region came to an end around 4000 BP.
References
Hu, Y. (胡耀武), Li, F. (李法军), Wang, C.(王昌燧), & Richards, M. P. (2010). Analysis of the C
and N stabilized isotopes of Liyudun Ruin in Zhenjiang City: A preliminary study on the
lifestyle of the Neolithic Age peoples of South China. Acta Anthropologica Sinica, 29(3), 264–
269 (《广东湛江鲤鱼墩遗址的C、N稳定同位素分析:华南新石器时代先民生活方式初
探》,《人类学学报》,2010,29(3), 264–269).
Ruan, W. (阮文好). (2006). The Da But Culture of Vietnam. In CASS (Ed.), Prehistoric
archaeology in South and Southeast Asia (pp. 341–346). Beijing: Cultural Relic Publishing
House (《越南的多笔文化》,中国社会科学院考古研究所:《华南及东南亚地区史前考
古》,北京,文物出版社,2006年。P341–346).
Tấn, H. V. (Ed.). (1998). KhảocổhọcViệt Nam I. HàNội: Nhàxuấtbản khoahọcxãhội.
Zhang, C.(张弛), & Hung, H.(洪晓纯). (2008). Neolithic fishing, hunting and gathering culture in
South China and its adjacent areas. In School of Archaeology and Museology of Peking
University (Ed.), Archaeological Research VII (pp. 415–434). Beijing: Science Press (《中国
华南及其邻近地区的新石器时代采集渔猎文化》,《考古学研究》(七),北京,科学出版
社,2008:415–434).
Chapter 8
The Late Neolithic to Early Bronze Age
on the Northeastern Coast of Vietnam
Abstract The peoples who inhabited the Northeast Coast of Vietnam during the
Late Neolithic are associated with the Hoa Binh Culture and other subsequent
cultures over time, including the Soi Nhu Culture (ca. 18,000-7,000 BP) and the Cai
Beo Culture (ca. 7,000-4,500 BP). Out of these came the Ha Long Culture
(ca. 5,000-3,500 BP), which developed throughout the Northeast and illustrates
cultural characteristics unique to the region. In particular, the discovery of a group
of jade ornament manufacturing workshop sites at a later stage of this culture has
contributed to the understanding of the Bronze Age transition in this area.
The Northeast Coast of Vietnam extends through Hai Phong and Quang Ninh
Provinces before reaching the northern border with China. This area, also known as
the Ha Long Bay UNESCO World Heritage Site, contains thousands of sea islands,
sand dunes and limestone caves. French archaeologist Madeleine Colani and
Swedish geologist and archaeologist Johan Gunnar Anderson conducted the first
systematic research of the region’s geology and archaeology in the late 1930s.
Between 1937 and 1938, Coloni discovered a series of prehistoric sites in the
islands of Hai Phong and Quang Ninh, conducting excavations at the Cai Beo Site
(Cat Ba Islands, Hai Phong Province) and the Ha Giat Site (or Ha Yart Site of Cai
Bau Island, Quang Ninh province) (Colani 1938). Meanwhile, Anderson surveyed
the Ha Long area in 1938 and discovered 7 caves containing archaeological
remains. He subsequently excavated the Xich Tho, Ngoc Vung (Danh Do La), and
Dong Mau (Dong Mang) Sites of Quang Ninh Province (Anderson 1939). Based on
the general characteristics of the excavated cultural material, Anderson proposed
K. D. Nguyen (&)
Vietnam Association of Archaeology, Hanoi, Vietnam
e-mail: kimdzungkc.kdn@gmail.com
the existence of a Neolithic culture named Danh Do La. These sites were later
categorized as belonging to the Ha Long Culture by Vietnamese archaeologists
(Nguyen 2003).
In 1954, Edmond Saurin conducted research on the Giap Khau Site, near the
border of Vietnam and China (Saurin 1956). These early foreign scholars believed
that the prehistory of Ha Long Bay illustrated a cultural association with the Bac Son
Culture, which exhibits archaeological continuity with the Hoa Binh Culture of the
Bac Son Karst Mountains, Lang Son Province. They also drew comparisons to the
Bac Son Culture, claiming that the stone tools in the caves of Ha Long Bay are
similar to Bac Son pebble stone tools. Therefore, they concluded, the age and
features of the Ha Long Bay caves offer a deeper understanding of the Bac Son
Culture. In addition, foreign scholars established a system of prehistoric develop-
ment for Ha Long Bay that progressed through two main cultural stages. Anderson
argued that there were two types of archaeological sites representing two continuous
stages of cultural development: Early Stage deposits in the caves, and Late Stage
deposits in open-air sand dune sites that include shouldered axes and edges, grooved
grinding stones and soft pottery whose temper contains large amounts of shell.
Since the 1960s, Vietnamese archaeologists began carrying out much of the
fieldwork and excavations on the Northeast Coast. A number of important surveys
were conducted starting in 1964, most notably the excavation of Soi Nhu Cave in
1967. This investigation introduced crucial new information on the prehistory of Ha
Long Bay, gathered from a rich assemblage of stone tools, fresh-water shell cultural
layers, animal teeth and bones, and human burials with maintained skulls and
skeletons. Other caves such as Ang Giua, Bai Tu Long, and Hang Duc were
subsequently excavated by Vietnamese archaeologists from 1980–1990.
Featuring dozens of archaeological excavations, these investigations led to the
creation of a new model for cultural development in this area based on three
periods: (1) Soi Nhu Culture (contemporary with Hoa Binh and Bac Son), whose
evidence is concentrated in 34 limestone caves containing shell layers and pebble
stone tools; (2) Cai Beo Culture, based on deposits from 5 sites; and (3) Ha Long
Culture, based on findings from 38 open air sites and the 4 sites of the Trang Kenh
Group. Large amounts of stone artifacts, animal bone tools and pottery were found
at all of these sites, including both the shell deposits of caves and the sand dunes.
The Soi Nhu, Cai Beo, and Ha Long cultural environments were all close to the sea,
while the limestone mountains offered only a limited variety of stone for making
tools. This scarcity of lithic material led to the use of all existing stone, even
limestone, for the manufacture of tools. The resulting stone tools are one of the
defining characteristics of archaeological finds from the northeastern coast of
Vietnam.
This paper focuses specifically on the Late Neolithic to Early Bronze Age sites
of the Northeast Coast, including the Ha Long Culture and the Trang Kenh Group
Sites. In particular, the cultural interaction between Ha Long Culture and the Trang
Kenh Group with other ancient peoples of North Vietnam and South China is one
of most prominent characteristics of the Late Neolithic in North Vietnam (Fig. 8.1).
8 The Late Neolithic to Early Bronze Age … 143
This culture has been identified in dozens of limestone caves, and represents the
First Stage of Ha Long Bay prehistory. The Soi Nhu Cave of Cam Pha District in
Quang Ninh Province contains some of the most characteristic features of this
culture and consists of 3 small areas: lower, middle, and upper. The middle area,
excavated in 1967, contained stone flaking tools, edge-polished pebble axes, and
grinding stones. The cultural layer from this area contains fresh-water mollusks
such as Cycrophorus, Melania, and a few animal bones and teeth as well as marine
mollusks (Nguyen 1997: 16–28; Trinh et al. 2000). The stone artifacts are made
from coarse chipped pebbles or limestone. Very few pottery fragments have been
found in the caves (Ha and Nguyen 1998).
Toward the end of the period associated with Soi Nhu Culture, some coarsely
made chipped pebble axes and flaked pebble scrapers were also found. Pebble and
limestone tools exhibiting the same features as those from Soi Nhu Cave have also
been excavated in many of the caves on Cat Ba Island in Hai Phong Province,
including Ang Ma, Mai Da Ong Bay, Gieng Nghoe, Tien Duc, and Ang Giua. In
research conducted on Cat Ba Island from 1999 to 2001, the author and her col-
leagues discovered 45 caves characterized by cultural layers full of shells from
fresh-water mollusks such as Cycrophorus and Melania, in addition to a few marine
mollusks (Nguyen 2002; Nguyen et al. 2005: 541–560).
On both the mainland and among the islands, the sites associated with the Soi
Nhu Culture only appear in limestone caves, and have never been found at open-air
sites. According to research conducted by Ha Huu Nga (1998), Soi Nhu Culture
can be divided into three periods: an Early Period dating back to 18,000 BP,
144 K. D. Nguyen
Five sites associated with the Cai Beo Culture have been investigated, including Cai
Beo, Ao Coi, Va Bac, Giap Khau, and Ha Giat. The Cai Beo Site is a cornerstone of
the Cai Beo Culture. The people from this culture occupied open-air sites, and
based on the Cai Beo Site the cultural deposit can be divided into two periods: Early
(7000 BP) and Late (5000 BP).
Cai Beo Site was discovered by Colani in 1938, during her survey on Cat Ba
Island. At the time, only a small-scale test pit was excavated. In 1956, Saurin
published his own studies on two other sites known as Giap Khau and Ha Giat
(Hayat). Then from 1974–1986 the Cai Beo Site was excavated four times by
Vietnamese archaeologists, covering a total area of 370 square meters (Luu et al.
1983; Nguyen et al. 1986…).
In 1999, the author conducted a survey on Cat Ba Island and discovered two
more sites belonging to Cai Beo Culture, Ao Coi and Va Bac (Nguyen et al. 2002).
A deposit measuring 2.5–3.5 meters thick from these investigations indicated that
the Cai Beo Site contains three cultural layers that can be divided into two stages, or
two separate cultures: Cai Beo and Ha Long.
Archaeological material from the Cai Beo Culture appears at a depth of 1.8 meters
and continues down to the bottom of the Cai Beo Site. This assemblage includes a
great number of stone tools including choppers, short axes, pointed tools made from
chipped pebbles, chipped scrapers, and partially polished axes and chisels. Simple
flaking techniques were applied in the manufacture of these tools, and they had not
been thoroughly retouched. The most common tools found here were intact chipped
pebbles which had been used as pointed tools (Fig. 8.2). The handmade pottery
excavated from this stage featured coarse, thick styles with small amounts of laterite
and quartz in the temper. The flat bottoms of these vessels were marked by
basket-impressions, and most either had no decoration or a simple “comb by a
paddle” decoration, in addition to some cord-marks on their bodies.
This stage dates to 7000-6500 BP or earlier and is associated with the Late Soi
Nhu Culture, or pre-Ha Long Culture.
8 The Late Neolithic to Early Bronze Age … 145
Fig. 8.2 Pointed tools and pottery from the Cai Beo Culture
The Ha Long Culture at the Cai Beo Site appears starting around 1.8 meters below
datum up to the surface layer. Fully-ground stone implements found in this
assemblage include quadrangular axes, quadrangular adzes, chisels, shouldered
axes, shouldered adzes, and stepped adzes. Especially abundant are the grinding
stones with grooves called “Ha Long Marks,” a particular object used to identify Ha
Long Culture sites (Fig. 8.3).
Pottery from this stage includes two types: soft, whose temper contains many
small shells; and hard, which is made from clay mixed with sand. Appliqué and
incised decoration techniques were commonly used on these artifacts.
This stage dates from 5000 to 4000 BP and is known as the Ha Long Culture, a
Late Neolithic culture in Vietnam. The most important findings from the Cai Beo
Site are hundreds of kilograms of very large fish bones and marine shells that
illustrate the emergence of a sea- oriented economy.
146 K. D. Nguyen
Fig. 8.3 Stone tools from the Ha Long Culture (left: main types of stone tools; right: pointed tools
from the Bai Cat Don Site)
Evidence of the Ha Long Culture appears at nearly 50 open-air sites and the upper
layers of several caves (Fig. 8.4). Three important sites associated with this culture
and previously excavated are Danh Do La (Ngoc Vung), Xich Tho and Dong
Mang. Anderson used the name Danh Do La to describe all sites with similar
cultural features (Anderson 1939: 104). These sites became associated with Ha
Long Culture due to the large number of similar sites later excavated by Vietnamese
archaeologists and identified as characteristic of Ha Long Culture (Figs. 8.5, 8.6).
Ha Long cultural sites appear across a broad range of mainland and island areas.
In some caves, the upper layer of cultural deposits are often associated with the Ha
Long Culture due to the presence of the same stone tools and pottery also found at
Ha Long cultural sites on sand dunes. In all more than one hundred Ha Long
cultural sites have been discovered, including dozens that have been excavated
(Nguyen Khac Su 1975, 1986, 1997; Ha 1998).
Evidence from the Early Stage of Ha Long Culture has been found in the Mong Cai
and Tien Yen areas of Quang Ninh Province at sites such as Thoi Gieng, Go Mung,
Quat Dong Nam, and Hon Ngo.
The Thoi Gieng Site (5500-4500 BP) is one example, and contains all the
elements of cultural transition from Cai Beo to Ha Long Culture. At a later stage,
the prehistoric population associated with Ha Long Culture lived throughout the
8 The Late Neolithic to Early Bronze Age … 147
Fig. 8.4 Map showing sites associated with the Ha Long Culture
Fig. 8.5 Stone and bone tools from the Ha Long Culture found at the Hon Hai Co Tien Site in
2004
148 K. D. Nguyen
Northeast Coast, especially on the bigger islands such as Cat Ba, Cai Bau, and Van
Don. Some 48 sites, including stone workshops and local cemeteries, have been
identified as late stage of Ha Long culture. Some of the C14 dates associated with
this stage of Ha Long Culture are from: Ba Vung, 4820/4520 ± 80 BP and
4470/4450 BP (Chun 2003); and Bai Ben, 4070 ± 50 BP and 3900 ± 80 BP
(Nguyen 2002).
Similar to the late layer of the Cai Beo Site, stone artifacts collected from Ha
Long cultural sites consisted of stone quadrangular axes and adzes, as well as
shouldered axes and adzes. The asymmetrically shouldered adze, in particular, was
found at many sites. Many of the tools from this assemblage are small. Ha Long
Marks, or the grinding stones made of fine sandstone blocks criss-crossed by
narrow grooves and originally identified by Colani, are widely found throughout the
Ha Long sites. More than 1000 artifacts have been found at the Ba Vung Site in
Van Don District, Quang Ninh Province, 500 of which belong to this particular
typology with Ha Long Marks (Nguyen et al. 2001).
All the Ha Long sites are located on fine coastal sand dunes, contain cultural
deposits of more than a meter thick, and extend over 1000 square meters each. As
for the cave sites, almost no distinguishable Ha Long Culture layers are present in
their stratigraphical sequence except for a thin layer on the surface that contains
characteristic artifacts such as pottery, stepped and shouldered stone tools, Ha Long
Marks, and pointed tools.
8 The Late Neolithic to Early Bronze Age … 149
At the Bai Ben Site on Cat Ba Island, “Ha Long Marks” and pointed tools
comprise 30% of the stone tools found. Pointed tools made from chipped pebble
appear beginning in the Early Stage of the Cai Beo Culture and continuing
throughout the existence of the Ha Long Culture. This particular type of tool may
be closely related to the exploitation of coastal shellfish such as oysters. By con-
trast, the Ha Long Mark is only associated specifically with the Ha Long Culture.
The great majority of the pottery from Ha Long sites is soft pottery made from
clay tempered with shells and made both by hand and on potters’ wheels. The body
section of the pottery is often thin and decorated primarily with cord-mark patterns,
though some are decorated with dotted lines and have a simple everted rim. Special
incised designs on the outer surface of the pottery combine linear appliqué on the
shoulder part with wavy incised motifs and dotted holes on the ring feet.
Late Ha Long Culture (4000-3500 BP) had close links with the contemporary
peoples living in the Hong River and Ma River Deltas in northern Vietnam. These
included the Phung Nguyen Culture, from the Hong River Delta and the Man Bac
Site on the coast of Ninh Binh Province, and the Hoa Loc and Con Chan Tien from
the Thanh Hoa coastal area. The excavation of sites associated with the Ha Long
Culture revealed many potsherds and nephrite ornaments that originally came from
these other cultures, suggesting a close relationship between them.
The ancient inhabitants of Viet Nam used nephrite for working tools and ornaments
for thousands of years, and nephrite objects have been found with the remains of
many Neolithic cultures of North Vietnam dating to between 5000 and 3500
BP. Nephrite manufacturing workshops are most heavily concentrated on the
Northeast Coast, and are well represented by the Trang Kenh Group. This group
includes 3 sites: Trang Kenh (Fig. 8.7), Dau Ram (Fig. 8.8) and Bo Chuyen, which
contains the most important Trang Kenh Workshop Site for not only this group but
the Neolithic and Early Bronze Age manufacturing of nephrite in Vietnam, as a
whole. The Trang Kenh Site is situated in Hai Phong Province, while Dau Ram and
Bo Chuyen are located in Quang Ninh Province. It is easy to envision the contact
between these three sites and the importance of the river system as well as the
limestone mountains surrounding the area.
The excavated material indicates that the Trang Kenh Site is one of largest
nephrite workshop sites in Northern Vietnam. Thousands of broken pieces of cut
nephrite have been unearthed, material closely related to thousands of sawing traces
150 K. D. Nguyen
Fig. 8.7 Excavation at the Trang Kenh Site, 1996 (photo courtesy of Tang Chung)
(Fig. 8.9) and large quantities of well preserved nephrite flakes, blanks, and discoid
cores from the cultural layers of this site (the cultural layers in all sites are 2 meters
deep). An excavation in 1996 unearthed 28,000 half-finished objects, discarded
objects with sawing marks and flakes in an area of just 21 square meters. This level
of concentration suggests that the people inhabiting Trang Kenh were well versed in
making nephrite ornaments. The nephrite discoid cores (Fig. 8.10) appear to have
been most often drilled into rings to make a series of large and small bracelets.
The remains of discoid cores in the Trang Kenh workshop group are classified
into ten different types based on their diameter (ranging from 0.3 to 5 cm), thick-
ness, and cutting trace. The apparent manufacture of rings by cutting them from a
discoid core is the most intriguing discovery. Traces of sawing can clearly be seen
on both sides of hundreds of nephrite blanks, some of which were made into fine
bangles, demonstrating the prevalence of sawing technology at the Trang Kenh
workshop group. This popular practice of cutting with a sandstone saw in Neolithic
Vietnam stands in contrast to the string technique widely used to cut nephrite in
South China in Yunnan prehistory.
Careful study of the circular lines and other traces left on the discarded material
suggests the existence of two different ways to cut nephrite discoid cores: (1) a
pottery making method using the rotation of a wooden wheel and rotary cutting by
the jasper drill with bamboo handle; and (2) a turning lathe method. There are, of
course, other ways to drill a discoid off of a nephrite blank; but the remaining traces
of cutting on the drilled objects suggest the specific use of these two techniques.
Geological analysis has revealed that the nephrite used at the Neolithic work-
shops of the Trang Kenh Group is tremolite nephrite, a type formed by a meta-
morphic process closely related to carbonate rock fields. The prehistoric
populations associated with the Trang Kenh Group Sites settled among large
limestone mountains, suggesting the possibility that this area became their source
for tremolite nephrite. In addition, the location of their Neolithic workshops near
transportation routes and rivers indicates that the river systems around these sites
played an important role in their operation.
8.6 Conclusion
Based on the archaeological evidence, two main points can be drawn about the Late
Neolithic to Early Bronze Age inhabitants of Vietnam’s Northeast Coast.
First, the Northeast Coast was one of the most important areas for Late Neolithic
and Early Bronze Age populations of Vietnam and South East Asia. Prehistoric
artifacts from this area include stone tools and pottery that support the existence of
frequent trade and cultural exchanges across peoples, and the Ha Long Bay area
appears to be a clear center for “migration flows” in this region.
The Ha Long Culture emerged directly from a blending of the Soi Nhu and Cai
Beo Cultures. The stepped adze and axe, as well as the shouldered axe and adze, are
found over vast regions of Northeast Vietnam, Taiwan, the Philippines, and in
Guangdong and Yunnan, China. Smaller numbers have also been found in Laos,
Thailand and Cambodia. Above all, these types of artifacts are most strongly
associated with the Ha Long Culture, from which a typology has emerged based on
early to late forms. The spread of these particular objects throughout such a large
area illustrates the broad connections between the people of the Ha Long Culture
and other communities outside of their own settlements.
8 The Late Neolithic to Early Bronze Age … 153
Second, Ha Long Culture is the earliest one in which nephrite rectangular axes
have been found in Vietnam. At the Cai Beo Site, nephrite with evidence of cutting
has also been discovered in the Late Stage of the period associated with Ha Long
Culture (Nguyen 2005, 2009). Throughout its existence, this culture is characterized
in part by the presence of nephrite ornaments such as bracelets and beads.
For example, nephrite objects have been found at 70 archaeological sites asso-
ciated with Phung Nguyen Culture in the middle of the Hong River Valley. In
addition, pre-Dong Son Sites in Thanh Hoa, associated with the Hoa Loc Culture of
the Con Chan Tien Group, and the Man Bac Site have also contained large numbers
of nephrite objects. These included a broad range of nephrite tools such as axes,
edges, chisels, and hammers, as well as and ornaments including bracelets, beads
and pendants. Among these findings, T-section bracelets are one of the most
remarkable ornaments of the Phung Nguyen Culture due to how complicated they
are. One of the most popular styles appears to have been T-section bracelets dec-
orated with 6 to 12 parallel lines on the outside (Fig. 8.11: 1).
Some of the simpler T-section rings found at the Xom Ren and Man Bac Sites
were discovered in their original articulations relative to human remains (Fig. 8.11:
2, 3), while similar T-section rings have been found at the Trang Kenh Workshop
Site (Fig. 8.12). Thus, the manufacture of nephrite tools and ornaments at the
4. Man Bac 5. PN 6. DR
Fig. 8.11 Evidence of regional cultural exchange, including: nephrite T-section ring from the
Phung Nguyen Culture (1 and 3); nephrite T-section ring from the Man Bac Site (2); and similar
decorative motifs on pottery from the Man Bac (4), Phung Nguyen (5), and Dau Ram (6) Sites
154 K. D. Nguyen
Fig. 8.12 Nephrite ornaments found at the Trang Kenh Workshop Site
Trang Kenh Group appears to have had direct links to other cultures along the
Vietnam Coast (Fig. 8.13).
In addition to this particular jade ring type T-section bracelet, a special slit jade
ring with 4 projections also characterizes the Phung Nguyen Culture of north
Vietnam (Fig. 8.12). This type of artifact continued to develop in the Dong Dau and
Go Mun Cultures, as well as the Dong Shan Culture of Vietnam’s Northeast Coast,
during the Bronze and Early Iron Age, from around 3500 to 2300 BP. Meanwhile,
the Sa Huynh Cultures of Vietnam’s South Coast are associated with the same slit
jade ring with 4 projections, as well as a similar type with 3 projections
(LingLing-O) and a style of jade earrings with animal heads at both ends
(Fig. 8.14). These findings illustrate the extent of economic and cultural exchange
8 The Late Neolithic to Early Bronze Age … 155
Fig. 8.13 Nephrite objects and grinding saws from the Dau Ram Workshop Site
Fig. 8.14 Bronze and Early Iron Age jade slit rings from the coast of Vietnam
156 K. D. Nguyen
from north to south along the Vietnam coast during the Bronze and Early Iron Age
(Nguyen 2010).
As a result, the population associated with the Ha Long Culture appears to have
communicated with the peoples of the Phung Nguyen Culture as well as those along
the coast, including the Man Bac and Hoa Loc populations in the Ninh Binh and
Thanh Hoa Provinces. In particular, one of the centers for interaction between the
Phung Nguyen and Ha Long Cultures appears to have been the nephrite workshop
site of the Trang Kenh Group, demonstrating the crucial importance of jade culture
in Vietnamese prehistory. Excavations in this area therefore confirm the exchange
of nephrite products between the Trang Kenh Group and the Phung Nguyen
Culture, as well as the significant role played by the Trang Kenh Group nephrite
workshop in the prehistory of Vietnam.
References
Anderson, J. G. (1939). Archaeological research in the Fai Tsi Long Archipelago, Tonkin. Bulletin
of the Museum of Far Eastern Antiquities, 11, 11–27.
Chun, C. W. (2003). Excavation at Ba Vung site in Quang Ninh Province (Ha Long culture). In
Vietnam Institute of Archaeology (Ed.), New archaeological discoveries in Vietnam, 2002
(pp. 123–124). Hanoi: Social Science Publishing House (Những phát hiện mới về Khảo cổ học
Việt Nam năm 2002. Hà Nội: Nhà xuất bản Khoa học xã hội.).
Colani, M. (1938). Decouvertes préhistoriques dans les parages de la Baie d’Along. Paris:
Institute Indochinois pour l’Homme.
Dang, V. N. (1968). Excavation at Soi Nhu cave site. Institute of Vietnam History (Ed.), Vietnam
Historical Research, 17, 57–61 (Khai quật di chỉ hang Soi Nhụ).
Ha, H. N. (1998). The results of archaeological research in Quang Ninh Province, early 1997. In
Vietnam Institute of Archaeology (Ed.), New archaeological discoveries in Vietnam 1997
(pp. 89–91) (Kết quả điều tra khảo cổ học đầu năm 1997 ở tỉnh Quảng Ninh).
Ha, H. N. & Bui, V. (1982). Excavation at Ang Giua Cave, Cat Ba Island. Hanoi: The Library,
Vietnam Institute of Archaeology (Khai quật hang Áng Giữa, đảo Cát Bà).
Ha, H. N. & Nguyen, V. H. (1998). Prehistory of Ha Long. Hanoi: The World Publishing House
(Hạ Long thời tiền sử. Nhà xuất bản Thế Giới, Hà Nội).
Le, T. T., & Trinh, C. (1983). Back to the Cai Beo archaeological site: results and findings. In
Vietnam National Historical Museum (Ed.), Science information (pp. 14–24. Hanoi: Vietnam
National Historical Museum (Trở lại di chỉ Cái Bèo: Kết quả và nhận thức. Thông Báo Khoa
học, Bảo tàng Lịch sử Việt Nam.).
Nguyen, K. S. (1975). Archaeological study on Cat Ba Island, Hai Phong City. Hanoi: Vietnam
Institute of Archaeology (Nghiên cứu khảo cổ học trên đảo Cát Bà, Hải Phòng).
Nguyen, K. S. (1986). The prehistoric site of Cai Beo on the northeast coast of Vietnam.
Archaeology, 2, 17–26 (Di chỉ Cái Bèo với tiền sử vùng ven biển Đông Bắc Việt Nam. Khao
Co Hoc, số 2, 17–26).
Nguyen, K. D. (1992). Stone tool manufacturing and its role in the economy of Bronze Age
Vietnam. Archaeology, 4, 12–18 (Kỹ nghệ sản xuất công cụ đá và vai trò của nó trong kinh tế
thời đại đồng thau Việt Nam, Khao Co Hoc, số 4, tr. 12–18).
Nguyen, K. S. (1997). Prehistoric maritime culture in Vietnam: a model and hypothesis.
Archaeology, 3, 16–28 (Văn hóa biển tiền sử Việt Nam: mô hình và giả thiết. Khao Co Hoc, số
3, 16–28).
8 The Late Neolithic to Early Bronze Age … 157
Nguyen, K. D. (2001). New insight from two excavations at the Bai Ben Stone Workshop Site, Cat
Ba Island. Archaeology, 4, 3–24 (Nhận thức mới về khảo cổ học Cát Bà qua hai lần khai quật
di chỉ Bãi Bến, Khao Co Hoc, số 3, tr 3–24).
Nguyen, K. D. (2002). More contributions to prehistoric research on Cat Ba Island from recent
C14 dating. In Vietnam Institute of Archaeology (Ed.), New archaeological discoveries in
Vietnam, 2001 (pp. 184–188). Hanoi: Social Science Publishing House (Từ kết quả niên đại
C14 gần đây ở một số di chỉ khảo cổ học Cát Bà, đóng góp thêm một vài suy nghĩ về tiền sử
đảo Cát Bà. Trong Những phát hiện mới về Khảo cổ học Việt Nam năm 2001, tr. 184–188).
Nguyen, K. D. (2003). Prehistoric techniques in the Ha Long Culture on Cat Ba Island:
J. G. Anderson’s discoveries and recent research. In A. Karlstróm, & A. Kállén (Eds.),
Southeast Asian Archaeology: Fishbones and Glittering Emblems (pp. 59–70). Stockholm:
Museum of Far Eastern Antiquities. Óstasiatiska Museet.
Nguyen, K. S. (Ed.). (2005). Archaeology on the Northeastern Coast, Vietnam. Hanoi: Social
Science Publishing House (Khảo cổ học vùng duyên hải Đông Bắc Việt Nam).
Nguyen, K. S. (2009). Cai Beo prehistoric site, Cat Ba Island. Hanoi: Social Science Publishing
House (Di chỉ khảo cổ học Cái Bèo, đảo Cát Bà).
Nguyen, K. D. (2010). Jade slit rings in Vietnam. In C. Tang, & C. Wu (Eds.), Southeastern
Archaeological Research, 4, 147–152 (越南出土的玉玦,Yuenan Chutu de Yujue, 《东南考
古研究》(第四辑), Dongnan KaoguYanjiu No. 4, Xiamen University Press).
Nguyen, V. H., Nguyen, G. D., Nguyen, T. L., & Dao, T. N. (1986). The report of excavation in
Cai Beo site 1986. Library of Vietnam Institute of Archaeology. (Báo cáo khai quật di chỉ Cái
Bèo năm 1986. Tư liệu lưu trữ tại Thư viện, Viện Khảo cổ học Việt nam.).
Nguyen, K. D., Trinh, N. C., & Nguyen, G. D. (2005). Archaeological survey from 1998 to 2000,
Hai Phong Province. In K. S. Nguyen (Ed.), Archaeology of the Northeastern Coast of Vietnam
(pp. 541–560). Hanoi: Social Science Publishing House.
Saurin, E. (1956). Outillage Hoabinhien à Giap Khau, port Courtbet (North Vietnam). Bulletin d’
Ecole Française d’Extrème Orient, XLVIII(2), pp. 581–592.
Trinh, N. T., Nguyen, K. D., & Nguyen, G. D. (2000). Archaeological caves discovered recently
on Cat Ba Island. In Vietnam Institute of Archaeology (Ed.), New archaeological discoveries
in Vietnam, 1999 (pp. 143–145). Hanoi: Social Science Publishing House (Những phát hiện
mới về Khảo cổ học Việt Nam năm 1999. Hà Nội: Nhà xuất bản Khoa học xã hội.).
Chapter 9
Why Rice Farmers Don’t Sail: Coastal
Subsistence Traditions and Maritime
Trends in Early China
Abstract The Lower Yangtze River Valley is a key region for the early devel-
opment of rice farming and the emergence of wet rice paddy field systems.
Subsistence evidence from Neolithic sites in this area highlights the importance of
freshwater wetlands for both plant and animal food resources. Early Neolithic rice
cultivators looked inland, especially to wetlands and nearby woodlands, for their
main protein sources. Links to the sea among these Neolithic populations are
notably scarce. Due to the high yields of wet rice, compared with other staple crops
as well as dryland rice, the wetland rice focused subsistence strategy of the Lower
Yangtze would have supported high, and increasing, local population densities.
Paddy agriculture demands labor input and water management on a large scale,
which would have stimulated and reinforced trends towards more complex soci-
eties, such as that represented by Liangzhu in the lower Yangtze region. Population
growth could have been largely absorbed locally, suggesting that population
packing, not migration, was the dominant trend. Other case studies of agricultural
dispersal, for the Korean Peninsula and Japan further illustrate the lack of corre-
lation between the spread of rice agriculture and wet rice cultivation. Although wet
rice cultivation was a pull factor that drew local populations towards increased
density and increased social complexity, it did not apparently push groups to
migrate outwards. Instead, the transition from wetland to rain fed rice cultivation
systems and/or the integration of rice with rain fed millet crops are much more
likely to have driven the demographic dynamics that underpin early farmer
migrations and crop dispersal.
L. Qin (&)
School of Archaeology and Museology, Peking University, Beijing, China
e-mail: qinling@pku.edu.cn
D. Q. Fuller (&)
Institute of Archaeology, University of London, London, UK
e-mail: d.fuller@ucl.ac.uk
School of Cultural Heritage, Northwest University, Xi’an, China
9.1 Introduction
of Nankuanli East confirms the presence of all three of these Chinese cereals (rice,
Panicum miliaceum and Setaria italica) in the earliest yet found archaeobotanical
assemblage on Taiwan, dating back to at least 4300 BP and perhaps as early as
5000 BP (Tsang et al. 2017).
Since at least the 1970s linguistic data for the Austronesian language family, the
most geographically dispersed language family in the world, have been traced back
to Taiwan, where all the basal branches in this tree are found among the indigenous
Formosan languages. Thus, from these derive the Malayo-Polynesian languages,
while other branches have spread through much of island Southeast Asia,
throughout the Pacific and even to Madagascar (Blust 1995; Pawley 2003; Spriggs
2011). The structure of this language family tree gave rise to the “express train”
model of population expansion and colonization that emanated out of Taiwan,
through island Southeast Asia and ultimately out into the Pacific via the Lapita
expansion starting ca. 3350 BP (Greenhill and Gray 2005; Spriggs 2011). Although
there are criticisms of this linguistic model (e.g., Donohue and Denham 2010), it
remains the dominant and most widely accepted explanation for how the far-flung
Austronesian languages came to be historically related.
Based on this model, the people of Neolithic Taiwan have been identified as
“proto-Austronesian.” One of archaeologist Peter Bellwood’s major contributions
was to synthesize archaeological evidence throughout island Southeast Asia,
highlighting cultural similarities in ceramics and other features that link the
Indo-Malaysian Neolithic cultures to those of the northern Philippines and Taiwan.
Drawing upon linguistic patterns and the cultural inferences of the archaeological
record he developed the “language farming” dispersal model, based on the idea that
a main demographic motor of expansion was the development of farming and the
seeking of new arable lands as agricultural populations expanded (Bellwood 1996,
2005). As these growing agricultural populations spread into the islands they lar-
gely replaced, and to some degree incorporated, pre-existing hunter-gatherer pop-
ulations. Archaeobotanical evidence for movement into the islands and the dispersal
of rice outside Taiwan remains limited (Paz 2003; Barton and Paz 2007; Fuller et al.
2010a). However, in the islands in particular a key transformation appears to have
taken place, as tuber crops like taro and yams ultimately became more important
than rice. This expanding Neolithic world of Austronesian farmers and sailors has
provided a narrative that unifies archaeological and linguistic histories of island
Southeast Asia and Taiwan for the later Holocene, despite the lack of hard evidence
for past agriculture.
This historical narrative can be questioned in three ways. First, we might ask:
“Why rice?” Why should rice agriculture have been central to the process of
demographic growth and the migration of farmers, and could other forms of food
production have been the driving force behind such movements, instead? Second, it
begs the question: “What kind of rice?” The range of potential forms of rice
cultivation cover a broad spectrum, from upland slash-and-burn systems to much
more intensive flood and irrigation systems (Fuller et al. 2011; Weisskopf et al.
2014). Among these various strategies, which forms of rice cultivation might have
driven the migrations to Taiwan and beyond? Scant attention has been paid to this
162 L. Qin and D. Q. Fuller
particular detail, although the research generally appears to assume it was more
intensive and productive forms of wet rice cultivation (e.g., Bellwood 1997: 208,
2005: 125). In fact, our research has shown that current evidence and logical
deductions suggest exactly the opposite. Third, we might reasonably ask: “Does the
empirical record, when assessed in terms of current hard evidence for agricultural
systems and their dispersal, actually support the maritime-based dispersal of rice
farming?
In response to these three questions, we propose that early wet rice farmers were
neither particularly expansive nor engaged in much maritime activity. Instead they
tended to be associated with a focus on freshwater wetland exploitation, with little
indication of engagement with the marine. This preference becomes clear in
reviewing the empirical record of archaeobotanical, faunal and settlement evidence
from the Lower Yangtze River Valley. Indeed, the highly productive systems of wet
rice agriculture supported population packing rather than geographical expansion.
Looking beyond the Lower Yangtze and the evidence for rice, other forms of food
production clearly need to be considered and compared, including millets, low
intensity dry rice, and vegeculture. In fact, when potential yields, labor demands,
land requirements and sustainability are taken into account it is much more likely
that millets and lower intensity forms of rice cultivation lent themselves to geo-
graphical expansion in search of new lands. In combination with coastal
forager-fisher traditions, this means that Neolithic Lower Yangtze rice farmers are
unlikely to have had anything to do with the spread of farming and farmers to
Taiwan and the Southeast Asian islands or mainland. Thus, established hypotheses
require either rejection or revision.
A key region for the early development of rice farming was the Lower Yangtze
River Valley, including northern Zhejiang, southern Jiangsu and the areas around
Shanghai (Fig. 9.1). Indeed, Neolithic cultures of this region such as Hemudu and
Majiabang have long featured in narratives about the emergence of rice agriculture
and the origins of the Austronesian Neolithic (e.g., Higham and Lu 1998; Bellwood
1997, 2005). Yet increasing numbers of Neolithic excavations in China have
documented additional regions and cultures that cultivated rice early on and that
likely contributed independently to one or more trajectories of rice domestication,
including Middle Yangtze cultures such as Pengtoushan, the Baligang of the middle
Yangtze Han River Valley, and the Jiahu and Shunshanji of the Huai River Valley
(e.g., Fuller et al. 2010a, b, 2011b; Qin 2012; Gross and Zhao 2014; Deng et al.
2015; Silva et al. 2015; Stevens and Fuller 2017). Nevertheless, the Lower Yangtze
is geographically the closest to Fujian and Taiwan, as well as one of the best
documented regions archaeologically and archaeobotanically. It therefore provides
9 Why Rice Farmers Don’t Sail … 163
Fig. 9.1 Map of Lower Yangtze River. 1. Kuahuqiao (跨湖桥), 2. Hemudu (河姆渡), 3.
Tianluoshan (田螺山), 4. Majiabang (马家浜), 5. Caoxieshan (草鞋山), 6. Chuodun (绰墩), 7.
Chenghu (澄湖), 8. Jiangli (姜里), 9. Liangzhu ancient city (良渚古城), 10. Maoshan (茅山)
a useful focus for considering the roles of freshwater and marine resources relative
to the evolution of rice cultivation.
In the Lower Yangtze region, cultural developments associated with the emer-
gence of wet rice agriculture can be identified through scrutiny of agricultural and
non-agricultural subsistence, technology, landscapes and diet. This region, in par-
ticular, has benefited from a large increase in archaeological fieldwork and the
practice of systematic archaeological science over the past two decades. With
regard to rice domestication, cultural change can be tracked through various traits.
The shattering versus non-shattering trait, which makes a crop dependent on
humans for successful reproduction, can be seen undergoing a rapid shift between
7000 and 6000 BP, during a period marked by the remains of the Hemudu and
Majiabang cultures. As for the bulliform phytoliths, directional linear changes in
size actually began around 6000 BP, suggesting the continued evolution of rice
plants (in terms of their leaves) under domestication. This shift parallels the evo-
lution of fatter grains, which began alongside non-shattering but continued after-
wards in both rice and other cereals (Fuller et al. 2010b; Stevens and Fuller 2017).
In addition, under domestication after 6000 BP rice grains split into two types, short
and long grain forms, which appear to have been quite stable varieties found in
different communities and settlements since 6000 BP. For example, these disparate
lineages of domesticated rice ultimately stabilized into today’s forms of tropical
versus temperate japonica rice (Zhao et al. 2011). The origins of such differenti-
ation may date back as far as the Late Neolithic in the Lower Yangtze, although
164 L. Qin and D. Q. Fuller
Fig. 9.2 Paddy fields and agricultural tools of the Lower Yangtze River a paddy fields of the
Chuodun site (Fuller et al. 2009) b paddy fields of the lower layer of the Maoshan site (Illustration
by L. Qin) c paddy fields of the top layer of the Maoshan site (Zhuang et al. 2014) d agricultural
tools of the Lower Yangzte (Illustration by L. Qin)
166 L. Qin and D. Q. Fuller
bones from wet sieved samples, freshwater wetland fish were clearly predominant,
such as snakehead (Channa), carp (Cyprinus), crucian carp (Carassius), and catfish
(Silurus). All of these species could have lived in or around rice stands or nearby
deeper water where Trapa or Euryale would grow. The carp and crucian carp in
particular have size ranges that indicate year-round fishing in freshwater wetlands,
while the snakeheads were targeted more in spring (Zhang 2018). In this assem-
blage a small quantity (0.7%) of Japanese sea bass indicates some coastal or
estuarine fishing, although this species also swims up into freshwater rivers when
not breeding. Despite a few large tuna vertebrae that were hand collected at the site
(e.g., Sun 2013) and a single dolphin bone from Kuahuqiao (see Eda et al. 2019:
Table 9.1), marine and coastal resources clearly appear to have been the exception;
a form of exotica set apart from the routine worlds of Neolithic inhabitants. Thus
these rice cultivators looked inland, especially to wetlands, for their main protein
sources.
The large mammal fauna include a wide range of deer and some pigs and
buffalo, likewise indicating an environment of wetlands and inland hill forests
(Zhang et al. 2011; Eda et al. 2019: Table 9.1). Significant numbers of water deer
(Hydropotes inermis) and water buffalo (Bubalus sp.) remains suggest the practice
of hunting in and around wetlands, while sika deer and sambar (Cervus spp.) point
towards woodland habitats. A significant minority of pig and boar bones (Sus
scrofa) has been interpreted as evidence of the early management of pigs and the
hunting of boars beginning sometime after 8000 BP (e.g., Liu and Chen 2012;
Zhang et al. 2011). Animal representations from the Liangzhu period also emphasis
wetland fauna alongside birds (Fig. 9.3a).
Taken together, the food resources discovered from Neolithic sites in the Lower
Yangtze allow us to reconstruct early land use and resource catchment in this area
(e.g., Qin et al. 2010; Fuller and Qin 2010; Zhang 2018). Material culture from the
Liangzhu Period also reflects the same catchment and resource management sys-
tems, in which birds, freshwater fish and turtles remain a recurrent theme (Fig. 9.3).
Neolithic inhabitants’ engagement with this landscape is further reflected in their
diet, which can be reconstructed through isotopic data (Fig. 9.4). In dietary terms,
the Lower Yangzte is characterized by a C3 terrestrial and wetland type, a signature
markedly distinct from either maritime hunter-gatherers, maritime millet farmers, or
terrestrial millet farmers (see Fig. 9.4). Two archaeological discoveries of canoes in
this region, at the Kuahuqiao Site (8000 BP) (Jiang 2013) and Maoshan (4500 BP)
(Zhao et al. 2013; Zhuang et al. 2014; see Fig. 9.3), also indicate the existence of
simple riverine and wetland boat technologies.
We therefore conclude that neither subsistence interests nor transportation
technologies link Lower Yangtze Neolithic populations to the sea. Instead, fresh-
water wetlands and nearby woodlands were the main landscape features exploited
by Lower Yangtze rice farmers. These communities appear to have looked inland,
and not towards the sea.
168 L. Qin and D. Q. Fuller
Fig. 9.3 Material culture reflects wetland management of the Lower Yangtze River a animal
images from Liangzhu jades and pottery decoration (from exhibition at the Liangzhu Museum)
b canoe from the Kuahuqiao site (Zhejiang Provincial Institute of Archaeology and Culture Relics
et al. 2004) c canoe from the Maoshan site (lower layer) (photograph by L. Qin)
9 Why Rice Farmers Don’t Sail … 169
Fig. 9.4 Different forms of landscape engagement are reflected in dietary stable isotopes.
5 different types can be recognized. Type 1 (lower left) is the Lower Yangtze type characterized by
C3 wild plants, freshwater wetland resources and terrestrial mammals (Tianluoshan, Minagawa
et al. 2011; Sanxingcun, Hu et al. 2007; Songze, Zhang 2003; Tangshan, Zhang et al. 2015; Jiahu,
Hu et al. 2006; Tanshishan, Wu et al. 2016). Type 2 (lower middle) is a mixed rice, millet and pig
based subsistence strategy represented by the Neolithic Qujialing culture in Hubei (Qinglongquan
site, Guo et al. 2011). Type 3 (lower right) is the typical Northern Chinese Neolithic diet focused
on millets (C4) and terrestrial mammals like pigs, represented here by Bianqian, a Shandong
Dawenkou Period site (Wang et al. 2012) and the Zongri site (Longshan Period) in Qinghai (Cui
et al. 2006). Type 4 (top left) is a maritime hunter-gatherer diet represented here by Liyudun (Hu
et al. 2010) on the south coast of Guangdong and typical of much of Jomon, Japan (Minagawa
et al. 2011). Type 5 (upper right) is a maritime millet agriculture signature represented by the early
Dawenkou Neolithic Period in the Changdao Archipelago of the Bohai Sea (Zhang 2003).
Numbers in brackets refer to the sample numbers
The idea that rice farmers migrated southwards from the Lower Yangtze, dispersing
out of this region by boat, was based on the underlying demographic logic of demic
diffusion. This theory supposes that a growing population splinters, with daughter
populations moving outwards in search of new land to settle and farm (Ammerman
and Cavalli-Sfroza 1971). Rindos (1980, 1984) explained that such migration
events will occur when local populations grow to or beyond their natural carrying
capacity. Carrying capacity itself will fluctuate between years due to factors like
variations in yield, and the extent of this instability may speed up or slow down
overall migration rates. Shennan’s (2018) recent synthesis of Neolithic datasets
from Europe took an explicitly demographic perspective, however, and identified a
tendency for dispersal to occur when regional populations were growing rapidly,
9 Why Rice Farmers Don’t Sail … 171
Fig. 9.5 Population growth and fission model. Schematic representation of population growth and
dispersal through fission. a Indicates population growth towards carrying capacity with dispersal of
“excess” population as carrying capacity is breached, or, alternatively in a scenario of underpro-
duction as rapid growth rates cross a threshold into decreasing returns. b Population growth and
dispersal scenarios given two contrasting productivity regimes with different carrying capacity
but before growth slowed; in other words, well before reaching carrying capacity.
The European data therefore imply that populations can disperse in search of new
agricultural territory not only when they reach their maximum size (as implied by
the Rindos model), but during an intermediate period of rapid growth.
This theory also makes sense in light of comparative ethnographic studies
indicating that many traditional small-scale societies operate well below carrying
capacity, in what Sahlins (1972) called “underproduction” or the “underuse of
resources” (42). Using data from a range of traditional production systems, their
populations and computed potential productive capacity, Sahlins found that all of
them appear to have under-produced. Only a couple of the groups produced at 65%
or 75% of their capacity, while the average rate of production was only about 45%
of their estimated capacity (Sahlins 1972: 42–48; cf. Carlstein 1980: 239). Thus, it
may not be carrying capacity per se that drives fission, but rather population growth
to a threshold at which increasing effort is needed to keep feeding more people. In
either case, the total potential carrying capacity will affect how quickly a population
grows and at what point migration is likely to begin (Fig. 9.5).
These observations raise two questions about the nature of early subsistence in
East and Southeast Asia. First, what specific and inherent differences in carrying
capacity (CC) and its associated underproduction (*60% CC) between different
regions or crops would have raised or lowered the ceiling to which populations
172 L. Qin and D. Q. Fuller
grew? And second, what similar differences determined the point at which daughter
populations dispersed? Based on the existing evidence, wet rice agriculture appears
less likely to propel population migration than alternative rainfed forms of agri-
culture, including both rainfed and upland rice and millets.
It is well known that rice productivity varies significantly based on water avail-
ability during the growing season, as well as varying demands for labor input and
potentially different outputs of greenhouse gases (e.g., Fuller et al. 2011a, 2016).
Previously we suggested that the higher labor demands of wet rice might have
restricted the appeal of its adoption by some societies, and there might even be a
threshold of social complexity below which wet rice cultivation was avoided (Fuller
and Qin 2009). Still more important, however, are the inherent differences in potential
carrying capacity that can be estimated in terms of the land necessary for rice culti-
vation to feed a self-sustaining village community or typical Neolithic community. In
order to estimate the amount of land needed to feed populations at Neolithic sites, we
have assembled a range of ethnographic and historic data on yield per hectare (ha) for
wet rice, dry rice and traditional millet agriculture. This can be converted into a
caloric yield and divided by the amount of cereal crop consumed per person per year
(assuming grains were the caloric staple) and the population of past communities. It
should be noted that population estimates are not meant to be precise, but rather
provide an order of magnitude approximation: thus the difference between 50 and 500
is significant, whereas that between 30 and 100 is less meaningful.
For population sizes we have taken empirical values from the size of archaeo-
logical sites as well as a few pre-existing estimates of population size. These
include Chengtoushan (Hunan) in the Middle Yangtze (6500-6000 BP), at ca. 8 ha,
Hemudu (7000-6300 BP), at ca. 4 ha, and Tianluoshan (7000-6300 BP), at ca. 3 ha,
in the Lower Yangtze (Zhejiang). All of these sites have quite reliable maximum
size estimates from their main periods of occupation. Previous population estimates
for Chinese Neolithic habitation sites agree on a ratio of approximately 50 persons/
hectare, including an estimate from Hemudu based on building numbers and floor
space (Sun 2013: 563). An independent estimate of 53.5 person/ha has also been
made for the millet-producing area of northern China, based on house areas and
burial numbers from the Early Yangshao site of Jiangzhai (Liu 2004: 79).
Modern data provide estimates of rice consumed per person, with *250 kg of
unhusked rice required for *2000 calories per person according to Grist (1975:
450), and 160 kg/person/year estimated for traditional Southeast Asia (Hanks 1972:
48). The typical intake observed for traditional coastal Odisha, India of 160 kg/
person/yr (Smith and Mohanty 2018: 1328) is similar, assuming this number rep-
resents dehusked rice, which weighs the equivalent of 60–70% of unhusked rice.
These modern estimates probably account for ca. 80% of total caloric intake (Grist
1975: 450), but we assume that Neolithic populations ate a more diversified diet, as
clearly indicated by the archaeobotanical data from sites like Hemudu, Kuahuqiao
and Chengtoushan. These deposits suggest a diet rich in other carbohydrates such as
acorns, Trapa water chestnuts (Fuller et al. 2007, 2009; Fuller and Qin 2010), and
in some cases millet, as observed at Chengtoushan (Nasu et al. 2007, 2012). We
have therefore assumed that rice in this context might account for roughly 50% of
the total diet (if, as in the modern diet, grains accounted for 75–80% then land need
estimates would need to be increased by 50–60%).
9 Why Rice Farmers Don’t Sail … 173
Past yields may be difficult to estimate, as they depend directly on land use
systems. Nor can modern traditional yields serve as perfect analogues for earlier in
prehistory. In general, wet rice is expected to yield better than rainfed rice; thus the
lower bounds of reasonable yields draw upon data from dry rice productivity. Dry
rice yields range from around 480 kg/ha to as much as some 1500 kg/ha, in some
modern systems (Fig. 9.6). The average of our comparative data on dry rice is
1062 kg/ha, although data from Palawan and Borneo swiddens alone average just
578 kg/ha, with yields as low as 229 kg (Barton 2012). The average of our com-
pilation of wet rice yields is 1897 kg/ha, with the lower end of recently reported
traditional wet rice yields standing around 1500 kg/ha. Historical data, however,
indicate that about 1300 kg/ha was achieved in 10th century Japan, while around
1000 kg/ha was observed in the Han Dynasty, at Hangzhou nearly 2000 years ago.
Thus the slightly lower yields of 830 and 950 kg, estimated from rice leaf phytolith
densities in paleosols of field surfaces around Neolithic Tianluoshan (ca. 6700 BP),
might be reasonable for early, unimproved wet rice yields (Zheng et al. 2009).
Rounding these down to 800 or 900 kg and taking into account the upper and lower
estimates of modern rice consumption per person, we can therefore bracket the land
area needed for rice production among a small selection of Chinese Neolithic sites
(see Table 9.1).
Based on the above calculations, we estimate that Neolithic rice producing sites
need between 6.25 and 9.75 hectares of rice cultivation land for every hectare of
settled land (or for every *50 persons), with a median estimate of about 8 ha of
rice cultivation land for each hectare of settlement land (see Table 9.1). Our pro-
ductivity estimates are also quite low, meaning that if 1000 kg or more rice per
hectare were produced, even less land would be needed per person and local
carrying capacity would exceed our existing estimates. Historical and ethnographic
data indicate that most fields are found within 3 km of settlements, while farm plots
over 4 km from settlements appear to have been more or less impossible due to the
need for daily travel, on foot, to work in the fields and return home (Carlstein 1980:
172). This suggests that about 2800 hectares of land could readily support a local
population on the order of 14,000 people.
This relatively high productivity estimate for wet rice can be contrasted with the
much lower expected estimates for rainfed rice or millet production (Figs. 9.6, 9.7).
Rainfed rice production has been well documented in Southeast Asia, and as
summarized by Barton (2012), the productivity of such rice in Borneo was quite
low (ranging from 229 to 1000 kg/ha). For Neolithic dry rice these yields would
have been, on average, about half that of wet rice, or between 400 and 500 kg/ha.
This low rate of productivity would have been further exacerbated by the need to
shift fields, as fertility decreased and weed competition with rice increased over
time. In other cases rainfed rice is grown in shifting cultivation systems unless an
external source of fertilizer can be employed, such as manure from domesticated
cattle, In the well-studied case of traditional agriculture amongst the Iban in the
Philippines, about 0.33 ha was cleared for rice per person per year, and a
174
Fig. 9.6 Traditional and historical rice yields, contrasting predominantly rainfed/dry (tan, at left) and wet/irrigated (blue, at right) Where multiple values are
reported from the same study the mean and standard deviation are shown. Sources from left to right: 1. Barton 2012; 2, 4, 5. Ruthenberg 1976: 52; 3, 20.
Geddes 1954: 68; 6, 7. Saito et al. 2006; 8, 9, 24, 32. Sherman 1990: 131; 10, 14, 26, 31, 33. Bray 1986; 11. Grigg 1974: 97; 12. Heston 1973; 13. Randhawa
1958; 15. Vincent 1954; 16, 17. Zheng et al. 2009; 18, 34. Ellis and Wang 1997; 19. Latham 1998: 22; 21, 22, 23, 29. Boomgaard and Kroonenberg 2015; 25,
27. Watabe 1967; 28. Leonard and Martin 1930; 30
L. Qin and D. Q. Fuller
9 Why Rice Farmers Don’t Sail … 175
Table 9.1 Estimated rice consumption, land requirements and carrying capacity for Yangtze
River Valley
Site Est. Lower Est. Higher Est. Lower Est. Higher Est. Median
Population Rice Needs Rice Needs Land Land Rice
(kg/yr, 68% (kg/yr, 80% Needs Needs Land
of diet) of diet) (900 kg/ha) (800 kg/ha) (ha)
Tianluoshan, ca. 150 16875 23437.5 18.75 29.29688 24.02344
6700 BP (3 ha)
Hemudu, ca. 200 22500 31250 25 39.0625 32.03125
6700 BP (4 ha)
Chengtoushan, 400 45000 62500 50 78.125 64.0625
ca. 6000 BP
(8 ha)
Hypothetical 50 5625 7812.5 6.25 9.765625 8.007813
1 ha Site
Maximum Size 14,000 1,575,000 2,187,500 1050 2242 1892
Based on Wet (based on
Rice Farming 1500 kg/
within 3 km ha yield)
(~280 ha)
Hypothetical 50 5625 7812.5 18.75 52.08 35.42
Dry Rice Site (based on (based on
(1 ha) 600 ka/ha, 300 kg/ha
+1/2 fallow yield)
yield)
long-house village of 140 people required 50 ha per year (Carlstein 1980). Based
on these figures the Iban could reside at a single settlement for a maximum of
14 years before needing to move, but ten years was considered a better estimate
given the unsuitability of some land in a given catchment as well as the shifting
age-sex demographics of the community over time (Carlstein 1980: 174).
The land needs of the Iban are therefore approximately four times those esti-
mated for the Yangtze Neolithic communities (see Table 9.1). This would mean
that carrying capacity for a given settlement catchment based on rainfed rice is
roughly one quarter what it would be for wet rice. Assuming uniform rates of
population growth, this predicts that community fission and migration in search of
new space would occur four times as often among dry rice farmers as among wet
rice farmers (see Fig. 9.5). Given dry rice farmers’ need to shift fields for fallowing,
or indeed their need to relocate altogether (e.g., every 10–15 years for a group like
the Iban), cultural traditions of mobility and the establishment of new settlements
are likely to have encouraged the kind of movement that underpinned long-term
sequences of migration. This also suggests that as wet rice productivity increased
over time, it allowed for more tightly packed populations.
176
Fig. 9.7 Traditional and historical rice yields, including data from South Asia (blue) and East Asia (red). Where multiple values are reported from the same
study the mean and standard deviation are shown. Sources from left to right: 1, 8. Weber 1991; 2. ICAR 1980: 828; 3. Randhawa 1958; 4, 5, 15. Rachie 1975:
16; 6. CSIR 1966: 226; 7. Heston 1973; 9, 10. ICAR 1980: 835–837; 11, 12. King 1927; 13, 14. Bray 1981
L. Qin and D. Q. Fuller
9 Why Rice Farmers Don’t Sail … 177
Table 9.2 Estimated millet consumption, land requirements and carrying capacity for Yellow
River Valley
Site Est. Lower Est. Higher Est. Lower Est. Higher Est. Median
Population Grain Needs Grain Needs Land Needs Land Needs Millet
(kg/yr, 68% (kg/yr, 80% (650 kg/ha, (500 kg/ha, Land
of diet) of diet) on rich on poor (ha)
loess, 1/3 soils, 2/3
fallow) fallow)
Banpo (Early 250 43,697 51,408 153.95 474.54 314.245
Yangshao),
(5 ha)
Wangchengang 1750 305,880 359,859 1077.7 3324 2200.8
(Longshan),
(35 ha)
Hypothetical 50 8739.44 10281.7 30.7907 94.9079 62.8493
1 ha Site
Maximum Size 2000 349,577 411,267 1232 3796 2514
Based on
Millet
Cultivation
within 3 km
(~40 ha)
sub-tropics and tropics, lands are likely to have been left fallow for two out of three
years, or even more. Thus, as millet cultivation spread to new communities beyond
its core area in the loess plateau it required increasing land areas in order to
maintain the same levels of productivity.
Based on the nature of cultivation systems, we can conclude that the wetland rice
focused subsistence strategy of the Lower Yangtze (and Middle Yangtze) would
have supported high, and increasing, local population densities. Thus, population
growth could have been largely absorbed locally, through the expansion and
intensification of production. In this sense wet rice agriculture was a factor that
drove the creation of larger, more concentrated populations, and also tended to
provide for non-subsistence specialists such as those practicing stone working,
ceramic production or ritual. The ultimate emergence of urban centers out of this
very process can be seen in the mega-sites of Liangzhu, in the Lower Yangtze, and
Shijiahe, in the Middle Yangtze. Both of these settlements were supported by local
hinterlands of wet rice cultivation, represented by paleosols and field systems such
as those discovered at Masohan, to the northeast of Liangzhu. Population packing,
and not migration, was the dominant trend among Neolithic populations focused on
wet rice cultivation.
The higher population densities made possible by wet rice agriculture were both
a product of and a promoter for engagement with wetlands. Thus, the wetland
landscapes of the Lower Yangtze and Taihu lake region included networks of
natural water ways that were expanded through rice cultivation, creating a geog-
raphy that fostered social networks, the capture and transportation of aquatic
resources such as fish, and larger, more sustainable populations. Wet rice produc-
tion required greater investments of labor, but the resulting social and economic
organization played a key role in the development of larger social and political
units. Thus the development of rice agriculture pulled people together. It also
provides a context for understanding how and why earthworks and water control
systems such as those discovered at Liangzhu, also known as the Peripheral Water
Conservancy System of the Liangzhu City Site, came into existence in this period
(Liu et al. 2017). This water control system helped to guarantee the development of
the Liangzhu economy, with its specialized jade artwork, as well as the agricultural
tool kits that subsequently drove further social complexity and more intensified wet
rice agriculture (Qin 2013; Renfrew and Liu 2018).
The following three cases of agricultural dispersal offer a contrast to the above case
in the Lower Yangzte, illustrating the lack of correlation between the spread of rice
agriculture and wet rice cultivation.
9 Why Rice Farmers Don’t Sail … 179
The Northeast Asian regions beyond China, including the Korean Peninsula and the
Japanese archipelago, came to agriculture relatively late and received their major
agricultural staple crops from China. The millets (Setaria italica and Panicum
miliaceum) and rice (Oryza sativa) spread as domesticated species from China to
Korea, and later to Japan. Evidence for millets on the Korean peninsula dates back
to the Middle Chulmun Period, or 5500 to 5000 BP (Crawford and Lee 2003; Lee
2011). Millet crops of similar date have been found at sites in southeastern Siberia,
in the Primorye region of far eastern Russia. Rice subsequently arrived in Korea
later, perhaps around 3500 BP, although some room remains for debating the
precise date (Ahn 2010; Lee 2011, 2015).
The migration of farmers was likely part of the process that brought millets and
agriculture to these regions. Archaeological evidence suggests a cultural origin in
northeastern China (from Jilin or Heiligong in the Chifeng region) (e.g., Miyamoto
2016), while recent research in historical linguistics traces Koreanic and Japonic
languages back to a hypothetical Transeurasian language family originating in
northeast China (Robbeets 2017a, b). The key point, however, is that these
migrations were driven by the lower productivity levels of dry millet crops, not wet
rice. Rice as a crop was adopted as an add-on to millet based subsistence and
presumably spread through adoption from the Shandong peninsula across to the
Liaodong peninsula, then south through the Korean peninsula and eventually to
Japan (Ahn 2010; Miyamoto 2016, 2019). Nor does the archaeobotanical evidence
from the Shandong and Liaodong peninsulas indicate any regional wet rice farming
dominance during the Bronze Age (Liu 2016). The selective adoption of rice
cultivation in wet paddy systems only became a characteristic component of Bronze
Age agriculture in Korea, alongside millets, soybeans and other crops (Lee 2015).
The emphasis on marine food evident in earlier Chulmun ceramics and shell
middens moreover indicate that maritime skills were prevalent in the region before
this shift began (e.g., Shoda et al. 2017). Indeed, marine foods remained a key part
of subsistence through the later Chulmum and Mumun Periods in Korea.
The advent of agriculture in Korea therefore took place gradually via adoption.
The transition from foraging to farming may indeed represent a farming dispersal,
and has been associated with a language/farming dispersal hypothesis associated
with the ancestry of Koreanic and Japonic languages as well as the Transeurasian
hypothesis (e.g., Whitman 2011; Miyamoto 2016; Robbeets 2017a, b). However
rice, whether wet or dry, was only adopted later as an add-on crop and not an
economic driver of cultural or demographic change.
180 L. Qin and D. Q. Fuller
The dispersal of rice and millet together into the tropical far south of China rep-
resents the passage of cereal agriculture, predominately rice with some foxtail
millet, into mainland Southeast Asia as early as 4500 to 4000 BP (Fig. 9.8). The
earliest directly dated crop in mainland Southeast Asia is foxtail millet (Setaria
italica) found at Non Pa Wai, in central Thailand, and dated to around 4400 to 4200
BP. The first evidence for rice, on the other hand, is not yet clearly older than about
4000 BP in Vietnam, Cambodia or Thailand (Castillo 2017; Silva et al. 2015).
182 L. Qin and D. Q. Fuller
Fig. 9.8 Map of sites with archaeobotanical evidence mentioned in the text or relevant to the
southward dispersal of rice and millets. Numbered sites: 1. Baligang; 2. Jiahu; 3. Shuanshanji; 4.
Pengtoushan; 5.Chengtoushan; 6. Shijiahe; 7. Nanshan; 8. Pingfengshan; 9. Huangguashan; 10.
Baitoushan; 11. Nankuanli East; 12. Chaolaiqiao;13. Baiyangcun; 14. Gantuoyan; 15. Non Pa Wai;
16. Phu KhaoThong; 17. Khao Sam Kaeo; 18. Ban Non Wat & Non Ban Jak ; 19. Rach Nui. *Dash
line in the lower Yangzte area shows the area with only rice agriculture. See Fig. 9.1 for details
Throughout Southeast Asia, transitions from dry to wet rice occurred in later
prehistory or in historical times. Recent research at Ban Non Wot and Non Ban Jak
provides a long regional sequence of archaeobotanical data in northeast Thailand
between 3000 and 1300 BP (Castillo et al. 2018b) (Fig. 9.8). During this period dry
rice weeds decline as wet rice weeds appear around 2100 BP. Wet rice subsequently
increases and dry rice weeds disappear by 1500 BP. This indicates that in the face
of increasing aridity, rice cultivation was bolstered by irrigation; but it also suggests
that increasingly hierarchical societies in the region were investing greater labor in
more intensive wet rice production. While rainfed rice has persisted in the hills of
Southeast Asia into recent times, throughout most of the plains wet rice cultivation
has long been the predominant cultivation system, responsible for supporting his-
torically known states and urban systems throughout the region (Scott 2009). This
indicates that wet rice cultivation in the Southeast was a secondary development
driven by the growth of social complexity and perhaps population growth, rather
than the primary force driving regional demographic change in the Early Neolithic.
Rice is not simply one thing. As a modern crop it illustrates a vast range of
ecological diversity, growing from nearly 40° North in latitude to the equator and
from sea level to over 2000 meters above sea level. Genetic evidence indicates the
influence of multiple wild populations and numerous trajectories of adaptation and
cultural selection over time (e.g., Fuller et al. 2016). Just as rice was transformed
ecologically as it came into new regions and responded to the genetic inputs of local
wild populations, the cultures that moved rice are also likely to have been trans-
formed through new cultural adaptations and interactions with local cultural tra-
ditions, including hunter-fisher folk and hypothetical tuber cultivators. This means
that the challenge for archaeology and archaeobotany through East and Southeast
Asia is to understand the beginnings of rice cultivation in its local context, in which
both the ecology of rice and its place in subsistence culture may have varied. It is no
longer sufficient to use a simplistic proxy like ceramic styles to indicate migration
and the spread of rice farming. Different subsistence strategies, including myriad
cultivation systems and disparate forms of rice, had variable demographic conse-
quences and impacts on community fission and movement in search of new land.
In terms of understanding the advent of rice agriculture, we can differentiate
three major modes. First, we can identify cases where wild rice was brought into
cultivation locally and evolved into the domesticated form. The data available from
the Lower Yangtze region clearly illustrates this process in which primary
domestication takes place, represented clearly in the evolution of non-shattering,
and is followed by post-domestication evolution in the form of continuing trends of
change in bulliforms and grain shape and size. The evidence from the Lower
Yangtze indicates that wet rice cultivation was a pull factor that drew local pop-
ulations towards increased density, increased social complexity and deeper
184 L. Qin and D. Q. Fuller
entanglements with inland freshwater wetland habitats. However it did not appar-
ently push groups to migrate outwards.
Second, rice was also brought into new regions as an already domesticated
crop. These introductions could have happened in two ways: either it was adopted
by local populations as an add-on to existing subsistence systems, or it was carried
by migrant farmers. Examples of the first form of rice adoption are evident in
Northern China, Korea and more broadly in northeastern Asia. In these areas rice
was added to local subsistence in places where the cultivation of domesticated
millets was already established. The extent to which wet rice or dry rice was
adopted would have been constrained by both environmental conditions (e.g., water
availability) and social conditions (e.g., labor availability), and these factors would
have driven the population’s engagement with intensive wet rice systems or low
input rainfed systems.
A third possibility is that rice was carried as a part of the migrant culture of food
producers. Wet rice is less likely to have spread this way due to its higher local
carrying capacity and relatively high labor demands. Instead, in cases where the
immigration of farmers with rice did occur, rainfed rice is likely to have been more
common. Thus, dry rice tends to push populations towards outward migration. This
in turn raises a key, unresolved question: “Where, when and how many times dry
rice cultivation systems evolve?” It is plausible that rainfed rice developed once in
Southeastern Shandong prior to its adoption in Korea, but it is likely to have
evolved separately, and perhaps more than once, in the hilly regions south of the
Yangtze River. For example, this could have occurred prior to the dispersal of rice
into Fujian or Guangdong. These arguments and the current evidence highlight the
importance of applying systematic archaeological science to both archaeobotanical
macro-remains and phytolith assemblages in order to recover and reconstruct
subsistence systems throughout southern China and Southeast Asia.
For too long the transition to rice farming has been a kind of “black box”
mechanism for driving population migrations and transforming the demography of
eastern Asian Neolithic societies. As we have argued, however, subsistence details
matter. Indeed, wet rice cultivation systems appear to have achieved the opposite of
what has been supposed, and are actually more likely to underpin local population
growth and the intensification of freshwater wetland exploitation rather than pro-
mote Neolithic migration. Instead, the transition from the original wetland rice
cultivation systems to rainfed rice and/or the integration of rice with rainfed lower
intensity millet crops are much more likely to have driven the demographic
dynamics that underpin early farmer migrations and crop dispersal. This is sup-
ported by rich archaeological evidence from the Hangzhou Bay region and the
Lower Yangtze, which indicates a decidedly inward, freshwater wetland focus
rather than a maritime turn. It is also substantiated by recent data highlighting the
importance of millets alongside rice in the Neolithic traditions of Fujian, Taiwan
and mainland Southeast Asia.
Thus, in Thailand the turn to intensive wet rice agriculture was late, dating to the
Iron Age, and is more likely to have been instrumental in urbanization rather than in
establishing Neolithic populations. The non-dispersing character of early wet rice
9 Why Rice Farmers Don’t Sail … 185
and the need for lower intensity dry rice and/or millet farming to become estab-
lished in sub-tropical South China prior to major Neolithic dispersals help to
explain the long lag time between early rice cultivation (>8000 BP), rice domes-
tication (by 6000 BP) and the beginnings of the cereal-based Neolithic phase in
Southeast Asia (<4500 BP). We have offered some explanatory factors, based on
the productivity of different cropping systems, that help to explain these patterns
and suggest that the lower intensity rainfed rice crop systems are more likely to
support community fission and Neolithic migration than the more productive wet
rice systems. Ultimately the less productive, rainfed cultivation of rice and millet
could be characterized as centrifugal forces that push populations outwards in
search of more land, in contrast to the more centripetal pull of wet rice agriculture.
Acknowledgements Research that contributed to this paper and research on early rice across
parts of Asia over the last decade has been supported by two research grants to Ling Qin, from the
National Social Science Foundation of China (Project number: 13BKG006) on “Archaeological
investigation of Yunnan prehistoric agriculture” and from the National Education Ministry of
China (Project number: 16JJD780004) on “Technology and civilization: the foundation of early
Chinese civilizations”; and three research grants from the Natural Environment Research Council
(U.K.), to DQ Fuller, most recently NE/N010957/1 on “The impact of intensification and dein-
tensification of Asian rice production: transitions between wet and dry ecologies.”
References
Boomgaard, P., & Kroonenberg, P. M. (2015). Rice, sugar, and Livestock in Java, 1820–1940:
Geertz’s Agricultural Involution 50 years on. In F. Bray, P. A. Coclanis, E. L. Fields-Black, &
D. Schafer (Eds.), Rice: Global networks and new histories (pp. 56–83). Cambridge:
Cambridge University Press.
Bray, F. (1986). The rice economies: Technology and development in Asian Societies. Oxford:
Blackwell.
Cao, Z. H., Ding, J. L., Hu, Z. Y., Knicker, H., Kögel-Knabner, I., Yang, L. Z., & Yin, R., et al.
(2006). Ancient paddy soils from the Neolithic age in China’s Yangtze River Delta.
Naturwissenschaften, 93(5), 232–236.
Carlstein, T. (1980). Time resources, society and ecology: On the capacity for human interaction
in space and time Part 1: preindustrial societies. Lund: Department of Geography, The Royal
University of Lund, Sweden.
Carson, M. T., & Hung, H. C. (2018). Learning from Paleo-landscapes: Defining the land-use
systems of the ancient Malayo-Polynesian homeland. Current Anthropology, 59(6), 790–813.
Castillo, C. (2017). Development of cereal agriculture in prehistoric Mainland Southeast Asia.
Man In India, 95(4), 335–352.
Castillo, C. C., Bellina, B., & Fuller, D. Q. (2016). Rice, beans and trade crops on the early
maritime Silk Route in Southeast Asia. Antiquity, 90(353), 1255–1269.
Castillo, C. C., Fuller, D. Q., Piper, P. J., Bellwood, P., & Oxenham, M. (2018a). Hunter-gatherer
specialization in the Late Neolithic of southern Vietnam: The case of Rach Nui. Quaternary
International, 489, 63–79.
Castillo, C. C., Higham, C. F., Miller, K., Chang, N., Douka, K., Higham, T. F., & Fuller, D. Q.
(2018b). Social responses to climate change in Iron Age north-east Thailand: New
archaeobotanical evidence. Antiquity, 92(365), 1274–1291.
Chang, K. C. (1986). The archaeology of Ancient China (4th ed., pp. 234–294). New Haven, CT:
Yale University Press.
Chang, K. C., & Goodenough, W. H. (1996). Archaeology of southeastern coastal China and its
bearing on the Austronesian homeland. Transactions of the American Philosophical Society,
86(5), 36–56.
Crawford, G. W., & Lee, G. A. (2003). Agricultural origins in the Korean peninsula. Antiquity, 77
(295), 87–95.
CSIR [Council for Scientific and Industrial Research]. (1966). The Wealth of India (Vol. 7). New
Delhi: Publications and Information Directorate.
Cui, Y., Hu, Y., Chen, H., Dong, Y., Guan, L., Weng, Y., & Wang, C. (2006). Stable isotopic
analysis on human bones from Zongri site. Quaternary Sciences, 26(4), 604–608.
Dai, J., Zuo, X., Cai, X., Wen, S., Jin, J., Zhong, L., & Xia, T. (2019). The mixed rice and millet
agriculture in Neolithic age lower Minjinag River: Phytolith evidence from Baitoushan site.
Quaternary Sciences, 39(1), 161–169 (DiSiJi YanJiu, in Chinese).
Deng, Z., Qin, L., Gao, Y., Weisskopf, A. R., Zhang, C., & Fuller, D. Q. (2015). From early
domesticated rice of the middle Yangtze Basin to millet, rice and wheat agriculture:
Archaeobotanical macro-remains from Baligang, Nanyang Basin, Central China (6700–500
BC). PLoS ONE, 10(10), e0139885.
Deng, Z., Hung, H. C., Carson, M. T., Bellwood, P., Yang, S. L., & Lu, H. (2018a). The first
discovery of Neolithic rice remains in eastern Taiwan: Phytolith evidence from the Chaolaiqiao
site. Archaeological and Anthropological Sciences, 10(6), 1477–1484.
Deng, Z., Hung, H. C., Fan, X., Huang, Y., & Lu, H. (2018b). The ancient dispersal of millets in
southern China: New archaeological evidence. The Holocene, 28(1), 34–43. http://doi.org/10.
1177/0959683617714603.
Denham, T., Zhang, Y., & Barron, A. (2018). Is there a centre of early agriculture and plant
domestication in southern China? Antiquity, 92(365), 1165–1179.
Diamond, J. (1997). Guns, Germs and Steel. New York: Random House.
Diamond, J., & Bellwood, P. (2003). Farmers and their languages: The first expansions. Science,
300(5619), 597–603.
9 Why Rice Farmers Don’t Sail … 187
Hanks, L. M. (1972). Rice and man: Agricultural ecology in Southeast Asia. Chicago: Aldine.
Heston, A. W. (1973). Official Yields Per Acre in India, 1886–1947: Some questions of
interpretation. The Indian Economic & Social History Review, 10(4), 303–332.
Higham, C. (2003). Languages and farming dispersals: Austroasiatic languages and rice
cultivation. In P. Bellwood & C. Renfrew (Eds.), Examining the farming/language dispersal
hypothesis (pp. 223–232). Cambridge: McDonald Institute for Archaeological Research.
Higham, C., & Lu, T. L. D. (1998). The origins and dispersal of rice cultivation. Antiquity, 72
(278), 867–877.
Higham, C. F., & Rispoli, F. (2014). The Mun Valley and Central Thailand in prehistory:
Integrating two cultural sequences. Open Archaeology, 1(1). http://doi.org/10.2478/opar-2014-
0002.
Hu, Y., Ambrose, S. H., & Wang, C. (2006). Stable isotopic analysis of human bones from Jiahu
site, Henan, China: implications for the transition to agriculture. Journal of Archaeological
Science, 33(9), 1319–1330.
Hu, Y., Li, F., Wang, C., & Richards, M. P. (2010). Carbon and Nitrogen Stable Isotope Analysis
of the Human Bones from the Liyudun Site, Zhanjiang, Guangdong: A Preliminary
Exploration of the Neolithic Human Lifestyle in South China. Acta Anthropologica Sinica,
29(3), 264–269.
Hu, Y., Wang, G., Cui, Y., Dong, Y., Guan, L., & Wang, C. (2007). The ancient diets study of
Sanxingcun site in Jintan, Jiangsu province. Chinese Science Bulletin, 52(1), 85–88.
Hung, H. C., & Carson, M. T. (2014). Foragers, fishers and farmers: Origins of the Taiwanese
Neolithic. Antiquity, 88(342), 1115–1131.
ICASS (Institute of Archaeology of the Chinese Academy of Social Sciences), Fujian Provincial
Museum and Mingxi County Museum. (2017). Excavation report for Cave 4 of the Nanshan
Site in Mingxi County, Fujian Province. Archaeology, 10, 3–22 (Kaogu, in Chinese).
Jiang, L. (2013). The Kuahuqiao site and Culture. In A. Underhill (Ed.), A companion to Chinese
archaeology (pp. 537–554). Oxford: Wiley-Blackwell.
Jiao, T. (2007). The Neolithic of Southeast China: Cultural transformation and regional
interaction on the coast. New York: Cambria Press.
Latham, A. J. H. (1998). Rice: The primary commodity. London: Routledge.
Lee, G. A. (2011). The transition from foraging to farming in prehistoric Korea. Current
Anthropology, 52(s4), s307–s329.
Lee, M. (2015). Rice in ancient Korea: status symbol or community food? Antiquity, 89, 838–853.
Leonard, W. H., & Martin, J. H. (1930). Cereal Crops. New York: Collier Macmillan.
Lin, H. (1930). Taiwanese original culture. National Academia Sinica Institute of Social Sciences.
Lin, H. (1955). Research on stone age remains in Taiwan (one of the research reports of the
Human Museum in 1955). Journal of Xiamen University (Social Science Edition), 4, 135–155.
Liu, B., Wang, N., Chen, M., Wu, X., Mo, D., Liu, J., & Xu, S., et al. (2017). Earliest hydraulic
enterprise in China, 5,100 years ago. Proceedings of the National Academy of Sciences, 114
(52), 13637–13642.
Liu, L. (2004). The Chinese Neolithic: Trajectories to Early States. Cambridge: Cambridge
University Press.
Liu, L. Chen, & X,. (2012). The Archaeology of China. Cambridge: Cambridge University Press.
Liu, X. (2016). The archaeological study on the distribution and assemblage of the crops of the
pre-Qin period and the Han dynasty, in Acta Archaeologica Sinica (China), 4, 465–465.
Matsumura, H., & Oxenham, M. F. (2014). Demographic transitions and migration in Prehistoric
East/Southeast Asia through the lens of nonmetric dental traits. American Journal of Physical
Anthropology, 155, 45–65.
Minagawa, M., Matsui, A., Nakamura, S., & Sun, G. (2011). Carbon and nitrogen stable isotope
analysis of human and animal remains from Tianluoshan site: to understand food resource
management of Hemudu culture. In Center for the Study of Chinese Archaeology, Peking
University and Zhejiang Province Institute of Archaeology and Cultural Heritage (Eds.),
Integrated Studies on the Natural Remains from Tianluoshan (pp. 262–269).
9 Why Rice Farmers Don’t Sail … 189
Miyamoto, K. (2016). Archaeological explanation for the diffusion theory of the Japonic and
Koreanic Language. Japanese Journal of Archeology, 4(1), 53–75.
Miyamoto, K. (2019). The spread of rice agriculture during the Yayoi Period: From the Shandong
Peninsula to the Japanese Archipelago via the Korean Peninsula. Japanese Journal of
Archeology, 6(2), 109–124.
Nasu, H., Momohara, A., Yasuda, Y., & He, J. (2007). The occurrence and identification of
Setaria italica (L.) P. Beauv. (foxtail millet) grains from the Chengtoushan site (ca. 5800 cal
BP) in central China, with reference to the domestication centre in Asia. Vegetation History
and Archaeobotany, 16(6), 481–494.
Nasu, H., Gu, H. B., Momohara, A., & Yasuda, Y. (2012). Land-use change for rice and foxtail
millet cultivation in the Chengtoushan site, central China, reconstructed from weed seed
assemblages. Archaeological and Anthropological Sciences, 4(1), 1–14.
Pawley, A. (2003). The Austronesian dispersal: Languages, technologies and people.
In P. Bellwood & C. Renfrew (Eds.), Examining the farming/language dispersal hypothesis
(pp. 251–274). Cambridge: McDonald Institute for Archaeological Research.
Paz, V. (2003). Island Southeast Asia: Spread or friction zone? In P. Bellwood & C. Renfrew
(Eds.), Examining the farming/language dispersal hypothesis (pp. 275–286). Cambridge:
McDonald Institute for Archaeological Research.
Qin, L., Fuller, D. Q., & Zhang, H. (2010). Modelling wild food resource catchments amongst
early farmers: Case studies from the Lower Yangtze and Central China. Quaternary Sciences,
30(2), 245–261 (DiSiJi YanJiu, in Chinese).
Qin, L. (2012). Archaeobotany in the research of agriculture origin in China, current status and
future prospects. In School of Archaeology and Museology, Peking University, Center for the
Study of Chinese Archaeology, Peking University (Eds), A collection of Studies on
Archaeology IX (pp. 260–315). Beijing: Cultural Relics Press (in Chinese).
Qin, L. (2013). The Liangzhu culture. In A. P. Underhill (Ed.), A Companion to Chinese
Archaeology (pp. 574–595). Malden: Wiley-Blackwell.
Qiu, Z., Jiang, H., & Ding, J. (2014) Phytolithic analysis of Majiabang period paddy fields at
Jiangli site of Kunshan, Suzhou. In Institute of Culture Heritage of Shandong University ed.
DongFang Kaogu (East Arcaheology) (Vol. 11, 376–384), Beijing: Science Press.
Qiu, Z., Shang, X., Ferguson, D. K., & Jiang, H. (2016). Archaeobotanical analysis of diverse
plant food resources and palaeovegetation at the Zhumucun site, a late Neolithic settlement of
the Liangzhu Culture in east China. Quaternary International, 426, 75–85.
Renfrew, C., & Liu, B. (2018). The emergence of complex society in China: The case of Liangzhu.
Antiquity, 92(364), 975–990.
Rindos, D. (1980). Symbiosis, instability, and the origins and spread of agriculture: A new model.
Current Anthropology, 21(6), 751–772.
Rindos, D. (1984). The origins of agriculture: An evolutionary perspective. New York: Academic
Press.
Robbeets, M. (2017a). Proto-Transeurasian: Where and when? Man in India: An International
Journal of Anthropology, 97(1), 19–46.
Robbeets, M. (2017b). Austronesian influence and Transeurasian ancestry in Japanese: A case of
farming/language dispersal? Language Dynamics and Change, 7(2), 210–251.
Ruthenberg, H. (1976). Farming systems in the tropics (2nd ed.). Oxford: Clarendon Press.
Sagart, L. (2005). Sino-Tibeto-Austronesian: An updated and improved argument. The peopling of
East Asia: Putting together archaeology (pp. 161–176). Routledge Curzon: Linguistics and
Genetics.
Sagart, L. (2008). The expansion of Setaria farmers in East Asia: A linguistic and archaeological
model. In A. Sanchez-Mazas, R. Blench, M. Ross, I. Peiros, & M. Lin (Eds.), Past human
migrations in East Asia: Matching archaeology, linguistics and genetics (pp. 133–157).
London: Routledge.
Sagart, L. (2011). How many independent rice vocabularies in Asia? Rice, 4(3–4), 121–133.
Sahlins, M. (1972). Stone age economics. London: Tavistock Publications.
190 L. Qin and D. Q. Fuller
Saito, K., Linquist, B., Keobualapha, B., Phanthaboon, K., Shiraiwa, T., & Horie, T. (2006).
Cropping intensity and rainfall effects on upland rice yields in northern Laos. Plant and Soil,
284(1–2), 175–185.
Scott, J. (2009). The art of not being governed: An anarchist history of upland Southeast Asia.
New Haven: Yale University Press.
Shanghai Cultural Heritage Bureau. (1985). Tang Miao Cun Site at Songjiang of Shanghai. Kaogu
(Archaeology), 7, 584–594.
Shennan, S. (2018). The first farmers of Europe: An evolutionary perspective. Cambridge:
Cambridge University Press.
Sherman, D. G. (1990). Rice, Rupees, and Ritual: Economy and society among the Samosir Batak
of Sumatra. Stanford: Stanford University Press.
Shoda, S., Lucquin, A., Ahn, J. H., Hwang, C. J., & Craig, O. E. (2017). Pottery use by early
Holocene hunter-gatherers of the Korean peninsula closely linked with the exploitation of
marine resources. Quaternary Science Reviews, 170, 164–173.
Silva, F., Stevens, C. J., Weisskopf, A., Castillo, C., Qin, L., Bevan, A., et al. (2015). Modelling
the geographical origin of rice cultivation in Asia using the rice archaeological database. PLoS
ONE, 10(9), e0137024.
Smith, M. L., & Mohanty, R. K. (2018). Monsoons, rice production, and urban growth: The
microscale management of ‘too much’ water. The Holocene, 28(8), 1325–1333.
Spriggs, M. (2011). Archaeology and the Austronesian expansion: Where are we now? Antiquity,
85(328), 510–528.
Stevens, C. J., & Fuller, D. Q. (2017). The spread of agriculture in Eastern Asia: Archaeological
bases for hypothetical farmer/language dispersals. Language Dynamics and Change, 7(2),
152–186.
Sun, G. (2013). Recent research on the Hemudu Culture and the Tianluoshan site. In A. Underhill
(Ed.), A companion to Chinese archaeology (pp. 555–573). Oxford: Wiley-Blackwell.
Tsang, C. H. (2005). Recent discoveries at the Tapenkeng culture sites in Taiwan: Implications for
the problem of Austronesian origins. In L. Sagart, R. Blench, & A. Sanchez-Mazas (Eds.), The
Peopling of East Asia: Putting together archaeology, linguistics and genetics (pp. 87–102).
London: Routledge.
Tsang, C. H., Li, K. T., Hsu, T. F., Tsai, Y. C., Fang, P. H., & Hsing, Y. I. C. (2017). Broomcorn
and foxtail millet were cultivated in Taiwan about 5000 years ago. Botanical Studies, 58(1), 3.
Vincent, V. (1954). The Cultivation of Floodland (Paddy) Rice. Rhodesia Agricultural Journal, 51
(4), 287–292.
Wang, F., Fan, R., Kang, H., Jin, G., Luan, F., Fang, H., Lin, Y., & Yuan, S. (2012).
Reconstructing the food structure of ancient coastal inhabitants from Beiqian village: Stable
isotopic analysis of fossil human bone. Chinese Science Bulletin, 57(12), 1037–1044.
Watabe, T. (1967). Glutinous rice in Northern Thailand. Kyoto: The Center For Southeast Asian
Studies, Kyoto University.
Weisskopf, A., Harvey, E., Kingwell-Banham, E., Kajale, M., Mohanty, R., & Fuller, D. Q.
(2014). Archaeobotanical implications of phytolith assemblages from cultivated rice systems,
wild rice stands and macro-regional patterns. Journal of Archaeological Science, 51, 43–53.
Weisskopf, A., Qin, L., Ding, J., Ding, P., Sun, G., & Fuller, D. Q. (2015). Phytoliths and rice:
From wet to dry and back again in the Neolithic Lower Yangtze. Antiquity, 89(347), 1051–
1063.
Whitman, J. (2011). Northeast Asian Linguistic Ecology and the Advent of Rice Agriculture in
Korea and Japan. Rice, 4, 149–158.
Wu, M., Ge, W., & Chen, Z. (2016). Diets of a late Neolithic maritime settlement: Carbon and
nitrogen stable isotope analysis of human bones from the Tanshishan site. Acta Anthropologica
Sinica, 35(2), 246–256.
Yang, X., Barton, H. J., Wan, Z., Li, Q., Ma, Z., Li, M., et al. (2013). Sago-type palms were an
important plant food prior to rice in southern subtropical China. PLoS ONE, 8(5), e63148.
Yang, X., Wang, W., Zhuang, Y., Li, Z., Ma, Z., Ma, Y., et al. (2017). New radiocarbon evidence
on early rice consumption and farming in South China. The Holocene, 27(7), 1045–1051.
9 Why Rice Farmers Don’t Sail … 191
Yang, X., Chen, Q., Ma, Y., Li, Z., Hung, H. C., Zhang, Q., et al. (2018). New radiocarbon and
archaeobotanical evidence reveal the timing and route of southward dispersal of rice farming in
south China. Science Bulletin, 63(22), 1495–1501.
Zhang, X. (2003). Study on the diet of ancient people by analyzing bone elements and isotopes.
Acta Anthropologica Sinica, 22(1), 75–84.
Zhang, Y. (2018). Exploring the Wetland: Integrating the Fish and Plant Remains into a Case
Study from Tianluoshan, a Middle Neolithic Site in China. In E. Pişkin, A. Marciniak, & M.
Bartkowiak (Eds.), Environmental archaeology (pp. 199–227). Current Theoretical and
Methodological Approaches: Springer.
Zhang, G., Jiang, L., Hu, Y., Si, Y., Song, G., Wang, C., & Richard, M. P. (2015). Carbon and
nitrogen stable isotope analysis of human and aninal bones from Tanshan site in Zhejiang
province. Hua Xia Kao Gu (Huaxia Archaeology), 2, 138–146.
Zhang, Y., Yuan, J., Huang, Y., Matsui, A., & Sun, G. (2011). A preliminary study of mammal
remains from Tianluoshan site in fiscal year 2004. In Center for the Study of Chinese
Archaeology, Peking University and Zhejiang Province Institute of Archaeology and Cultural
Heritage (Eds.), Integrated studies on the natural remains from Tianluoshan (pp. 79–124).
Beijing: Wenwu Press (in Chinese).
Zhao, K., Tung, C. W., Eizenga, G. C., Wright, M. H., Ali, M. L., Price, A. H., & Norton, G.J.,
et al. (2011). Genome-wide association mapping reveals a rich genetic architecture of complex
traits in Oryza sativa. Nature Communications, 2, 467.
Zhao, W., Tian, G., Wang, B., Forte, E., Pipan, M., Lin, J., et al. (2013). 2D and 3D imaging of a
buried prehistoric canoe using GPR attributes: A case study. Near Surface Geophysics, 11(4),
457–464.
Zhejiang Provincial Institute of Archaeology and Culture Relics and Xiaoshan Museum. (2004).
Kua Hu Qiao. Beijing: Wenwu Press.
Zhejiang Provincial Institute of Archaeology and Culture Relics, Huzhou Museum. (2006). Pi
Shan. Beijing: Wenwu Press.
Zheng, Y., Sun, G., Qin, L., Li, C., Wu, X., & Chen, X. (2009). Rice fields and modes of rice
cultivation between 5000 and 2500 BC in east China. Journal of Archaeological Science, 36
(12), 2609–2616.
Zhuang, Y., Ding, P., & French, C. (2014). Water management and agricultural intensification of
rice farming at the late-Neolithic site of Maoshan, Lower Yangtze River, China. The Holocene,
24(5), 531–545.
Part II
Prehistoric Seafaring and Exchange: From
Coastal Waters to the Open Sea
Chapter 10
Social Reciprocity Facilitated Overseas
Exchange in Early Austronesian
Cultures
Barry V. Rolett
10.1 Introduction
The European historian Ferdinand Braudel, whose concept of deep time (the longue
durée), is widely invoked in the study of Oceanic prehistory, defined two
contrasting modes of maritime trade (Braudel 1972: 103–108) useful for
understanding the ancient Austronesian world. First, Braudel’s “tramping” model is
based on the notion of coastal travel with limited open-sea voyaging. Ships engaged
in tramping typically made circuitous and opportunistic trading voyages with many
B. V. Rolett (&)
Department of Anthropology, University of Hawaii, Honolulu, HI, USA
e-mail: rolett@hawaii.edu
stops. The composition of a tramping ship’s cargo changed significantly during the
course of a single voyage. By contrast, his second model of destination-conscious
commerce consists of shipping commodities, such as tea or porcelains, from their
place of origin to one or more specific destinations. Unlike tramping,
destination-conscious commerce implies planned trade routes, the transport of large
quantities of goods, and regular contacts at a frequency based on supply and
demand. Braudel’s models were formulated to describe Mediterranean maritime
trade during the medieval era. However, they are also relevant for the analysis of
seafaring and overseas exchange in other parts of the world. In this chapter, I use
these models to help interpret archaeologically-documented examples of Neolithic
and Metal Age maritime trade centered on the Taiwan Strait.
Let us begin with two case studies, both of which rely upon the chemical
characterization of stone artifacts as evidence for ancient voyaging networks. The
first study, by Hung and colleagues (Hung et al. 2007), demonstrates that nephrite
jade from the Fengtian Quarry in Taiwan was exchanged widely across Southeast
Asia from around 500 BC to AD 500. Their analysis suggests that the transport of
small quantities of Taiwanese jade across a vast area, spanning from Island
Southeast Asia (ISEA) to southern Vietnam, occurred within the context of inter-
action between culturally and linguistically-related Austronesian peoples. Another
well documented example of the overseas transport of lithic artifacts in the Taiwan
area involves a type of stone tool that was mass-produced on Qimei, a volcanic
island in the Penghu Archipelago (Taiwan Strait) (Figs. 10.1, 10.2). Around 2500
to 1500 BC, tools made of Qimei basalt were transported in large quantities to
Taiwan, where they are quite common in Neolithic sites dating to that time (Rolett
et al. 2000). Qimei lies about 45 km off the coast of Taiwan, in the Taiwan Strait.
On the face of it, these examples involving the quarrying and overseas transport
of artifacts seem quite different. For instance, the jade artifacts were transported
from Taiwan to other locations, while the basalt artifacts were transported from
Qimei Island to Taiwan. Moreover, Fengtian jade is a precious material used mainly
for producing prestige goods such as lingling-o ear ornaments (Fig. 10.3), while the
Qimei basalt was used for making utilitarian tools for activities like digging or
woodworking. Jade was moved from Taiwan to many different and distant loca-
tions, while it appears the Qimei basalt was moved to a single location, the
southwest coast of Taiwan. Finally, the jade artifacts were moved in small numbers
and they are rare where they are found, while the basalt tools were moved in
strikingly large numbers.
Yet despite these differences, there is one significant similarity - both case
studies apparently document exchange within the context of Austronesian-speaking
communities. This is especially clear for the Qimei basalt, as the Neolithic settle-
ment of Taiwan and the Penghu Islands is closely associated with the origins of
Austronesian language and culture (e.g., Bellwood 2005). The Fengtian jade
artifacts date to a later period, after Austronesian farmers had dispersed across
10 Social Reciprocity Facilitated Overseas Exchange … 197
Fig. 10.1 Overview of the Nankang Quarry Site on Qimei, Penghu Islands (Taiwan Strait). Cliffs
in left middle ground show the source of the quarried basalt. The ground surface is covered by
dense accumulations of debitage from the manufacture of stone tools. Another volcanic island in
the Penghu Archipelago is visible on the horizon. Photo by B. Rolett
ISEA and had thoroughly displaced the original hunter-gatherer inhabitants of that
region (Hung et al. 2007). Notably, the Fengtian jade artifacts discovered in
Vietnam derive from historic-era coastal sites associated with Austronesian ethnic
groups. This reinforces the notion of an ancient interaction sphere that linked
Austronesian-speaking communities in ISEA with related populations located on
the mainland (Hung et al. 2007; Yamagata and Matsumura 2017).
198 B. V. Rolett
Fig. 10.2 Profile view of the dense accumulation of debitage at the Nankang Quarry Site on
Qimei, Penghu Islands (Taiwan Strait). The deposit consists almost entirely of chipped stone.
A tool preform (flat and rectangular in shape) protrudes from the profile to the right of the tape
measure at the 10 cm mark. Photo by B. Rolett
In the context of Braudel’s models for maritime trade, the exchange of Taiwanese
jade documented by Hung et al. (2007) most closely resembles tramping due to the
evidence of small cargoes and coastal travel with many stops. The clearly different
trade in Qimei basalt represents a form of destination-conscious shipping based on
the apparent movement of large quantities of stone tools from a single place of
origin (the Penghu Islands) to a specific destination. The landscape of that desti-
nation, on the southwest coast of Taiwan, stands out because of its particular
composition: a mass of uplifted limestone with no naturally occurring volcanic rock
of the kind found in abundance throughout the Penghu Islands (Rolett et al. 2000).
Is it possible to further reconstruct the exchange systems involved in these
remarkable examples of ancient Austronesian seafaring? What was the social and
economic context in which jade ornaments, basalt tools and other objects changed
hands? Scholars generally view the jade and basalt as evidence for “exchange,”
although what this actually means has been left largely unspecified. In reference to
jade, Hung and colleagues suggest that “the ear ornaments of Fengtian nephrite in
Southeast Asia (outside Taiwan) were made by a small number of highly skilled
10 Social Reciprocity Facilitated Overseas Exchange … 199
Fig. 10.3 Nephrite lingling-o ear ornaments excavated from archaeological sites in the
Philippines and Vietnam (Hung et al. 2007)
and perhaps itinerant jade craftsmen” (2007: 19749). They further propose that the
craftsmen, “with or without the help of transporting middlemen, carried or acquired
their raw materials from Taiwan, then traveled and/or resided along the shorelines
of the South China Sea to produce extremely similar jade ornaments to suit the
demands of local elites” (2007: 19749). Although the “itinerant craftsmen” model is
an appealing explanation for the exchange of jade, it cannot likewise account for the
exchange of Qimei basalt tools since those artifacts were manufactured on Qimei
itself before being transported to Taiwan (Rolett et al. 2000).
Instead, an alternative but complementary model that helps explain the exchange
of both jade and basalt derives from an integrated historical and comparative
approach. It requires the acceptance of two assumptions. First, we must assume that
the exchange in question occurred within the cultural context of Austronesian
societies. As noted above and explained in more detail by Bellwood (2005), there is
a strong basis for this assumption. The second assumption is that “true money” as
we know it, or as defined as a universal measure of value and a universal means of
payment (Pospisil 1963), did not exist in ancient Austronesian societies. This
second assumption is also well supported. For example, in explaining semantic
evidence for a reconstruction of the linguistic term PMP (Proto-Malayo-
Polynesian) *beli, Blust and Trussel (2010) argue: “There is no known evidence
of any kind that speakers of PMP (circa 3,000 BC) were familiar with a money
economy.” True money was unknown throughout the Austronesian world until
colonization and the introduction of a money economy. Moreover, the antiquity of
200 B. V. Rolett
colonization in the history of this area (early in the second millennium for Taiwan
and during the 1500s for ISEA), together with the paucity of early written records,
effectively precludes the use of textual sources in reconstructing pre-colonization
economies. Even the role of barter, defined here as “the behavior of persons who
exchange goods on the assumption that each makes the most of it” (Polanyi 1977:
42), is unknown.
How was exchange facilitated in the pre-money Austronesian economies of
ISEA and Taiwan? In the lack of archaeological evidence to help answer this
question, the disciplines of historical linguistics and comparative ethnography
suggest some answers. As explained by Blust (2017):
The inferences of historical linguistics… are not limited to those aspects of material culture
that survive into the present… In addition, the entire realm of non-material culture,
including systems of kinship, marriage rules, social organisation, and ideas about the spirit
world are accessible to inferences based on cognate vocabulary, but lie beyond the pale of
confident archaeological investigation (275–276).
Such inferences are made by finding cognate words in languages that represent
primary branches of the family or subgroup whose immediate ancestor is being
reconstructed. For example, an innovative study of ancient Austronesian societies
by Kirch and Green (2001) uses historical linguistics to infer the presence of chiefs
and spiritual priests in the Ancestral Polynesian homeland. In the same way, Blust
and Trussel (2010) reconstruct the notion of reciprocity (*bales: meaning to answer,
retaliate; reciprocate good or evil) as a Proto-Austronesian concept, demonstrating
that it existed in the earliest Austronesian societies around 5000 years ago. Also of
interest, and dating to the earliest Austronesian dispersal into ISEA, is the term
PMP* be Ray, which Blust and Trussel (2010) gloss as “give, present gifts to; gift.”
Both of these concepts are closely related to the practice of gift exchange, which
was integral to traditional Polynesian societies. The specific term for gift exchange
(*sau: return for services or to a gift) can be attributed to Ancestral Polynesian
society around 2000 to 3000 years ago (Kirch and Green 2001: 221). Notably,
Polynesian cultures, which evolved among Austronesians who settled remote
islands in the East Pacific, derive from the same ancestral foundation that gave rise
to Austronesian cultures in Southeast Asia. However, Polynesians and their
immediate ancestors were largely isolated from other Austronesians for around
3000 years ago prior to the time of Western contact. Because of this isolation,
Polynesian cultures preserved certain Austronesian traditions that may have been
lost due to the influence of contact with other Asian cultures in ISEA. Thus,
Polynesian cultural practices documented at the time of Western contact can pro-
vide insight into early Austronesian cultural practices for which there are no written
records.
10 Social Reciprocity Facilitated Overseas Exchange … 201
Unlike Taiwan and ISEA, most Polynesian cultures were not colonized until the
early nineteenth-century. Thus, a rich body of historic literature provides descrip-
tions of contact-period cultures throughout Polynesia. Comparative ethnography
draws upon this literature and can complement and enhance the results of historical
linguistics, as Kirch and Green (2001) have illustrated with their reconstruction of
Ancestral Polynesian society. Indeed, comparative ethnography allows detailed
reconstructions of ancient Polynesian economies and social systems.
For understanding gift exchange, in particular, there is no better starting point
than Douglas Oliver’s (1989) unparalleled ethnographic study of ancient Tahitian
society. Oliver concludes that goods-focused trade was rare in ancient Tahiti, and
what little existed “took place, not at central marketplaces (for there were none) but
mainly within the contexts of ties of kinship and ‘friendship’ (i.e. simulated kin-
ship) - in other words, in the form of gift exchange” (566). Raymond Firth, another
eminent ethnographer of Polynesia, described the essence of Polynesian gift
exchange as the principle of gift and counter-gift, meaning that “for every gift
another of at least equal value should be returned” (Firth 1959: 412–413). The
notion of obligation was quite clear, and it also required that a gift must be accepted
graciously. Failure to do so, or negligence in making an acceptable counter-gift, led
to a loss of prestige (Firth 1959: 423). Traditional Polynesian gift exchange served
as a mechanism for the distribution of resources and as a means for establishing and
maintaining both social bonds and hierarchical relationships. It occurred among
people of all social ranks and on a variety of occasions, from the obviously cere-
monial to the economically motivated. Although the specific nature of gift
exchange varied among Polynesian cultures, the practice itself played an important
role throughout Polynesia.
Oliver (1989: 588) distinguishes among egalitarian, competitive and coercive
gift exchange. Egalitarian exchanges involve reciprocity in which the gift and
counter-gift are perceived to be roughly equal in value. By contrast, competitive
exchanges are intentionally imbalanced, through the effort of one party to give with
such extravagance that the receiver cannot command the resources necessary to
reciprocate with a gift of equal value. Oliver associates these exchanges with the
motivation to establish political influence by imparting a sense of unfulfilled
obligation upon the receiver. Coercive exchanges differ in that one party offers a
present as a way of coercing the receiver to reciprocate, and the typical context is
one in which the “coercer” desires a specific item.
Firth’s analysis of gift exchange among New Zealand’s Maori draws a broad
distinction between economic gift exchange, motivated by the desire “to acquire
something of practical utility,” and ceremonial gift exchange, in which “the
transaction fulfils some wider social purpose, the acquisition of goods not being the
principal motive” (1959: 402). Economic exchange was stimulated in part by craft
specialization and the uneven distribution of natural resources. For example,
202 B. V. Rolett
interaction among coastal and inland Maori involved the exchange of seaweed and
shells for forest products. Evidence of economic exchange is also shown by the
widespread transport, in New Zealand as well as throughout tropical Polynesia, of
adzes made of fine-grained rock (e.g., Rolett 1998; Walter et al. 2017) (Fig. 10.4).
New Zealand jade only occurs naturally in a restricted area of the South Island.
As in Taiwan, the stone was highly valued for both adzes and ornaments and it was
transported to even the most distant communities by means of exchange. Maori
living outside the source regions organized expeditions to those areas, where they
acquired nephrite in exchange for food delicacies, fine mats and garments (Firth
1959: 407). Some of the finest jade ornaments, including amulets (hei tiki) and
whale-tooth-shaped pendants (rei niho), were treasured as heirlooms and exchanged
on ceremonial occasions such as those honoring the death of a chief (Firth 1959:
414–415).
In summary, traditional Polynesian gift exchange can be defined as a form of
reciprocity where the primary value lies in creating and maintaining social bonds
rather than in immediate economic gain. The pattern of exchange in this case is
asymmetrical, since the gift and counter-gift are often qualitatively different.
Exchange objects included food and prestige goods (e.g., ornaments and other
emblems of high status), as well as materials and utilitarian implements. Gift
exchange played a crucial role when individuals crossed community boundaries,
but it was also likely practiced within communities. Three elements of gift
Fig. 10.4 A Polynesian stone adze discovered at the Hanamiai archaeological site (Rolett 1998),
Marquesas Islands (French Polynesia). Adzes made of fine-grained volcanic rock figured
prominently in interisland exchange. Photo by D. Hazama
10 Social Reciprocity Facilitated Overseas Exchange … 203
10.4 Discussion
The broad distribution of gift exchange suggests that it, like certain other geo-
graphically widespread aspects of Polynesian culture, may derive from ancestral
traditions rooted in the distant past (Kirch and Green 2001). Although the specific
institution of gift exchange, as it is described here, is unknown outside Polynesia,
an extraordinary system with striking parallels was recorded by Bronislaw
Malinowski (1922) in the Trobriand Islands. The Austronesian-speaking Trobriand
Islanders live on small islands off the coast of New Guinea. Malinowski’s
book-length description and analysis of the Trobriand Kula exchange demonstrates
that social reciprocity and exchange were as inextricably intertwined in the
Trobriands as in Polynesia.
Kula exchange linked communities on islands separated by sailing distances of
hundreds of miles, and was sustained by “permanent and lifelong” partnerships
between men in different communities (Malinowski 1922: 83). It involved
preparing for and carrying out major overseas expeditions on a regular basis, along
pre-planned routes. The central act involved the ceremonial presentation of gifts
consisting of shell necklaces and bracelets. Although reciprocity was required for
the sake of honor and for maintaining the relationship between trading partners, the
presentation of a return gift could be and often was delayed (Malinowski 1922: 85,
95). This Kula exchange of ceremonial valuables went hand-in-hand with barter
and ordinary trade, so that” every outbound canoe carries a cargo of such things as
are known to be most desirable in the overseas district” (Malinowski 1922: 100).
Kula exchange differs from Polynesian gift exchange in that it is essentially
symmetrical—reciprocity involves the circulation of the same kind of ceremonial
valuables. However, the two systems are remarkably similar in the way that acts of
ceremonial exchange create bonds of social reciprocity. In the complete absence of
“true money,” these bonds of social reciprocity set the stage for, and ultimately
facilitated, myriad forms of utilitarian exchange.
The cumulative body of evidence for social reciprocity in Austronesian
exchange systems can be summarized as follows: (1) Historical linguistics recon-
structs reciprocity as a social concept in the Proto-Austronesian world (centered on
Taiwan) and demonstrates that a specific term for gift giving existed by the time of
the earliest Austronesian dispersal into ISEA; (2) Lexical terms for gift exchange
and richly detailed records of the practice collected from multiple island groups
suggest that the characteristically Polynesian system of gift exchange existed in
Ancestral Polynesian society; and (3) Similar to Polynesian gift exchange, the Kula
204 B. V. Rolett
with limited resources for agriculture, maintaining social bonds with communities
on Taiwan would have been more important than narrowly defined motivations
associated with economic gain.
References
Beaglehole, J. C. (Ed.). (1967). The Journals of Captain James Cook on his Voyages of Discovery.
Volume III. The Voyage of the Resolution and Discovery 1776–1780 (2 vols.). Cambridge:
Cambridge University Press.
Bellwood, P. (2005). First farmers: The origins of agricultural societies. Malden, MA: Blackwell.
Blust, R. (2017). Historical linguistics and archaeology: An uneasy alliance. In P. J. Piper, H.
Matsumura, & D. Bulbeck (Eds.), New perspectives in Southeast Asian and Pacific Prehistory,
Terra Australis 45 (pp. 275–291). Canberra: Australian National University Press.
Blust, R., & Trussel, S. (2010) (revision 1/14/2018). Austronesian Comparative Dictionary. www.
trussel2.com/ACD.
Braudel, F. (1972). The Mediterranean and Mediterranean World in the age of Philip II (Vol. I).
New York: Harper and Row.
Dening, G. (Ed.). (1974). The Marquesan Journal of Edward Robarts, 1797–1824. Honolulu:
University of Hawaii Press.
Firth, R. (1959). Economics of the New Zealand Maori (2nd ed.). Wellington: R. E. Owen,
Government Printer.
Hung, H. C., Iizuka, Y., Bellwood, P., Nguyen, K. D., Bellina, B., Silapanth, P., & Dizon, E., et al.
(2007). Ancient jades map 3,000 years of prehistoric exchange in Southeast Asia. Proceedings
of the National Academy of Sciences 104(50), 19745–19750.
Kirch, P. V., & Green, R. C. (2001). Hawaiki, Ancestral Polynesia: An essay in historical
anthropology. Cambridge: Cambridge University Press.
Malinowski, B. (1922). Argonauts of the Western Pacific: An account of native enterprise and
adventure in the archipelagoes of Melanesian New Guinea. London: Routledge & K. Paul Ltd.
Oliver, D. L. (1989). Oceania: The native cultures of Australia and the Pacific Islands (Vol. 1).
Honolulu: University of Hawaii Press.
Polanyi K. (1977). The Livelihood of Man. In H. W. Pearson (Ed.). Cambridge, MA: Academic
Press.
Pospisil, L. (1963). Kapauku Papuan Economy, Yale University Publications in Anthropology 67.
New Haven: Yale University Press.
Rolett, B. V. (1998). Hanamiai: Prehistoric colonization and cultural change in the Marquesas
Islands (East Polynesia), Yale University Publications in Anthropology 81. New Haven: Yale
University Press.
Rolett, B. V., Chen, W. C., & Sinton, J. M. (2000). Taiwan, Neolithic seafaring and Austronesian
origins. Antiquity, 74, 62–74.
Walter, R., Buckley, H., Jacomb, C., & Matisoo-Smith, E. (2017). Mass migration and the
Polynesian settlement of New Zealand. Journal of World Prehistory, 30(4), 351–376.
Yamagata, M., & Matsumura, H. (2017). Austronesian migration to Central Vietnam: Crossing
over the Iron Age southeast Asian sea. In P. J. Piper, H. Matsumura, & D. Bulbeck (Eds.), New
perspectives in Southeast Asian and Pacific Prehistory, Terra Australis 45 (pp. 333–355).
Canberra: Australian National University Press.
Chapter 11
Perspectives on Early Holocene
Maritime Ethnic Groups of the Taiwan
Strait Based on the “Liangdao Man”
Skeletons
Chun-Yu Chen
Abstract Two “Liangdao Man” skeletons from the Early Holocene, found on
Liangdao (亮岛) Island of the Matsu (马祖) archipelago, are the oldest human
remains ever found in the Taiwan Strait. This discovery has attracted considerable
academic interest, especially in the fields of archaeology and anthropology.
Scholars in several different disciplines have conducted studies of “Liangdao Man.”
Among these were the author and the Matsu Archaeological Team, who carried out
three fruitful field seasons of excavation. The most notable outcome of this field-
work was the successful extraction of both mitochondrial and Y chromosome DNA
from “Liangdao Man 1.” Results from the study of the ethnic origins of “Liangdao
Man” are presented here, along with contemporaneous maritime activities in the
Taiwan Strait.
11.1 Introduction
Located near the estuary of the Min River of Fujiang (福建) Province, mainland
China, the Matsu Islands, fall under the jurisdiction of Taiwan. On 2011, the author
and the Matsu Archaeological Team discovered the Liangdao Daowei (岛尾) Site,
one of four local clusters on northeastern part of Liang Island (Chen et al. 2012: 1,
2). Two human skeletons unearthed in the excavations of the Liangdao Daowei I
Site were named “Liangdao Man 1” and “Liangdao Man 2.” Based on C14 dating,
“Liangdao Man 1” dates to about 8,320-8,160 BP, while “Liangdao Man 2” dates to
about 7,590-7,560 BP (Chen 2013: 120). DNA studies identified “Liangdao Man 1”
with haplogroup E, and “Liangdao Man 2” with haplogroup R9 (Chen 2013: 51–
58). Haplogroups E and R9 have also been associated with some indigenous groups
Although part of the skull of “Liangdao Man 1” is damaged, the remainder of the
skeleton is well preserved, or about 70% intact. The skull, teeth, limbs, and bone
structure all appear to be in good condition. In the burial, the head was oriented to
the north and the body lay on its side with limbs flexed (Fig. 11.1). The flexed
burial position has been identified as a special cultural tradition of the Early
Neolithic Age on the southeast coast of China and Southeast Asia. Thus, the
presence of this local cultural phenomenon indicates that some branches of
indigenous peoples were living in this region. In Guangxi (广西), most of the
burials associated with the Dingshishan (顶狮山) Culture of Yongning (邕宁)
County are flexed position and date to 8000-7000 BP. The same burial custom was
found in excavations of the Qiujiang (秋江) Site in Hengxian (横县) and the
Beidaling (北大岭) Site in Du’an (都安) County, Guangxi (GXAI CASS 1998; Qin
2010). In the coastal region of Guangdong, the Xiankezhou (蚬壳洲) Site in
Gaoyao (高要) County,the Liyudun (鲤鱼墩) Site in Suixi (遂溪) County, and
other Neolithic sites dating to 6000-5000 BP also exhibited this special flexed burial
position (GDPM GYCAB 1990; GDPICRA et al. 2015). Archaeological evidence
indicates that the same custom was continued by indigenous peoples in Taiwan
including the Paiwan (排湾), Amei (阿美), and Yamei (雅美) and appears to have
been a primary cultural characteristic of early Austronesians (Qin 2010).
A deep, narrow sciatic notch, a particular skull appearance, and stout limbs are all
common characteristics of male individuals, and based on bone morphology
“Liangdao Man 1” appears to have been male. The upper and lower jaw show com-
plete third molar eruption and noticeable molar wear; the epiphysis indicates healing
of bone damage; and both the length of the limbs and the parietal sagittal suture
condition indicate adulthood. The age at the time of death appears to have been about
30 years. The ridge line and the thickness of the bone suggest medium-sized,
well-developed skeletal muscles and show that this individual was active with a strong
body and especially strong limbs. The man’s height of 160 ± 3.59 cm (Mo 1983: 84,
85) is within the height range of modern Han (汉) Chinese males.
The “Liangdao Man 2” skeleton is also well preserved, about 85–90% intact, but
was buried in the extended position (Fig. 11.2). Preliminary examinations revealed
that this individual did not have particularly well developed limbs and had a small
greater sciatic notch angle. “Liangdao Man 2” had a somewhat strong physique, but
the skull has no apparent male characteristics and the muscular ridges of the limbs
are not particularly well developed. Compared to “Liangdao Man 1,” this skeleton
seems underdeveloped and most likely belonged to a female. The skull sutures and
the extent of tooth wear indicate that this individual was approximately 30 years old
at the time of death. The individual’s height appears to have been about 165 to
169 cm.
DNA extraction for the “Liangdao Man” remains was done by Professor Mark
Stoneking and his team at the Max Planck Institute (MPI) Division of Evolutionary
Genetics in Germany. According to their analysis, the best haplogroup match for
“Liangdao Man 1” is haplogroup E, specifically between E and E1. Haplogroup E is
not found in today’s mainland Chinese population, however it has been identified
among indigenous Taiwanese and in inhabitants of the Philippines, Indonesia, and
some South Pacific islands. Of the locations where haplogroup E is found, Taiwan
is closest to Liangdao geographically as well as genetically, in terms of a complete
haplogroup type match. Indeed, indigenous Taiwanese peoples show a higher
frequency of lineages close to the root of haplogroup E than the other groups do,
including E1a and E1a1. All the haplogroup data currently available from the
Philippines and Indonesia show either substitutions or more differences. This dis-
covery supports the idea that coastal Fujian and Liangdao were source areas for
pre-Austronesian development about 8,300 years ago. More specifically, it offers
substantial evidence as to the timing and location of the separation between
pre-Austronesian and ancestral Chinese groups according to maternal lineage.
Ancient mtDNA was extracted from the phalanx of “Liangdao Man 2” and was
identified as belonging to haplogroup R9, with the specific sublineages R9b and
R9c. The haplogroup R9b occurs in about 4% of the indigenous Tsou (邹) Tribe of
Taiwan, but also occurs in 20-30% of mainland Chinese Dai (傣) and Hmong-Mien
(苗) populations and is occasionally found among Han Chinese. While the hap-
logroup R9c has its highest frequency (20%) in the Tsou Tribe, it also occurs in
other indigenous groups from Taiwan including the Bunun (布农), Puyuma (卑南),
and Rukai (鲁凯) Tribes. Populations of the Philippines, Indonesia, Malaysia, and
the Chinese Daic (傣) people illustrate similar frequencies of R9c (3–7%), but it is
rare among Han Chinese and Hmong-Mien peoples (Chen 2013: 50–59).
11 Perspectives on Early Holocene Maritime Ethnic Groups … 211
On the interface between East Asia and the Pacific Ocean, the coastline of China
stretches more than 18,000 km long. Aside from the mainland, this region also
includes 6,500 coastal islands of various sizes. Encompassing a range of different
geological strata including gravel deposits, the coastal region has a mild climate due
to its location between temperate and the sub-tropical zones.
Living in such a desirable ecological environment, the ancient coastal inhabi-
tants may have developed cultures comparable to the ancient inland cultures. As
early as 8,000 to 5,000 BP, prehistoric cultures were developing, from north to
south, in the gulf of Bohai (渤海), the Shandong (山东) Peninsula, the coast and
estuary of the Yangtze (长江) River, Hangzhou (杭州) Bay, the estuary of the
Minjiang (闽江) River in Fujian, the Pearl River Delta (珠江三角洲), and the
Beibuwan (北部湾) Gulf of Guangxi. The presence of ancient oceanic groups in
China’s coastal regions might moreover have started as early as the beginning of the
post-glacial era, or the Early Holocene. Given these dates and the local environ-
ment, subsistence was probably based on marine resources such as fish and shellfish
and supplemented with gathered terrestrial plants. The discovery of the “Liangdao
Man” individuals may provide insight on this question.
Even today, the Tanka (蜑家) of coastal China depend upon trading and fishing
much as ancient seafaring people once did. The ancient inhabitants probably stayed
close to the shore, erecting houses on stilts, as did other sea nomads along China’s
coasts. Indeed, Tanka is a term used more generally to refer to China’s ethnic
groups of sea nomads, although almost every area along the coast has distinct ethnic
groups or subgroups with different names. Thus, the group living in the estuary of
the Min River in Fujian Province are known as the Kedeh (曲蹄); the group in the
estuary of the Jiulongjiang (九龙江) River are the Bosui (白水); and the group in
the Pearl River Delta of Guangdong Province, also the largest maritime ethnic
group, are known as the Tanka (蜑家).
Fuzhou (福州), the capital of Fujian Province and home to the Kedeh, is located
in the estuary of the Minjiang River. According to ethnographic records, the Kedeh
have been isolated from Han Chinese society for long time. Before the Republic of
China was established in 1912, they were not allowed to live on land and they were
only allowed to wear clothes made from coarse materials, never fine silk clothing.
Kedeh children and adolescents were not allowed to receive formal education and
were forbidden from taking the official examinations for government jobs. Only
after 1912 did these restrictions gradually began to change.
Kedeh usually live in nuclear family units, and when young people get married
they build a new boat to establish their household. In the Minjiang River Basin, the
average inland boat is 5–6 meters long and 2–3 meters wide. The stern of these
boats is slightly narrower while, the middle is flat. A bamboo shelter encloses a
small living space that serves as a cabin. The Kedeh make their living ferrying
11 Perspectives on Early Holocene Maritime Ethnic Groups … 213
passengers, transporting freight, or fishing the river. Their lives are often extremely
difficult due to their low income level and poor living conditions. Some Kedeh who
live near the sea in the Minjiang River Estuary make their living fishing in the
ocean. Due to the demands and rewards of sea fishing, their vessels are larger, with
more compartments, and are shared by an extended family of brothers whose wives
and children help to provide the manpower necessary to operate a sea fishing vessel.
The Bosui (白水) are another maritime ethnic group that live in the Jiulongjiang
River Estuary and Basin of southern Fujian Province. Their customs, habits, living
conditions and social situation are similar to those of the Kedeh, although these two
groups fish two separate regions of the Taiwan Strait. Both the ethnographical
Kedeh and Bosui reveal the different cultural sub-traditions which might be orig-
inated from variety of maritime cultures along the coast of southeast China (Chen
2017).
11.4 Conclusion
Neolithic sites dating to more than 8,000 years ago have been discovered along the
coast of China. Canoes and wooden paddles have been found, including those with
sophisticated designs excavated at the Kuahuquiao (跨湖桥) Site, near Hangzhou
Bay in Zhejang Province (ZJPICRA and XSM 2004: 42–48). These discoveries
support the hypothesis that human activity near the coast began in the Early
Holocene. The scope of maritime activities practiced by these populations would
have been limited by their production of sailing tools. They must have developed
navigation skills gradually, initially staying within sight of land and later relying on
the locations of the sun and the constellations to find their way. The distance they
could travel was limited by sailing techniques and the amount of food and water
they were able to carry with them.
Analysis of DNA from the “Liangdao Man” skeletons shows that these indi-
viduals belonged to two different oceanic ethnic groups and indicates that two
separate ethnic groups were active on or living by the sea in the Taiwan Strait as
early as 8,000–7,500 years ago. Honglin Chiu has used morphometric analysis of
the cranium of “Liangdao Man 1” to illustrate an affinity to the “Minato-gawa Man”
from Okinawa, Japan (Chiu et al. 2015: 15–26), while Hirofumi Matsumura sug-
gests an affinity between “Liangdao Man 1” and the “Austrian-Papuan” group
(Matsumura et al. 2019). Although “Minato-gawa Man” was found on one of the
Pacific Islands of the eastern Asian continent and the “Austria-Papuan” group was
discovered on the islands of Southeast Asia, both of them belong to maritime ethnic
groups.
Meanwhile, Chiu and Matsumura consider “Liangdao Man 2” to have an affinity
to the southern Mongoloid group (Chiu et al. 2015: 15–26; Matsumura et al.
2017). Thus, it appears that different complex ethnic groups had emerged on the
southeastern coast of China and in the Taiwan Strait by about 8,300–7,500 years
ago.
214 C.-Y. Chen
In addition, DNA studies of the two “Liangdao Man” skeletons show that they
have maternal blood relationships with some modern Austronesians, illustrating the
origin of “Proto-Austronesians” in the Taiwan Strait. Influenced by the strong
farming culture of mainland China (Bellwood 2004), “Proto-Austronesian” mar-
itime groups multiplied and dispersed widely across the South Pacific, all the way
to Easter Island in the east and Madagascar, off the coast of East Africa, in the west.
Indigenous legends about Taiwan and Southeast Asia also document the dispersal
of these peoples from mainland China to Southeast Asia. According to the Moken
in Myanmar (Burma), their ancestors came from China 4,000 years ago (Wikipedia:
Moken), while similar legends are found among the Sama-Banju in Sarawak of
Sabah, Malaysia; among coastal dwellers in Mindanao, the Philippines; as well as
in northern and eastern Borneo, the Celebes, and throughout the eastern islands of
Indonesia (Wikipedia: Sama Bajau). All of these sea nomads may be descendants of
the Neolithic maritime groups that once lived along the southeast coast of
China (Due to short of time, this part better all deleted).
Acknowledgements Thanks to Edna and Simon Litten for their editorial assistance.
References
Bellwood, P. (2004). The earliest farmers: the origins of agricultural societies. Oxford: Blackwell
Publishing.
Chen, C. Y. (2013). Matsu Liangdao Man DNA studies. Lienjiang County Government Press
《
( 亮島人DNA研究》, 連江縣南竿鄉).
Chen, C. Y. (2017). On the original ethnic groups of ‘Koeh-da’ and ‘Bek-sui.’ In Proceedings of
International Symposium on Fujian Qihe Cave Site, sponsored by Fujian Museum, Changping
County, Fujian Province. 〈談福建曲蹄族與白水族的源流〉
( , 福建省漳平縣, 福建博物院主
辦,《福建奇和洞遺址國際學術研討會論文集》).
Chen, C. Y., Chiu, H. L., You, K. S., Yin, Y. C., Lin, F. Y. (2012). Test excavations at the
Liangdao-Daowei-I Site. Lianjiang, Nangang: Lianjiang County Government《馬祖亮島島尾 (
I遺址試掘》, 連江縣政府出版).
Chen, C. Y., Chiu, H. L., You, K. S., Yin, Y. C., Lin, F. Y. (2013). Project report: Excavations of
the Matsu Liangdao-Daowei Sites Cluster and sorting for the “Liangdao Man” skeletons.
Lianjiang, Nangang: The Lienjiang County Government Press《馬祖亮島島尾遺址群發掘及
(
『亮島人』修復計畫》, 連江縣政府出版).
Chiu, H. L., Chen, C. Y., Yu, C. F. (2015). Comprehensive studies of the Liangdao-Daowei Sites
Cluster. Hsinchu, Taiwan: Institute of Anthropology, National Chinghwa University 《馬祖亮 (
島島尾 遺址群綜合研究計畫》 ).
GDPICRA (Guangdong Provincial Institute of Cultural Relics and Archaeology), ZJMM
(Zhanjiang Municipal Museum), & SXCM (Suixi County Museum). (2015). Report on the
excavation of the Liyudun Neolithic Shell Midden, Suixi, Guangdong. Cultural Relics, 7, 4–18
〈广东遂溪鲤鱼墩新石器时代贝丘遗址发掘简报〉
( ,《文物》7).
GDPM GYCAB (Guangdong Provincial Museum, Gyaoyao Cultural Affairs Bureau). (1990).
Neolithic shell midden discovered at Xiankezhou, Gaoyao County, Guangdong. Archaeology,
6, 565–568 〈广东高要县蚬壳洲发现新石器时代贝丘遗址〉
( 《考古》6).
11 Perspectives on Early Holocene Maritime Ethnic Groups … 215
GXAI CASS (Guangxi Team of Archaeology Institute, Chinese Academy of Social Science).
(1998). The excavation of Dingshishan Site in Yongning County, Guangxi. Archaeology, 11,
11–33 (< 广西邕宁县顶蛳山遗址的发掘 >,《考古》第11期).
Matsumura, H., Nguyen, L. C., Li, Z., Hung, H. C., & Huang, Y. Z. (2017). The origins of
EarlyHolocene hunter-gatherers at Huiyaotian and Liyupo in Guangxi, southern China:
Craniometricperspective. In H. Matsumura, H. C. Hung, Z. Li, & K. Shinoda (Eds.),
Bio-anthropologicalstudies of Early Holocene hunter-gatherer sites at Huiyaotian and Liyupo
in Guangxi, China (pp. 155–170). Tokyo: National Museum of Nature and Science No. 47.
Matsumura, H., Hung, H. C., Higham, C., Zhang, C., Yamagata, M., Nguyen, L. C., & Li, Z.,
et al. (2019). Craniometrics reveal ‘two-layers’ of prehistoric human dispersal in eastern
Eurasia. Scientific Reports 9, 1451.
Minagawa, M., & Wada, E. (1984). Stepwise enrichment of 15N along food chains: further
evidence and the relation between d15N and animal age. Geochimica et Cosmochimica Acta,
48, 1135–1140.
Mo, S. T. (莫世泰). (1983). The regression equation of a southern Chinese male adult from the
length of the long bone. Acta Anthropologica Sinica, 2(1), 80–85 (〈華南地區男性成年人由
長骨長度推算身長的回歸方程〉《人類學學報》2(1):80-85).
Qin, F. (覃芳). (2010). An analysis of the prehistoric flexed burial and dismembered burial in the
Neolithic culture of the Dingshishan Site, Guangxi. Cultural Relic of South China, 3, 74–80,
73 (〈广西邕宁顶狮山史前屈肢葬与肢解葬的考察〉,《南方文物》2).
Zhang, X. (2003). Study on the diet of ancient people by analyzing bone elements and isotopes.
Acta Anthropologica Sinica, 22(1), 75–84.
ZJPICRA (Zhejiang Provincial Institute of Cultural Relics and Archaeology) & XSM (Xiaoshan
Museum). (2004). Kuahuqiao: Archaeological Report of Puyang River Stage I. Beijing:
Cultural Relics Publishing (《跨湖橋》(浦陽江流域考古報告之一),文物出版社).
Chapter 12
Coast to Coast: The Spread of Cereal
Cultivation in the Taiwan Strait Region
Before 3500 BP
Tuukka Kaikkonen
12.1 Introduction
In China, research suggests that cereals such as domesticated rice (Oryza sativa),
common millet (Panicum miliaceum), and foxtail millet (Setaria italica) originated
in the Yangzi and Yellow River regions during the Early Holocene, after which
they began to spread outward from these core areas (He et al. 2017; Wang et al.
2016; Zhao 2011). One direction was to the south, as observed through the gradual
expansion of rice cultivation from the Yangzi River to coastal southern China
between 7000 and 5000 BP (He et al. 2017; Wang et al. 2016; Zhao 2011).
Although the intensity and extent of cereal cultivation varied between regions, it has
been argued that the adoption of cultivation was a major process that contributed to
cultural, demographic, and environmental changes over a wide area of the
Asia-Pacific region (Bellwood 2005, 2017). A key location in this process was the
T. Kaikkonen (&)
Canberra, Australia
e-mail: tuukka.kaikkonen@anu.edu.au
Fig. 12.1 Map of representative sites in the Taiwan Strait region with plant remains dating to ca.
5000-3500 BP (except for Liangdao, which dates to an earlier period, ca. 8000-7000 BP). Map
generated by the author in QGIS using data available at naturalearthdata.com
Taiwan Strait region, defined here as a stretch of ocean and islands bracketed by the
coastlines of present-day Fujian and Taiwan. According to present knowledge,
cereal cultivation first appeared in this region between 5000-4000 BP (Hung and
Carson 2014; Zhang and Hung 2010) (Fig. 12.1). Yet instead of an instant or
wholesale replacement of existing subsistence practices, the introduction and
adoption of cultivated cereals appears to have occurred in variable patterns that
reflect a sensitivity to the changing environmental conditions on both sides of the
Strait.
12.2 Background
12.2.1 Terminology
as well as in much of the world, these crafts and practices did not emerge or spread
as a single package, but rather were assembled over time, in regionally variable
patterns (Cohen 2014; Liu and Chen 2012; Zhang and Hung 2010, 2012). For
example, in southern China, pottery was already being manufactured in the
Pleistocene by non-sedentary communities without demonstrable cereal cultivation
practices (Cohen et al. 2017). Likewise, ground and polished stone tools emerged in
southern China in the Pleistocene and became gradually more common in the
southeast from 8000 BP onwards (Cohen 2014; Zhao et al. 2004). Sedentism,
identified through physical evidence of longer and greater investment in fixed
settlements, also developed gradually over time, without a clear-cut beginning (Liu
and Chen 2012). And as will be demonstrated below, agriculture, defined here as a
system of food production with a significant reliance on domesticated resources
(Smith 2001), also emerged gradually over time, spread slowly through space, and
was often combined with food procurement strategies that predate the adoption of
cultivation.
The above examples illustrate how the ‘Neolithic’ in its conventional meaning
fails to capture the gradual, cumulative, and multifaceted nature of the cultural
changes it intends to label and describe. To circumvent some of the problematic
assumptions embedded in the established terminology, the following discussion
uses ‘Neolithic’ solely as shorthand for the three millennia (6500-3500 BP) under
examination, without assuming any of the particular forms of culture (including
residence or subsistence strategies) usually associated with the term. To help orient
the reader, names of culture phases from the Chinese-language literature are
retained and used together with calibrated radiocarbon dates (BP), where available.
Prior attempts to characterize the onset of cereal cultivation in the Taiwan Strait
region have prompted a multitude of perspectives on when, where, how, how fast,
and why it took place. The first reconstructions appear to have been informed by an
explicitly diffusionist paradigm. Focusing on Taiwan and working with limited
materials, Chang (1969) envisaged a transition from primary foraging to primary
agriculture that was marked by changes in material culture. The earlier cord-marked
pottery cultures were equated with forager groups practicing root and tree horti-
culture, while incised and painted pottery-producing groups were equated with rice
farmers arriving from the mainland at a later time. Subsistence and material culture
were tightly associated with notions of distinct ethnic groups, and change was
explained as a movement of these groups from the mainland to Taiwan.
Since then, the diffusionist paradigm has remained influential in the archaeology
of the Taiwan Strait. However, the process of subsistence change has come to be
viewed in a more gradualistic light. Recently, Bellwood (2005, 2017) has
hypothesized that the shift from foraging to farming may have occurred through an
incremental process of “demic expansion” (e.g., Cavalli-Sforza 1997). According to
220 T. Kaikkonen
The Taiwan Strait region (Fig. 12.1) has experienced dramatic changes to its land-
and seascapes during the Holocene. These have been the consequence of changes to
climate, sea levels, geomorphology, and vegetation.
Today, Taiwan and coastal Fujian are separated by 130 kilometers or more of
sea and islands. The Fujian coastline is steep, winding, and dotted with bays, deltas,
and offshore islands (Rolett et al. 2011), while the Taiwan seaboard is characterized
by extensive coastal plains in the west and narrow plains and mountain valleys in
the east (Carson 2017). Both inland Fujian and Taiwan are mountainous, creating
barriers for movement and limiting the availability of arable land. The climate is
subtropical monsoonal and supports broadleaf to coniferous vegetation and the
cultivation of cereals, legumes, roots and tubers, and fruit.
12 Coast to Coast: The Spread of Cereal Cultivation … 221
In the past, the Strait region would have looked radically different than it does
today. During the Last Glacial Maximum, sea levels were between 100 and 150
meters lower than at present (Wang and Sun 1994), and Taiwan was connected to
the mainland via a land bridge. Modelling suggests that this connection lasted until
ca. 11,500 BP (Guedes et al. 2016), after which Taiwan was separated from the
mainland as the warming climate led to a rise in sea levels. A sea-level high stand of
ca. 2 meters above present sea levels lasted probably until the Mid-Holocene (Zong
2004). During this time, the sea reached far inland, coastal shelves were narrow,
arable land scarce, and settlements were located in elevated positions on hilltops
and offshore islets. This state of affairs lasted until the Late Holocene, when a
cooling climate led to falling sea levels that, together with isostatic uplift and
erosional sedimentation, contributed to a seaward expansion of coastlines on both
sides of the Strait. On Taiwan, it is estimated that the expansion of the western
coastal plain began ca. 4800 BP, while that of the eastern plains began ca. 3500 BP
(Carson 2017). In contrast, on the central Fujian coast the formation of bays,
beaches, and deltas is estimated to have begun only ca. 1900 BP (Rolett et al.
2011). Intriguingly, it has been suggested that the seaward expansion of coastlines
may have been further compounded by erosion resulting from vegetation clearance
(Carson 2017), raising questions about the extent to which human activities con-
tributed to geomorphological changes in the region.
Evidence for anthropogenic and climate-driven vegetation changes can be seen
in the regional pollen and charcoal records collected from highland bogs, riverine
settings, and archaeological sites. Vegetation in the warm and humid Early
Holocene was dominated by subtropical broadleaf vegetation. In the Late Holocene,
as the climate took a turn towards cooler and drier conditions, there was a shift
towards coniferous and herbaceous taxa (Lee et al. 2010; Liew et al. 2006; Ma et al.
2016a; Yue et al. 2012, 2015; Zhao et al. 2017). In Fujian, these changes are
paralleled by increased charcoal in both near-site and off-site records, possibly
reflecting anthropogenic disturbance that appears to have remained minimal until
3500-900 BP (Ma et al. 2016a, b; Yue et al. 2012, 2015; Zhao et al. 2017).
A gradual shift towards herbaceous taxa in central Taiwan beginning ca. 5000 BP
has also been taken to indicate human disturbance (Liew et al. 2006; Tsukada
1967). However, no similar changes are observed in southern Taiwan during this
period (Lee et al. 2010).
To summarize, geomorphological and vegetation changes in the Taiwan Strait
region appear to have been asynchronous and variable in their extent and magnitude
during the Late Holocene. Although human activities may have contributed to
vegetation change and erosion, the effects appear to have been limited or indis-
tinguishable from other changes until later in prehistory. By contrast, changes to the
land- and sea-scapes may have exerted a stronger influence on subsistence choices,
making these an important contextual factor in examining the archaeological
record.
222 T. Kaikkonen
Since Chang’s (1969) pioneering work, significant progress has been made on the
culture histories of the Taiwan Strait region. But while radiometric dating and
typological studies have helped to reconstruct cultural chronologies, reports
addressing subsistence practices, especially plant use, have been scarce. This sec-
tion reviews available evidence for plant use (with a focus on domesticated cereals)
up to 3500 BP. A brief overview of representative sites in southern China will
provide background for the subsequent review of evidence from Fujian and Taiwan.
Recent syntheses of the evidence for plant use in the northern regions (including the
Yangzi) can be found in, for example, He et al. (2017) and Wang et al. (2016).
The earliest plant remains in southern China have been reported from a number of
Pleistocene and Holocene cave sites. Analyses of sediments from Xianrendong and
Diaotonghuan (Jiangxi), Yuchanyan (Hunan), and Zengpiyan (Guangxi) have
produced small quantities of rice phytoliths, but it is not clear whether the rice was
cultivated or whether it was wild or domesticated (Nakamura 2010; Zhang and
Hung 2012). Apart from rice, these sites have also yielded seeds and nuts from a
range of plants including Chinese gooseberries (Actinidia sp.), hickory nuts (Carya
sp.), Chinese hackberries (Celtis sinensis), plums (Prunus sp.), and grapes (Vitis
sp.). At Zengpiyan, excavations also recovered the charred remains of tubers,
possibly from either taro (Colocasia sp.) or yam (Dioscorea sp.) (Wang et al. 2016;
Zhang and Hung 2012).
Later in the Holocene, a broad range of plants is evident at the open-air Xincun
Site (ca. 5300-4420 BP) in the Pearl River Delta of Guangdong Province (Yang
et al. 2013). Starch granules and phytoliths from grinding stones found here have
been attributed to sago palms (e.g., Caryota sp.), bananas (Musa sp.), freshwater
roots and tubers (e.g., lotus [Nelumbo nucifera], Chinese arrowroot [Sagittaria sp.],
water chestnut [cf. Eleocharis dulcis]), fern (Angiopteris sp.), acorns (Quercus sp.),
Job’s tears (Coix sp.), and rice (Oryza). The rice phytoliths at Xincun are interpreted
as originating from wild rather than domesticated rice, and as with the earlier cave
sites, it is not clear what role rice may have played in the local economy.
At present, the earliest directly dated evidence for domesticated rice (and, by
inference, rice cultivation) in southern China comes from the Shixia Site, also
located along the Pearl River in Guangdong (Yang et al. 2016). One rice grain,
identified as belonging to the Oryza sativa subspecies japonica, was directly dated
to ca. 4300-4100 BP. Based on contextual evidence, the authors place the arrival of
domesticated rice in the region between 5000-4100 BP, long before the estab-
lishment of intensive cultivation in the Pearl River Delta around ca. 2500 BP (Yang
et al. 2016). Phytoliths attributed to domesticated rice at the even earlier date of
12 Coast to Coast: The Spread of Cereal Cultivation … 223
ca. 5600 BP have also been recently reported from Hainan, but it is unclear whether
this conclusion is supported by the data and analysis reported in the same study.
Until more evidence becomes available, it may be safer to favor the later date,
especially as it more closely reflects previous reconstructions of the spread of rice
cultivation from the Yangzi River southwards (Silva et al. 2015; Zhang and Hung
2010).
In its current state, the archaeobotanical record allows some broad generaliza-
tions to be made about plant use in southern China. Although rice may have been
utilized from the Late Pleistocene onwards at various locations, it seems that rice
cultivation was introduced to the southern China coast only after 5000 BP, after
which it took another 2500 years for intensive rice agriculture to become estab-
lished there (Yang et al. 2016). Prior to this time, the broad range of fruits, nuts, and
vegetative crops indicates at least a long tradition of plant use and management, if
not quite an entirely separate center for plant domestication (see Zhao 2011).
However, the introduction of exotic domesticates and cultivation practices did mark
the beginning of a prolonged transition from foraging to farming. A similar
transformation, albeit with local characteristics, began to unfold in the Taiwan Strait
region around the same time.
The archaeological record in Fujian begins with cave sites that were occupied in the
Pleistocene and the Early Holocene (Jiao 2013). However, little information about
plant remains has emerged from these sites, and the earliest plant remains date only
to the Mid-Holocene during the Neolithic.
The Neolithic begins in Fujian by at least 6500 BP. At this date, ceramics and
polished stone tools appear in fully developed form at coastal sites, having arrived
there as probable introductions from source areas to the north and south (Jiao 2013;
Lin 2005; Zhang and Hung 2012). This period is also marked by the emergence and
growth of open-air sites, first in the form of shell middens and later as ‘villages’,
though most of these remained only a modest 1–2 ha in size (Jiao 2013; Rolett et al.
2011). These were scattered on offshore islets and promontories along coastlines
that would have extended far inland during the time of the sea-level high stand
(Rolett et al. 2011). The overall number of sites appears to have remained low,
implying low population densities throughout the Neolithic (Hosner et al. 2016;
Jiao 2013).
The Neolithic period has been subdivided into several cultural phases, some of
them regional. Although recent reports have begun to fill in gaps in the record of
inland Fujian, the region is still poorly understood (Deng et al. 2018; Jiao 2013),
and here the main focus will be on the central and northern Fujian coast, followed
by a brief overview of sites in southern and inland Fujian (Table 12.1).
224 T. Kaikkonen
Table 12.1 Neolithic culture phases of central and northern coastal Fujian (based on Lin 2005)
Phase Culture Region Dating (BP)
Early Neolithic Keqiutou Central coast 6500-5500
Middle Neolithic Tanshishan Central to northern coast 5500-4300
Late Neolithic Huangguashan Central to northern coast 4300-3500
from these sites, their application to the clearance of vegetation appears to have
been limited.
The modest scale of rice cultivation at Tanshishan and Zhuangbianshan may
be explained in the light of palaeo-landscape reconstructions. Modelling suggests
that when the sites were occupied, the coastline of the Fuzhou Basin reached
75–80 kilometers further inland than at present, limiting the availability of arable
wetlands and thus the productivity of cereal cultivation (Ma et al. 2016b; Rolett
et al. 2011). Wild resources such as shellfish, fish, and terrestrial fauna appear to
have been the preferred sources of food (Jiao 2013), a conclusion also supported by
analyses of stable carbon and nitrogen isotopes. It is interesting to speculate
whether stable isotopes might reveal changes in diet over time that would reflect the
diachronic reduction in shell midden densities and the increase in rice phytoliths
observed at Zhuangbianshan (Ma et al. 2016b).
Despite the limited scale of food production documented in the Upper Fuzhou
Basin, cereal cultivation appears to have persisted over time, becoming established
further north on the coast. This is recorded at Huangguashan (4500-3800 BP) and
Pingfengshan (3800-3500 BP), both located on low hills within 8 kilometers of the
present-day coastline in northern coastal Fujian (Deng et al. 2018). The material
culture at these sites is mostly associated with the Huangguashan Phase and
comprises ceramics, stone and bone tools, and bones of marine and terrestrial
animals, including pigs (Jiao 2013).
In contrast to the Upper Fuzhou Basin sites, flotation and sediment samples at
Huangguashan and Pingfengshan revealed macroscopic plant remains and phy-
toliths of fruits and cereals dated to at least 4000-3500 BP (Deng et al. 2018).
Included among these are a few carbonized grains and spikelet bases from
domesticated rice that, despite their small number, still vastly outnumber the very
limited remains recovered from the Tanshishan Phase sites. Perhaps more impor-
tantly, material from these sites represents the earliest known instance of cultivated
millets in coastal Fujian (Deng et al. 2018). Carbonized foxtail millet grains at both
locations, as well as the Panicoid-type millet husk phytoliths found at
Huangguashan, indicate that these grains were cultivated together with rice.
Together, these findings probably signal the dispersal of a crop cultivation package
that ultimately can be traced to regions further to the north. However, many gaps
exist in the prehistory of mixed cereal cultivation in southern China, including the
Taiwan Strait region. It remains to be seen whether sampling from the deeper layers
at Huangguashan and other sites could push the dates back to rival those recently
reported from inland Fujian, examined below.
The sites discussed above contain some of the earliest material evidence for cereal
cultivation in southeastern coastal China and are important for tracing the spread of
crop cultivation from the Yangzi to the southern seaboard and onwards into
226 T. Kaikkonen
Taiwan. However, questions remain concerning when and by what route (inland or
coastal) these cereals spread to coastal Fujian. Despite the presence of rice in the
Pearl River Delta in 5000 BP (Yang et al. 2016), no equivalent results have been
reported for southern coastal Fujian sites such as Damaoshan (5500-4500 BP) (Jiao
2013). However, new evidence from inland Fujian is beginning to fill in other gaps
in the record. Recent reports from the Nanshan site in western Fujian document the
presence of rice, common millet, and foxtail millet (as well as other plant remains)
in significant quantities at an early date of ca. 5000 BP (Zhao 2017; Fig. 12.2). Rice
and foxtail millet have also been reported from Hulushan (4000-3500 BP) in
northwestern Fujian (Fujian Provincial Museum et al. 2016), although more
extensive reports have yet to be published (Deng et al. 2018). Forthcoming results
from these sites will hopefully offer insight on the timing, extent, and variability of
cereal cultivation in Fujian and the wider region.
Fig. 12.2 View of the Nanshan archaeological site, Mingxi County, Fujian Province on
November 5, 2017. Photo by the author
12 Coast to Coast: The Spread of Cereal Cultivation … 227
As in Fujian, the archaeological record of Taiwan begins with cave sites that were
occupied from the Pleistocene until the Mid-Holocene (Hung and Carson 2014).
Plant remains have not been reported at these sites, and with the exception of
indeterminate fibers and possible acorn starch extracted from the dental calculus
of the Liangdao burials (ca. 8000-7500 BP) (Chiu et al. n.d.), the nature of
pre-Neolithic plant subsistence on Taiwan and the Strait islands remains poorly
understood (Hung and Carson 2014).
The beginning of the Neolithic on Taiwan is conventionally dated to ca.
6000 BP, when the arrival of coarse cord-marked ceramics, polished stone adzes,
and other novel material culture at open-air sites mark a distinct break from the
preceding cultural traditions (Hung and Carson 2014). Chronology and typology
suggest that these were introduced from the southern coast of China, including
coastal Fujian (Hung and Carson 2014). While the Neolithic Age in Taiwan has
been subdivided into several culture phases (Table 12.2), the focus here is on the
Early and Late Tapenkeng (or TPK; Dabenkeng in modern pinyin), as well as the
Middle Neolithic. Despite the presence of rice at Late Neolithic sites and its
implications for tracing the long-term history of cultivation on Taiwan, findings
from this period are beyond the scope of this chapter.
Early TPK, 6000-4800 BP
The Early TPK (ca. 6000-4800 BP) is known from thinly layered shell middens and
sand dune sites from across Taiwan that were located on promontories near aquatic
resources during their occupation (Hung and Carson 2014). Only two of the ten
known sites have been radiocarbon dated, with results ca. 6500-5300 BP (Hung and
Carson 2014). This leaves the chronology of Taiwan’s earliest Neolithic popula-
tions rather weakly resolved.
Perhaps owing to the small number of poorly preserved sites, the Early TPK has
produced little direct evidence for plant use or cultivation. Phytoliths of what might
be wild rice from pottery sherds have been reported from the Dapenkeng Site, but
no other plant remains have been reported (Hung and Carson 2014). Stone adzes are
still few and harvesting knives are absent (Hung and Carson 2014), while no signals
Table 12.2 Neolithic culture phases of Taiwan and the Straits islands (after Hung and Carson
2014)
Phase Culture Region Dating (BP)
Early Neolithic Early Tapenkeng Island-wide 6000-4800
Late Tapenkeng Island-wide 4800-4500/4200
Middle Neolithic Several Regional subdivisions 4500/4200-3500
Late Neolithic Several Regional subdivisions 3500-2400
228 T. Kaikkonen
12.5 Conclusion
As suspected for some time (e.g., Chang 1969), there are strong reasons to believe
that the onset of cereal cultivation was not the result of indigenous innovation but
rather one stage in the spread of domesticates and cultivation practices from the
Yangzi River and northern China to the south. Although the lack of directly dated
plant remains makes it difficult to estimate the speed of dispersal, the available
chronologies agree with reconstructions of the gradual southward spread of rice
cultivation (Silva et al. 2015; Zhang and Hung 2008, 2010). The existing
chronology further shows that crossing the Taiwan Strait did not pose a major
barrier to cross-Strait crop dispersal. Indeed, maritime-oriented technology and
networks had already been established prior to the expansion of cultivation, as
documented by the cross-Strait movement of Penghu basalt, ceramics, stone tools,
and other novel material culture (Rolett et al. 2007; Zhang and Hung 2010, 2012).
Similar growth conditions (climate, daytime length) probably also aided this pro-
cess, as did the practice of mixed rice and millet farming, which may have enabled
the more efficient utilization of available land and swidden farming (Bellwood
2011).
The discovery of a mixed cereal package of rice and millets on both sides of the
Taiwan Strait is significant for helping to trace how cereal farming may have moved
together with people and languages (e.g., Bellwood 2005). Prior to the recent
discoveries of mixed cereals in coastal and inland Fujian, the rice and millets from
Nanguanli and Nanguanlidong were the earliest known case of a Neolithic mixed
cereal package in the region. This co-occurrence has in part been used to suggest
that the Early Neolithic period in Taiwan was an extension of the mixed-cereal
farming societies in the Shandong–Jiangsu region of northern China (Fuller 2011;
Sagart 2008). Although the recent findings of millets in Fujian cannot disprove the
Shandong-Jiangsu connection, it now seems reasonable that the closest source area
for mixed-cereal cultivation in Taiwan should be looked for in present-day Fujian
Province (Deng et al. 2018; see also Sagart 2008).
230 T. Kaikkonen
Why and how cultivation expanded across the Taiwan Strait at the particular
time and in the way that it did probably has no simple answer. Barring trade,
founder lineage, and other social hypotheses, it has been suggested that the dis-
persal of cereal cultivation was likely due to a combination of environmental and
demographic factors (Bellwood 2005, 2017). As described above, land- and seas-
capes across the Taiwan Strait underwent major changes during the Late Holocene,
and the asynchronous development of coastal alluvia on Taiwan and in the river
deltas of Fujian may have contributed to the spread and establishment of cereal
cultivation. A higher reproductive rate among farmers, relative to foragers, may also
have played a role (Bellwood 2005; Bocquet-Appel 2011). However, this
hypothesis is complicated by low population estimates and the limited evidence for
cereal cultivation and anthropogenic disturbance observed in Fujian during this
period (Hosner et al. 2016; Jiao 2013; Zhao et al. 2017). An alternative explanation,
namely the proposal of an expansionist “frontier mentality” associated with shifting
dry land cultivation (Bellwood 2011), remains to be evaluated through further
studies of cultivation regimes and vegetation histories that address the typology of
early cultivation systems. Above all, the spread of cereal cultivation was part of a
wider process, but its local manifestations emphasize the importance of paying
attention to the local conditions under which it occurred. Whatever further research
may reveal, the dynamic land- and seascapes of the Taiwan Strait should not be
ignored when discussing the spread of cereal cultivation from continental Asia to
the doorstep of the Pacific.
Acknowledgements I wish to thank Wu Chunming and Barry Rolett for the opportunity to
participate in the conference and to contribute to these proceedings. I also wish to express my
gratitude to Hung Hsiao-chun for constructive feedback and suggestions that greatly helped to
improve this paper. Any errors remain my own. I acknowledge the Australian Government
Research Training Program scholarship as the source of funding for my Ph.D. candidature at the
Australian National University (2017–2018).
References
Bellwood, P. S. (2005). First farmers: The origins of agricultural societies. Malden, MA:
Wiley-Blackwell.
Bellwood, P. S. (2011). The checkered prehistory of rice movement southwards as a domesticated
cereal—From the Yangzi to the Equator. Rice, 4(3–4), 93–103. https://doi.org/10.1007/
s12284-011-9068-9.
Bellwood, P. S. (2017). First islanders: Prehistory and human migration in island Southeast Asia.
Malden, MA: Wiley-Blackwell.
Bocquet-Appel, J. P. (2011). When the world’s population took off: the springboard of the
Neolithic Demographic Transition. Science, 333(6042), 560–561. https://doi.org/10.1126/
science.1208880.
Carson, M. T. (2017). Coastal palaeo-landscapes of the Neolithic. In P. S. Bellwood (Ed.), First
islanders: Prehistory and human migration in island Southeast Asia (pp. 240–244). Malden,
MA: Wiley-Blackwell.
12 Coast to Coast: The Spread of Cereal Cultivation … 231
Cavalli-Sforza, L. L. (1997). Genes, peoples, and languages. Proceedings of the National Academy
of Sciences, 94(15), 7719–7724. https://doi.org/10.1073/pnas.94.15.7719.
Chang, K. (1969). Fengpitou, tapenkeng, and the prehistory of Taiwan. New Haven: Department
of Anthropology, Yale University.
Chiu, H. L., Chen, C. Y., & You, Z. (n.d.). Research on the Liangdao-Daowei site group on
Liangdao, Mazu islands: Final report (pp. 1–274). Unpublished report on file at National
Tsinghua University, Hsinchu (Mazu Liangdao Daowei Yizhiqun Zhonghe Yanjiu Jihua:
Chengguo Baogao 《馬祖亮島島尾遺址群綜合研究計劃:成果報告》).
Cohen, D. J. (2014). The neolithic of Southern China. In C. Renfrew & P. Bahn (Eds.), The
Cambridge world prehistory 3 volume set (pp. 765–781). Cambridge: Cambridge University
Press. https://doi.org/10.1017/CHO9781139017831.053.
Cohen, D. J., Bar-Yosef, O., Wu, X., Patania, I., & Goldberg, P. (2017). The emergence of pottery
in China: Recent dating of two early pottery cave sites in South China. Quaternary
International, 441, Part B, 36–48. https://doi.org/10.1016/j.quaint.2016.08.024.
Deng, Z., Hung, H., Carson, M. T., Bellwood, P. S., Yang, S., & Lu, H. (2017). The first discovery of
Neolithic rice remains in eastern Taiwan: phytolith evidence from the Chaolaiqiao site.
Archaeological and Anthropological Sciences, 1–8. https://doi.org/10.1007/s12520-017-0471-z.
Deng, Z., Hung, H., Fan, X., Huang, Y., & Lu, H. (2018). The ancient dispersal of millets in
southern China: New archaeological evidence. The Holocene, 28(1), 34–43. https://doi.org/10.
1177/0959683617714603.
Fujian Provincial Museum, Xiamen University, Wuyishan Museum. (2016). Preliminary report on
the 2014 excavation of the Hulushan site, Wuyishan City, Fujian Province. Southeast Culture,
2, 19–36 (Fujian Wuyishanshi Hulushan Yizhi 2014 Nian Houjue Jianbao 《福建武夷山市葫
芦山遗址2014年发掘简报》東南文化 2016 年).
Fuller, D. Q. (2011). Pathways to Asian civilizations: Tracing the origins and spread of rice and
rice cultures. Rice, 4(3–4), 78–92. https://doi.org/10.1007/s12284-011-9078-7.
Fuller, D. Q., & Lucas, L. (2017). Adapting crops, landscapes, and food choices: Patterns in the
dispersal of domesticated plants across Eurasia. In N. Boivin, M. Petraglia, & R. Crassard
(Eds.), Human dispersal and species movement (pp. 304–331). Cambridge: Cambridge
University Press. https://doi.org/10.1017/9781316686942.013.
Guedes, J. A., Austermann, J., & Mitrovica, J. X. (2016). Lost foraging opportunities for East
Asian hunter-gatherers due to rising sea level since the Last Glacial Maximum.
Geoarchaeology, 31(4), 255–266. https://doi.org/10.1002/gea.21542.
He, K., Lu, H., Zhang, J., Wang, C., & Huan, X. (2017). Prehistoric evolution of the dualistic
structure mixed rice and millet farming in China. The Holocene, 27(12), 1885–1898. https://
doi.org/10.1177/0959683617708455.
Hosner, D., Wagner, M., Tarasov, P. E., Chen, X., & Leipe, C. (2016). Spatiotemporal distribution
patterns of archaeological sites in China during the Neolithic and Bronze Age: An overview.
The Holocene, 26(10), 1576–1593. https://doi.org/10.1177/0959683616641743.
Hung, H. (2017). Neolithic cultures in Southeast China, Taiwan, and Luzon. In P. S. Bellwood
(Ed.), First Islanders: Prehistory and human migration in island Southeast Asia (pp. 234–
240). Malden, MA: Wiley-Blackwell.
Hung, H., & Carson, M. T. (2014). Foragers, fishers and farmers: Origins of the Taiwanese
Neolithic. Antiquity, 88(342), 1115–1131. https://doi.org/10.1017/S0003598X00115352.
Jiao, T. (2013). The neolithic of Southeast China. In A. P. Underhill (Ed.), A companion to
Chinese archaeology (pp. 599–612). Chichester, West Sussex: John Wiley & Sons Inc.
Jiao, T. (2016). Toward an alternative perspective on the foraging and low-level food production
on the coast of China. Quaternary International, 419, 54–61. https://doi.org/10.1016/j.quaint.
2015.06.060.
Lee, C. Y., Liew, P. M., & Lee, T. Q. (2010). Pollen records from southern Taiwan: Implications
for East Asian summer monsoon variation during the Holocene. The Holocene, 20(1), 81–89.
https://doi.org/10.1177/0959683609348859.
232 T. Kaikkonen
Liew, P. M., Huang, S. Y., & Kuo, C. M. (2006). Pollen stratigraphy, vegetation and environment
of the last glacial and Holocene: a record from Toushe Basin, Central Taiwan. Quaternary
International, 147(1), 16–33. https://doi.org/10.1016/j.quaint.2005.09.003.
Lin, G. (2005). Neolithic cultures in coastal Fujian. In C. C. Chung-yu & J. Pan (Eds.), The
archaeology of southeast coastal islands of China (pp. 75–90). Mazu: Lianjiang County
Government (Fujian Yanhai Xinshiqishidai Wenhua Zongshu 《福建沿海新石器時代文化綜
述》連江省政府 2005年).
Liu, L., & Chen, X. (2012). The archaeology of China: From the late paleolithic to the early
bronze age. Cambridge: Cambridge University Press. http://ebookcentral.proquest.com/lib/
anu/detail.action?docID=907107. Accessed 8 May 2018.
Ma, T., Tarasov, P. E., Zheng, Z., Han, A., & Huang, K. (2016a). Pollen- and charcoal-based
evidence for climatic and human impact on vegetation in the northern edge of Wuyi
Mountains, China, during the last 8200 years. The Holocene, 26(10), 1616–1626. https://doi.
org/10.1177/0959683616641744.
Ma, T., Zheng, Z., Rolett, B. V., Lin, G., Zhang, G., & Yue, Y. (2016b). New evidence for
Neolithic rice cultivation and Holocene environmental change in the Fuzhou Basin, Southeast
China. Vegetation History and Archaeobotany, 25(4), 375–386. https://doi.org/10.1007/
s00334-016-0556-0.
Nakamura, S. (2010). The origin of rice cultivation in the Lower Yangtze Region, China.
Archaeological and Anthropological Sciences, 2(2), 107–113. https://doi.org/10.1007/s12520-
010-0033-0.
Rolett, B. V., Guo, Z., & Jiao, T. (2007). Geological sourcing of volcanic stone adzes from
Neolithic sites in Southeast China. Asian Perspectives, 46(2), 275–297. https://doi.org/10.
1353/asi.2007.0018.
Rolett, B. V., Zheng, Z., & Yue, Y. (2011). Holocene sea-level change and the emergence of
Neolithic seafaring in the Fuzhou Basin (Fujian, China). Quaternary Science Reviews, 30(7–8),
788–797. https://doi.org/10.1016/j.quascirev.2011.01.015.
Sagart, L. (2008). The expansion of Setaria farmers in East Asia: A linguistic and archaeological
model. In A. Sanchez-Mazas, R. Blench, M. D. Ross, I. Peiros, & M. Lin (Eds.), Past human
migrations in East Asia: Matching archaeology, linguistics and genetics (pp. 133–157).
London: Routledge.
Silva, F., Stevens, C. J., Weisskopf, A., Castillo, C., Qin, L., Bevan, A., et al. (2015). Modelling
the geographical origin of rice cultivation in Asia using the Rice Archaeological Database.
PLoS ONE, 10(9), e0137024. https://doi.org/10.1371/journal.pone.0137024.
Smith, B. D. (2001). Low-level food production. Journal of Archaeological Research, 9(1), 1–43.
Tsang, C., & Li, K. (2013). Archaeological heritage in Tainan Science Park of Taiwan, 1. Taidong
City: National Taiwan Museum of Prehistory.
Tsang, C., Li, K. T., Hsu, T. F., Tsai, Y. C., Fang, P. H., & Hsing, Y. I. C. (2017). Broomcorn and
foxtail millet were cultivated in Taiwan about 5000 years ago. Botanical Studies, 58(1), 3.
https://doi.org/10.1186/s40529-016-0158-2.
Tsukada, M. (1967). Vegetation in subtropical Formosa during the Pleistocene glaciations and the
Holocene. Palaeogeography, Palaeoclimatology, Palaeoecology, 3, 49–64.
Wang, C., Lu, H., Zhang, J., He, K., & Huan, X. (2016). Macro-process of past plant subsistence
from the Upper Paleolithic to Middle Neolithic in China: A quantitative analysis of
multi-archaeobotanical data. PLoS ONE, 11(2), e0148136. https://doi.org/10.1371/journal.
pone.0148136.
Wang, P., & Sun, X. (1994). Last glacial maximum in China: Comparison between land and sea.
CATENA, 23(3), 341–353. https://doi.org/10.1016/0341-8162(94)90077-9.
Wu, M., Ge, W., & Chen, Z. (2016a). Diets of a late Neolithic maritime settlement: carbon and
nitrogen stable isotope analysis of human bones from the Tanshishan site. Acta Anthropologica
Sinica, 35(2), 246–256 (Haiyangxing Juluo Xianmin De Shiwu Jiegou: Tanshishan Yizhi
Xinshiqishidai Wanqi Rengu De Tandan Wending 《海洋性聚落先民的食物结构:昙石山遗
址新石器时代晚期人骨的碳氮稳定同位素分析》人類學學報 2016年).
12 Coast to Coast: The Spread of Cereal Cultivation … 233
Wu, Y., Mao, L., Wang, C., Zhang, J., & Zhao, Z. (2016b). Phytolith evidence suggests early
domesticated rice since 5600 cal a BP on Hainan Island of South China. Quaternary
International, 426, 120–125. https://doi.org/10.1016/j.quaint.2016.01.008.
Yang, X., Barton, H. J., Wan, Z., Li, Q., Ma, Z., Li, M., et al. (2013). Sago-type palms were an
important plant food prior to rice in southern subtropical China. PLoS ONE, 8(5), e63148.
https://doi.org/10.1371/journal.pone.0063148.
Yang, X., Wang, W., Zhuang, Y., Li, Z., Ma, Z., Ma, Y., et al. (2016). New radiocarbon evidence
on early rice consumption and farming in South China. The Holocene, 27(7), 1045–1051.
https://doi.org/10.1177/0959683616678465.
Yue, Y., Zheng, Z., Huang, K., Chevalier, M., Chase, B. M., Carré, M., et al. (2012). A continuous
record of vegetation and climate change over the past 50,000 years in the Fujian Province of
eastern subtropical China. Palaeogeography, Palaeoclimatology, Palaeoecology, 365–366,
115–123. https://doi.org/10.1016/j.palaeo.2012.09.018.
Yue, Y., Zheng, Z., Rolett, B. V., Ma, T., Chen, C., Huang, K., & Lin, G., et al. (2015). Holocene
vegetation, environment and anthropogenic influence in the Fuzhou Basin, Southeast China.
Journal of Asian Earth Sciences, 99(Supplement C), 85–94. https://doi.org/10.1016/j.jseaes.
2014.12.004.
Zhang, C., & Hung, H. (2008). The Neolithic of southern China: Origin, development, and
dispersal. Asian Perspectives, 47(2), 299–329. https://doi.org/10.1353/asi.0.0004.
Zhang, C., & Hung, H. (2010). The emergence of agriculture in southern China. Antiquity, 84
(323), 11–25. https://doi.org/10.1017/S0003598X00099737.
Zhang, C., & Hung, H. (2012). Later hunter-gatherers in southern China, 18,000-3000 BC.
Antiquity, 86(331), 11–29. https://doi.org/10.1017/S0003598X00062438.
Zhao, C., Wu, X., Wang, T., & Yuan, X. (2004). Early polished stone tools in South China
evidence of the transition from Palaeolithic to Neolithic. Documenta Praehistorica, 31, 131.
https://doi.org/10.4312/dp.31.9.
Zhao, L., Ma, C., Leipe, C., Long, T., Liu, K., Lu, H., et al. (2017). Holocene vegetation dynamics
in response to climate change and human activities derived from pollen and charcoal records
from southeastern China. Palaeogeography, Palaeoclimatology, Palaeoecology, 485, 644–
660. https://doi.org/10.1016/j.palaeo.2017.06.035.
Zhao, Z. (2011). New archaeobotanic data for the study of the origins of agriculture in China.
Current Anthropology, 52(S4), S295–S306. https://doi.org/10.1086/659308.
Zhao, Z. (2017). Discussion on ancient ocean passage based on flotation results from the Nanshan
site. Presented at the International Conference of Prehistoric Archaeology in Southeast China
through the Pacific, Jiangle County, Fujian Province, 3 November 2017 (Cong Nanshan Yizhi
Fuxuan Jieguo Tan Gudai Haiyang Tongdao 《從南山遺址浮選結果談古代海洋通道》,
Zhongguo Dongnan Ji Huan Taipingyang Diqu Shiqian Kaogu Guoji Xueshu Yanjiuhui 《中
国东南及环太平洋地区史前考古国际学术研讨会》 2017年).
Zong, Y. (2004). Mid-Holocene sea-level highstand along the Southeast coast of China.
Quaternary International, 117(1), 55–67. https://doi.org/10.1016/S1040-6182(03)00116-2.
Chapter 13
A Study of Geometric Stamped Pattern
Pottery and Early Maritime Cultural
Interactions Between Mainland China
and Taiwan
Lin Fu
Abstract Geometric stamped pattern pottery is one of the most distinctive cultural
features from the Neolithic to the Early Iron Age in South China and Southeast
Asia. Professor Huixiang Lin and Rongfang Lv have already noted the importance
of this pottery type in the prehistoric and ancient cultural development sequences of
southeastern China (Wu 1992; Lv 1959). A special seminar called “Stamped
Pattern Pottery in the Jiangnan (江南, South of Yangtze River) Region” and held in
Lushan (庐山), Jiangxi Province, in 1978 focused on geometric stamped pattern
pottery and the related cultural history of the Jiangnan region (ECCR 1981). Many
other scholars, such as Boqian Li and Shifan Peng, have also done in-depth studies
on this topic (Li 1981; Peng 1987). In South China the Minjiang (闽江) River
region, in Fujian, and Taiwan Island have been grouped into one district due to
cultural similarities based on the typological analysis of geometric stamped pattern
pottery (Li 1981). However, the geometric stamped pattern pottery across the
Taiwan Strait has different development sequences during the early history of
ancient China (Pre-Qin and Han Dynasties). Further typological comparison of
geometric stamped pattern pottery on both sides of the Taiwan Strait can offer
greater insight on early historical cultural contact and exchange in this region.
The geometric stamped pattern pottery first developed in the Minjiang River region
during the early cultural period of the Pre-Qin and Han Dynasties. According to
typological analysis tracking cultural changes in these archaeological remains, the
development of geometric stamped pattern pottery in the Minjiang River region can
be divided into three stages. The First Stage marks the beginning of this style and
L. Fu (&)
Department of History, Xiamen University, Xiamen, China
e-mail: fulinyaoming@163.com
dates to about 4,800 to 3,500 years ago. The Second Stage is when it flourished,
from about 3,500 to 3,000 years ago. The Third Stage is the period of decline, from
about 3,000 to 2,000 years ago.
Regionally, the cultural character of geometric stamped pattern pottery in the
Minjiang River region illustrates two different sub-regions that correspond to the
inland upper reaches and the coastal lower reaches of the Minjiang River. The
cultural development sequence from the Neolithic to the Early Iron Age in the
upper reaches of the Minjiang River includes a series of culture types including
Niubishan (牛鼻山), Maling (马岭), and Baizhuduan (白主段) as well as the Zhou
Dynasty remains found at the mounded tombs in Guanjiu (管九) Village and the
remains of the Minyue State from the Han Dynasty Site in Chengcun (城村)
Village. The contemporaneous cultural development sequence in the lower reaches
of the Minjiang River, along the coast of the Taiwan Strait, includes such cultural
affiliations as the Keqiutou (壳丘头) Culture, the lower Tanshishan (昙石山)
remains, the Tanshishan Culture, the Huangguashan (黄瓜山) type, and the
Huangtulun (黄土仑) type. From the Zhou to the Han Dynasties, the cultural
affiliation in this region was the same as that of the mounded tomb remains and
Chengcun Culture in the upper reaches of the Minjiang River.
Among the remains associated with the Keqiutou Culture and those from the
lower Tanshishan, which date to 6500 to 5000 years ago, the pottery is decorated
mainly with incised lines, punctured dots, cowrie shell impressions, stamped cords
and dense dots. The punctured-dot pattern is moreover not a typical kind of geo-
metric stamped pattern. At this stage, then, stamped geometric patterns had not yet
appeared.
The geometric stamped pattern pottery in the Minjiang River region emerged
between 5000 and 4000 years ago. There are a few stamped checked, laddered, leaf
vein and matted patterns on the pottery associated with the Tanshishan Culture
(Fig. 13.1: 1–8). According to chronological studies of the tombs from Tanshishan
Culture (FJPM AND FJTSM 2015), a small amount of the geometric stamped
pattern pottery first appeared during the Second Stage and continued through the
Third, Fourth and Fifth Stages of this culture. Based on the generally recognized
dates for the existence of the Tanshishan Culture (IACASS 2010), geometric
stamped pattern pottery appears to have emerged in the Minjiang River region
around 4,800–4,500 years ago. Several Hu (壶) jars and Guan (罐) jars of the
Niubishan type with stamped laddered and leaf vein patterns have also been dis-
covered in the upper reaches of the Minjiang River (Fig. 13.1: 9, 10; FJPM 1996).
These date to about 4,800 years ago and include remains from the Haochuan (好川)
Culture, from which similar geometric stamped pattern pottery artifacts have been
collected (Fig. 13.1: 11, 12; ZJPCRARI and SCCRMC 2001). Generally speaking,
the geometric stamped pattern pottery in the region Minjiang River appeared
around the same time as that of the lower layer types from Shanbei (山背) Culture
and Zhuweicheng (筑卫城) Cultures, in the Ganpo (赣鄱) region of Jiangxi (江西)
Province, and the Shixia (石峡) Culture from the Lingnan (岭南) region of
Guangdong (广东) Province, where the earliest geometric stamped pattern pottery
13 A Study of Geometric Stamped Pattern Pottery … 237
Fig. 13.1 Tanshishan, Niubishan and Haochan geometric stamped pattern pottery including: 1,
Tanshishan M109:1; 2, Tanshishan M119:5; 3, Tanshishan M137:1; 4, Tanshishan M130:14; 5,
Tanshishan M22:4; 6, Tanshishan M101:2; 7, Tanshishan M104:2; 8, Tanshishan M126:2; 9,
Niubishan M16:6; 10, Niubishan M2:5; 11, Haochuan M71:7; and 12, Haochuan M52:9. Patterns
include: 1, 2, 5, checked pattern; 3, 7, matted pattern; 4, 8, 10, laddered pattern; 6, 9, 11, leaf vein
pattern; and 12, zigzag pattern
originated. At this early stage, geometric stamped patterns were irregular and hasty,
and may continue the incised and impressed patterns of the previous stage.
Both the Huangguashan type in the lower reaches and the Maling type in the
upper reaches of the Minjiang River date to the First Stage in the development of
geometric stamped pattern pottery, or about 4,000 to 3,500 years ago. Although
stamped patterns are not the main decoration for these two types, they steadily
continued to increase overall (Fig. 13.2: 1–10). The checked pattern, matted pattern
and cloud/thunder pattern are the three main decorations on the Maling type pot-
tery, along with woven and zigzag patterns, while black and dark brown slip is the
most distinctive feature of this type (Huang 2015). In some cases, black or red slip
was also applied to the stamped geometric patterns (Fig. 13.2: 1, 2; FJPM et al.
2016). The Huangguashan type pottery was decorated with stamping patterns such
as checked, leaf vein, trellised and cloud/thunder, while the most distinctive fea-
tures of this type were the painted geometric patterns of checked, grid, cloud/
thunder, compound line, linked-hook, parallel-line, diagonal triangle patterns and a
large number of complicated patterns combining different designs (Fig. 13.2: 7–
238 L. Fu
Fig. 13.2 Geometric stamped and painted patterns on Maling and Huangguashan Types’ pottery
from the Hulushan (葫芦山) Site in Wuyishan City (1–6) and the Tanshishan Site in Minhou
County (7–20). Examples of the stamped pattern include: 1, zigzag pattern; 2, grid pattern and leaf
vein pattern; 3, 9, leaf vein pattern; 4, 10, cloud/thunder pattern; 5, matted pattern; 6, 8, checked
pattern; and 7, trellised pattern. Examples of the painted pattern include: 11, cloud/thunder and
grid pattern; 12, checked pattern; 13, vertical stripe pattern, doubling-line triangle pattern and
checked pattern with an inner dot; 14, zigzag pattern and cloud/thunder pattern; 15, cloud/thunder
pattern and parallel-diagonal line pattern; 16, linked-hook pattern; 17, zigzag pattern and linear
pattern; 18, grid pattern; 19, parallel-diagonal line pattern; and 20, zigzag pattern and checked
pattern
10). Thus, during this period geometric stamped pattern pottery in the Minjiang
River region developed several new stamped and painted geometric designs that
subsequently became the basis for rapid growth of this style in the next stage.
The geometric stamped pattern pottery of Huangtulun and Baizhuduan types
flourished during the Second Stage, from about 3,500 to 3,000 years ago in the
Shang and early Zhou Dynasties. During this period black slip and painting
declined, and stamped geometric patterns became the primary type of pottery
decoration in the Minjiang River region. Patterns observed from this stage include
the stamped cloud/thunder pattern, modified thunder-squared, checked, matted,
fretted, zigzag and diamond, as well as a series of compound patterns including
cloud/thunder, checked, cord, basket, incised and carved patterns (Fig. 13.3). In the
third layer of the Jigushan (积谷山) Site in Guangze (光泽) County, Fujian
Province, 47% of the excavated pottery bore geometric stamped patterns, signifi-
cantly more than that with cord and basket patterns (FJPM 2003). And these pottery
may be made at Jiaoshan (角山) Kiln Site in Yingtan (鹰潭) County, Jiangxi
Province (JXPICRA et al. 2017). Remains of the Huangtulun type in the lower
reaches of the Minjiang River also illustrated a large quantity of stamped pottery
with the cloud/thunder and checked patterns (FJPM 1984). The fourth stage of the
13 A Study of Geometric Stamped Pattern Pottery … 239
Fig. 13.3 Geometric stamped patterns on Huangtulun and Baizhuduan Types’ pottery from the
Huangtulun Site (1, 2) and the Jigushan Site (3–14), including examples of: 1, 3, checked pattern;
2, 5, modified thunder pattern; 4, cloud/thunder pattern; 6, grid pattern; 7, matted pattern; 8, zigzag
pattern; 9, cloud/thunder pattern and checked pattern; 10, cloud/thunder pattern and cord pattern;
11, cloud/thunder pattern and basket pattern; 12, thunder pattern and incised triangle pattern; 13,
matted pattern and basket pattern; and 14, zigzag pattern and cord pattern
Tanshishan Site, also associated with the Huangtulun type, shares similar charac-
teristics including cloud/thunder and checked pattern stamped pottery. Geometric
stamped pattern pottery has also been found on the offshore islands near the estuary
of the Minjiang River. For example, several archaeological sites associated with the
Huangtulun type were investigated on Pingtan (平潭) Island (FMTCRA
AND DHXU 1995). The geometric stamped pattern pottery excavated at the
Guishan (龟山) Site, Donghuaqiu (东花丘) Site, and Fengmei (凤美) Tomb Site
also share characteristics with the Huangtulun type.1
Starting in the early Western Zhou (西周) Dynasty, the Wuyue (吴越) Culture
began to spread from the lower reaches of the Yangtze River into the Minjiang
River region. This shift is illustrated by the Guanjiu (管九) Tomb Site in Pucheng
(浦城) County, in the upper reaches of the Minjiang River (FJPM AND FJMSCM
2007), and by M7 and M8 of the Luohanshan (罗汉山) Site in Fuzhou City in the
lower reaches of the Minjiang River (FMTCRA 2014). Thus began an important
cultural stage in the Minjiang River, when both bronze and proto-porcelain vessels
spread into the region. The proto-porcelain artifacts discovered include not only
small vessels for holding and eating food, such as Dou (豆) bowls and Yu (盂)
basins, but also large vessels for ritual use or storage such as Guan (罐) jars. This is
the Third Stage in the development of geometric stamped pattern pottery, charac-
terized by a decline from the late Western Zhou to the Spring and Autumn (春秋)
Periods, about 3,000 to 2,000 years ago. During this time only a few artifacts
1
Based on the 2017 excavation material from the Institute of Cultural Relics and Archaeology,
Fujian Museum.
240 L. Fu
Fig. 13.4 Geometric stamped pattern pottery of the Minjiang River region from Zhou to Han,
from Gangtou (岗头) M1 of Wuyishan (1, 2), M1 of the Hulushan Site (3, 4), and the Chengcun
Site in Wuyishan (5–8), including: 1, proto-porcelain Guan jar with matted pattern; 2,
proto-porcelain Guan jar with checked pattern; 3, 4, stone ware Guan jar with matted pattern;
5, stone ware Weng (瓮) jar with small checked pattern; 6, tile brick with geometric pattern; and 7,
8, tile with diamond grid pattern
Fig. 13.5 The stamped, painted and incised geometric patterns of pottery from the Middle
Neolithic Age in Taiwan, including examples of: 1, 2, stamped checked pattern; 3, 4, stamped grid
pattern; 5, painted checked pattern; and 6, incised checked pattern. Examples are drawn from: 1, 2,
Xuntangpu Culture; 3, Niumatou Culture; 4, 5, Niuchouzi Culture; and 6, Eastern Cord Pattern
Culture
pottery from the Fengbitou Site and the second E’luanbi (鹅銮鼻) Site in south
Taiwan (Liu 2002; Tsang 1995; Wu 1999). At this stage most of the pottery from
Yuanshan Culture in the north, Yingpu Culture in the center, Dahu Culture in the
south, and Beinan Culture in the east and southeast is undecorated. Indeed, by now
the cord pattern had almost disappeared and only limited amounts of geometric
stamped pottery were present. Following the Yuanshan Culture large amounts of
checked, rectangular grid, laddered, zigzag, leaf vein and other stamped geometric
pattern pottery emerged in the Zhiwuyuan Culture of north Taiwan (Fig. 13.6: 9,
10; Chen and Kuo 2004).
The Early Iron Age began in Taiwan around 2,000 BP, and lasted until about
200 BP. In contrast to Fujian, the geometric stamped pattern pottery from this
period in Taiwan continued to develop over time. The red brown sandy paste
pottery from the Shisanhang Culture in north Taiwan was usually decorated with
stamped geometric patterns that are more diverse and complicated than those from
the Zhiwuyuan Culture. For example, the checked pattern in the Shisanhang
Culture illustrates a rounded angle and dot in the center of the check, as well as a
compound with different stamped and impressed patterns (Fig. 13.7; Tsang 2001;
Tsang and Liu 2001). According to discoveries at the Shuiweixipan (水尾溪畔)
Site (Song and Chang 1954) and the Luliao (鹿寮) Site (AGNMNS 2005) in
Taizhong (台中) County, the Fanzaiyuan Culture in the central region of Taiwan
also included large amounts of geometric stamped pattern pottery. The most
common patterns from this culture are the stamped checked pattern, leaf vein (or
13 A Study of Geometric Stamped Pattern Pottery … 243
Fig. 13.6 Stamped and painted geometric pattern pottery from the Late Neolithic in Taiwan,
including examples of: 1, painted parallel-line crossed grid pattern; 2, painted triangle filling lines
and circle pattern; 3, painted parallel-line woven pattern; 4, painted parallel lines and leaf pattern;
5, painted parallel thin-line crossed grid pattern; 6, stamped diamond grid pattern; 7, 9, stamped
rectangular grid pattern; and 8, 10, stamped checked pattern. Examples are drawn from: 1-8,
Zhishanyan Culture; and 9, 10, Zhiwuyuan Culture
Fig. 13.7 Geometric stamped pattern pottery from the Shisanhang Culture found at the
Shisanhang Site, including examples of: 1, 2, 5, checked pattern; 3, geometric pattern; 4, face
image and geometric pattern; 6, rectangular grid pattern; 7, leaf vein (fishtail lines) pattern; and
8-12, geometric pattern
244 L. Fu
Fig. 13.8 Geometric stamped pattern pottery from the Qiwulan Site, including examples of: 1,
zigzag pattern; 2, doubling-line diamond grid pattern; 3, waving pattern; 4, laddered grid pattern,
vertical stripe pattern, and zigzag pattern; 5, vertical stripe pattern, grid pattern and horizontal
stripe pattern; 6, grid pattern, vertical stripe pattern, horizontal stripe pattern, and zigzag pattern; 7,
wooden laddered pattern and zigzag pattern; 8, flowery pattern, laddered grid pattern and zigzag
pattern; and 9, eyelash pattern and zigzag pattern
13 A Study of Geometric Stamped Pattern Pottery … 245
during this period. In addition, one ring foot pottery bowl collected in the upper
cultural layer of the site was the same type as the blue and white porcelain bowls
and was decorated with the stamped zigzag pattern, indicating some cross cultural
exchange between the Gamalan indigenous peoples and immigrants from the Han
Culture. Geometric stamped pattern pottery declined and ultimately disappeared
after large quantities of glazed pottery and ceramic were imported to Taiwan by
Han ethnicity immigrants from mainland China over the last 200 years.
2
Guo Suqiu argues that the Xuntangpu Culture first appeared around 5000 years ago. However,
the stamped checked pattern was observed at a later stage of this culture, as indicated by the
Dazhuyuan Site in Yilan County.
246 L. Fu
13.4 Conclusion
The distribution of geometric stamped pattern pottery on both sides of the Taiwan
Strait offers important archaeological insight into the cultural characteristics of the
original inhabitants of south coastal China and maritime Southeast Asia. The dif-
fering processes of development and cultural dissemination of geometric stamped
pattern pottery on the west and east coasts of the Taiwan Strait emerged from
distinct geographical and cultural backgrounds. Most notably, on the west coast this
style flourished from 3,500-3,000 BP with a focus around the Minjiang River
Valley, then declined between 3,000 and 2,000 BP due to the arrival of
proto-porcelain and glazed pottery following cultural communications with north-
ern China. In Taiwan, geometric stamped pattern pottery developed during the
Early Iron Age, starting around 2,000 BP, and continued to about 200 BP due to
13 A Study of Geometric Stamped Pattern Pottery … 247
References
FJPM (Fujian Provincial Museum), & FJMSCM (Fujian Minyue State City Museum). (2007). The
Mound Tombs in Guanjiu Village, Pucheng County, Fujian. Archaeology, 7, 28–37. (Fujian
Puchengxian Guanjiucun Tudun Muqun《福建浦城县管九村土墩墓群》, Kaogu 《考古》).
FJPM (Fujian Provincial Museum), & FJTSM (Fujian Tanshishan Site Museum). (2015).
Tanshishan Site: Excavation of the Fujian Tanshishan Site from 1954–2004. Fuzhou: The
Straits Publishing House (Tanshishan Yizhi: Fujiansheng Tanshishan Yizhi 1954*2004 Nian
Fajue Baogao《昙石山遗址:福建省昙石山遗址1954*2004年发掘报告》).
FJPM (Fujian Provincial Museum), & FMTCRA (Fuzhou Municipal Team of Cultural Relics and
Archaeology). (2015). Excavation of Western Han Dynasty remains from the Pingshan
Subway Station Site, Fuzhou City, Fujian Relics and Museology, 3, 16–25. (Fuzhoushi Ditie
Pingshan Yizhi Xihan Yicun Fajue Jianbao《福州市地铁屏山遗址西汉遗存发掘简报》,
Fujian Relics and Museology(FujianWenbo《福建文博》).
FJPM (Fujian Provincial Museum), & WMM (Wuyishan Municipal Museum). (2017). The
Excavation of the Gangtou Western Zhou Mound Tombs in Xingtian Town, Wuyishan City.
Fujian Relics and Museology, 2, 9–14. (Wuyishanshi Xingtianzhen Gangtou Xizhou
Tundunmu Fajue Jianbao《武夷山市兴田镇岗头西周土墩墓发掘简报》, FujianWenbo
《福建文博》).
FMTCRA (Fuzhou Municipal Team of Cultural Relics and Archaeology). (2014). The 2008
Excavation of Shang and Zhou Dynasty Sites at Luohan Mountain, Xindian County, Fujian.
Fujian Relics and Museology, 4, 24–33. (Fuhzhoushi Xindian Luohanshan Shangzhou Yizhi
2008 Niandu Fajue Jianbao《福州市新店罗汉山商周遗址2008年度考古发掘简报》,
FujianWenbo《福建文博》).
FMTCRA (Fuzhou Municipal Team of Cultural Relics and Archaeology), & DHXU (Department
of History of Xiamen University). (1995). New Findings from the 1992 Archaeological Survey
on Pingtan Island, Fujian. Archaeology, 7, 577–584. (1992 Nian Fujian Pingtandao Kaogu
Diaocha Xinshouhuo《1992年福建平潭岛考古调查新收获》, Kaogu《考古》).
Huang, S. (1984). Excavation of the Zhishanyan Site, Taipei. Taipei: The Committee of Taipei
Historical Records (Taibei Zhishanyan Yizhi Fajue Baogao《台北芝山岩遗址发掘报告》,
台北市文献委员会).
Huang, Y. (2015). Preliminary analysis of the Maling Culture. In C. Wu, et al. (Ed.), Study of
Baiyue (Vol. 4, pp. 252–272). Xiamen: Xiamen University Press (MalingWenhua De Chubu
Fenxi《马岭文化的初步分析》, BaiyueYanjiu 4《百越研究》第四辑).
IACASS (Institute of Archaeology, Chinese Academy of Social Sciences). (2010). Chinese
archaeology: The Neolithic Age. Beijing: China Social Sciences Press (Zhongguo Kaoguxue
Xinshiqi Shidai Juan 《中国考古学▪新石器时代卷》).
JXPICRA (Jiangxi Provincial Institute of Cultural Relics and Archaeology), & YMM (Yingtan
Municipal Museum). (2017). Jiaoshan Kiln Site: 1983–2007 Excavation Report. Beijing:
Cultural Relics Press (JiaoshanYaozhi——1983*2007 Nian Kaogu Fajue Baogao《角山窑
址——1983*2007年考古发掘报告》).
Kokubu, N. (1981). Taiwan Archaeological Ethnography. Tokyo: Keitomo Press (Taiwan
KaoguMinzuzhi《台灣考古民族誌》, 东京庆友社).
Kuo, S. (2015). The cultural deposits of the middle Neolithic Age, Taiwan. In Fujian Provincial
Museum (Ed.), Special archaeological journal between Taiwan Straits (Vol 1, pp. 3–12).
Fuzhou: Fujian Education Press (Taiwan XinshiqiShidaiZhongqi De WenhuaYangxiang《台
湾新石器时代中期的文化样相》, Haixia Kaogu Jikan (1) 《海峡考古辑刊》一).
Kuo, S. (2016a). Discussion of the Xuntangpu Culture of North Taiwan. In Y. Liu, (Ed.),
Monograph on Taiwan prehistory (pp. 185–246). Taipei: Linking Publishing Company
(Taiwan BeibuXuntangpuWenhua De NeihanTantao《台湾北部讯塘埔文化的内涵探讨》,
Taiwan Shiqianshi Zhuanlun《台湾史前史专论》), 联经图书).
Kuo, S. (2016b). Connotations of Dabenkeng Type Pottery in Central Taiwan: Case study from the
Niupu Site, Zhanghua County. Field Archaeology, 18(2), 1–36. (Zhongbu Diqu Dabenkengshi
Taoqi De Neihan:Yi Zhanghuaxian Niupu Yizhi Weili《中部地区大坌坑式陶器的内涵——
以彰化县牛埔遗址为例》, Tianye Kaogu《田野考古》.
13 A Study of Geometric Stamped Pattern Pottery … 249
Li, B. (1981). The division and chronological sequence of geometric stamped pattern pottery in
South China. Journal of Peking University Philosophy & Social Sciences, 1, 39–57. (Woguo
Nanfang Jihexing Yinwentao Yicun De Fenqu, Fenqi Jiqi Youguan Wenti《我国南方几何形
印纹陶遗存的分区、分期及其有关问题》, Beijing Daxue Xuebao Zhexue Shehui Kexue
Ban 《北京大学学报(哲学社会科学版)》).
Liu, Y. (1982). The Goutishan Site in Shulin Town, Taipei County. Unpublished Masters Thesis,
Department of Anthropology, National Taiwan University (Taibeixian Shulinzhen Goutishan
Yizhi《台北县树林镇狗蹄山遗址》).
Liu, Y. (2002). The Indigenous Prehistory of Taiwan. Taipei: Taiwan Historica
(TaiwanYuanzhuminshi Shiqianpian《台湾原住民史史前篇》, 国史馆台湾文献馆).
Lv, R. (1959). One of the Neolithic Features of Southeastern China: Stamped Pattern Pottery.
Journal of Xiamen University, 2, pp. 45–56. (Zhongguo Dongnanqu Xinshiqi Wenhua Tezheng
Zhiyi:Yinwentao《中国东南区新石器文化特征之一:印纹陶》, Xiamen DaxueXuebao《厦
门大学学报》).
Peng, S. (1987). Ancient stamped pattern pottery in South China. Beijing: Cultural Relics Press
(Zhongguo Nanfang Gudai Yinwentao《中国南方古代印纹陶》).
Qiu, S., & Li, Z. (2006). Illustrations of the Qiwulan Pottery Guan-jars. Yilang: Lanyang Cultural
Foundation (Qile Taotao——QiwulanTaoguanTushuo《淇乐陶陶——淇武兰陶罐图说》,
兰阳文教基金会).
Song, W., & Chang, K. C. (1954). Excavations of the Prehistoric Site along Shuiwei Creek,
Taizhong County. Journal of Archaeology and Anthropology, 3, 26–38. (Taizhongxain
Shuiweixipan Shiqian Yizhi Shijue Baogao《台中县水尾溪畔史前遗址试掘报告》, Kaogu
Renlei Xuekan《考古人类学刊》).
Tsang, C. (1995). Taiwan Archaeology. Taipei: Council for Cultural Affairs (Taiwan Kaogu《台
湾考古》, “行政院”文建委).
Tsang, C. (2001). The Shisanhang Prehistorical People. Taipei: Taipei County Shisanhang
Museum (Shisanhang De Shiqian Jumin《十三行的史前居民》).
Tsang, C., & Liu, Y. (2001). Shisanhang Site: Rescue and preliminary research. Taipei: Cultural
Affairs Bureau of Taipei County Government (Shisanhang Yizhi: Qiangjiu Yu Chubu Yanjiu
《十三行遗址:抢救与初步研究》, 台北县政府文化局).
Wu, C. (1992). Lin Huixiang’s Archaeological life and scholarly reflections. Cultural Relics
World, 4, 2–4. (Linhuixiang De Kaogu Shengya Ji Xueshu Sixiang《林惠祥的考古生涯及学
术思想》, Wenwu Tiandi《文物天地》).
Wu, C. (1999). Archaeological survey of the history and culture of the Indigenous Peoples of
Southeast China. Xiamen: Xiamen University Press (Zhongguo Dongnan Tuzhu Minzu Lishi
Yu Wenhua De Kaoguxue Guancha《中国东南土著民族历史与文化的考古学观察》).
ZJPCRARI (Zhejiang Provincial Cultural Relics and Archaeology Research Institute), &
SCCRMC (Suichang County Cultural Relics Management Committee). (2001). Haochuan
Cemetery. Beijing: Cultural Relics Press (Haochuan Mudi《好川墓地》).
Chapter 14
Tapa Beaters from 5000 to 4200 BP
in Taiwan
Su-Chiu Kuo
Abstract Tapa beaters for making bark cloth first appeared in Neolithic Taiwan
during the Tapenkeng Period, around 5000 to 4200 BP, and continued to be used by
indigenous peoples throughout Taiwan’s later history. The Neolithic tapa beaters
are made of stone and they fall into two categories: (1) Tapa beaters with handles.
These developed during the early Tapenkeng Period, with a wide variety of shapes;
and (2) Tapa beaters without handles (compound beaters). These developed in the
later Tapenkeng Period and have only been found in small numbers. The compound
tapa beater appears to have emerged as a later variation of the earlier type. Evidence
suggests that tapa beaters and bark cloth culture developed locally in Taiwan. Tapa
beaters are not present on the southeast coast of mainland China during the same
time period or in earlier archaeological deposits from the mainland. Similar tapa
beaters, with handles and without handles, are known from Southeast Asia but they
are from a later date than those found in Taiwan.
14.1 Introduction
Until the first documentary records on Taiwan’s history appear in the seventeenth
century, the island appears to have had a long and arduous prehistoric age. Taiwan
entered the Palaeolithic Age by at least 30,000 BP, when the Changbin (长滨)
Culture emerges on the east coast. The Neolithic Age followed around 5600 BP,
and then the Metal Age around 1800 BP. Thousands of years of prehistoric cul-
tural development and evolution illustrate a rich, diverse, and complex cultural
system among Taiwan’s indigenous peoples that marks a unique historical devel-
opment since the prehistoric age.
This study explores the cultural landscape of Taiwan during the Early Neolithic
Age based on the analysis of a unique stone tool, the tapa beater. Tapa beaters first
appeared in Neolithic Taiwan during the Tapenkeng Period, around 5000 to 4200
Many archaeologists had previously thought that the “Tapenkeng Culture” began as
early as 7000-6000 BP (Chang 1969; Tsang 2004; Kuo 2013). However, more
recent typological and chronological research on a series of the excavated sites
indicate that deposits previously associated with “Tapenkeng Culture” actually
represent the Shuntanpu (讯塘埔) Early Culture in north, the Niumatou (牛骂头)
Early Culture in the center, the Kuoyeh (菓叶) Culture in the south and on the
western coast of Taiwan, and other regional cultures in eastern Taiwan.
“Tapenkeng-style pottery,” which is characterized by circle protrusions on the neck
of jars and the comb incised pattern decoration, was discovered in the different
stages of the “Tapenkeng Culture” dated mostly to 5000-4200 BP, while evidence
from a few sites including the bottom layer of the Anhe (安和) Site in Taichung
was dated to 5640-5490 BP (National Museum of Natural Science 2016: 284,
Chart 31). In this study, the term “Tapengkeng Period” includes the different
sub-regional types associated with the Shuntanpu Early Culture, the Niumatou
Early Culture, the Kuoyeh Culture and others whose deposits had previously been
identified as belonging to the “Tapenkeng Culture” (Table 14.1).
Archaeological sites of the Tapenkeng Period are scattered across much of
Taiwan and Penghu (Pescadores), including a series of new sites on the west coast
discovered in the last few years. These recently excavated sites are generally large
and were obviously inhabited over long periods, with evidence of cultivated rice
and millet farming. The most important sites among these include the Talungtung
(大龙峒) Site in Taipei, the lower layer of the Botanical Garden Site (植物园), the
Anhe Site in Taichung, the Nankuanli (南关里) Site in Tainan, the
East Nankuanli (南关里東) Site, the Kuoyeh Site in Penghu, the mid and lower
layer of the Huakangshan (花冈山) Site in Hualien (花莲), and the lower layer
of the Chungkuang (重光) Site in Hualien. Among these sites, the mid and lower
layers of the Huakangshan Site and the lower layer of the Chungkuang Site on the
east coast were dated to the Early Tapenkeng Period, but no Tapenkeng-style
pottery was discovered.
Archaeological discoveries have revealed the appearance of many important
cultural elements during the Tapenkeng Period, primarily from 5000-4200
BP. Material remains from this time point to drastic changes in social structure,
culture, livelihoods and settlement patterns. These shifts are especially evident in
14 Tapa Beaters from 5000 to 4200 BP in Taiwan 253
the northern, central, and southern regions of western Taiwan, where the highest
concentration of tapa beaters were recovered, although beaters were also found in
the lower layer of the Chungkuang Site, next to nephrite mines in eastern Taiwan
(Fig. 14.1). Tapa beaters from this period have mostly thick, heavy bodies and fall
into two categories: (1) Tapa beaters with handles, which emerged and developed in
the earlier stage of this period, illustrate a wide variety of shapes and have been
found in variable quantities; and (2) Tapa beaters without handles, which developed
in the later stage of this period and have only been found in small numbers.
Tapa beaters with handles, which date to between 5000 and 4200 BP in Taiwan, all
exhibit thick bodies. Archaeological sites where this type of tapa beater has been
found include: the Talungtung Site in northern Taiwan; the lower layer of the
Botanical Garden Site; the lower layer of the Huilai (惠来) Site in Taichung; the
Nankuanli Site in Tainan; and the lower layer of the Chungkuang Site in Hualien.
14 Tapa Beaters from 5000 to 4200 BP in Taiwan 255
Fig. 14.2 The dwelling and well remains of a residential complex at the Talungtung Site,
including: a plan of the households (Chu 2010: 10); and b E6-T3 section with ash pits next to well
remains (Chu 2012: 154)
A large number of tapa beaters were unearthed from the Talungtung Site, which
otherwise contained deposits representative of the Shuntanpu Early Culture. In
order to understand the significance of these tapa beaters and the cultural landscape
from which they came, their distribution and the systematic cultural composition of
the site are also important to consider.
Large-scale excavations took place at the Talungtung Site in 2009–2010. In
addition to the tapa beaters, a rich residential deposit at the site includes ash pits,
wells, ditches, and pillar hole clusters (Chu 2012). As shown in Fig. 14.2a, the
houses were a rectangular shape and were clustered together with a few of parallel
ditches built along the boundaries of each house. These ditches run primarily east–
west, with only a few oriented north to south, and their shapes are generally linear
with the longest one measuring 20 meters. The existence of wells (Fig. 14.2b)
indicates that the people associated with Shuntanpu Early Culture had learned how
to create a reliable water source. Most of the ash pits were located around the
buildings, suggesting their use for dumping garbage next to residences or on vacant
lands near villages.
Based on these discoveries, it appears that the people associated with Shuntanpu
Early Culture had already developed the necessary knowledge and techniques for
ground water exploitation, as illustrated by the presence of individual wells near
residences; water diversion, as shown by the existence of ditches in lines or grid
patterns; and planned settlements. Based on the remains of cultivated
rice (Fig. 14.4y) and large number of stone hoes, stone knives, and other farming
tools also found at the site, the Talungtung settlement pattern may relate specifically
to rice farming and agricultural livelihoods.
Shuntanpu Early Culture illustrates a wide variety of pottery types. The most
common are jars made of brown sandy paste with cord marked patterns, a low rim,
bulging belly and round bottoms. Some of these jars also have a low ring foot, and
most were coated in red slip. Most of the vessels with orange sandy paste or fine
red clay paste are Dou plates (豆) with wide rims and a ring foot, either plain with
no pattern or decorated with cord marks or red slip on a burnished surface. There is
also some pottery with a perforated rim and a circle embossed decoration around
256 S.-C. Kuo
the belly. In addition to those without any pattern decoration, many of the vessels
have cord marked patterns and a small number have a comb incised pattern, checker
impressed pattern and red line painting.
The pottery from the Talungtung Site falls into two categories: fine clay paste
and sandy paste. The fine clay paste potteries are mostly orange, and some have
mixed red particles inside the paste. The vessels in this category include jars, Dou
plates and spindle whorls. Most of the sandy paste vessels are jar-shaped con-
tainers, but this category also includes Dou plates, bottles, spindle whorls, lids and
cooking pot pedestals (Fig. 14.3; Chu 2012).
(1) Jars
Jars accounted for the majority of the pottery unearthed at the Talungtung Site
and includes two types of paste, dark brown coarse sandy and fine reddish brown
sandy. Most of the dark brown coarse sandy vessels are jars with an outward
opening rim, narrow and thick folding neck, bulging belly, and cord marked pattern
decorations on the belly. The fine reddish brown sandy pottery jars feature a
folding neck which is thinner than that of the coarse sandy jars, and some are
decorated with the cord marked pattern.
(2) Dou Plate
Most of the Dou plates are fine reddish brown clay paste pottery with a wide rim,
straight belly, round bottom, low ring foot and are usually decorated with the cord
marked pattern. One puzzling find was the discovery of a group of Dou plates in the
bottom layer of the E5-T5P3 ash pit.
(3) Bottle
Bottles from the Talungtung Site are made of reddish brown sandy paste and
have a long neck, folding bulging belly and low ring feet.
(4) Spindle Whorls
All of the spindle whorls from this site are cone shaped with a flat bottom. Most
are made from fine reddish clay paste, though some are sandy coarse paste.
(5) Pedestals
Most of the pedestals, which were used as the foot for cooking pots, have a bird
head shape and a small handle. They are made of dark brown sandy coarse paste.
Similar to other Neolithic cultures from this period in Taiwan, most of the
pottery from Shuntanpu Early Culture can be characterized as having a bulging
belly and round bottom. However, some Dou plates with a wide rim and
basin-shape also appeared in the Shuntanpu Early Culture. Considering the mar-
itime cultural contact across the Taiwan Strait, Dou plates and basins with a wide
rim and low ring foot from Shuntanpu Early Culture in northern Taiwan bear a
marked resemblance to typical pottery from the Tanshishan (昙石山) Culture in
Fujian. From the same period, many of the fine red clay paste Dou plates and basins
14 Tapa Beaters from 5000 to 4200 BP in Taiwan 257
Fig. 14.3 Collection of Pottery from the Shuntanpu Early Culture, Talungtung Site, including: a,
b jars; c wide rim Dou plates and basins; d group of Dou plates in the bottom layer of E5-T5P3 ash
pit (Chu 2010: 11, 18); e lids; f long neck bottle (Chu 2012: 268, 289); g spindle whorls;
h pedestals (Chu 2010: 19, 20)
258 S.-C. Kuo
with a round or irregularly perforated high ring foot from the Niumatou
Early Culture in central Taiwan are similar to the characteristic Dou pottery from
Late Liangzhu (良渚) Culture in Zhejiang (浙江).
Shuntanpu Early Culture left a rich and diverse deposit of stone implements
including tools such as chipped and polished axes, hoes, adzes, chisels, spears,
arrowheads, knives, choppers, scrapers, pointed tools, grinders, hammers, concave
anvils, fishnet weights, spindle whorls, tapa beaters, and jade bracelets and pen-
dants. The techniques for manufacturing these stone implements included chipping,
polishing, straight-cutting, circular rotary cutting, and hole-drilling.
The most typical stone tools from the Shuntanou Early Culture, as represented
by the Talungtung Site deposit, include the tapa beater and vary in mineral and
type. The minerals present include stone, shale, andesite, nephrite and flint, while
the types include stone hoes, adzes, chisels, arrowheads, knives, chopping tools,
flake tools, fishnet weights, stone hammers and tapa beaters (Chu 2012: 30).
(1) Stone Hoe
A total of 84 stone hoe fragments were collected, the largest concentration of any
stone implement found at the site. These hoes were mostly andesite, with a few
sandstone, and the majority were polished, with a tongue-shaped blade. One frag-
ment was identified as a shouldered stone tool (Fig. 14.4c). The discovery of so
many farming tools at the site suggests a local livelihood associated with agriculture.
The presence of a shouldered stone tool is interesting. In addition to this artifact
from the Talungtung Site, two shouldered stone tool fragments were unearthed in
the lower layer of the Botanical Garden Site in Taipei (Fig. 14.4e, f). Similar tools
were also found at the Nankuanli Site in southern Taiwan. All of the shouldered
14 Tapa Beaters from 5000 to 4200 BP in Taiwan 259
Fig. 14.4 Stone tools from the regional cultures of Tapengkeng Period, including: a hoes,
Talungtung Site (Chu 2012: Chart 28); b hoes, Talungtung Site (Chu 2012: Chart 29);
c shouldered axe/hoe (Chu 2012: Chart 194-2); d shouldered axe/hoe, Nankuanli Site (Tsang
2004: Chart 4-27); e shouldered axe/hoe, Botanical Garden Site (T55P4L35, excavated by author);
f shouldered axe/hoe, Botanical Garden Site (T59 P4L56, excavated by author); g adzes,
Talungtung Site (Upper Right Corner stepped adze Chu 2012: Chart 30); h knives, Talungtung
Site (Chu 2012: Chart 31); i stepped adzes, Tachiapingting Site (author’s material); j stepped adze,
Fengpitou Site (Chang 1969: Fig. 34-3); k spear and arrowhead, Talungtung Site (Chu 2012:
Chart 32); l chipped chopper, Talungtung Site (Chu 2012: Chart 33); m tapa beater with handle,
Talungtung Site (Chu 2012: Chart 204); n tapa beater with handle, Talungtung Site (Chu 2012:
Chart 205); o tapa beater with handle, Talungtung Site (Chu 2012: Chart 206); p tapa beater with
handle, Nankuanli Site (Tsang 2004: Chart 4-22); q tapa beater with handle, Huilai Site (Chu
2009: 82); r, s tapa beater with handle, Botanical Garden Site (T46 P1-2 L52, excavated by the
author); t tapa beater with handle, Chungkuang Site (Kuo 2016: Fig. 21); u tapa beater with
handle, Chungkuang Site (Kuo 2016: Fig. 21); v tapa beater without handle, Talungtung Site (Chu
2012: Chart 221); w fishnet weight, Talungtung Site (Chu 2012: Chart 36)
260 S.-C. Kuo
hoes found in northern Taiwan were made of andesite, while those from the
Nankuanli Site in southern Taiwan were made of olivine and basalt from Penghu,
indicating the use of local mineral rocks or stone from Penghu for the manufacture
of all the shouldered stone tools from this period. The shouldered stone tool dis-
appeared very quickly in this period and reappeared later in the Yuanshan Culture,
between 3200 and 2300 BP (Kuo 2014c).
(2) Stone Adze
The adze was the second most common stone implement found at the site, totaling
46 pieces. Most of these were nephrite from Hualien, though a few were argillite and
one fragment from a large adze was made of andesite. These artifacts were fully
polished, with straight and slanted blades. Some nephrite adzes were polished into
stepped adzes with a crossing concave (Fig. 14.4g, top right). A small number of
similar stepped stone adzes from this period were also discovered in central
(Fig. 14.4i) and southern Taiwan (Fig. 14.4j). Large quantities of stone stepped
adzes reappeared in the Yuanshan Culture from 3200 to 2300 BP (Kuo 2014c).
(3) Stone Arrowheads and Small Pointed Implements
A total of 7 pieces of fully ground shale or nephrite arrowheads and pointed
implements were discovered.
(4) Stone Knives
In all, 11 pieces of fully polished sandstone or shale knives with a semi-lunar or
rectangular shape were discovered. Some of them were perforated.
(5) Tapa Beaters
A total of 15 fragments of thick, heavy tapa beaters with handles were unearthed
from the Talungtung Site. Most of these had broken handles or beating parts and
only two were fully intact. Made of grey sandstone or siltstone, the tapa beaters
found at the Talungtung Site consisted of two parts, the front beater and the back
handle. The front beater section was usually cut into horizontal grooves on one side,
with only one fragment cut into horizontal grooves on two sides (Fig. 14.4m, top).
The front beaters are polished into an oval shape with a flat, round or pointed
top. In addition to the horizontal grooves, other parts of the beaters bear apparent
scars and scratches caused by hammering (Fig. 14.4m–o). Similar observations
were made about the tapa beaters with handles found at other sites from the same
period (Fig. 14.4p–t and 14.5).
Tapa beaters without handles first appeared in Taiwan around 4300 to 4200 BP.
They have been found associated with the Shuntanpu Early Culture (Fig. 14.6a) at
the Shuntanpu Site in Taipei, and also appeared at the Chungleng (中冷) Site
14 Tapa Beaters from 5000 to 4200 BP in Taiwan 263
Fig. 14.5 Remains of cultivated rice seeds from Tapenkeng Period Sites, including: a Nankuanli
Site (Tsang 2004: Chart 4-11); and b Talungtung Site (Chu 2012: Chart 46)
(Fig. 14.6b; Liu 2007), the lower layer of the Dingchiaotzu (顶桥仔) Site
(Fig. 14.6c; Chu 2011a), and the Huilai Site (Chu 2011b) in Taichung. These tapa
beaters all have the same shape and thick, heavy body, slightly slanted cut grooves
on their wider sides for beating, and a concaved circle around their narrow sides for
attaching a wooden handle (Fig. 14.6).
Another, quite different, piece of sandstone tapa beater without a handle was also
found at the Talungtung Site. It is a heart-shaped implement with a thin, light body,
no cut grooves on the wider side for beating, and a concaved circle around the
narrow side for attaching a handle (Fig. 14.4v). Similar stone tapa beaters without
Fig. 14.6 Tapa beaters without handles from Taiwan dated to 4300 to 4200 BP, from:
a Shuntanpu Culture (Liu 2008: Fig. 4-52-5); b Chungleng Site (Tang 2012: Lower Left Figure);
and c Dingchiaotzu Site (Chu 2011a: Fig. 4-3)
264 S.-C. Kuo
handles that have thin, light bodies were discovered at the Baojingwan (宝镜湾)
Site in the Pearl River Delta of mainland China.
Except for the lack of handle, this type of beater with a thick, heavy body is quite
similar to the front beating part of beaters with handles. As a result, the beaters
without handles appear to have been developed from the ones with handles. More
specifically, however, the presence of this beater without a handle in Taiwan might
illustrate the maritime cultural influence of the inhabitants of the Pearl River Delta.
The main pottery vessels from this period were jars and bowls, most of which
feature a bulging belly and round bottom, and in some cases low ring feet, Dou
plates or tripods. Pottery lids, knobs, pedestals, and spindle whorls also existed. The
decoration of this pottery was characterized by cord marked patterns, red slip and
painting, while a few had incised comb patterns, check and shell rim patterns. The
stone tools included hoe-axes, shouldered tools, adzes and chisels, stepped adzes,
spears and arrows, knives, and tapa beaters.
Based on the latest comparative study of the prehistoric cultures of Taiwan and the
southeast coast of mainland China from 5000 to 4200 BP, the regional cultures of
the Tanpenkeng Period on Taiwan were mainly affected by the Late Liangzhu
Culture in Zheijiang, Tanshishan Culture in Fujian (福建), the culture associated
with the Baojingwan Site in the Pearl River Delta, and other archaeological cul-
tures on the southeast coast of mainland China. After these Neolithic cultural
disseminations across the Taiwan Strait, the indigenous cultural landscape of
Taiwan Island was drastically altered. One primary aspect of this change was the
apparent reliance on cultivated rice and millet, which were found in high concen-
trations at the archaeological sites along Taiwan’s west coast (Fig. 14.5a, b). The
discovery of large quantities of hoe-axes and other farming tools likewise indicates
the development of early agriculture, which supported large-scale settlement and
long-term habitation. The new, stable livelihoods based on growing crops and
farming had a significant impact on prehistoric peoples and their villages during
this period. The pottery vessels most closely linked to these residential settlements
were jars, bowls, and Kuei (簋) cups, especially the bowl and Dou with wide rims,
the Dou with a high ring foot, and the long-necked bottle. New artifacts also
associated with these villages were shouldered stone tools, stepped stone adzes,
stone knives, jade earrings, and the core remains from use of the circular rotary
cutting technique (Kuo 2014a, b, 2015).
The most prominent feature of this period is undoubtedly the appearance of
cultivated rice and millet in Taiwan. These two crops were unearthed in large
quantities at the Nankuanli Site, the Anhe Site, the Botanical Garden Site, and the
Talungtung Site. In addition to the import of the seeds from these crops to Taiwan,
the farmers equipped with the knowledge and tools to cultivate these crops might
have simultaneously migrated and arrived in Taiwan via the Taiwan Strait. As the
local indigenous peoples were then taught how to grow crops or adopted the
practice, a substantial change took place among the native Neolithic cultures. The
appearance of rice and millet cultivation promoted the establishment and devel-
opment of large scale residential settlements along the west coast of Taiwan. These
266 S.-C. Kuo
new settlements, in turn, exhibit more comprehensive spatial planning. For exam-
ple, the planned villages and households with boundaries marked by stone lines
found at the Talungtung and Niupu (牛埔) sites in central Taiwan; concentrated
cemeteries with 48 tombs containing extended burials and stone constructed
boundaries at the Anhe Site in Taichung; and the specific location of disposed waste
at the edge of residences in the lower layer of the Huilai and Niupu Sites in
Taichung (Kuo 2016). Yet the spread of cultivated agriculture across the Taiwan
Strait and its cultural impact on prehistoric settlements appears to have occurred
only on Taiwan’s west coast during the Tapenkeng Period. Thus the indigenous
cultures in the eastern region of the island continued their traditional liveli-
hood practices, such as horticulture, hunting, gathering and fishing, and continued
to live in relatively small residential villages.
These cultural elements which appear to have originated on the southeast coast
of mainland China during the Tapenkeng Period and spread to Taiwan via ocean
crossing brought drastic social-cultural changes to prehistoric Taiwan. Yet the tapa
beaters endured as a specific and native cultural feature of the island. Indeed, tapa
beaters made of stone from the Tapenkeng Period and dating to 5000-4200 BP in
Taiwan are the oldest bark cloth related tool to be discovered in eastern Asia. The
tapa beater and related bark cloth culture therefore appears to have been developed
locally in Taiwan, rather than introduced through the spread of foreign cultural
elements from mainland China or Southeast Asia. According to modern ethno-
graphic studies of Taiwan’s indigenous people, these kinds of tools were used for
beating bark cloth using a “non-woven fabrics” system that developed over thou-
sands of years in Taiwan.
14.4 Conclusion
Tapa beaters made of stone from prehistoric Taiwan and dating to between 5000
and 4200 BP include both those with handles and those without handles. The front
beating section of the tapa beaters with handles includes both one sided grooves and
double sided grooves, and can have a round, flat, or pointed top. The tapa beat-
ers without handles have the same shape in the front beating section as the handled
beater, except for the concaved circle around the narrow side for attaching a
wooden handle. The tapa beater without a handle therefore appears to have
developed as a variation of the tapa beater with a handle.
Archaeological discoveries show that cord marked pattern pottery, clay spindle
whorls, and tapa beaters made of stone co-existed during Taiwan’s Tapenkeng
Period, from 5000 to 4200 BP. The cord marked pattern of pottery and clay spindle
whorls validates the existence of a “weaving textile” culture, while the appearance
of tapa beaters proves the existence of a “non-woven fabrics” culture. In other
words, the traditions of “weaving textile” and “non-woven fabrics” have
co-existed for thousands of years in Taiwan, since the Early Neolithic Age, and
14 Tapa Beaters from 5000 to 4200 BP in Taiwan 267
References
Chang, K. C. (1969). Fengpitou, Tapenkeng and the prehistory of Taiwan (p. 73). New Haven:
Yale University Publications in Anthropology.
Chu, W. L. (2009). Archaeology of cities. Taichung: National Museum of Natural Science.
Chu, W. L. (2011a, May 28–30). Report on excavation of the Dingchiaotsai Site, National Chung
Hsing University. Paper presented at 2010 Annual Conference of Taiwan Archaeologists,
National Museum of Prehistory, Taitung.
Chu, W. L. (2011b, May 28–30). Combing through Civilizations-A Second Glance into the
Features of the Niumatou Culture, Hsitun Village. Paper presented at 2010 Annual
Conference of Taiwan Archaeologists, National Museum of Prehistory, Taitung.
Chu, C. Y. (2010). Mid-report on the restoration, excavation, construction, and supervision of the
Talungtung Site. Taipei City: Taipei City Government Department of Cultural Affairs.
Chu, C. Y. (2012). Report on the restoration, excavation, construction, and supervision of the
Talungtung Site. Taipei City: Taipei City Government Department of Cultural Affairs.
Kuo, S. C. (2013). Taiwan cord-marked pottery from the same period as the Jōmonjidai and its
cultural associations. In K. Imamura & T. Izumi (Eds.), Archaeology Courses in Japan 3: The
Jōmonjidai (Vol 1, pp. 684–702). Tokyo, Japan: Aoki Bookstore.
Kuo, S. C. (2014a, September 27–28). Cultural affiliations in Taiwan and Southeast China, 4000
Years BP. Paper presented at the seminar, From the Matsu Islands to Asia’s Southeast
Coast, Institute of History and Philology, Academia Sinica, Taipei.
Kuo, S. C. (2014b, December). Exploratory research on circular rotary cutting and its rotary
mechanics. In C. Tang (Ed.), International Conference on the Prehistoric Wheels and Axles
(pp. 268–299). Macau: Praia de HacSá.
Kuo, S. C. (2014c). Discussion of the significance of the Yuanshan culture in Northern Taiwan.
Journal of Austronesian Studies, 5(2), 69–152.
Kuo, S. C. (2015). Discussion of the cultural significance of the Shuntanpu culture in Northern
Taiwan. In Y. Liu (Ed.), Institute of history and philology, Academia Sinica 80th
268 S.-C. Kuo
Abstract Barkcloth, also known as tapa, is a non-woven fabric that took on great
importance in cultures of the Pacific Islands. Archaeological evidence traces the
origins of barkcloth culture to East Asia. Stone beaters are the major tools asso-
ciated with barkcloth production. The double-shouldered barkcloth beater
(Hainan-type beater) is the only club beater found on both the East Asian mainland,
as well as in Hainan Island, Taiwan and Island Southeast Asia. It is an important
archaeological indicator of Late Neolithic and Early Bronze Age seafaring between
South China and the neighboring regions. Double-shouldered stone beaters first
appeared in mainland South China during the Late Neolithic. They began to spread
across the ocean by 4,000 BP and by 3,500 BP they existed on both the mainland
and in Island Southeast Asia. Thus evidence suggests that ancient seafarers carried
their barkcloth culture across the ocean as early as the Late Neolithic. The distri-
bution of double-shouldered stone beaters allows the identification of three oceanic
interaction spheres. One reached from the Vietnam coast of the Gulf of Tonkin to
Hainan and the Leizhou Peninsula. The second interaction sphere extended across
the South China Sea, from Hainan and the mainland to northern Luzon in the
Philippines. This likely marks the beginning of systematic seafaring across dis-
tances of over one thousand kilometers. The third interaction sphere, occurred in the
islands of western Pacific, may have reached northward from Luzon to the northeast
coast of Taiwan and possibly the Ryukyu Islands. The abundant archaeological data
on barkcloth stone tools have created a new foundation for the studies of prehistoric
maritime link of the region.
15.1 Introduction
A large amount of Early Neolithic barkcloth stone beaters were recently discovered in
the region of South China (Tang and Hayashi Tang2017). Barkcloth culture is nor-
mally believed to have close ties to the oceanic expansion of the Austronesian people
into the Pacific (Bellwood 1979; Kirch 2002). Since the 1990s, significant devel-
opments have been made in the archaeological study of the origins of barkcloth
culture in East Asia. In particular, the discovery of abundant prehistoric barkcloth
stone tools in South China and Southeast Asia have made these stone tools one of the
most important archaeological indicators in the study of prehistoric seafaring.
Beginning in the Paleolithic, seafarers used their voyaging knowledge to travel
the vast ocean and explore many unknown territories. The oldest evidence for these
early sea voyaging peoples comes from Southeast Asia (Glover and Bellwood
2004). The reconstruction of prehistory in this area has become a multidisciplinary
project, drawing upon scholars from fields such as archaeology, linguistics, and
genetics (Anderson 2005; Bellwood 1991, 1995; Chang et al. 2015; Jiao 2010). In
linguistics, study of the Austronesian languages have offered fundamental insight
into human expansion, dispersal, and interactions in the Pacific and Indian Oceans
(Bellwood et al. 1995; Blust 1985; Deng 1994; Tryon 1995). The compilation of
genetic data has made it possible to map potential routes of Austronesian expansion
that can then be compared to the proposals of other disciplines (Soares et al. 2016).
In archaeology, the discovery of material remains has challenged the course of sea
crossings; for example, the finding of Neolithic obsidian from New Britain in
northern Borneo, or the excavation of Iron Age pottery assemblages from
Austronesian speakers in Central Vietnam (Bellwood 1997; Yamagata &
Matsumura 2017). The material culture shared by these ancient seafarers can shed
light on the rich history of open-sea voyage as well as the common social systems
and identities of early seafaring communities.
Although the early cultures associated with barkcloth have been interpreted as a
reflection of Austronesian traditions, archaeological study of the relationship
between barkcloth culture and prehistoric seafaring is scant. Barkcloth, also known
as tapa, is a non-woven fabric that flourished particularly in the Pacific Ocean
(Ewins 1982; Howard 2006; Neich and Pendergrast 1997). Composite and club
stone tools were two of the major tools associated with barkcloth production
(Tolstoy 1963).
During the transition from the Neolithic to the Bronze Age, we begin to see
evidence of barkcloth production spreading from Lingnan into the rest of South
China and Southeast Asia. These prehistoric stone beaters can be divided into
composite and club and then further sorted into subtypes. The resulting evidence
illustrates that stone beaters of the same type have very distinct spatiotemporal
15 Double-Shouldered Barkcloth Beaters … 271
Fig. 15.1 Cultural spheres for barkcloth stone beaters in South China and Southeast Asia
Since the emergence of club beaters in the Late Neolithic, this type of beater
largely dominated the island regions of South China and Southeast Asia. Among
these, the Hainan type club beater (also known as the double-shouldered stone
beater) is the only club beater found on both the mainland and the islands. Yet, little
research has been carried out on the relationship of the Hainan beater to prehistoric
social networks between land and sea. Based on this central case study, this chapter
aims to shed light on prehistoric oceanic voyaging between South China and its
neighboring regions during the Late Neolithic and Early Bronze Age. It is our belief
that the abundant archaeological data on stone beaters makes barkcloth stone tools
one of the most reliable sources of evidence for prehistoric maritime links
throughout this region.
As early as the 1960s, scholar Paul Tolstoy (1963) categorized barkcloth stone
beaters into two major categories: (1) composite; and (2) club. However, the
timeline for the emergence of these types and their possible relationship remained
poorly understood. In the last two decades, archaeological discoveries in South
China have shown this region to be the earliest known birthplace of barkcloth in the
world, while composite beaters appear to predate club beaters by about 3,000 years
(Tang 2013). Thus, a discussion of the early composite beaters of the Pearl River
Delta is necessary in order to understand the emergence of the club beater and its
influence on the rest of Southeast Asia.
Composite beaters are divided into three subcategories: (1) racquet; (2) hammer;
and (3) Phung Nguyen (Tang 2012). This type of beater often has a deep channel
running along its perimeter for hafting, since it required an additional handle for
beating. Hafting materials included rattan and wood. The earliest composite beaters
were found at the Early Neolithic site of Xiantouling, located in the Pearl River
Delta in South China, and date to 6,800 BP (Tang 2013). Similar stone beaters
subsequently flourished widely in the region for over 3,000 years, a period divided
into five developmental stages from 6,800 to 3,500 BP. The earliest stages of the
composite type (such as the beaters from the Tai Wan site, in Hong Kong) were
generally thinner and lacked a channel, raising questions as to how these tools were
hafted. The earliest beater with a hafting channel was found at the Longxue site, in
Guangdong, and dates to about 6,200 to 6,000 BP. More broadly, the Early
Neolithic stone beaters in South China indicate that such technology was already
fairly mature, suggesting that the tool’s origins could go very far back and perhaps
date to the Paleolithic (Tang 2003).
The composite beaters excavated in South China are the earliest known ancestor
to the material barkcloth traditions later found in East Asia, the Pacific Islands, and
Mesoamerica. A very sophisticated barkcloth tradition with ritualistic significance
existed in the Pearl River Delta from as early as four to five thousand years ago.
15 Double-Shouldered Barkcloth Beaters … 273
At the Baojingwan site in Zhuhai, a special stone beater from more than 4,000 years
ago was found with a raised relief of four arcs and reddish purple pigment,
depicting what resembles a rainbow on both its striking surfaces. The symbolic
motif on this stone beater adds further evidence that such barkcloth traditions were
highly developed in the Pearl River Delta during the Neolithic Age. Stone beaters
with special motifs are widely found in the ethnographic record from Southeast
Asia, the Pacific Islands, and Mesoamerica (Kooijman 1972; Tolstoy 1963). The
Baojingwan composite type stone beaters also have the same typical racquet shape
later seen in Southeast Asia and Mesoamerica.
The trends we see in the earlier composite stone tools of the Pearl River Delta
appear to have influenced the double-shouldered club beaters later found in the Late
Neolithic. Double-shouldered club beaters and composite beaters share the fol-
lowing characteristics: (1) two striking platforms; (2) vertical grooved lines; and
(3) similar use-wear resulting from the prolonged beating of the contact substance.
In contrast, the biggest changes introduced by club beaters were: (1) an increase in
the beaters’ weight, from an average of 300 to 400 g to between 700 and 1000 g;
and (2) the addition of an attached handle.
The transition to club beaters in the Late Neolithic cannot easily be explained,
but plant materials may have played a factor. According to ethnographic studies,
various plants have been used as barkcloth source material in the world; most
notably paper mulberry (Broussonetia papyrifera), of the Moraceae family, has
been used extensively in South China, Taiwan, Indonesia, and across the Pacific
(Chang 2011). Breadfruit (Artocarpus) and wild fig (Ficus) have also been used in
some areas (Neich and Pendergrast 1997). A shift in the choice of plant material
(perhaps to a thicker bark) might have driven a greater preference for club type
beaters. Greater study, analysis, and archaeological experimentation are needed to
answer this question.
The emergence of club type stone beaters marks an important point in the history of
the world’s barkcloth traditions. Club beaters appeared roughly 3,000 years after
the earliest found Xiantouling composite beaters. Thus, at the same time that
composite beaters with special motifs were being used in the Late Neolithic, an
entirely different kind of stone beater was also emerging. This new type of club
stone beater is characterized by the incorporation of a stone handle, making the
striking platform and handle into one continuous implement. There are often one or
two steps between the working surface and the handle, and unlike composite
beaters, the striking surfaces of club beaters are often located on the sides of the
head. The striking surface is also sometimes carved with vertical, grid, or
cross-hatched lines; and in some rare cases, the striking surface is smooth. The
Sumu of Honduras still use club beaters for barkcloth production, today (Roth and
Lindorf 2002; Tolstoy 1991).
274 M. Hayashi Tang et al.
It is interesting to note that once club beaters departed from the mainland, such
technology had no lasting impact on the mainland region. Indeed, only a limited
number of club beaters have been found on the mainland, and composite beaters
comprised the majority of these barkcloth production tools. In contrast, the club
type beater quickly became the dominant production tool in the island regions, and
various islands appear to have had their own preferences for particular club beaters.
These different types of club beater are sorted into subcategories according to their
morphology, the presence or absence of shoulders, their striking surface charac-
teristics, the quantity of grooves, and other techno-typological characteristics (Tang
2012). Five major club types have been documented thus far: (1) Hà Jiang;
(2) Hainan; (3) Horned; (4) Palawan; and (5) Peinan (see Fig. 15.1). These sub-
categories can be described, thus:
1. Hà Jiang Type: a club beater with two grooved surfaces and grid-patterned
grooved lines. The head is usually rectangle-shaped, and there are no visual
steps between the striking platform and the handle. Large in size, with an
average length of 30 cm. Found in Hà Jiang Province, northern Vietnam.
2. Hainan Type: a double-shouldered beater with two steps between the striking
surface and the handle. The head is usually rectangle-shaped, and the striking
platforms are located on the sides. Has two grooved surfaces, and grooved lines
are mostly vertical. Found both on the mainland and in island regions of China
and Southeast Asia, including Vietnam, Yunnan, Guangdong, and Hainan, in
addition to the Philippines and Taiwan.
3. Horned Type: a single grooved surface beater with a step underneath the striking
platform. Grooved lines are often vertical, but the striking surface can be
smooth. Characterized by an elongated corner on the opposite side of the
striking platform. Coined “The Philippines Type” by anthropologist Otley Beyer
(1948) due to its abundance in the Philippines. Also found in Taiwan and
Borneo.
4. Palawan Type: a straight-back beater with a single grooved surface and only one
step underneath the striking platform. Lines are vertical. Found on Palawan, in
the Philippines, as well as throughout Taiwan.
5. Peinan Type: similar to the Palawan beater but with grooved lines cross-hatched
instead (Lien 1979). Found extensively only at the Peinan site in southern
Taiwan.
As mentioned earlier, the Hainan type beater, also known as the double-shouldered
stone beater, is the only kind of club beater found on both the mainland and
the islands. So far, sixteen specimens have been systematically documented.
15 Double-Shouldered Barkcloth Beaters … 275
These beaters have been found in the mainland regions of Yunnan, Guangdong and,
in Lạng Sơn Province in northern Vietnam, and in the island regions of Hainan,
Taiwan, and in northern Luzon in the Philippines. The production technology
required to make these prehistoric double-shouldered stone beaters could provide
crucial information regarding the possible oceanic interactions between ancient
seafarers in this region. A more detailed description of the double-shouldered stone
beater is therefore in order.
The most remarkable feature of a double-shouldered stone beater is the two steps
between the striking surfaces and the handle. In most cases, the head of the beater is
rectangular, and the plain surface of the head takes up the most area while the striking
platform (grooved surface) is located on the sides. Grooved lines are usually carved
vertically, however one grid-patterned beater has also been found (Yunnan Haidong
beater artifact no. TG1②79). The beater can have two striking platforms, but only
one grooved surface (Hongling beater artifact no. BFH④B85). The amount and
depth of the grooved lines are relatively even on both striking surfaces, although
certain exceptions have been found such as the Qituo beater of Hainan (artifact no.
BYQ①B94) and the Arku Cave beater of the Philippines. The presence of grooved
lines or lack thereof as well as the density and depth all contributed to the different
functions necessary to macerate plant fibers during the multi-stage process of bark-
cloth production. Furthermore, shoulders between the head and the handle can be
prominent or subtle; although the majority of them are the former. All handles are
polished into an oval-shape for a comfortable grip. The oval-shaped handle also
suggests that these club type beaters were used directly by hand, without the addition
of other hafting materials. The length of the handle varies and can have a 1:1, 1:2, or
2:1 ratio to its head, but handles with a 1:1 ratio to the head are most common. The
terms used to denote different sections of the beater can be seen in Fig. 15.2.
By conducting careful lithic analysis, we will be able to observe the fine com-
monalities and differences between these Hainan type beaters, details that could
potentially shed light on the dispersal of barkcloth cultural traditions through the
mainland and the islands. With this in mind, an in-depth comparative lithic and
technological analysis on the double-shouldered stone beaters of South China and
Southeast Asia is currently being prepared for future publication.
Like other club beaters, the double-shouldered stone beater appears mostly in the
island regions. Of the sixteen beaters that were systematically documented, ten were
found on the islands of Hainan, Taiwan, and Luzon, while the remaining examples
were found in mainland China and northern Vietnam. Eight of these specimens
were found in Hainan. The following section details the geographical situation of
these related sites, their excavation background, related cultural assemblages, and
the characteristics of the beaters found there.
276 M. Hayashi Tang et al.
Yunnan
Yunnan is key to the study of the origins of barkcloth culture, but unfortunately the
academic literature on prehistoric barkcloth in this region is severely limited. Situated
in southwestern China, excavations in Yunnan have yielded both composite and club
type beaters. The composite beaters have been found mainly in the western region of
the province, whereas the club beaters came mostly from the eastern side. This
distinctive distribution is similar to the distribution pattern of composite and club
beaters in the rest of Southeast Asia; a topic discussed further below.
15 Double-Shouldered Barkcloth Beaters … 277
Thus far, three double-shouldered stone beaters have been either collected or
excavated in Yunnan. One of these was excavated at the Neolithic site of Haidong,
a shell midden site located in northern Tonghai County (23°55′–24°14′N, 102°30′–
102°54′W) (Yunnan Institute of Cultural Relics and Archaeology, Yuxi City
Department of Relics Preservation, and Tonghai County Culture Bureau 1999).
A total of 372 square meters were excavated using a grid 37.2 meters wide and 10
meters long. Stratigraphy was divided into seven layers, of which six were iden-
tified as cultural layers. Among the artifacts found were pottery, bone and stone
tools (including stone adzes and axes), stone rings, net sinkers, stone and baked clay
spindle whorls, and personal accessories (Fig. 15.3, column III). Human bones
from burial M1 (Layer 2) and M13 (Layer 3) were taken to Peking University’s
archaeological laboratory for radiocarbon dating and yielded results of 3945 ± 100
and 4235 ± 150 BP, respectively. Samples from the lower layers were insufficient
for testing, but the date for the lowest layer is estimated to be around 5,000 BP. This
site was therefore inhabited around 4,000 to 5,000 BP. Similar cultural assemblages
have been found at four other shell midden sites in nearby areas, with estimated
dates similar to those associated with the Haidong site.
Unlike the typical double-shouldered stone beater, the head of the Haidong
beater (TG1②:79) is square-shaped. Moreover, its grooved surface is carved with
grid-patterned lines. The visible steps between the working surface and the handle
have led to the categorization of this beater as double-shouldered. It measures
19 cm long by 4.9 cm thick, with approximately 11 to 12 vertical and 12 to 13
horizontal grooved lines on each striking platform.
Two other beaters have been collected from Yunnan, in Pingbian County and the
township of Huangmaoling (Yunnan Institute of Cultural Relics and Archaeology,
Wenshan Zhuang and Miao and Honghe Hani and Yi Autonomous Prefecture
Departments of Relics Preservation 2008). With steps that are rather subtle, these
beaters do not have the classic typology of a double-shouldered stone beater. As a
result, closer observation is needed for further analysis. The Pingbian and
Huangmaoling beaters measure 28 cm long by 6.7 cm wide by 2.2 cm thick and
29 cm long, respectively.
Leizhou Peninsula, Guangdong
One fully intact double-shouldered stone beater has been found to the east of
Yunnan, in Guangdong, in the Yinglinating Village of the Leizhou Peninsula. It
measures 28.5 cm long by 7.6 and 7.4 cm wide by 4.1 cm thick and weighs 1,533
g, the heaviest of all the specimens. The striking surface and the handle are 15.1 cm
and 13.4 cm respectively, which means it has a nearly 1:1 ratio between the head
and the handle.
This beater exhibits the kind of common use-wear found across composite and
club beaters, with visible horizontal and vertical breakages on the grooved surfaces
and damage on the upper, outer, and middle sections of the striking platforms. The
Leizhou stone beater was also discovered with a double-shouldered stone adze and
a large stone spade (Fig. 15.3; column II). This type of large stone spade, also
known as the Guinan Stone Spade, has been found at over 150 sites in the
278 M. Hayashi Tang et al.
Fig. 15.3 Cultural assemblages unearthed across associated sites. *Double-shouldered stone adze
was unearthed at the Xuntangpu site (from Liu et al. 2008). **Personal ornaments were unearthed
from Dazhuwei, a late Xutongpu cultural site in Taiwan (from Liu et al. 2001). ***All items in
chart are not in scale
neighboring province of Guangxi (Jiang and Peng 1992; Li 2011). These associated
assemblages could therefore provide information on the time period in which the
Leizhou beater may have been used; a topic discussed further below.
15 Double-Shouldered Barkcloth Beaters … 279
Fig. 15.4 Chi Lăng beater from Northern Vietnam (artifact no. BTLS519)
possibly made from schist. This beater was also subjected to refitting. Force appears
to have been exerted on one side, resulting in a large flake scar that reveals the
cleavage of the handle.
15 Double-Shouldered Barkcloth Beaters … 281
Hainan Island
In the south of Guangdong, at the place where it is divided by the South China Sea,
lies the island of Hainan. A total of ten stone beaters have been found here.
Astonishingly, eight of these are doubled-shouldered, making Hainan the region
with the highest apparent density of this particular type of stone beater, to date. For
this reason this type of beater, with two shoulders between the striking platform and
the handle, has been coined the “Hainan Beater” (Tang 2012). The two remaining
beaters found in Hainan have yet to be categorized.
The majority of the Hainan beaters have been found in the Wuzhi Mountains, a
range located in the center of the island with five ridges. The second ridge boasts
the highest mountain peak, at 1,876 meters, and the rivers that form across these
mountains eventually lead back to the South China Sea. Hainan Island illustrates
the most variety in its double-shouldered stone beaters, such as beaters with various
handle lengths, grooved lines with drastic differences in density, and even a beater
with two striking platforms but only one grooved surface. Hainan is also the only
island where beaters have been discovered in the mountainous areas; the beaters
found on other islands have all been unearthed near the coasts. The average mea-
surement for the Hainan beaters is 22.7 cm long by 6.73 wide by 3.22 thick, and
their average weight is 907.5 g.
Arku Cave Site, Northern Luzon, The Philippines
Approximately 1,284 kilometers east of Hainan Island is Penablanca, in the
Cagayan Province of northern Luzon, the Philippines. Here, at the Arku Cave,
another double-shouldered stone beater was unearthed with a unique top (Thiel
1986). Dated to between 2,200 and 50 B.C., Arku Cave is a burial site for
approximately 57 people. Measuring 60 meters long, the large limestone cave is
situated at a transitional zone between the Cagayan Valley and the Sierra Madre
Mountains. Habitation sites have been found at seventy-one caves and seven rock
shelters in the surrounding area (Thiel 1986).
In the fall of 1976, Professor Barbara Thiel, from Northern Kentucky University,
excavated nine 2 2 meter squares in Arku Cave and discovered rich grave goods
including pottery, personal ornaments, and tools (see Fig. 15.3, column VI).
Among all the cultural material only one barkcloth beater was found, a
double-shouldered sandstone beater discovered in Level 4 of square H4/5. Square
H4/5 contains soft brown, yellowish-brown, and red sediments. Apart from the
beater, 645 bones (including skull bones stained with red ochre), six child bones, a
282 M. Hayashi Tang et al.
spindle whorl, and a jade slit earring were found in Level 4. Charcoal discovered in
the ashy area of this level was dated to 935 B.C (Thiel 1986).
Unlike the double-shouldered stone beaters found elsewhere, the Arku beater has
its own unique morphology. It measures 13.67 cm in length, the shortest of all of
the intact Hainan beaters, and its top is flat, with a notch in the middle. This beater
also contains two grooved surfaces with differing line densities: one grooved sur-
face has 22 shallow vertical lines, whereas the other surface has 5 deep vertical
lines. The top corners of both grooved surfaces show severe damage. Similar to the
Chi Lăng beater (artifact no. BTLS519), a great force appears to have been applied
to this beater from the side, creating a flaking scar on the plain surface.
Dalongdong Site, Taiwan
Turning now to Taiwan, hundreds of beaters have been unearthed across 38
archaeological sites across the island, yet only one double-shouldered stone beater
has been found. This specimen was excavated at the site of Dalongdong, a Middle
Neolithic site located in Taipei, Taiwan (25°04′26″N, 121°31′00″W) (Tree Valley
Foundation 2012). Excavations began in 2009 at the Taipei Municipal Dalong
Elementary School in the Datong District. A total of 7575.88 square meters were
excavated, and according to radiocarbon dating and the cultural assemblage, this
site appears to be a settlement of the Middle Neolithic Xuntangpu Culture. Six
charcoal samples taken for radiocarbon dating yielded a date of approximately
4,800 to 4,200 BP for this site (Tree Valley Foundation 2012). The 2.5 meter thick
stratigraphic sequence is separated into 12 strata, and strata 9 and 8 contained
cultural remains associated with Xuntangpu. Stratum 8 (30 cm thick) is the main
cultural layer of this site, comprised of yellowish-brown silt.
Abundant archaeological remains were found, including 13 intact and fractured
stone beaters, pottery shards, lithic tools, and spindle whorls (see Fig. 15.3, column
IV). Also found here were Taiwan’s earliest flint sickles, tools shaped in a
semi-ellipse with horizontally aligned double holes and used for agricultural pur-
poses. Although no personal ornaments were discovered at Dalongdong, jade and
stone rings were found at the Xuntangpu cultural site of Dazhuwei, in northeastern
Taiwan (Liu et al. 2001). Elaborate water systems also appear to have existed at
Dalongdong, based on the discovery of wells and ditches (Tree Valley Foundation
2010). Thus, by this period the people of Dalondong already had a broad under-
standing of how to extract groundwater (Kuo 2015), a vital resource in the pro-
duction of barkcloth.
Of the 13 stone beaters discovered at Dalongdong, only two were fully intact,
and one of these intact beaters was a double-shouldered stone beater (Kuo 2015;
Tree Valley Foundation 2012). It is made of siltstone and measures 22.4 cm long.
This beater also has a very unique arched top, a regional variation unheard of in
both the composite and club types. The Dalongdong beater also shows noticeable
damage on the outer and arch portions of the grooved surface. This damage is
limited, however, and the tool is still in good condition for the manufacture of
barkcloth. The other intact stone beater from Dalongdong is a horned beater, a type
widely adopted in the Philippines and also found elsewhere in Taiwan (Beyer 1948;
15 Double-Shouldered Barkcloth Beaters … 283
Lynch and Ewing 1968). The remaining eleven stone beaters are severely fractured.
Due to the large number of fractured beaters found at this site, and given the
aforementioned presence of water systems at Dalongdong, we believe the inhabi-
tants of this area likely favored the production of barkcloth as their culture
flourished.
15.6 Discussion
During the second millennium BC, a very specific kind of barkcloth stone tool
spread through a distinct spatiotemporal region to form cultural spheres based
around certain techno-typological traditions, revealing distinct regional patterns of
oceanic interaction (Fig. 15.5). To date, sixteen double-shouldered stone beaters
have been systematically documented in an area confined to the northern region of
the South China Sea, with a northernmost distribution in Taiwan, westernmost in
Yunnan, and easternmost in the northern Philippines. In the first half of the last
century, Beyer (1948) observed that beaters with “grooving on two sides of blade
[head]—sometimes fine on one side and coarse on the other, and sometimes coarse
on both sides” had been found in the Philippines. He also included an illustration of
Fig. 15.5 Distribution of double-shouldered stone beaters in and around South China
284 M. Hayashi Tang et al.
same beater type has been found in both of these regions clearly illustrates the
prehistoric sea crossings that must have taken place here. In Hainan today, the Li
People still practice barkcloth traditions, and their barkcloth production resembles
that of the Hali People in Yunnan (Tang 2012). Baisha, Changjiang, Lingsui,
Ledong, and Wuzhishan in Hainan are all within the barkcloth cultural region,
while the Lingnan region is currently accepted as the birthplace of the world’s
barkcloth. A meticulous ethnological comparison of barkcloth technologies found
on the Southeast Asian continent and among its islands could reveal significantly
more about the origins and spread of this cultural tradition.
In terms of timing, the double-shouldered stone beater was used in this region
from the Late Neolithic to the Bronze Age. In mainland China, the cultural
assemblage found together with the Leizhou beater offers some idea of when the
beater may have been used. Thus the large stone spade found in association with
this beater provides important clues about its possible time period. This spade is a
typical Guinan Stone Spade, or a large ceremonial object believed to be associated
with agriculture and commonly found in the neighboring province of Guangxi. For
example, Dalongtan is a Late Neolithic site where a total of 231 intact Guinan Stone
Spades have been found. Two samples taken from this site for radiocarbon dating
yielded results of 4750 ± 100 BP (a tree-ring date of 5320 ± 135) and
4735 ± 120 BP (a tree-ring date of 5300 ± 150) (Jiang and Peng 1992). The
Leizhou stone spade is moreover particularly similar to those unearthed at
Dalongtan, as well as the Late Neolithic site of Nalintun (Li 2011). Thus, the
Leizhou beater could had been in use around the same time, in roughly 4,000 BP. In
addition to southern Guangxi, these large stone spades have also been found
extensively in Guangdong and Hainan, and one has even been discovered in
Vietnam (Jiang and Peng 1992). As a result, definite cultural interaction appears to
have taken place in this region during the Late Neolithic, and it is quite possible that
future archaeological excavations may discover double-shouldered stone beaters in
the coastal region of Guangxi. Elsewhere in mainland China, a double-shouldered
stone beater was found in Yunnan that also dates to approximately 4,000 BP. Thus,
by then this type of club beater had already spread beyond the mainland with the
help of ancient seafarers. Indeed, by approximately 3,500 BP double-shouldered
stone beaters were already being used in both mainland and maritime Southeast
Asia.
Although the largest number of double-shouldered stone beaters have thus far
been found in Hainan, the prehistory of this island is poorly understood due to a
lack of systematic research and documentation. However, three coastal sites in
southeastern Hainan have recently shed new light on this area. (The First South
China Archaeology Team, CASS and Hainan Museum 2016). Archaeological
assemblages from the sites of Yingdun, Lianziwan, and Qiaoshan date to the
Neolithic to the Early Bronze Age, and double-shouldered stone adzes and axes
unearthed at Lianziwan have been dated to approximately 5,000 BP.
In addition, large quantities of baked-clay spindle whorls and a
quadrangular-sectioned stone adze, were excavated at the Qiaoshan site, date to
approximately 3,500 to 3,000 BP. Spindle whorls began to appear in the Yellow
286 M. Hayashi Tang et al.
River and the Yangtze River regions between 5,000 and 3,000 BC (Kuhn 1988),
but did not appear in the Pearl River region until roughly 4,000 to 5,000 years ago.
Spindle whorls were also found across the sites of Haidong, Mai Pha, Dalongdong,
and the Arku Cave (see Fig. 15.3). This evidence suggests that woven and non-
woven technology co-existed in these regions during the Late Neolithic through the
Bronze Age. The co-existence of these technologies at all four of these sites also
suggests that a similar situation may have existed at Hainan around 3,500 to 3,000
BP. While spindle whorls have been found in the coastal areas of Hainan, stone
beaters have only been found in the mountainous areas of the island. Since the
archaeological assemblages from Hainan’s coastal sites also share similarities with
the sites discussed in this chapter, they might offer further insight into when the
Hainan beaters were most likely used. By contrast, by the Late Neolithic, woven
technology dominated the barkcloth culture that had flourished in the Pearl River
Delta for over 3,000 years.
We now have a picture of the broad time frame: double-shouldered stone beaters
first appeared in mainland China (perhaps in the region of Yunnan, Guangxi, and
Guangdong) during the Late Neolithic. They then began to spread, and had dispersed
across the ocean by 4,000 BP. By the Bronze Age, or 3,500 BP, beaters of this kind
already existed in both mainland and maritime Southeast Asia. This evidence chal-
lenges earlier suggestions that stone beaters began to spread during the Bronze Age,
instead illustrating how ancient seafarers in fact brought their barkcloth technology
and knowledge across the ocean as early as the Late Neolithic.
Comparing the cultural sphere of Hainan type beaters with the other barkcloth
stone beaters, a fascinating distribution pattern emerges in the rest of South China
and Southeast Asia. As mentioned earlier, Yunnan yielded both composite and club
type beaters, and the distributions of these beaters are drastically divided within that
province. On the one hand, large numbers of hammer composite beaters have been
discovered in western Yunnan. This type of beater was also later found to the south,
in western Thailand, on the Malay Peninsula, and as far as Java, but has never been
found east of the Mekong River in Southeast Asia. The Phung Nguyen beater is
another type of composite beater found extensively in the Red River Delta. Both the
Hammer and Phung Nguyen composite beaters appeared later than the
Baojingwang beaters of Zhuhai in the Pearl River Delta, which are dated to 4,200
BP. Nowadays, Sulawesi is one of the only remaining places in the world where
barkcloth is still being produced with primarily composite beaters similar to the
Phung Nguyen type.
On the other hand, club beaters including double-shouldered and Hà Giang
beaters have been found in eastern Yunnan. Although these beaters have also been
discovered in northern Vietnam, in this region club beaters in general are found
much less frequently than the composite types. Meanwhile, as noted above,
double-shouldered stone beaters have been found in Hainan, Guangdong, Taiwan,
and the Philippines, where the composite types have either never been found
(Hainan) or have only been found in limited numbers (Taiwan and the Philippines).
Thus, based on archaeological records the distribution of barkcloth stone tools in
South China and Southeast Asia reveals a distinct dispersal pattern: composite
15 Double-Shouldered Barkcloth Beaters … 287
dispersion had perhaps later expanded far into Polynesia, where barkcloth culture is
still being practiced today.
Interestingly, jade objects, which were regarded as one of the finest symbolic
vehicles in the East, did not emerge in the coastal regions of Southeast Asia in
South China Sea until 2,000 BC (Hung et al. 2012). The sudden emergence of
precious objects from the north and the wide adaptation of barkcloth stone beaters
during the second millennium BC signifies that this period was a crucial transition
in the history of maritime navigation and interactions for this region, and the
extensive open-sea voyage of this era had continued to flourish throughout Oceania
later in the Bronze Age. Intriguingly, far in the West, a double-shouldered stone
beater was found in Central America (Tolstoy 1963). Evidence has shown that
barkcloth production was practiced in Mesoamerica in approximately 2,000
BP. Comparative analysis on the production technology of barkcloth and stone
beaters across the East and the West may provide crucial insight into the human
adaptive strategies of material production, and allow us to compare how they relate
to the cultural characteristics of these ancient societies. The abundant existing
archaeological data on barkcloth beaters have created a new foundation for the
study of prehistoric marine networks that tied together inhabitants of the mainland
and the islands, opening a brand new chapter in the rich history of ancient seafaring
in the Pacific Ocean.
Acknowledgements This research was supported by the Hong Kong Research Grants Council’s
General Research Fund Project Nos. CUHK431000 and 450413. The authors thank Professor
Tang Lingling, Professor Zhou Weimin, Deputy Director Iksam Djahidin Djorimi, Research
Fellow Kuo Su-chiu, and the Vietnam Institute of Archaeology, Hainan Museum, and Hainan
Baisha Museum for providing important archaeological data regarding barkcloth beaters in these
regions. We would also like to thank Professor Wu Chunming for the opportunity to publish here
and for his kind patience during this entire process.
References
Anderson, A. (2005). Crossing the Luzon Strait: Archaeological chronology in the Batanes
Islands, Philippines and the regional sequence of Neolithic dispersal. Journal of Austronesian
Studies, 1(2), 27–48.
Bellwood, P. (1979). Neolithic and early metal age cultures on the Southeast Asian Mainland.
Man’s conquest of the pacific: The prehistory of Southeast Asia and Oceania (pp. 153–202).
Oxford: Oxford University Press.
Bellwood, P. (1991). The Austronesian dispersal and the origin of languages. Scientific American,
265, 88–93. https://doi.org/10.1038/scientificamerican0791-88.
Bellwood, P. (1995). Austronesian prehistory in Southeast Asia: Homeland, expansion and
transformation. In P. Bellwood, J. J. Fox, & D. Tryon (Eds.), The Austronesians: Historical
and comparative perspectives (pp. 103–118). Canberra: ANU Press.
Bellwood, P. (1997). Ancient Seafarers: New evidence of early Southeast Asian Sea Voyages.
Archaeology, 50(2), 20–22.
15 Double-Shouldered Barkcloth Beaters … 289
Bellwood, P., Fox, J. J., & Tryon, D. (1995). The austronesians in history: Common origins and
diverse transformations. In P. Bellwood, J. J. Fox, & D. Tryon (Eds.), The Austronesians:
Historical and comparative perspectives (pp. 1–16). Canberra: ANU Press.
Beyer, H. O. (1948). Philippine and East Asian archaeology, and its relation to the origin of the
Pacific islands population. National Research Council Bulletin, 29. Quezon City: National
Research Council of the Philippines.
Blust, R. (1985). The Austronesian Homeland: A linguistic perspective. Asian Perspective, 26(1),
45–67.
Chang, C. S., Liu, H. L., Moncada, X., Seelenfreund, A., Seelenfreund, D., & Chung, K. F. (2015).
A holistic picture of Austronesian migrations revealed by phylogeography of Pacific paper
mulberry. In P. V. Kirch (Ed.), Proceedings of the National Academy of Sciences, 112(44),
13537–13542. http://doi.org/10.1073/pnas.1503205112.
Chang, C. S. (2011). Austronesian and tapa culture. In Chang, C. S. (Ed.), Felting bark to make
cloth: Catalogue of the tapa collections of the national Museum of Prehistory, Taiwan
(pp. 9-15). Taitung: Taiwan National Museum of Prehistory (南島語族與樹皮布文化, 《打樹
成衣: 南島語族的樹皮布及其文化》,臺東:國立臺灣史前文化博物館,2011年).
Colani, M. (1933). Céramique, procédésanciens de décoration. Bulletin de
l’Ecolefrançaised’Extrême-Orient, 33, 349–355.
Deng, X. (1994). The Ancient Austronesian languages in the Dialect of Southern Chinese.
Minority Languages of China, 3, 36–40 (Nanfang Hanyuzhong de Gu Nandaoyu Chengfen, 南
方漢語中的古南島語成分,《民族語文》,第三期,1994年, 頁36-40).
Ewins, R. (1982). Fijian Artefacts. Hobart: Tasmanian Museum and Art Gallery Collection.
Glover, I., & Bellwood, P. (Eds.). (2004). Southeast Asia: From Prehistory to history. Oxford:
Routledge Curzon.
Howard, M. C. (Ed.). (2006). Bark-Cloth in Southeast Asia. Bangkok: White Lotus.
Hung, H. C., Yang S. L., Nguyen K. D., Iizuka Y., & Bellwood P. (2012). Jade objects of Taiwan
in Oversea and Cultural Elements of Beinan. Field Archaeology of Taiwan, 15(1), 19–40
(Haiwaichutu de Taiwan yujiqi beinanwenhuayaosu海外出土的台灣玉及其卑南文化要素,
《田野考古》,第十五卷第一期,2012年,頁19-40).
Jiang, Y., & Peng, S. (1992). Guinan large stone spade. Cultural Relics in Southern China, 1, 19–
24 (Guinan DashichanYanjiu桂南大石鏟研究,《南方文物》,第一期,1992年,頁19-24).
Jiao, T. L. (2010). Prehistoric China and Polynesia: An Austronesian link. In Zhiyue Wu (Ed.),
Splendor of Hawaii and Polynesia (pp. 3–18). Fuzhou: Fujian Education Press (史前中國與波
利尼西亞:南島語族的聯繫, 《尋找夏威夷》,福州:福建教育出版社,2010年).
Kirch, P. V. (2002). On the road of the winds: An Archaeological history of the Pacific Islands
before European contact. Berkeley: University of California Press.
Kooijman, S. (1972). Tapa in Polynesia. Bernice P. Bishop Museum Bulletin, 234. Honolulu:
Bishop Museum Press.
Kuhn, D. (1988). Science and Civilization in China. Volume 5, Chemistry and Chemical
Technology; part 9, Textile Technology: Spinning and Reeling. In J. Needham (Ed.). New
York: Cambridge University Press.
Kuo, S. (2015). A Discussion on the contents of the Shuntanpu culture in Northern Taiwan. In L.
Yichang (Ed.), Monograph on Taiwan’s Prehistory (pp. 185–246). Taipei: Academia Sinica,
Linking Publishing (Taiwan Beibu Xuntangpu Wenhua de Neihan Tantao臺灣北部訊塘埔文
化的內涵探討, Taiwan Shiqianshi Zhuanlun《台灣史前史專論》,台北:中央研究院、聯經
出版公司,2015年).
Li, Z. (2011). Shell Mound, Large Stone Spade, and Burial Cave: The evolution of prehistoric cultures
in the Nanning Region. Journal of National Museum of China, 7, 58–68 (貝丘、大石鏟、岩洞葬
——南寧及其附近地區史前文化的發展與演變,《中國國家博物館館刊》,2011年).
Lien, C. M. (1979). Grooved beaters of Taiwan. The Continent Magazine, 58(4) (pp. 164–178).
(Taiwan de youcaoshibang台灣的有槽石棒,《大陸雜誌》,臺北市:大陸雜誌社,1979年).
Liu, I. C., Chung, I. H., & Yan, Y. C. (2008). Rescue excavation and data analyses of Xuntangpu
site due to the construction of Xindian line of the Taipei Metro. Taipei: NEWASIA Ltd
(Dongxixiang Kuaisu Gonglu Bali Xindianxian Bali Wuguduan Gongcheng Yingxiang
290 M. Hayashi Tang et al.
Tolstoy, P. (1963). Division of Anthropology: Cultural parallels between Southeast Asia and
Mesoamerica in the Manufacture of barkcloth. Transactions of the New York Academy of
Sciences, 25, 646–662. https://doi.org/10.1111/j.2164-0947.1963.tb01485.x.
Tolstoy, P. (1991). Paper Route: Were the manufacture and use of bark paper introduced into
Mesoamerica from Asia? Natural History, 6, 6–14.
Tree Valley Foundation. (2010). Preliminary report on the rescue excavation and construction
monitoring of the Dalongdong site. Taipei: Department of Cultural Affairs, Taipei City
Government (Dalongdong Yizhi Qiangjiu Fajue Ji Shigong Jiankan Jihua Qichu Baogao《大
龍洞遺址搶救發掘及施工監看計畫期初報告》,臺灣市政府文化局委託財團法人樹谷文
化基金會執行之報告,2010年).
Tree Valley Foundation. (2012). Report on the rescue excavation and construction monitoring of
the Dalongdong site. Taipei: Department of Cultural Affairs, Taipei City Government
(Dalongdong Yizhi Qiangjiu Fajue Ji Shigong Jiankan Jihua Chengguo Baogao《大龍洞遺址
搶救發掘及施工監看計畫成果中報告》,臺灣市政府文化局委託財團法人樹谷文化基金
會執行之報告,2012年).
Tryon, D. (1995). Proto-Austronesian and the major Austronesian Subgroups. In P. Bellwood,
J. J. Fox, & D. Tryon (Eds.), The Austronesians: Historical and comparative perspectives
(pp. 19–41). Canberra: ANU Press.
Yamagata, M., & Matsumura, H. (2017). Austronesian migration to central Vietnam: Crossing over
the Iron Age Southeast Asian Sea. In P. J. Piper, H. Matsumura, & D. Bulbeck (Eds.), New
perspectives in Southeast Asian and Pacific Prehistory (pp. 333–356). Canberra: ANU Press.
Yunnan Institute of Cultural Relics and Archaeology, Yuxi City Department of Relics Preservation,
and Tonghai County Culture Bureau. (1999). Excavation Report on TonghaiHaidong Shell
Midden Site. Cultural Relics of Yunnan, 2, 11–27 (Tonghai Haidong Beiqiu Yizhi Fajue Bagao
通海海東貝丘遺址發掘報告,《雲南文物》,第二期,1999年,頁11-27).
Yunnan Institute of Cultural Relics and Archaeology, Wenshan Zhuang and Miao and Honghe
Hani and Yi Autonomous Prefecture Departments of Relics Preservation. (2008).
Archaeological report of the border region of Yunnan province. Kunming: Yunnan
Technology Publishing House (Yunnan Bianjing Diqu Kaogu Diaocha Baogao《雲南邊境
地區(文山州和紅河州)考古調查報告》, 昆明:雲南科技出版社,2008年).
Chapter 16
Prehistoric Migration and Cultural
Change in the Philippine Archipelago
Eusebio Z. Dizon
Abstract Perhaps the earliest migration and cultural change in the Philippine
archipelago happened between 4500 and 4000 years ago, during the Neolithic Age.
The initial crossing from the mainland was probably Austronesian speakers from
southern Taiwan who traveled to Batanes and northern Luzon in the Philippines.
After this first step, boat building technology developed and sea voyaging became
more convenient, allowing these early settlers to return to where they came from
and also continue to explore and colonize other distant islands of the Philippine
archipelago such as Palawan, the Visayas and Mindanao. One of the leading signs
of cultural change in the Neolithic Period, or the New Stone Age, is the shift in
lithic or stone tool technology from the crude flaking technique to the grinding
technique. In particular, the manufacture of ground adzes and axes becomes evident
in the archaeological record from this period, while settlements began to relocate
from cave to open sites. Whereas hunting and gathering characterized the lives of
local inhabitants during the Palaeolithic, or Old Stone Age, the New Stone Age was
a time of increasingly sedentary populations whose livelihoods were based on the
domestication of animals and the cultivation of plants. Pottery emerged about
3,000 years ago and continued to develop into the Metal Age.
16.1 Introduction
The role of the Philippine archipelago in the early migrations and movements of
people in Island Southeast Asia and the Pacific has been underestimated. Indeed,
before the most recent archaeological research on the Batanes area (Dizon 2007b;
Hung et al. 2007; Bellwood and Dizon 2005; Bellwood et al. 2003), it was believed
that the early peopling of the Philippines was through a “wave of migrations,” first
over land bridges from mainland Asia starting about 500,000 years ago, and then
from Indonesia and Malaysia starting about 10,000 years ago (Beyer 1947, 1948).
E. Z. Dizon (&)
National Museum of the Philippines, Manila, Philippines
e-mail: drbongdizon@yahoo.com
Mainland and Island Southeast Asia, including Madagascar to the west, to New
Guinea, New Zealand and some remote Pacific Islands to the east and south. The
origins and homeland of the Austronesians has been a topic of debate (Dizon
2007a). For example, Solheim proposed an Austronesian homeland in Mindanao,
Philippines and/or in northeastern Indonesia, which would have meant that speakers
of the language moved along a northern route before spreading throughout
Southeast Asia and the Pacific. Bellwood (1997) disagrees, arguing for the exis-
tence of an Austronesian homeland somewhere in southern China and Taiwan. His
“Out of Taiwan” hypothesis involves a north to south movement to the Philippines,
across Southeast Asia and into the Pacific.
In order for people to move and occupy the vast islands of both Southeast Asia
and the Pacific, one very important and significant technology is needed: the
capacity and knowledge to build voyaging boats. We believe that after the initial
Austronesian crossing from Taiwan to Batanes and the rest of Luzon, their
boat-building technology was greatly improved in the Philippines by learning from
their experience. This development allowed them to build better boats that could
carry them further in all directions: southward, eastward, westward and also
northward, back to where they had come from. Based on this logic, the early
peoples who occupied the Philippine archipelago were likely responsible for
moving out into Indonesia, Malaysia, the Marianas and, to some extent, the
south-easterly islands of Polynesia by around 3,500 to 3,000 years ago. Indeed, for
a nation of great voyagers, this history resonates for the Filipinos of the present.
The links between the peopling of Taiwan, Batanes, Luzon and the rest of the
Philippine archipelago, Island Southeast Asia and the Pacific become clear through
comparisons of archaeological materials such as: earthenware sherds, or broken
pots; stone and shell adzes used as basic tools for boat building; bark cloth beaters;
stone net weights for fishing; fish hooks; and other material culture such as body
ornaments, bracelets, earrings, and design elements. Scientific analysis of excavated
artifacts and ecofacts from carefully documented archaeological contexts using
dating techniques such as Carbon14 and Accelerated Mass Spectroscopy
(AMS) have demonstrated a strong link between the movement of these cultural
material remains and Austronesian speakers. In addition, the animal and plant
remain associated with these artifacts included the bones of pigs, dogs, chickens
and fish, as well as the rice impressions found in pot sherds and the botanical
remains of taro and other root crops found on the soot residues of pot sherds. Strong
similarities have also been observed among the prehistoric burials and associated
practices of these cultures. Certain types of pottery have been associated with
Austronesian speakers based on diagnostic characters such as circle stamp designs
and red slip ware. The Lapita pottery first found in the Bismarck Archipelago has its
own particular characteristics as well, some of which suggest a close affinity with
the pottery linked to Austronesian speakers. Based on these archaeological findings,
the Philippines appear to have played a significant role in peopling the Pacific and
Island Southeast Asia.
Based on anatomical evidence, there is no doubt that modern humans inhabited
both Mainland and Island Southeast Asia, including the Australian continent, at
296 E. Z. Dizon
least from the Late Pleistocene about 60,000 years ago to the Holocene Period, and
during this time there was always the possibility they could have been sailing small
boats or dugout canoes (Doran 1981). These early Palaeolithic populations had their
own lithic traditions, including chopper-chopping tools and flake industries, and
they survived largely through hunting and gathering. Yet even during the
Mesolithic Period, about 10,000 years ago, they must have had their own language
and culture prior to the appearance of Austronesian languages. By around
6,000 years ago, after the arrival of the Holocene, a proto-Austronesian language
evolved among the southern-most inhabitants of China, and may have moved
through Taiwan. The culture associated with the Austronesian language is char-
acterized by a particular Neolithic assemblage including: grinding, drilling and
polishing stone technology, which was a completely new lithic industry adapted to
boat-building and sea faring; developed pottery making; horticulture, agriculture
and domestication of animals; and of course a distinctive language that eventually
evolved into much of what we hear spoken in most of Southeast Asia and the
Pacific today.
Fig. 16.1 Shell adzes made from the giant clam Tridacna gigas, including shell adzes made from
giant clams (left), and a reconstructed shell adze hafted to a wooden handle with rattan (right)
clam and two polished stone tools, an adze and a gouge. Shell samples were used to
generate radiocarbon dates, the earliest of which was 8760 ± 130 BP (6810 BC)
(Ronquillo et al. 1993). This C14 date may be too early for the Neolithic Period,
however, since C14 dates taken from shell samples often contain enough carbon and
oxygen intake to cause an error of at least 1000–2000 years.
The Philippines has always been a crossroads in Asia, Island Southeast Asia and
the Pacific. Since the beginning of sailing and maritime activities in Neolithic
Southeast Asia, probably around 5000 years ago or more, Austronesian speakers
from southern Taiwan were able to cross the Balintang Channel between Taiwan
and the Island of Itbayat, a group of islands in the Batanes in the northern
Philippines (Bellwood 1997, 2005; Bellwood and Dizon 2005, 2008), and may
have continued all the way to Nagsabaran and Magapit, in the Lallo area of the
Cagayan Valley (Hung 2005, 2008). From the Batanes region and northern Luzon,
the boats being used by these people likely evolved into more advanced watercraft
that permitted crossings to the rest of the Philippine archipelago through Palawan,
all the way south to Mindanao and Sulu; Borneo and Sulawesi (Indonesia) to the
west and the Marianas Islands to the east. The widespread distribution of similar
Neolithic stone artifacts in both Mainland and Island Southeast Asia has previously
served as evidence for cultural movements of this kind (Bellwood 2005: 141). Yet
some suggest they might just as well indicate possible trade, especially with respect
to pottery (Solheim 2002). In view of the studies on the ability of Austronesian
canoes to make multi-directional voyages (Doran 1981), as opposed to single- or
uni-directional voyages, return voyages may have begun during the Neolithic and
“small time” maritime trading in Southeast Asia in this early period may well have
occurred. Such exchange during Southeast Asia’s Neolithic Period would have
involved the simple trading of goods and resources that were not available locally,
rather than the large-scale trade that took place around the eighth century AD in
both Mainland and Island Southeast Asia.
298 E. Z. Dizon
The Filipinos who currently inhabit the Philippine archipelago may find their
roots or origins among the Neolithic Austronesian speaking peoples who set out
from southern Taiwan to Batanes and northern Luzon in the Cagayan Valley by
3000 BC or earlier. The initial crossing of these Austronesians may have been with
simple dug-out canoes with sails (Blust 1995, 1999; Bellwood 1997, 2005). They
were coastal people who developed further maritime culture and building tech-
nology in the Philippines before sailing on to other coastal areas of the archipelago,
and to Borneo, Sulawesi, Malaysia, Indonesia, Malaysia, Thailand, Vietnam,
Cambodia and even as far as the Marianas Islands in the Pacific, a distance of more
than 2000 kilometers from the Philippines.
During the Metal Age these Austronesian speakers kept on sailing and colonizing
the rest of Southeast Asia and the Pacific, from about 500 BC to 1000 AD, or at
least 2500 years ago. Solheim (1975, 1984–85) called this the Sa-Huynh-Kalanay
Culture, whose archaeological remains illustrate direct contact and the sharing of
artifacts between central Vietnam and the central Philippines. These artifacts
include decorated and footed earthenware vessels and lingling-O earrings whose
source material is nephrite from southern Taiwan (Iizuka et al. 2005; Hung et al.
2007).
The period of the Metal Age in Southeast Asia marked the development of social
complexity. Pots became more elaborately decorated, and a new practice emerged
of reburying the dead in burial jars with offerings. Secondary and multiple burials in
jars, curved limestone burial urns, or wooden coffins also took place. By way of
clarification: a primary burial is usually an extended burial of the whole cadaver; a
secondary burial is normally just the re-burial of some of the skeletal remains after
the flesh has decomposed; and a multiple burial consists of a secondary burial of
more than one individual. The Manunggul Jar and Maitum jars are a leading
example of secondary and multiple burial jars.
The Manunggul Jar is a secondary burial earthen jar and lid with painted, incised
and impressed designs. The late American anthropologist Robert B. Fox (1970)
described the Manunggul Jar as a significant find because it offered a fine example
of the level of pottery skill and mastery attained by a Late Neolithic or Iron Age
culture (Fig. 16.2). Upon closer examination, certain details that may relate to
16 Prehistoric Migration and Cultural Change … 299
burial beliefs and practices in the ethnographic present can be seen; for example, the
practices recorded among the Ngaju of Borneo (Evangelista 2001; Dizon 2011).
On the lid are two moulded figures on a boat, evoking a scene described by Fox
as similar to the “ship of the dead” theme (Fox 1970: 112). The interpretation of
ships as such do not hold true in all cases, but it seems appropriate here. With two
hands, the figure in the rear holds the steering paddle, the blade of which is missing,
and steers rather than paddles the “ship.” The other missing piece is the mast in the
center of the boat, against whose base the rear figure braces his or her feet. Both
figures appear to have bands tied over the crown of their heads and under their jaws,
a custom encountered today in funerary preparations among the Tagbanua group in
Palawan and in many rural areas of the Philippines. The manner in which the hands
of the front figure are folded across the chest is also a position favoured during
funerary preparations for the dead among the Ifugao and Ibaloi, in the northern
Philippines, as well as the Ngaju people of Borneo.
The branched-curl designs on the upper portion of the jar show evidence of
having been dabbed with hematite, or red ochre, in patterns suggesting sea waves.
The carved prow on the spirit boat represents a head with eyes, a nose, and mouth
resembling the features of either a sea snake or a hornbill. These motifs still occur
on the watercrafts of both the Sama in the Sulu Archipelago (Philippines) and the
Iban in Borneo (Malaysia). Similar characteristics have also been observed in the
300 E. Z. Dizon
eyes, ears, nose, and mouth on numerous types of wooden figures carved by the
Austronesian peoples of Taiwan, northern groups in the Philippines such as the
Ifugao and the Ibaloi, and elsewhere in Southeast Asia.
In his ethnographic studies among the Ngaju of Borneo, Scharer (1963)
observed that the ship of the dead has been used in Ngaju burial practices and that
the boat steerer is identified there as Templong. If the dead person was a female, the
horn bill motif was used on the boat, while if they were male then the sea snake
motif was used.
Numerous cultures believe that the souls of their deceased return to their place of
origin on a boat. The Manunggul Jar documents one such example. Another
instance can be seen among the Batak, who bury high-ranking men and women in
boat-shaped sarcophagi. Further examples include the carved and colorfully painted
boat coffins, literally “ships of the dead,” that carry the Ngaju Dayak to the afterlife,
and in the ceremonial “ship” cloths from the Lampung region of southern Sumatra
that are used mainly in gift exchanges during marriage. These cloths are woven
from cotton and bear the image of a large ship, figures, fish in the water, birds
above, and sometimes a tree.
The Manunggul Jar has not been dated by AMS, leaving only the “associated
charcoal” sample found outside the jar for use in dating based on conventional C14
analysis. Given the limitations of this method, an AMS measurement from the
Manunggul Jar itself may offer some key insights on the date of this vessel. Indeed,
it could be a contemporary of the Maitum jars, or even later.
Fig. 16.3 Anthropomorphic earthenware jars from Ayub Cave in Maitum, Mindanao
The local or vernacular name for the Lebak limestone urns is “Lugfing Selaman.”
These limestone funerary urns were found in caves and rock shelters in Salangsang,
Salaman, Lebak, formerly part of the Cotabato Province but now included within
the Sultan Kudarat Province in Mindanao, Philippines. The urns are from the Metal
Age (500 BC to the tenth century AD), and these particular artifacts date to the sixth
century AD. They come from the Austronesian Culture, which was animistic and
generally practiced jar burial.
The Lebak Jars are the quadrilateral type of limestone funerary urns considered
to be one of the early types. A C14 date of 585 ± 85 years AD was taken from the
collagen of human bones found associated with the burial, a date that belongs to the
Metal Age in the Philippines (Dizon 1983, 1988, 1998). The urns are characterized
by a fluted design and basal flanges. In their research, Kurjack and Sheldon (1970)
found that the earliest burial urns have quadrilateral shape basal flanges and excised
decorations, while the urns from a later period have the same shape but with fluted
decorations, then subsequently shift to a fluted quadrilateral shape without basal
flanges and finally, in the latest stage, become circular with fluted designs and basal
projections (Briones and Chiong 1977: 208).
From perhaps the third century AD to at least 1000 AD, greater movements of
people may have been occurring in the Magsuhot area, Bacong in Negros, and
Kalanay Cave in Masbate in today’s Philippines, as well as central Vietnam,
Indonesia and Cambodia. There is still relatively limited archaeological evidence
302 E. Z. Dizon
from this period, and that which does exist is primarily decorated pottery such as
presentation vessels and material that references Hindu-Buddhistic culture such as
the Nandi, the sacred cow of Brahma found in Kalanay Cave (Solheim II 1964,
2002: 246 and Plate 8).
One example is the Bacong Jar burials, first reported by Tenazas in 1974 as the
most important Iron Age site in the Philippines (Tenazas 1974, 1982). These are
earthenware jar burials located Magsuhot, Bacong, Negros Oriental in the Visayas
area of central Philippines. They exhibit both trunconical and cylindrical types and
are elaborately decorated, including some decorations based on anthropomorphic
figures. There are also some pottery coffins with lids shaped like a roof. These were
associated with human remains, metal artifacts, and glass bracelets and beads of
various colours, sizes, and shapes. Some of the pottery has two openings, one on
top and the other on the bottom, suggesting a kind of chimney lid for other pots.
The style and decorations of the Bacong Jars share characteristics with the
Sa-Huynh-Kalanay types of pottery that were widely distributed through the
Philippines, Vietnam and Thailand (Solheim II 1964).
The Agusan Image of the Golden Tara, which is now housed at the Field Museum
in Chicago, was recovered from the riverbank of the Agusan River and suggests a
similar link between regions (Fig. 16.4). In the ninth century AD, the Laguna
Copper Plate Inscription (LCI) provides evidence for such connections between
Java and Laguna in Luzon (Postma 1991). The town of Pila, in Laguna Province,
could have also formed a part of this connection. The American pioneering
anthropologist and archaeologist H. Otley Beyer (1947 Corrected], 1952) has
referred to this maritime connection of “Men of Champa” or “Orang Dampuans”
from southern Annam (Vietnam) who had contact with Cambodia and China as
well as Sulu. The association or relationship between the Austronesian animists and
Hindu-Buddhistic peoples does not appear to have flourished since there is no
archaeological evidence of the kind of monumental architecture in the Philippines
during this period as existed in Java; for example, Prambanan and Borobodur,
Funan and Champa in central Vietnam and Angkor Wat in Cambodia.
Nevertheless, there may well have been early Filipino contract workers who worked
in these places, as suggested by Scott (1984, 1989, 1995). A strong similarity can
be seen in the pottery motifs of women’s breasts from the Philippines and those
present on the stone work at MySon in Vietnam from the eighth to the tenth
centuries AD.
16 Prehistoric Migration and Cultural Change … 303
The Calatagan Pot, an earthenware footed presentation vessel with syllabary script
inscribed on its shoulder, is the best example of a funerary decorated pot as a burial
offering, or “Pang-alay” (Dizon 2003a). The Calatagan Pot is the first evidence of a
syllabary writing system in the Philippines known as baybayin, a very significant
find. In addition to decorated pots, glass beads, paste and opaque beads are found in
the early period, followed by translucent beads in the later period of the Metal Age.
Real metal artifacts such as copper, bronze and iron implements were also present,
not all of which may have been manufactured in the Philippines.
304 E. Z. Dizon
The Philippine archipelago was actively involved in the thriving Southeast Asian
Maritime Trade and Shipping Network from the ninth century AD until the Spanish
colonial period, from the sixteenth to the nineteenth centuries AD. Early archaeo-
logical evidence from both land and water illustrate the presence of stone and shell
adzes used for dug-out boat-making as well as a number of plank-built and
edge-pegged wooden boats, the earliest of which dates to as early as the fourth
century AD and was found in the Philippines. These archaeological finds, as well as
Chinese Annals and Spanish historical accounts, established the essential setting
and milieu for the subsequent success and expansion of the Southeast Asian
Maritime Trade and Shipping Network (Blake 1994) as well as the later
long-distance maritime trade with Europe via Acapulco on the Manila galleons
(Dizon 1997, 1998, 2003b, 2005).
The tenth century AD marks the establishment of a large-scale maritime trading
system and network where goods like high-fired ceramics, more varieties of metals,
and perhaps clothing and spices were moved around Mainland and Island Southeast
Asia in exchange for forest products, pearls, beeswax and other goods (Brown
2002; Flecker 2002, 2003, 2005). The high volume of foreign trade centering on
Malacca began to decline following the attack of the Cola Dynasty from Southern
India in 1025 AD, which eventually caused the fall of the Sri Viajayan Empire.
With the demise of the Sri Vijaya the trade route along the Western edge of the
South China Sea began to wane, necessitating a shift in regional trade patterns that
came to dominate the area for the next two centuries. Arab and Chinese traders
embarked on trading missions in Southeast Asia seeking the point of origin for the
region’s famous goods, especially the spices from Java (Hall 1985: 123). This
development resulted in the rise of a new political power in Java centering around
the Brantas River, which had direct foreign trade contacts as illustrated by the
Kamalagyan inscription (1037 AD) translated by Jan Wiessman (1977). This trend
made a significant impact on the routes linking China with the spice trade. Instead
of taking the earlier route through the western South China Sea, traders began
moving along the eastern edge of the South China Sea, which made the Philippines
a way-station between Java and China. The emergence of a zone in the Java Sea
region as a commercial power from the eleventh to the twelfth centuries drew
Chinese traders in the Sulu Sea region to trade for spices and marine products. This
newly established trade route made the Philippines and Borneo the fourth com-
mercial trading zone in the region. With the arrival of Chinese trade, an extensive
and intensive trade network grew up in the Philippines to meet the demands of
foreign exchange (Hutterer 1974).
16 Prehistoric Migration and Cultural Change … 305
Large thriving maritime communities are known to have existed from the tenth to
the twelfth century AD. Butuan, called Fu’tuan in the Chinese Annals, is located in
northeastern Mindanao and existing evidence indicates that it sent tributary trade to
China via Champa. Large volumes of high-fired Chinese trade ware ceramics from
to the eleventh century AD have been recovered, as well as large earthenware
basins, high-fired crucibles for reworking beads, wooden implements used as fish
lures, toys and a set used for gold-working, bronze gongs and shields, and the bones
of chickens, pigs and deer, some of which were made into implements and orna-
ments. In recent studies of the Butuan archaeological record, Bolunia (2013, 2015,
2016) has linked it to the maritime trade network in the tenth to the thirteenth
century AD. Primary and secondary burials in large wooden coffins and the practice
of skull reformation were also encountered at these sites.
One of the most significant finds from Butuan are plank-built and edge-pegged
wooden boats discovered in a muddy anaerobic environment. A total of nine such
boats are known to exist, but to date only three have been excavated. The available
C14 dates for the three excavated boats are 320, 1250 and 990 AD. All three of them
average fifteen meters in length and three meters across the beam (Ronquillo 1985,
1987, 1990, 1992). However, a more recent study and re-analysis of these Butuan
boats by Lacsina (2014, 2015) suggests that they were consistently made in the
ninth and tenth centuries AD.
All the excavated wooden boats illustrate the characteristic edge-pegged method
of construction which is typical of Southeast Asian boat-making technology. The
planks are one continuous piece, carved to shape, and are made of the hardwood
timber Dungon (Heritieralittoralis). The planks are pegged to the keel every
12 centimeters by hardwood pins or dowels. These pins are 19 centimeters long
and are driven into holes on the edge of each board.
The most distinctive feature of the wooden planks is a succession of flat and
rectangular protrusions, or lugs, which are carved from the same wooden plank on
the upper side, or inside, of the boats. These lugs are placed exactly opposite one
another on each plank and are 78 centimeters apart, with holes along their edges
and tops through which cords or lashings can be passed. The use of these lugs was
confirmed by the recovery of cordage of the cabo negro palm fiber (Arenga pinata),
and their presence indicates an older ship-building technology (Scott 1984). New
archaeological data, analyses and dates taken by Lacsina from wood samples of the
Butuan Boats suggest a median date of 870–904 AD (Lacsina 2015: 129). The
exceptional preservation of these Butuan boats is due in part to their abandonment
along the river banks. Rather than being shipwrecked, they were buried by the
geomorphological changes along the old river beds of Butuan.
306 E. Z. Dizon
Marco Polo’s arrival in China via land in the thirteenth century, sometime around
1274 AD, inaugurated the “Silk Route” trade and the documentation of a range of
Chinese cultural traditions and technologies, including “boat building” and naval
architecture. In his book, The Description of the World: The Book of Marvels,
Marco Polo wrote:
certain vessels, and these are the largest of them, also have thirteen bulkheads, that is to say,
internal compartments, made of strong planks that are well joined. Thus should misad-
ventures befall the vessel, and it is pierced in several places… the skipper will find the place
in which the ship had been damaged and the goods contained in the bulkhead which is
flooded will be removed and placed in others; for water cannot pass from one to another,
they are so firmly closed. (Goddio 2002: 26)
wreck is a good example for illustrating how not just Chinese junks, but Southeast
Asian trading vessels were conducting commercial activities in the Philippines.
The Lena Shoal wreck was discovered by local fishermen off the waters of
Busuanga, Northern Palawan, the Philippines while doing spear fishing activities.
Investigations by National Museum personnel revealed a wreck and its cargo
located approximately 6.5 nautical miles from the nearest island of Calauit,
Busuanga, Palawan. The shipwreck is located at a depth of between 48 to 50 meters
below the sea surface on a sandy and rocky seabed. The boat is an edge-pegged
plank-built type measuring approximately 25 meters long and 15 meters wide. This
type of boat was identified as a Chinese trading vessel.
The discovery of the Lena Shoal wreck was made in February of 1997 off the
northeast tip of Palawan Island when the fishermen found Chinese ceramics at a
depth of 48–50 meters below the water surface. The underwater archaeological
research was conducted jointly by the Far Eastern Foundation for Nautical
Archaeology (FEFNA), headed by Franck Goddio, and the National Museum of the
Philippines. The Lena Shoal wreck has been dated to roughly the end of the
fifteenth century AD based on the identification of ceramic cargoes from China,
Vietnam and Thailand. With more than 3000 ceramics and other artifacts, it is
suspected that the vessel was bound for the ports of Hormuz and Aden. Some of its
cargo would also have been traded for exotic products in the Moslem Sultanates of
the Philippines, Borneo or the Moluccas. The underwater archaeological excava-
tions revealed a portion of the bottom of the hull of the vessel which had been
preserved due to the concretion of the iron ingots and the sedimentation of a sand
layer to about a meter in thickness. The vestiges of the wood were partially
observed over a length of 18.3 meters and a maximum width of 5 meters.
The Santa Cruz wreck site was located approximately six nautical miles from the
main town of Santa Cruz, Zambales, on the west coast of the China Sea. The wreck
site was discovered by local fishermen while spear fishing. The vessel was found at
a depth of 32 m below the level of the sea surface and measured about 25 meters
long and 12 meters wide. The seabed topography was relatively flat and composed
of silt and clay mud. The underwater visibility at the time of diving activities was
generally good, ranging from 5 to 6 meters when undisturbed. The average water
current then was about one knot, felt only at a depth of between 6 and 15 meters;
practically insignificant at the bottom or seabed (Orillaneda 2008).
Between the thirteenth and fifteenth centuries AD, real long-distance maritime
trade routes were discovered and frequented between China, Southeast Asia and all
the way to India, Africa and the Middle East including Egypt, Iran and Iraq. For
example, an Egyptian high-fired ceramic sherd has been found in Laurel, Batangas
Province in Luzon. The bulk of the porcelain and other ceramic trade wares coming
from the shipwrecks of Pandanan, Lena Shoal, Palawan and Santa Cruz, Zambales,
including the archaeologically undocumented wreck in Jolo, Sulu, were from the
Hongzhi Period during the Ming Dynasty. This was a time when the Emperors of
the Ming Dynasty in China were Muslims and so the ceramic designs featured
Islamic motifs. Most of these ceramics were actually designed for the Muslim world
and Chinese eunuchs chose the designs for their market economy. For example,
308 E. Z. Dizon
Admiral Zheng He came from a Muslim family from Kunyang, Central Yunan in
China. His father, who claimed to be a descendant of an officer of Genghis Khan,
died in battle in 1381. The young Zheng He was captured by the Ming Army and
then castrated to become a eunuch. He was eventually associated with Emperor
Yong Le of the Ming Dynasty in 1402 AD (Dery 1996).
H. Otley Beyer (1948) mentioned an Italian Franciscan monk by the name of
Odoric who was in China and was able to sail with a merchant’s ship to Bolinao,
Pangasinan as part of this “Maritime Trade” ca. 1384 AD, and continue all the way
to Sulawesi and other parts of Southeast Asia. It was in this context of maritime
trade during the thirteenth and fourteenth centuries that traders from China, who
were themselves Muslims, were able to introduce Islam in Luzon, and mainly in
Manila and Batangas. Meanwhile, Islam may have independently reached
Tawi-tawi and Jolo in the southern-most parts of the Philippines through the Arab
traders from Malaysia and Indonesia at a much later period in the fifteenth century.
While the Land Silk Route through the Mongolian Dessert involved a long and
tedious voyage whose camels could only carry a limited quantity of trade ware
ceramics, the boats and ships of the Maritime Route could carry large quantities of
trade wares and other trade goods. Thus, the Maritime Route became more and
more profitable, and ceramic artifacts recovered from the Lena Shoal, Santa Cruz
and Pandanan shipwrecks have the same type of Islamic motifs observed in the
collections of the Topkapi Museum in Iran (Crick 2001; Goddio 2002; Carswell
2000).
It has been speculated that since traders of great Middle Eastern empires like
Egypt, Iran and Iraq might have been charged higher taxes in China, their Arabian
Dow, or sewn type, boats may not have sailed directly through the South China Sea
but instead taken a different route, as suggested by the Belitung wreck found in the
Java Sea in Indonesia. In fact, it is more likely that they depended on the
well-established Southeast Asian trade network. In this situation, the Philippine
archipelago came to serve as an entrepot from the redistribution of trade goods
throughout the region. Large quantities of trade goods and different types of
ceramics such as porcelains, celadon and stoneware jars, and possibly silk, metals
and beads could have switched boats or been transferred to Arabian boats in Brunei
(L’Hour 2001). This hypothesis is supported by the Arab or Indian shipwrecks in
Belitung, Indonesia (Flecker 2002, 2003, 2005) and Jolo in Sulu (Dizon 2003a, b)
from the ninth to the fourteenth centuries AD. Indeed, this could have characterized
the social and economic situation of Brunei and Sulu when they became
“Sultanates.” Trade goods bound for the Middle Eastern empires were controlled by
the head men, or barangay, which were the basic political and social unit of
Philippine societies (Dizon and Mijares 1999). A hegemony of barangay would
become a Sultanate. The shipwreck materials from Brunei and Looc in Jolo, Sulu
were similar to those found on the Lena Shoal and Santa Cruz. Most of the
porcelain cargoes show Islamic motifs, and were therefore designed for the Muslim
world.
From the fifteenth to the sixteenth century AD, maritime trade was heightened in
Southeast Asia with the participation of Europeans, including Portuguese and
16 Prehistoric Migration and Cultural Change … 309
Spanish traders in search of spices in the “Spice Islands.” During this period the
world, which was still believed to be “flat,” was divided between Spain and
Portugal by the Treaty of Tordesillas in 1494. The Portuguese were the first to reach
Malacca, East Timor and Sulawesi, and then Macao and Taiwan. The Portuguese
explorer Ferdinand Magellan was already in Malacca by 1509 and might have
sailed with the Southeast Asian maritime traders when he joined “Enrique de
Malacca,” who also claimed to be “Enrique de Carcar” (a town in Cebu) to visit the
Philippine archipelago even before he (Magellan) claimed his famous “discovery of
the Isles of St. Lazarus for Spain in 1521” (Quirino 1991). Christianity was then
officially introduced to the Philippines with the celebration of the “First Mass in
Limasawa” on March 16, 1521. Ruy Lopez de Villalobos subsequently changed the
name from St. Lazarus to Yslas Felipenas, or the Philippines, after King Philip II of
Spain, in 1542. It was Villalobos who set the boundary between the Philippines and
Indonesia, where the spice islands of Sulawesi were already under the control of the
Portuguese. This period saw the expansion of European culture in Southeast Asia,
and by the seventeenth century AD a battle between the Dutch and the Spanish off
Fortune Island near Manila Bay, in the area of Nasugbu, Batangas, the Philippines,
resulted in the sinking of the San Diego under the command of the Spanish captain
Antonio de Morga on the 14th of December, 1600 (Dizon 1993, 1995, 2016;
Goddio, 1996).
Conducted jointly by the National Museum of the Philippines and the French
organization, World Wide First (WWF), from 1991–93, the underwater archaeo-
logical excavation of the Spanish ship San Diego has brought to light some possible
previous mistakes and given archaeologists the opportunity to examine a great
volume of material culture remains from the vessel. The trade ceramics, in par-
ticular, have offered insight since most of the recovered items were complete pieces
of Kraak and Swatow wares from the Wanli Period (1573–1619 AD) of the Ming
Dynasty (1368–1644 AD). A number of trade jars from Burma, China, Thailand
and Spain were also recovered. Earthenware materials from both the New World
and local production sources were quite unusual. The San Diego wreck site was like
a time capsule with all the artifacts from the Philippines, China, Southeast Asia,
Japan, Spain, Peru and Mexico combined in one underwater site. There was a total
of 14 European model cannons recovered at the wreck site.
In 1985 the remnants of an East India Company vessel, the Griffin, were found
off Basilan Island in the southern Philippines (Dizon 2003a, b). The Griffin was one
of three British vessels plying the maritime trade route and was returning to
England when it floundered and sank. All on board were saved by the two
accompanying vessels and made their way back to England. The archaeological
materials recovered from the site include ceramics, most of which is porcelain and
comprises the bulk of the cargo; octagonal and oblong dishes; a group of tea or
coffee cup sets; underglazed blue and celadon-glazed semi-detached models of a
couple with broken heads; clay pipes; ivory fan sticks; and glass from different
shaped bottles. A few metal objects were also discovered, including “iron ingots
used as ballast, iron tools such as adzes, cannon balls, lead sheets used to line the
wooden tea crates, lead musket balls, tea pots, a Chinese coin of copper alloy, shoes
310 E. Z. Dizon
and belt buckles of copper alloy and gilt bronze, and a few other objects used for
daily life on board ship” (Goddio 1988).
From the eighteenth to the nineteenth centuries more European frigates, sailing
vessels and steam ships visited the Philippines, and most of the shipwrecks from
this period and found in the southwestern part of Palawan and northwestern
Mindanao were of British and Dutch origin. This is a strong indication that the
British Empire, which had established itself in Malaysia and Singapore, might have
extended its sea lane activities into the southern regions of the Philippine
archipelago.
To date numerous Southeast Asian boats and European galleons have been
reported and worked on by the National Museum, the primary institution mandated
by law to undertake underwater archaeological research activities in Philippine
territorial waters.
16.6 Conclusion
The interpretation of maritime trading activities in Southeast Asia from the ninth to
the nineteenth centuries AD presented here is based on archaeological, archival,
historical, linguistic, oral history and ethno-archaeological evidence. The excava-
tion of material remains, many of which have been saved from vandalism and
destruction by pothunters, treasure hunters and gold seekers, provides a fascinating
story of the thriving early Southeast Asian maritime trade and shipping networks as
well as the later long-distance trading activities of the Manila galleons between the
East and the West and across the vast Pacific Ocean. The resulting data offers
crucial insight on this important early period of “globalization” leading up to the
modern-day era.
References
Beyer, H. O. (1947). Outline review of Philippine archaeology by Island and Provinces. Philippine
Journal of Science, 77(3–4), 205–374.
Beyer, H. O. (1948). Philippine and East Asian archaeology and its relation to the origin of the
Pacific Islands population (p. 29). Manila: National Research Council of the Philippines.
Beyer, H. O. (1952). Philippines Saga. A pictorial history of the archipelago since time began (3rd
ed.). Manila: Capitol Publishing House Inc.
Blake, W. A. (1994). A preliminary survey of a Southeast Asian wreck in Phu Quoc Island,
Vietnam. International Journal of Nautical Archaeology, 23(2), 73–91.
Blust, R. (1995). The prehistory of the Austronesian-speaking peoples. Journal of World
Prehistory, 9, 453–510.
Blust, R. (1999). Subgrouping, circularity and extinction: some issues in Austronesian
comparative linguistics. In E. Zeitoun & P. J. K. Li (Eds.), Selected papers from the 8th
international conference on Austronesian linguistics (pp. 31–94). Taipei: Symposium Series of
the Institute of Linguistics, Academia Sinica.
Bolunia, M. J. L. A. (2013). Linking Butuan to the Southeast Asian Emporium in the 10th–13th
Centuries C.E.: An Exploration of the Archaeological Records and Other Source Materials.
Ph.D. Dissertation. Diliman, Quezon City: University of the Philippines.
Bolunia, M. J. L. A. (2015). Archaeological excavations in Butuan: Uncovering evidence of
maritime trade in the 10th–13th centuries. Orientations, 48(7), 62–67.
Bolunia, M. J. L. A. (2016). The Beads, the Boats, the Bowls of Butuan: Studying old things to
generate new knowledge. Journal of History, 68, 48–60.
Briones, S., & Chiong, L. (1977). Salangsang Urn Burials. In Alfredo Roces (Ed.), Filipino
heritage: The making of a nation (pp. 205–209). Singapore: Lahing Pilipino Publishing Inc.
Brown, R. (2002). Maritime Archaeology and Shipwreck Ceramics in Malaysia. Kuala Lumpur:
Department of Museum and Antiquities.
Carswell, J. (2000). Blue and White Chinese Porcelain around the World. Chicago: Art Media
Resources Ltd.
Crick, M. (2001). The Santa Cruz Wreck 2001: Preliminary archaeological report and Ceramic
Cargo study. Manila: Far Eastern Foundation for Nautical Archaeology.
Dery, L. C. (1996). A tale of treasure ships: A century before Magellan, the pearl road was already
leading to the Philippine Islands. In C. Loviny (Ed.), The Pearl Road: Tales of Treasure Ships
in the Philippines (pp. 110–121). Makati City, Philippines: Asiatype Inc.
Dizon, E. Z. (1983). Metal age in the Philippines: An archaeometallurgical investigation.
Anthropological Paper Number 12. National Museum. Manila. Philippines. (Also a M.Sc.
Thesis at the University of Pennsylvania.
Dizon, E. Z. (1988). An iron age in the Philippines?: A critical examination. Ph.D. Dissertation,
Department of Anthropology, University of Pennsylvania. University Microfilms International
8816166. Ann Arbor, Michigan.
Dizon, E. Z. (1993). War at sea: Piecing together the San Diego puzzle. In C. Valdez (Ed.), Saga
of the San Diego (A.D. 1600) (pp. 21–26). Manila, Philippines: Concerned Citizens of the
National Museum, Inc. and Vera-Reyes, Inc.
Dizon, E. Z. (1996). Anatomy of a Shipwreck: Archaeology of the 15th-century Pandanan
Shipwreck. In C. Loviny (Ed.), The Pearl Road: Tales of Treasure Ships in the Philippines
(pp. 63–94). Makati City, Philippines: Asiatype, Inc.
Dizon, E. Z. (1997). Philippines. In J. P. Delgado (Ed.), British Museum Encyclopedia of
Underwater and Maritime Archaeology (pp. 303–305). London: British Museum Press.
Dizon, E. Z. (1998). The Merchants of Prehistory. In J. Dalisay, Jr. (Ed.), Kasaysayan: The story
of the Filipino People (Vol 2, pp. 145–155). Hong Kong: Asia Publishing Co. Ltd.
Dizon, E. Z. (2003a). A second glance at the Calatagan Pot. In C. Valdes (Ed.), Pang-Alay, Ritual
Pottery in Ancient Philippines (pp. 39–42). Makati City, Philippines: Ayala Museum.
Dizon, E. Z. (2003b). Underwater and Maritime Archaeology in the Philippines. Philippine
Quarterly of Culture & Society, 31, 1–25.
312 E. Z. Dizon
Dizon, E. Z. (2005). The role of the Philippines as an entrepot during the 12th–15th century’s
Chinese and Southeast Asian trade network. In P. K. Cheng, G. Li, & C. K. Wan (Eds.),
Proceedings of the International Conference: Chinese Export Ceramics and Maritime Trade,
12th–15th Centuries (pp. 280–301). Hong Kong: Chinese Civilization Centre, City University
of Hong Kong.
Dizon, E. Z. (2007a). Austronesians. In A. M. Semah, S. Kasman, F. Semah, F. Detroit, D.
Griamaud-Herve, & C. Hertler (Eds.), First Islanders, Human Origins patrimony in Southeast
Asia (HOPSsea) (pp. 102–104). Paris: Impremeur Scriptolaser.
Dizon, E. Z. (2007b). The archaeological relationship between the Batanes Islands (Philippines),
Lanyu Island (Taiwan) and the Okinawan Islands (Japan). In M. Masako (Ed.), Archaeological
studies on the cultural diversity in Southeast Asia and its Neighbors (pp. 87–96). Japan:
Yuzamkaku Co. Ltd.
Dizon, E. Z. (2011). Maritime images and the Austronesian Afterlife. In P. Benitez-Johannot (Ed.),
Paths of origins: The Austronesian Heritage in the collections of the National Museum of the
Philippines, the Museum National Indonesia and the Netherlands Riksmuseum VoorVolenkund
(pp. 54–63). Singapore: Artpostatia Incorporated.
Dizon, E. Z. (2016). Underwater archaeology of the San Diego, a 1600 Spanish Galleon in the
Philippines. In C. Wu (Ed.), Early Navigation in the Asia-Pacific Region: A maritime
archaeological perspective (pp. 91–102). Singapore: Springer.
Dizon, E. Z., & Mijares, S. B. (1999). Archaeological evidence of a Baranganic culture in Batanes.
Philippine Quarterly of Culture & Society, 27, 1–10.
Dizon, E. Z. (1995). Archaeology and history of the San Diego, a 1600 A.D. Spanish galleon,
located offshore Fortune Island, Philippines. In International symposium on underwater
archaeology (pp. 93–135). Seoul. Korea.
Dizon, E. Z., & Santiago, R. A. (1996). Faces of Maitum. Quezon City, Philippines: Capitol Press.
Doran, E. D. (1981). Wangka: Austronesian Canoe Origins. Texas: Texas A&M University Press.
Evangelista, A. E. (2001). Soul Boat: A Filipino Journey of Self Discovery. Manila: National
Commission for Culture and the Arts.
Flecker, M. (2002). The archaeological excavation of the 10th-Century Intan Shipwreck. British
Archaeological Report International Series 1047. Archaeopress. Oxford.
Flecker, M. (2003). Cargo of the Zhangzahou porcelain found off BihnThuan Province. Vietnam.
Oriental Art, 48(5), 57–63.
Flecker, M. (2005). The advent of Chinese Sea-Going Shipping: A look at the Shipwreck
Evidence. In P. K. Cheng, G. Li, & C. K. Wan (Eds.), Proceedings of the International
Conference: Chinese Export Ceramics and Maritime Trade, 12th–15th Centuries (pp. 143–
162). Hong Kong: Chinese Civilization Centre, City University of Hongkong.
Fox, R. B. (1970). The Tabon Caves: Archaeological explorations and excavations on Palawan
Island, Philippines, 1. Manila: National Museum of the Philippines.
Goddio, F. (1996). The Treasures of the San Diego. New York: Association d’Action Artistique
and Fondation Elf and Elf Aquitane International Foundation, Inc.
Goddio, F. (2002). Lost At Sea: The strange route of the Lena Shoal junk. London: Periplus.
Goddio, F. & Jay E. (1988). 18th Century Relics of the Griffin Shipwreck. World Wide First.
Avenue de Rumine 20, Lausanne 1003. Switzerland.
Hall, K. R. (1985). Maritime trade and early state development in early Southeast Asia. Honolulu:
University of Hawaii Press.
Hung, H. C. (2005). Neolithic interaction between Taiwan and northern Luzon. Journal of
Austronesian Studies, 1(1), 109–133.
Hung, H. C. (2008). Migration and cultural interaction in Southern Coastal China, Taiwan and
the Northern Philippines, 3000 BCB to AD 100: The early history of the
Austronesian-speaking Populations. Ph.D. Dissertation. Canberra: The Australian National
University.
Hung, H. C., Iizuka, Y., Bellwood, P., Nguyen, K. D., Bellina, B., Silapath, P., et al. (2007).
Ancient jade maps: 3,000 years of prehistoric exchange in Southeast Asia. Proceedings of the
National Academy of Sciences, 104(50), 19745–19750.
16 Prehistoric Migration and Cultural Change … 313
Naoko Kinoshita
17.1 Introduction
The Ryūkyū Archipelago occupies the southernmost part of the Japanese archi-
pelago and consists of a chain of 188 islands extending over 1,300 kilometers from
Kyūshū to Taiwan. The subtropical climate and fringing coral reefs around many of
N. Kinoshita (&)
Faculty of Humanities and Social Science, Kumamoto University, Kumamoto, Japan
e-mail: kinon@gpo.kumamoto-u.ac.jp
the islands helped to support the formation of a distinctive culture that has differed
from Japanese mainland culture since prehistoric times.
Many of the shell mounds found in the Ryūkyū Archipelago contain artifacts
and bones in generally good condition. Pottery made in Kyūshū was brought to the
Ryūkyū Archipelago intermittently throughout prehistoric times (Takamiya 1978;
Shinzato and Takamiya 2014), though original pottery made by Ryūkyū inhabitants
has also been found whose style is often shared across several neighboring islands.
Thick stone axes with blades on one or both sides for woodworking are very
common at Ryūkyū archaeological sites, and one dugout canoe from about 2000
BC has been found in Okinawa.1 Based on this evidence, people appear to have
moved around the islands by boat during prehistoric times.
This chapter explores first how the seafaring of prehistoric Ryūkyūans was
influenced not only by the geographic relationship between islands but also by their
economic relations. Second, it examines the influence of seafaring on the cultural
formation of the islands based on archaeological observations. Third, it identifies
cases where cultural impacts are superior to the difficulties of seafaring in deter-
mining contact between island groups. The particular use of the word “seafaring”
stresses a repeated action that must be recognized as an historical, rather than an
accidental or occasional, behavior.
Instead of “Neolithic,” the word “prehistory” is used in order to avoid the
connotations relating to the former term. People of the Ryūkyūs depended on
fishing, hunting, and gathering, and did not accept agriculture until the twelfth
century AD; however, iron artifacts were generally used after the sixth century AD.
In addition, Kyūshū’s Neolithic culture is discussed with reference to the Ryūkyū
culture, a case in which the word “prehistory” is the most suitable.
The Ryūkyū Archipelago is a chain of islands separating the East China Sea from
the Pacific Ocean, a string of geanticline peaks that rise above the ocean and
together form a bow shape (Fujioka 1985). The chain is divided into three groups
because of two marine valleys, the Tokara Gap and the Miyako Depression, that
split this bow of geanticlines into the North Ryūkyū (NR), Middle Ryūkyū (MR),
and South Ryūkyū (SR) Islands (Fig. 17.1; Kizaki 1985).
The islands of NR and MR are mutually visible from north to south, whereas the
northernmost islands of SR are separated from these groups by 220 kilometers of
open sea and are therefore beyond the horizon (Fig. 17.2). The Ryūkyū
Archipelago is thus well connected through NR and MR, but disconnects from MR
to SR.
1
The broken tip of a Chinquapin boat was found at the Mēbaru Site on mainland Okinawa.
Calibrated radio carbon dates for this deposit are 2120–2080 and 2050–1945 BC (Tina 1999).
17 Prehistoric Ryūkyūan Seafaring … 317
The geological conditions responsible for this relationship appear to have had a
direct impact on the cultural relations between the prehistoric Ryūkyū Islands. Thus
pottery shapes and patterns indicate that NR and MR are closely related and often
betray influences from Kyūshū. Artifacts of stone, bone, and shell from these two
areas are also similar, and archaeological deposits from both groups contain deep
318 N. Kinoshita
Fig. 17.2 Distances between the islands of Kyushu and Fujian, China
pots with flat bottoms. On the other hand, the pottery from SR mostly consists of
shallow pots with round bottoms and handles, a style that appears to be isolated and
have no connection to that of Kyūshū (Fig. 17.3). It therefore appears that the
17 Prehistoric Ryūkyūan Seafaring … 319
Fig. 17.3 Pot types from Kyushu and the Ryukyu Archipelago, dating to around 3500 BP
people of MR and SR met for the first time only in the twelfth century AD (Kin and
Kinjyō 1986: 129–156; Ishigaki City History Editorial Committee 2008b).
Thus, the similarity of the material culture between islands relates closely to
seafaring, while the visibility of nearby islands served as a basic condition for
contact during the prehistoric period. As far as sailing is concerned, visibility not
only depends on the distance but the size of the island and most importantly the
height of its mountains. When an island is not high enough to be seen, it may
become difficult to reach even though it may not actually be very far away.
Fortunately, the Ryūkyūs have many mountainous islands which helped to
encourage sailing between most of the neighboring islands of NR and MR, as well
as within SR.
Geologist S. Mezaki has classified the height of the Ryūkyū Islands based on
two geomorphological types: “High Islands” (HI) and “Low Islands” (LI) (Mezaki
1980: 91–101, 1985) (Fig. 17.4). His classification seems simple, but it is based on
not only the terrain but also the systems of human subsistence relating to the
environment. HI have mountains and volcanoes, small rivers, and often large forests
of broadleaf evergreens like camellia and spinal trees. Inhabitants of this environ-
ment could find clay for pottery and metamorphic rock for stone implements, gather
nuts and hunt wild boars, and sometimes fish in reefs. Most LI, on the other hand,
have springs instead of rivers and are usually surrounded by large coral reefs where
people could find various kinds of seafood throughout the year. In many ways,
subsistence on HI was more comprehensive than on LI, which required the most
focus on fishing. There are 54 HI and 52 LI spread throughout the Ryūkyūs, and
their respective geomorphological types represent a crucial background for the
cultural characteristics that emerged on each island.
320 N. Kinoshita
Fig. 17.4 High Islands and Low Islands in the Ryukyu Archipelago (based on Mezaki 1980)
In two exceptional cases, one island can see its neighboring island but cannot
actually be seen by that neighbor. To begin with the first of these case studies,
Yonaguni Island is a HI located in the southwestern part of SR (see Fig. 17.1,
enlarged map), at the western tip of the Yaeyama Islands. When the weather
conditions are good, one can see the east coast of Taiwan from the southernmost tip
of Yonaguni Island, a distance of 108 kilometers, since the mountains of Taiwan
rise up 3000 meters above sea level. In contrast, the highest point of Yonaguni
Island is only 231 meters above sea level, and thus the island can hardly be seen
from Taiwan. Assuming that initial contact begins with people who can see another
island, any traces of prehistoric cultural exchange between the Yaeyama Islands and
Taiwan probably began on the initiative of Yaeyama’s inhabitants.
Between 4300 and 3550 cal BP, people who made Shimotabaru style pottery
lived in the coastal areas of the Yaeyama Islands (Ishigaki City History Editorial
Committee 2008a; Kin and Kinjyō 1986). They lived on fishing, hunting, and
gathering, and used partly-ground stone axes, various kinds of shell artifacts, bone
needles, tools of wild boar bone, and pendants of shark teeth (Figs. 17.5, 17.6, 17.7,
17.8). This population lived primarily on Ishigaki Island and Iriomote Island, both
large HI, and expanded their territory to include the neighboring, relatively small LI.
During the same period, people associated with the Suntangpu Culture were
living in Northern Taiwan and making cord-marked red-painted pottery
(Shisan-Hang Museum of Archaeology 2011). According to Dr. Su-Chiu Kuo, this
culture emerged around 4800 BP and ended around 3500 BP, when rice agriculture
was adopted in this area.2 People associated with Suntangpu Culture used polished
stone axes and adzes, tools, points and ornamental goods made of nephrite jade.
Large amounts of jade waste has been found in their cultural deposits. Dr. Kuo has
pointed out that this culture exhibited pottery with specific features and jade pro-
cessing techniques that did not exist in former periods but suddenly appear with the
emergence of Suntangpu Culture. Furthermore, similar pottery and jade artifacts
have been found in South China. Based on a comparative analysis, Kuo argued that
Suntangpu Culture appeared under the indirect influence of the Late Liangzhu
Culture that spread through the southeast coast of China and into Taiwan (Kuo
2014: 138–219, 2016: 185–246).
From about 4300 BP to 3550 BP, people were living in two places separated by
only 108 kilometers of water, and the difference between these two cultures is very
clear. On the one side, the people of Taiwan knew and practiced rice cultivation, and
exhibited advanced skills in the production of pottery and jade artifacts. On the other,
the people of the Yaeyama Islands lived solely from fishing and gathering, and
produced technically simple pottery and stone implements. The inhabitants of the
2
The earliest evidence of rice, broomcorn and foxtail millet in Taiwan dates to about 5000 BP
(Tsang et al. 2017: 1–10).
322 N. Kinoshita
Fig. 17.5 Shimotabaru type pottery (diameter of right pot is 18.1 cm)
Fig. 17.6 Stone implements (length of stone axe on the bottom right is 11.4 cm)
17 Prehistoric Ryūkyūan Seafaring … 323
Fig. 17.7 Shell, bone and teeth implements (length of bone implement on the left end, front row
is 20.8 cm)
Fig. 17.8 Artifacts excavated from Shimotabaru site in Hateruma Island 3630 ± 80 BP (Photo
Courtesy of Okinawa Prefecture Buried Cultural Property Center)
Yaeyama Islands likely knew about rice agriculture, the variety of prey and the
beautiful jade accessories in Taiwan; attractive features that might have enticed the
islanders travel. But to date no remains have been found in Yaeyama indicating
cultural contact with Taiwan, except some small Conus shell beads found only at the
Shimotabaru Site. After some study, these shell beads collected from the
Shimotabaru Site date to the same period as some found at the Kending (墾丁) Sites
324 N. Kinoshita
in southern Taiwan.3 Since the technical skill required to make shell beads appears to
have been much higher than the level of expertise attained by the people living at the
Shimotabaru Site, judging from their artifacts, the technique for making shell beads
appears to have been brought to Yaeyama from Taiwan (Kinoshita 1999: 315–354).
Although this remains an isolated example of this kind of contact, the Yaeyama
case may play a key role in further investigations into contact with Taiwan.
Meanwhile the Shimotabaru type pottery also appears to have been isolated in the
Ryūkyu Archipelago and its origins remain a mystery. For the Yaeyama people
Taiwan was the only visible neighbor, but replicating their journey to Taiwan
would not have been easy.
After 3500 BP, agricultural villages began to appear more widely in Taiwan.
Social complexity grew, and highly characteristic cultures emerged in several areas
along the east coast of Taiwan (Liu 2002). By contrast, in Yaeyama a unique
culture without any pottery appeared around 2800 BP and continued until the
seventh century AD. Confirmed interaction between Taiwan and the Yaeyama
Islands only begins in the first half of the twelfth century AD.
Professor Youbei Chen has pointed out that the reason for the mutual inde-
pendence of these two places was the result of their disparate subsistence patterns
(Chen 2004, 2014); a reasonable assumption given their archaeological records.
The roots of these different systems in the varying geographic character of the two
places is also important to note. The Yaeyama Islands are relatively small islands
surrounded by fringing reefs, while Taiwan is characterized by a plateau with large
rivers and a long coast of cliffs. Such extreme differences due to contrasting natural
environments may be one of the reasons for so little contact between the prehistoric
populations of these islands.
The second case study relates to three islands of SR (see Fig. 17.1, enlarged map).
Ishigaki Island is a HI that lies at the eastern end of the Yaeyama group, close to the
Miyako Island group. Only 35 kilometers to the east is Tarama Island, the
western-most island of the Miyako Island group. Continuing another 56 kilometers
to the east, Miyako Island lies at the eastern tip of this group. Both Tarama and
Miyako are LI, and Ishigaki and Tarama are mutually visible. In addition, from the
viewpoint of Tarama, in the middle of the three islands, one can see the large, low
hulk of Miyako Island even though it is almost impossible to spot the small, low
island of Tarama from Miyako.
3
Shell beads have been found at two sites in Taiwan, the Nan-Kuan-Li Site and the Nan-Kuan-Li
East Site, though the author has not seen them personally. Similar shell beads from
Strombusluhuanus have been found in E-luanbiin, in southern Taiwan.
17 Prehistoric Ryūkyūan Seafaring … 325
Cultural relations based on seafaring are usually based in some form of economic
relations, mutual help between inhabitants of neighboring islands, trade linking
remote places, or cults such as the Kula Cult. The archaeological record of pre-
historic Ryūkyū offers insight on such economic relations between the islands.
A few examples of these appear below.
Since the beginning of the first millennium BC, economic relations between
Northern Kyūshū and the Okinawa Islands in MR can be seen in the archaeological
4
To date, no human remains from 3300 to 2800 BC have been found in the Yaeyama Islands or on
Miyako Island.
5
Axes and adzes made from Tridacna shell are the most characteristic artifacts of the Non-Pottery
Culture on Miyako Island, where the oldest examples of this culture have been found. Dr. Kaishi
Yamagiwa has suggested that these shell tools were first created in the Miyako Islands as a result
of environmental adaptation (Yamagiwa 2017: 19–34).
6
Based on her excavations of the Arafu Site in the eastern part of Miyako Island, Professor
Tomoko Egami has suggested that the people who practiced stone-boiling cooking may have
arrived on Miyako Island via a route that did not touch on the Yaeyama Islands (Egami 2017: 169–
185). Dr. Yamagiwa further notes that the mutual independence between the people of the
Yaeyama Islands and Miyako Island continued throughout the period associated with the
Non-Pottery Culture (Yamagiwa 2015: 153–170).
326 N. Kinoshita
record (Fig. 17.9). People of the Yayoi Culture7 in Northern Kyūshū produced
religious items from the large conch shells that lived in the coral reef seas around
Okinawa, and permanent, long lasting exchange between Northern Kyūshū and
Okinawa was quickly established (Fig. 17.9: 12) (Kinoshita 1989). They also
acquired both Strombus sp. (Fig 17.9: 4) and Conus sp. (Fig 17.9: 6) shells in
exchange for rice, beans, and sometimes iron tools or glass beads (Fig. 17.9: 9).
Three groups of people were involved in this trade network. People from the
Yayoi culture in the north were mainly consumers (Fig. 17.9: 1, 10), while
inhabitants of southern MR (Okinawa) became the suppliers who gathered the
shells (Fig. 17.9: 3). Transport of these goods was carried out by a third population,
presumably fishermen, from the Amami Island group that lived along the island
coasts from Northern and Southern Kyūshū as well as from the northern part of MR
(Fig. 17.9: 2). Thus, trade goods were carried over one thousand kilometers from
north to south, and vice versa.
This relationship is documented by the following kinds of archaeological
material: (1) Bracelets made of Strombus and Conus, which were only available in
coral reefs, became a widespread item found in burials in Northern Kyūshū starting
in the first half of the Yayoi Period (Fig. 17.9: 7, 8); (2) in Okinawa, Yayoi pottery
from Northern Kyūshū has been found in sites from the same period in MR (or
Okinawa), while shell clusters of Strombus and Conus also appear frequently;
(3) roughly processed shells, or half-made bracelets of the kind shown in Fig. 17.9:
5, have been found in Okinawa that resemble the half-finished bracelets found in
North Kyūshū; and (4) stone cist burials, which were typical on the west coast of
Kyūshū at this time, were occasionally deposited in Okinawa and other MR islands
during the same period (Fig. 17.9: 11).
Both Strombus and Conus shells are large enough to provide material for two to
three shell bracelets, but in the Yayoi Period a single shell was used for just one
bracelet. A total of 662 bracelets made from Strombus or Conus have been found at
Yayoi sites, mostly in the burials of Yayoi men or women, where they were attached
to individuals’ arms. Meanwhile, in Okinawa, 138 shell clusters have been found
containing a total of 1,505 shells. Assuming that many clusters still remain to be
unearthed and that a significant number of them were also lost to sea travel, it would
appear that an extremely large quantity of shells was being traded at this time.
The shell trade lasted for 1600 years, with a short decline in the middle of this
period, and ended in the seventh century AD. By the time a unified Japanese state
was established in the eighth century, it had been completely abandoned. But then in
the ninth century trade relations started up again, as people from the mainland began
demanding Great Green Turban shells and Big Trumpet shells as raw material for
mother of pearl and use in the manufacture of Buddhist tools (Kinoshita 2000).
7
The Yayoi Culture is a Neolithic culture of Japan that began around 2900 BP and ended around
the third century AD. It was initiated by a series of migrations from the southern Korean peninsula
to Northern Kyūshū, when societies consisting of immigrants and natives established a culture
based on wet-rice cultivation that drew upon the Huan-he Civilization (黄河文明). The result was
the first hierarchically divided society based on agriculture. It also included the first priest officials.
17 Prehistoric Ryūkyūan Seafaring … 327
This stage of the shell trade was carried out by merchants who travelled on well-built
ships, and continued until the fourteenth century. By the time the Ryūkyū Kingdom
was established in the fifteenth century, the shell trade had been incorporated into the
national trade of the kingdom, bringing an end to the independent exchange of shells
(Ishigaki City History Editorial Committee 2008b).
328 N. Kinoshita
Above all, one key observation emerges from this historical trajectory: the rise
and fall of trade exchange depends on the demand of the consumer. Thus, the long
distance route that connected the islands during the shell trade fell out of use in the
same moment that the demand for particular non-local products disappears.
The Ryūkyū Islands shell trade was established by people from the temperate climate
zone who were demanding large conch shells from the subtropical zone. Despite a shift
in the type of shell being sought, this trade lasted over 2,300 years. During this time,
trade goods brought to Okinawa were mostly rice and metal goods, and the people
carrying goods appear to have flowed in only one direction, from north to south. Thus
people with a need for shell traveled south to the region where these shells were caught,
then brought them home to the north. No evidence has been found to indicate that people
from Okinawa might have traveled north to Kyūshū and back again.
According to the archaeological data, the inhabitants of the Ryūkyūs rarely visited
the main consumption areas. Although they had an ongoing relationship with the
consumer population from this area, they did not visit it. It also appears that the goods
they received in turn, such as cereals and iron tools, may not have been valuable
enough to actively establish trade by themselves. Thus, the decision of whether or not
to conduct sea voyages largely appears to have been a cultural one (Kinoshita 2012).
The earliest prehistoric cultural interaction between China and Taiwan may date to the
beginning of the Tapenkeng Culture.8 Based on a comparative study of pottery, Professor
Shihchiang Huang has concluded that piscatorial people from Fujian with a long tradition
of fishing and seafaring came to Taiwan in order to expand their territory. They subse-
quently decided to stay, and ended up creating the Tapenkeng Culture (Huang 1985).
Based on archaeological evidence, the most reasonable route for this migration appears to
have been along the coastal area of Fujian, through Jinmen via Penghu to Taiwan
(Liu 2011: 140). Notably, Fujian and Jinmen are intervisible, while Taiwan and Penghu
are not mutually visible. However, one can see Taiwan from Penghu, since it is a big
island with high mountains, but Penghu cannot be seen from Taiwan. We may therefore
assume that seafaring in the Taiwan Strait was still difficult in this time.9
Some cultural interaction between China and Taiwan is known to have taken place
during the period associated with the Tapenkeng Culture. Thus, the excavation of
Nankuanli East in Taiwan revealed that people began cultivating dry rice, foxtail and
broomcorn around 5000 BP, and the mixed farming of millet and rice was similar to
agricultural styles observed in central and north China (Tsang et al. 2017: 1–10).
Thus, seafaring on the Taiwan Strait seems to have become easier by this period.
8
Tapenkeng Culture is thought to have emerged around 6500 BP (Huang 1974).
9
The distance between Jinmen and Penghu is 145 kilometers, while that between Penghu and the
Taiwan mainland is 44 kilometers.
17 Prehistoric Ryūkyūan Seafaring … 329
Significant cultural influences appear to have been shared between China to Taiwan
from 4500 to 4300 BP, including jade technology, new styles of pottery and high level
kiln technology. Based on these features, Dr. Suchiu Kuo pointed out that the roots of
these cultural characteristics lie in three coastal areas of Southeast China: Zhejiang,
downstream of the Minjiang River in Fujian, and the Zhujiang Delta in Guangdong. In
addition, a contemporaneous shift was taking place as objects and people associated
with the Late Liangzhu Culture spread southward (Kuo 2014: 138–219). As a result,
cultural interaction based on seafaring appears to have been very popular between
China and Taiwan during this period, at exactly the same time that the Shimotabaru
Culture emerged in the Yaeyama Islands, only 108 to 211 kilometers east of Taiwan.
Why did people on both sides of the Strait, from both Taiwan and China,
succeed in sailing across it despite being invisible to each other? Ship building that
enables long distance sailing may be the most likely answer. Through years of
experience people may have found new ways and techniques for building ships
with greater stability that allowed them to avoid dangers at sea. Their triumph in
sailing a distance of 145 kilometers before 5000 BP was just the first step in their
pursuit of this challenge.
A deeper understanding of their seafaring successes can also be gained by
looking at the similarities between the geographical environments of China and
Taiwan. Both places have comparable big rivers, plateaus, hills, and mountains.
Even more importantly, Taiwan’s larger coastal plain relative to coastal
Southeastern China may have encouraged Chinese farmers to settle in Taiwan.
The similar characteristics of the geographical environments on either side of the
Strait likely made the people of both sides, but particularly those from China, feel
more comfortable. Their encounter with familiar geographic features may have
therefore served as a basic motivation for both sides to make and continue cultural
interactions across the Taiwan Strait. In the end, important insights can be found in
looking at the contrasting cultural interactions occurring on either side of Taiwan
around 4000 BP, and namely in the comparatively broad cultural influences passing
between Taiwan and China, on Taiwan’s east coast, and the much more limited
ones passing between Taiwan and the Yaeyama Islands, on the west coast.
17.7 Conclusion
References
Chen Y. (2004). Cultural relations between Taiwan and the Sakishima Islands from the viewpoint
of subsistence in prehistoric times. The Journal of the Okinawa Archaeological Society, 23,
31–42. (《生業の視点で捉えた台湾と先島諸島との先史文化関係》《南島考古》第23号,
沖縄考古学会).
Chen, Y. (2014). A reappraisal of prehistoric relations between Taiwan and the Ryukyu Islands:
From the perspective of geographic context. Journal of Archaeology and Anthropology, 81,
3–28. (《琉球列岛与台湾史前关系的再研究:从古代地理意识之角度》《考古人类学
刊》第81期).
17 Prehistoric Ryūkyūan Seafaring … 331
Egami, T., Mabuchi, K., & Matsuba, T. (Eds.). (2003). Research study of Arafu site on Miyako
Island, Naha City: Archaeological investigations of the Arafu Site. Tokyo: RokuIchi Shobō.
(《アラフ遺跡調査研究I》アラフ遺跡発掘調査団,六一書房).
Egami, T. (2015). The meaning of the non-pottery period on Miyako Island. Research and
investigation reports of the Okinawa Archaeological Society, 2015, 4–12. (《無土器期の位置
づけ》《沖縄考古学会2015年度研究発表会資料》).
Egami, T. (2017). The culture of the non-pottery period on Miyako Island from the viewpoint of
stone boiling remains. The Journal of the Okinawa Archaeological Society, 36, 169–185.
(《集積遺構から見た無土器期の様相》《南島考古》,沖縄考古学会).
Fujioka, K. (1985). New seminar on the topography of Japan IV: Kyūshū. Tokyo: Taimei-dō.
(《新日本地誌ゼミナールVI九州地方》,大明堂).
Huang, S. (1974). Investigation report on the Bajia Village Site, Guiren District, Tainan. Bulletin
of the Department of Anthropology of Taiwan National University, 37/38, 44–67. (《台南縣歸
仁鄉八甲村遺址調查》《國立臺灣大學考古人類學刊》, no.37/38, 國立臺灣大學出版委
員會).
Huang, S. (1985). Relations between the Neolithic culture of Southeast China and the prehistoric
culture of Taiwan. National Taiwan University Humanitas Taiwanica, 34, 1–24. (《试论中国
东南地区新石器时代与台湾史前文化的关系》《台湾大学文史哲学报》, no. 34).
Ishigaki City History Editorial Committee. (2008a). Prehistoric life in the Shimotabaru period: The
earliest pottery and culture in the Yaeyama Islands. Archaeological Visual Version of Ishigaki
City History, 2, 1–68. Ishigaki City: Ishigaki City History Editorial Committee. (《下田原期の
くらし-八重山諸島最古の土器文化》《石垣市史ビジュアル版2》, 石垣市).
Ishigaki City History Editorial Committee. (2008b). History of exchange based on ceramics.
Archaeological Visual Version of Ishigaki City history, 5, 1–54. Ishigaki City: Ishigaki City
History Editorial Committee. (《陶磁器から見た交流史》石垣市史ビジュアル版5, 石垣市).
Kin, S., & Kinjyō, K. (Eds.). (1986). Simotabaru Shell Mound, Ōddomari-bama Shell Mound.
Cultural Property Survey Report of Okinawa Prefecture, 74, 1–156. Naha City: Okinawa
Prefectural Board of Education. (《下田原貝塚・大泊浜貝塚》,沖縄県教育委員会).
Kinoshita, N. (1989). Shell trade between Okinawa and Kyūshū in the Yayoi Period. In City:
Publishing Association of Retirement Festschrift for Professor Yokoyama Kōichi (Ed.),
Archaeology on Production and Distribution I: Retirement Festschrift for Professor Yokoyama
Kōichi (pp. 203–249). Fukuoka City: Publishing Association of Retirement Festschrift for
Professor Yokoyama Kōichi. (《南海産貝輪交易考》,《生産と流通の考古学》,横山浩一
先生退官記念論文集刊行会).
Kinoshita, N. (1999). Study of shell beads in East Asia. In Obata, H. (Ed.), Research Study of
Prehistory and Archaeology (Vol. 3, pp. 315–354). Kumamoto City: Tatsuda Archaeology
Society (《東亜貝珠考》《先史学・考古学論究III》,龍田考古会).
Kinoshita, N. (2000). Kaiyuan-Tongbao and green snails: shell trade between Ryūkyū and China
in the 7th to 9th centuries. In Publishing Association of Retirement Festschrift for Professor
Takamiya Hiroe (Ed.), Man and Culture in Ryūkyū and East Asia, 1 (pp. 187–219). Naha City:
Publishing Association of Retirement Festschrift for Professor Takamiya Hiroe (《開元通宝と
夜光貝-7*9世紀の琉・中交易試論-》《琉球・東アジアの人と文化(上巻)》.高宮廣衞先生
古希記念論集刊行会).
Kinoshita, N. (2012). Formation of culture and movement of people in prehistoric Ryūkyū, with
focus on the mutual human-geographical relations between islands. Kumamoto Journal of
Culture and Humanities, 103, 13–27 (《琉球列島における先史文化の形成と人の移動-島嶼
間の人文地理的関係に注目して-》《『文学部論叢》第103号. 熊本大学文学部).
Kishimoto, Y. (1992). Investigation report on the Soedō Site, Tarama Island. In Y. Kishimoto
(Ed.), Report on the investigation of cultural property in Tarama (Vol. 11). Tarama Village:
Board of Education of Tarama Village (《多良間添道遺跡発掘調査報告》,《多良間村文
化財調査報告書第11集》,多良間村教育委員会).
Kizaki, K. (Ed.). (1985). The Geology of the Ryūkyū Arc. Naha City: Okinawa Times (《琉球弧の
地質誌》,沖縄タイムス社).
332 N. Kinoshita
Kuo, S. (2014). Archaeological cultures of Taiwan and Southeast China around 4,000 B.P. In
Cheng, Z. (Ed.), From the Matsu Archipelago to the Coastal Area of Southeast of Asia:
International Academic Symposium of Prehistoric culture and Physical Remains (pp. 138–
219). Taipei City: Institute of History and Philology, Academia Sinica.
Kuo, S. (2016). Study of Suntangpu culture. In Liu, Y. (Ed.), Prehistoric Taiwan (pp. 185–246).
Taipei: Lianjing Press (《台湾史前史专论》联经出版事业公司).
Liu, Y. (2002). History of the Taiwan natives: Edition of Taiwan Natives history, Taiwan. Taipei
City: Taiwan Historica (《台彎原住民史 史前篇》, 國史館台彎文獻館).
Liu, Y. (2011). Whole history of the Taiwan: The third edition, Taiwan Natives history and
archaeology, Taiwan. Taipei City: Taiwan Historica (《台彎全史 巻三 住民志・考古
篇》, 國史館台彎文獻館).
Lu, J. (2014). Comparison between Shimotabaru Culture in Ryukyu Sakishima Islands and
Neolithic Cultures near the Hualian River, Taiwan’s East Coast: New Technical Data. Journal
of Archaeology and Anthropology, 81, 29–84 (《琉球先岛群岛下田原期与台湾东海岸花莲
溪口新石器时代遗址之文化内涵探讨:新资料及技术选择观点的尝试》《考古人类学
刊》第81期).
Mezaki, S. (1980). Parallel zonation of high and low Islands in the Ryukyu Island Arc.
Geographical Laboratory, 5, 91–101 (《『琉球列島の地質学研究』》第5).
Mezaki, S. (1985). Explorations of the Ryukyu Arc. Ginowan City: Aki Shobō (《琉球弧をさぐ
る》,あき書房).
Shinzato, T., & Takamiya, H. (Eds.). (2014). Empirical study on the transition of the environment and
culture in the prehistoric Ryukyu Archipelago, 1. Tokyo: Rokuichi Shobō (《琉球列島先史時代・
原史時代における環境と文化の変遷に関する実証的研究 研究論文集第1集》, 六一書房).
Shisan-Hang Museum of Archaeology (Ed.). (2011). Ciphers of red-painted pottery: A special
exhibition of the Suntangpu Culture. New Taipei City: Shisanhang Museum of Archaeology
(《塗紅陶器密碼:訊塘埔文化特展専輯》).
Takamiya, H. (1978). Tentative chronology of the Neolithic period, Okinawa Islands. The Journal
of the Okinawa Archaeological Society, 6, 11–22 (《沖縄諸島における新石器時代の編年(試
案)》《南島考古》第6号).
Tina, T. (1999). Mebaru Site: Report of the investigation caused by improvement of the prefectural
Kan’na-Matsuda route. In T. Tina (Ed.), The cultural property of Ginoza, 14, 1–302 Okinawa:
Board of education of Ginoza village. (《前原遺跡-県道漢那松田線道路整備工事に伴う発
掘調査報告書-》《宜野座村乃文化財14集》,宜野座村教育委員会).
Tsang, C., Li, K., Hsu, T., Tsai, Y., Fang, P., & Hsing, Y. (2017). Caroline Broomcorn and Foxtail
millet were cultivated in Taiwan about 5000 years ago. Botanical Studies, 58(3), 1–10.
Springer Open.
Yamagiwa, K. (2015). Regional archaeological variation based on natural resource use during the
Mudoki (Non-Pottery) Period, Miyako-Yaeyama Islands. Material Culture, 95, 153–170
(《「宮古・八重山諸島・無土器期における地域間変異と生体資源利用》《物質文化》95).
Yamagiwa, K. (2017). Cultural relationships of shell adze usage between the Southern Ryukyu
Islands and Southeast Asia. Journal of Southeast Asian Archaeology, 37, 19–34 (《南琉球地
域からみた東南アジア〈貝斧利用文化〉北上の可能性》《東南アジア考古学》, no. 37).
Yamagiwa, K., & Kugai, A. (2017). Chronology of Neolithic shell adzes in the Miyako-Yaeyama
Islands, Okinawa, Japan: An analysis based on the Urasoko Site, Miyako Island. Material
Culture, 95, 113–131 (《先史時代における沖縄県宮古島を中心としたシャコガイ製貝斧の展開-
浦底遺跡出土貝斧の分析を基にした時空間的変異の検証-》《物質文化》97).