Piecing Together Sha Po: Archaeological Investigations and Landscape Reconstruction [1 ed.] 9789888313952, 9789888208982

Hong Kong boasts a number of rich archaeological sites behind sandy bays. Among these backbeaches is Sha Po on Lamma Isl

155 43 44MB

English Pages 279 Year 2016

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Piecing Together Sha Po: Archaeological Investigations and Landscape Reconstruction [1 ed.]
 9789888313952, 9789888208982

Citation preview

15mm

5

172mm

12mm

19.7mm

12mm

172mm

5

15mm

15mm 5

Piecing Together Sha Po ‘A singular effort in the field of Hong Kong archaeology, Piecing Together Sha Po adopts a social landscape approach to chart the development of a single site over millennia of occupation, revealing as it does the untapped potential which careful field investigations hold for generating a better understanding of the region’s rich past.’ —Francis Allard, Department of Anthropology, Indiana University of Pennsylvania ‘This volume is the best overview of the early history of Hong Kong that I know. The authors have articulated patterns of human settlement at Sha Po in a masterly way that informs us not only of Lamma Island, or greater Hong Kong, but of Lingnan as a whole. I welcome it as the key source for specialists and the interested public alike.’ —Charles Higham, Department of Anthropology and Archaeology, University of Otago, New Zealand ‘It is rare indeed for a multi-period study of a region to not only synthesise a vast range of archaeological material but also include incisive points of theory alongside that narrative, such as the need to understand evidence at a landscape level and questioning the utility of “Neolithic” and “Bronze Age” categories. This is such a book.’ —Steve Roskams, Department of Archaeology, University of York

The artefacts unearthed from the site—some of them unique to the region—reveal a vibrant past which saw the inhabitants of Sha Po interacting with the environment in diverse ways. Evidence showing the mastery of quartz ornament manufacture and metallurgy in the Bronze Age suggests increasing craft specialisation and the rise of a more complex, competitive society. Later on, during the Six Dynasties–Tang period, Sha Po turned into a centre in the region’s imperially controlled kiln-based salt industry. Closer to our time, in the nineteenth century the farming and fishing communities in Sha Po became important suppliers of food and fuel to urban Hong Kong. Ultimately, this ground-breaking work tells a compelling story about human beings’ ceaseless reinvention of their lives through the lens of one special archaeological site.

Archaeology / Hong Kong

Printed and bound in Hong Kong, China

Archaeological Investigations and Landscape Reconstruction Mick Atha and Kennis Yip

Mick Atha and Kennis Yip

Mick Atha teaches archaeology and landscape studies at the Chinese University of Hong Kong. Kennis Yip is an archaeological consultant. They are married with a daughter and live near Sha Po on Lamma Island.

Piecing Together Sha Po

260mm

Piecing Together Sha Po presents the first sustained analysis, framed in terms of a multi-period social landscape, of the varieties of human activity in Sha Po spanning more than 6,000 years. Synthesising decades of earlier fieldwork together with Atha and Yip’s own extensive excavations conducted in 2008–2010, the discoveries collectively enabled the authors to reconstruct the society in Sha Po in different historical periods.

Archaeological Investigations and Landscape Reconstruction

Hong Kong boasts a number of rich archaeological sites behind sandy bays. Among these backbeaches is Sha Po on Lamma Island, a site which has long captured the attention of archaeologists. However, until now no comprehensive study of the area has ever been published.

Piecing Together Sha Po

Archaeological Investigations and Landscape Reconstruction

5 15mm

Piecing Together Sha Po

Piecing Together Sha Po

Archaeological Investigations and Landscape Reconstruction

Mick Atha and Kennis Yip

Supported by the Lord Wilson Heritage Trust

Hong Kong University Press The University of Hong Kong Pokfulam Road Hong Kong www.hkupress.org © 2016 Hong Kong University Press ISBN 978-988-8208-98-2 (Hardback) All rights reserved. No portion of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage or retrieval system, without prior permission in writing from the publisher. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.

Front cover image: Lamma Island today and artist’s impression of late Qing to early twentieth-century landscape of Sha Po. Illustration by Dina B. Knight.

10 9 8 7 6 5 4 3 2 1 Printed and bound by Paramount Printing Co., Ltd. in Hong Kong, China

This book is dedicated to the memory of Mick’s father, Reg Atha (29 September 1929–18 February 2015) and Kennis’ father Yip Hing Lun (葉慶倫) (19 September 1947–8 November 2015). Also for the scholar who first highlighted Sha Po’s archaeological potential, Father Daniel Finn, SJ, on the 130th anniversary of his birth and 80th of his untimely death, in appreciation of his all too brief but seminal contribution to Hong Kong archaeology.

Contents

List of Maps viii List Figures ix List of Plates xii List of Tables xv Preface xvi Acknowledgements xvii Part I:  Contextualising Sha Po 1.  Sha Po Tsuen: Hong Kong Archaeology in Microcosm 2.  How We Know about Ancient Sha Po 3.  Social Landscapes and Ancient Environments

3 16 29

Part II:  Sha Po’s Human Narrative 4.  Sha Po’s First People: Neolithic Fisher-Hunter-Foragers 5.  Bronze Age: Technology, Trade, and Increasing Social Complexity 6.  On the Edge of Empire: Han–Yuan Sha Po 7.  A Time of Great Change: Sha Po during the Ming–Colonial Era

39 63 86 134

Part III:  Exploring the Multi-period Social Landscape 8.  Reconstructing Sha Po’s Landscapes and Lifeways 9.  Conclusions and Reflections on Sha Po

161 177

Appendices Appendix 1: Catalogue of Selected Finds Appendix 2: Glossary

185 236

Bibliography 240 Index 257

Maps

Map 1 Yung Shue Wan Site of Archaeological Interest (YSWSAI), local and regional location maps

4

Map 2 Late sixteenth-century Chinese map showing Lamma Island as ‘Pok Liu’

6

Map 3 Map of YSWSAI study area showing plateau, backbeach, and former lagoon

9

Map 4 Geological map of Yung Shue Wan

11

Map 5 General plan of previous archaeological investigations at Sha Po

17

Map 6 1968 1:1200 scale map of Yung Shue Wan

33

Map 7 Neolithic–Bronze Age physical landscape

45

Map 8 Six Dynasties–Song physical landscape

95

Map 9 1905 map of Yung Shue Wan together with map highlighting features of late Qing to early twentieth-century landscape

137

Map 10 Composite map showing changing coastal landform through time

162

Map 11 Middle to Late Neolithic social landscape

166

Map 12 Bronze Age social landscape

167

Map 13 Six Dynasties–Tang social landscape

171

Map 14 Detail of 1905 map centred on Sha Po study area

174

Figures

Figure 1 1972 hard geometric pot

19

Figure 2 Middle Neolithic stone tools

47

Figure 3 Hafted adze reconstruction

48

Figure 4 Later Neolithic artefacts relating to fishing, hunting, foraging, and food processing

50

Figure 5 Later Neolithic domestic artefacts

51

Figure 6 Later Neolithic stone adzes for wood cutting and shaping

53

Figure 7 Later Neolithic tool, weapon, and ornament manufacture

55

Figure 8 Later Neolithic spindle whorls

55

Figure 9 Father Finn’s animal figurine body found in the 1930s and head found in 1995

69

Figure 10 Composite plan of Bronze Age post-holes found on the Sha Po plateau

72

Figure 11 Bronze Age fishing and hunting gear

74

Figure 12 Bronze Age pottery

76

Figure 13 Bronze Age stone artefacts

77

Figure 14 Plan of Grave 1

78

Figure 15 Speculative reconstruction of bronze casting at Sha Po (based on objects and residues)

80

Figure 16 Han pottery

88

Figure 17 Generic kiln design

93

Figure 18 Sketch of Jin skeleton in Grave 2 based on site photo

101

Figure 19 Lattice brick and stamped wine jar sherd

110

Figure 20 Plan of K6

111

Figure 21 Six Dynasties finds

113

Figure 22 Tang pottery

114

Figure 23 Plan of working floor K7

116

Figure 24 Plan and south-east-facing section K2

118

Figure 25 Plan and south-east-facing section K3

120

Figure 26 Plan of K4

122

Figure 27 Song–Yuan pottery

125

x   Figures

Figure 28 Ming pottery

140

Figure 29 Late Qing rubbish dumping and granite revetment

144

Figure 30 Qing to early twentieth-century ceramics

146

Figure 31 Rubbings of Chinese characters on Qing ceramics

146

Figure 32 60 m wide east–west transect across backbeach

164

Figure 33 Artist’s impression of Bronze Age landscape

169

Figure 34 Artist’s impression of Six Dynasties–Tang landscape

172

Figure 35 Artist’s impression of late Qing to early twentieth-century landscape

175

Figure 36 Rubbing of ‘tomb’ brick with impressed lattice decoration

189

Figure 37 Rubbing of guan ‘tomb’ brick

192

Figure 38 Iron hoe (cha) ‘A’

195

Figure 39 Iron hoe (cha) ‘B’

195

Figure 40 Silver hairpin from Jin dynasty Grave 2

197

Figure 41 Rubbing of Northern Song Mingdao Yuanbao coin

199

Figure 42 Later Neolithic coarse corded cooking pot

200

Figure 43 Later Neolithic pot-stand

201

Figure 44 Possible Early Bronze Age grave goods: (a) globular jar and (b) bowl

202

Figure 45 Possible Bronze Age grave goods: (a) hard pottery jar and (b) stem cup

205

Figure 46 Bronze Age hard pottery lid

206

Figure 47 Han hard pottery jar

207

Figure 48 Six Dynasties four-lugged storage jar

208

Figure 49 Southern Dynasties green glazed bowl from Grave 3

209

Figure 50 Sui dynasty green glazed lid

210

Figure 51 Southern Dynasties green glazed lotus bowl

211

Figure 52 Eastern Jin green glazed bowl

212

Figure 53 Sui to early Tang cup

213

Figure 54 Sui dynasty green glazed stem cup with flower motifs

214

Figure 55 Tang dynasty green glazed ink palette

215

Figure 56 Tang dynasty green glazed bowl

216

Figure 57 Northern Song wine jar sherd with stamp

217

Figure 58 Qing blue-and-white spouted wine jar

218

Figure 59 Later Neolithic rotary drill or polisher

219

Figure 60 Bronze Age bivalve mould for fish-hooks

220

Figure 61 Later Neolithic adze

221

Figure 62 Bronze Age grave goods from Grave 1: four stone knife rough-outs

222

Figures   xi

Figure 63 Later Neolithic pebble tool

224

Figure 64 Bronze Age polishing stone or mortar

226

Figure 65 Bronze Age polishing stone

227

Figure 66 Later Neolithic–Bronze Age ring ornaments and core

229

Figure 67 Later Neolithic adze rough-out

230

Figure 68 Later Neolithic–Early Bronze Age stepped adze

231

Figure 69 Soapstone ‘net-weights’ (pre-Tang, possibly even Bronze Age)

234

Figure 70 Qing ink-stone with melon motif and inscription on reverse

235

Plates

Plate 1

Yung Shue Wan seafront, Lamma Island

3

Plate 2

Sha Po’s modern semi-urbanised setting

5

Plate 3

1960s’ overview of Yung Shue Wan and Sha Po Old Village with terraced plateau in background

7

Plate 4 1960s’ view of Yung Shue Wan fishermen’s pier and waterfront

7

Plate 5

Father Finn working at Tai Wan in the 1930s

18

Plate 6

Sha Po Old Village: photo montage of 1972 HKAS and Museum of History excavation

20

Plate 7

Photo montage of the 2000–2001 AMO excavation beside Yung Shue Wan Back Street

22

Plate 8

Photo montage of the 2002 excavations around Sha Po Old Village and Yung Shue Wan Back Street

23

Plate 9

Photo montage of the 2008–10 excavations around Sha Po Old Village and Yung Shue Wan Back Street

25

Plate 10 Typical New Territories rice-farming village setting

31

Plate 11 Later Neolithic corded cooking pot

52

Plate 12 Bronze Age pottery animal figurine from the AMO 2004 excavation

70

Plate 13 Artefactual evidence for a white quartz earring workshop on the Sha Po plateau

73

Plate 14 Bronze Age axe and fish-hook moulds

79

Plate 15 Han corded hard pottery cooking pot

88

Plate 16 Han socketed iron axe

89

Plate 17 Iron hoe (cha) 89 Plate 18 Early historical pig cranium, dugong rib, and photos of dugong

96

Plate 19 In situ green turtle carapace in early historical midden deposits

97

Plate 20 Silver hairpin from Grave 2

102

Plate 21 Southern Dynasties green glazed bowl from Grave 3

102

Plate 22 Kiln furniture

104

Plate 23 Guan ‘tomb’ brick

107

Plates   xiii

Plate 24 Kiln K1

108

Plate 25 Northern Song coin and Tang ink palette from K1 backfill

109

Plate 26 Kiln K6 showing radial bars

112

Plate 27 Flat slabs of lime with linear bamboo-wood impressions from K6 backfill

115

Plate 28 Working floor K7

116

Plate 29 Kiln K2

118

Plate 30 Kiln K3

121

Plate 31 Kiln K4

122

Plate 32 Kiln K5

124

Plate 33 Ming pottery

141

Plate 34 Structures S1and S2

142

Plate 35 Old and new houses at Sha Po Old Village

143

Plate 36 Village children eating lunch in Sha Tin

145

Plate 37 Typical late Qing to early twentieth-century ceramics

147

Plate 38 Qing blue-and-white porcelain

148

Plate 39 Qing teapot made at Yixing kilns

149

Plate 40 Ink-stone, penholder, and ink bottle

150

Plate 41 Opium paraphernalia and glass medicine bottles

152

Plate 42 Modern stilt-houses at O Tsai, Yung Shue Wan

153

Plate 43 Ceramic net-weights

154

Plate 44 Carved and perforated fish vertebrae beads

188

Plate 45 Lattice ‘tomb’ brick

190

Plate 46 Guan ‘tomb’ brick

192

Plate 47 Socketed iron axe-head

193

Plate 48 Iron hoes (cha) 196 Plate 49 Jin silver hairpin and finger-rings from Grave 2

198

Plate 50 Northern Song coins

199

Plate 51 Later Neolithic coarse corded cooking pot

200

Plate 52 Later Neolithic pot-stand

201

Plate 53 Possible Early Bronze Age grave goods

203

Plate 54 Possible Bronze Age grave goods

205

Plate 55 Bronze Age hard pottery lid

206

Plate 56 Han hard pottery jar

207

Plate 57 Three Kingdoms–Southern Dynasties four-lugged storage jar

208

Plate 58 Southern Dynasties green glazed bowl from Grave 3

209

xiv   Plates

Plate 59 Sui dynasty green glazed lid

210

Plate 60 Southern Dynasties green glazed lotus bowl

211

Plate 61 Eastern Jin green glazed bowl

212

Plate 62 Sui to early Tang cup

213

Plate 63 Sui dynasty green glazed stem cup with flower decorations

214

Plate 64 Tang dynasty green glazed ink palette

215

Plate 65 Tang dynasty green glazed bowl

216

Plate 66 Northern Song wine jar sherd with stamp

217

Plate 67 Qing blue-and-white spouted wine jar

218

Plate 68 Later Neolithic stone rotary drill or polisher

219

Plate 69 Bronze Age bivalve mould for fish-hooks

220

Plate 70 Later Neolithic adze

221

Plate 71 Bronze Age grave goods from Grave 1: stone knife rough-outs

223

Plate 72 Later Neolithic pebble tool

224

Plate 73 Later Neolithic whetstone

225

Plate 74 Bronze Age polishing stone or mortar

226

Plate 75 Bronze Age polishing stone

227

Plate 76 Later Neolithic–Bronze Age stone rings and core

229

Plate 77 Later Neolithic adze rough-out

230

Plate 78 Later Neolithic–Early Bronze Age polished stepped adze

231

Plate 79 Bronze Age stone sceptre (zhang) 232 Plate 80 Soapstone ‘net-weights’ (pre-Tang, possibly even Bronze Age)

234

Plate 81 Qing ink-stone with melon motifs and inscription on reverse

235

Tables

Table 1 Summary of kiln characteristics

105

Table 2 Kiln structure and kiln debris thermoluminescence (TL) testing results

106

Preface

This book and the synthetic research it presents was the work of two archaeologists who, both separately and together, excavated and monitored various of the more recent parts of Sha Po’s multi-period archaeological ‘mosaic’. It was as a result of our practical involvement in Sha Po’s journey of archaeological discovery that we came to recognise just how diverse, interesting, and important its human past really was. We also realised that if we were to have any hope of doing justice to the rich—but also huge and highly variable—archaeological archive, we would need to invest a significant amount of our time and find some serious funding to support the research and eventual publication of the results. We were therefore incredibly fortunate that the Lord Wilson Heritage Trust were able to generously support the project and wholeheartedly ‘bought into’ our overriding goal of sharing our findings with the widest possible scholarly and general readership. That sounds—and is—a wonderful ideal, but its practical realisation has been a far bigger challenge than we at first, naively, imagined it would be. Nevertheless, we sincerely hope that learned scholars and inquisitive students alike will find archaeological depth and detail more than sufficient for their needs. While for the non-specialists, in particular those with a more general interest in history and archaeology, we hope our ‘story’ is also an informative, interesting and, above all, entertaining read that helps ‘bring to life’ the succession of communities that inhabited ancient Sha Po and, by extension, the entire coastal region of Hong Kong, the Pearl River Delta, and Lingnan.

Acknowledgements

Many people and organisations have helped make this book possible. Firstly, we extend our sincere gratitude to the Lord Wilson Heritage Trust, which provided the significant funding necessary to support the research for and production of this substantial work of synthesis. We also thank our editors at Hong Kong University Press (HKUP), initially Chris Munn and then later Eric Mok, for keeping faith with the project even when our busy lives at times limited our capacity to drive the thing forward. Such a richly illustrated book required the creation of many new figures and maps, the taking of many new photographs, as well as the securing of permissions to use those images we could not source or make entirely by ourselves. The cartographic data used in Maps 1 and 3–14 inclusive are reproduced with the kind permission of the Hong Kong SAR Director of Lands under Licence No. 59/2015). Map 2 is reproduced courtesy of the University of Hong Kong Libraries and the help of the Special Collections team is kindly acknowledged. The three reconstruction drawings in Figures 33, 34, and 35 were produced by Lamma-based artist Dina B. Knight and we are extremely grateful that she accepted the challenge of bringing the ancient landscape to life. The site plans combined in Figure 10 were provided by the Antiquities and Monuments Office (AMO) of the Hong Kong SAR government, which owns the copyright, and are reproduced with their permission. The traditional Chinese rubbings reproduced in Figures 36 and 37 were done with the expert guidance of Mr. Andy Yiu at the AMO Repository and his kind help is gratefully acknowledged. The photographs used in Plates 20, 21, 23, 25, 45, 46, 48–78, 80, and 81 were taken by Wan Yiu-ming and then further processed by Kathy Chan and we gratefully acknowledge their help and support on this and other projects. Plates 3 and 4 are two of a series of wonderful black-and-white photographs that appear framed on the walls of many of the restaurants lining Yung Shue Wan Main Street. They are believed to have been taken in the 1960s by Mr. Lee Kwan Wing (李坤榮) and are reproduced with our gratitude. The wonderful image of Father Daniel Finn in Plate 5 was kindly provided by the Irish Jesuit Archives in Dublin, Ireland. The Hong Kong Archaeological Society (HKAS) kindly agreed to our reproduction of images from the 1972 excavation shown in Plate 6. Plates 7, 12, 13, 14, 15, 16, 24, 47, 54, 69, and 79 are reproduced with the kind permission of the AMO, which is the copyright owner. The photomontages of 2003 and 2008–10 excavations in Plates 8 and 9 were prepared from archive photographs kindly provided by Archaeological Assessments Limited (AAL), who also allowed us to reproduce many site and artefact photographs as well. Plate 10 is one of the many high-quality black-andwhite photographs of Hong Kong taken by Hedda Morrison in 1946–47, the originals of which are held in the Harvard-Yenching Library of the Harvard College Library, Harvard University, and permission to reproduce the photograph was kindly granted by the librarian James K. M. Cheng. The photographs of Red Sea dugongs in Plate 18 were taken by Amina Cesario and Agnese Mancini and forwarded to us by their co-researcher Dave Baker of Hong Kong University’s School of Biological Sciences. The picture of

xviii   Acknowledgements

Sha Tin village kids in Plate 36 is used with the kind permission of FormAsia Books, Hong Kong, who own the copyright. The research for this volume involved many thousands of hours trawling through a diverse collection of paper, digital, and material archives held and made available to us by AAL, the HKAS, and AMO. We would like in particular to acknowledge the help of the following people: from AAL Pamela Rogers, Julie van den Bergh, and Ellen Cameron; from HKAS the Chairman Steven Ng and Honorary Secretary Patrick Lai; and from the AMO in alphabetical order: Doris Chan, Nelson Chow, Hong Fung, Raymond Lee, Rachel Leung, Coral So, Kevin Sun, Michael Tang, Tsang Chi-hung, Carmen Wong, and Ada Yau, as well as those in their teams who helped us in the background. We are also grateful to Eddy Leung and Karen Fong of the Central Conservation Section of the LCSD for providing us with a laboratory report on the three soapstone ‘net-weights’ that appear in our Catalogue of Selected Finds. We are also grateful to AAL for making desk space and facilities available to us during our research on the Sha Po archives. Various drafts of chapters were read and commented upon at different times by learned and interested friends and colleagues and we hereby offer heartfelt thanks to Geoff Arnott, Patrick Hase, Stuart Heaver, Phil Stride, and Kevin Sun for their constructive criticism and helpful suggestions which, collectively, have much improved the text. We thank Stephen Davies for advice on prehistoric and historical boats. We also wish to thank Peter Lam (former director of CUHK Art Museum) and Liu Chengji (deputy director of Guangdong Provincial Institute of Cultural Relics and Archaeology) for their valuable comments on some of the objects in the Catalogue of Selected Finds. The completed draft manuscript was also sent out for peer review by Hong Kong University Press to Chau Hing-wah of the Hong Kong Heritage Museum and Francis Allard of the Department of Anthropology, Indiana University of Pennsylvania, USA. We very much appreciated their constructive and positive reviews, which led to further improvement and tightening of the manuscript. Any remaining errors of fact or interpretation are entirely our own responsibility.

Part I Contextualising Sha Po

1 Sha Po Tsuen: Hong Kong Archaeology in Microcosm

Introduction For many people living in or visiting Hong Kong, archaeology is not something they readily associate with the affluent commercial metropolis. Profound surprise therefore often greets the news that Hong Kong’s riches also extend to archaeology. Despite its more rural setting, the same is also often true of the territory’s second largest outlying island, the tourist ‘hotspot’ of Lamma Island. On the north-west coast of Lamma Island lies the popular weekend and holiday destination of Yung Shue Wan1 (榕樹灣, Banyan Tree Bay), which is well-known for, among other things, its green environment, ethnically diverse Chinese-foreign community, eclectic range of shops, bars, and restaurants, and its contrastingly laid-back atmosphere when compared with the frenetic pace of downtown Hong Kong. It is also often the starting point for scenic walks to the island’s other main settlement at Sok Kwu Wan (索罟灣)—itself famous for its seafood—via the bathing beach of Hung Shing Yeh2 (洪聖爺) and, with short diversions, other beaches near the villages of Tai Wan (大灣) and Lo So Shing (蘆鬚城). However, few residents and fewer still among the many thousands of visitors passing through Yung Shue Wan each year will know that—like the three beaches to the south—it also gives its name to an Antiquities and

Plate 1:  Yung Shue Wan seafront, Lamma Island

Sham Wan

Lo So Shing

Hung Shing Ye

Yung Shue Wan

Lamma Island

Hong Kong Island

Kowloon

Map 1: Yung Shue Wan Site of Archaeological Interest (YSWSAI) with smaller maps showing study area location on Lamma Island and in Hong Kong region. Source: Survey and Mapping Office (SMO). 1:1000 Scale Topographic Map, Sheet Nos. 14-NE-10D and 14-NE-15B. Hong Kong: Lands Department, 2015. Reproduced with permission of the Director of Lands. © The Government of the Hong Kong SAR. Licence No. 59/2015.

Tai Wan

Lantau Island

New Territories

Shenzhen, PRC



Sha Po Tsuen: Hong Kong Archaeology in Microcosm   5

Monuments Office (AMO) Site of Archaeological Interest.3 At the heart of that site lies the village of the book’s title, Sha Po Tsuen (沙埔村)—literally, ‘sand spit village’—which is very appropriately named given that the settlement’s Old Village was built on top of an ancient storm beach or ‘backbeach’ full of archaeological remains.4 As early as the 1930s, Hong Kong’s archaeological pioneers recognised that the backbeaches were some of the richest sites in the region and had been a magnet for maritime-focused prehistoric peoples.5 A myriad of post-war investigations—some research-driven but many more recent ones in the form of predevelopment ‘impact assessment’ excavations—eventually confirmed that many backbeach sites, including those on Lamma, had in fact been used from prehistory right through into later historical periods, albeit discontinuously. The Hong Kong Archaeological Society was particularly active during the 1970s to early 1990s and carried out a number of backbeach excavations on Lamma, most famously at Sham Wan under the direction of Solomon Bard and later William Meacham, but also at Lo So Shing, and, of course, Sha Po Tsuen.6 In 1991, a Chinese University of Hong Kong team also revisited the important backbeach site at Tai Wan, which had been first highlighted by Father Daniel Finn in the 1930s.7 Sha Po is actually one of a small number of rich backbeach sites within areas of expanding village development,8 which have consequently seen a relatively high frequency of often small-scale pre-development investigations but with only limited publication and a general absence of synthetic analysis and interpretation. Their archaeological stories thus remain largely untold. It was while conducting archaeological investigations at Sha Po between 2008 and 2010 that the writers came to appreciate the area’s true research potential and archaeological significance.9 A review of the results of almost forty trenches excavated in Sha Po since the early 1970s confirmed that the Old Village backbeach had deeply stratified archaeological deposits reflecting some 6,500 years of human history spanning the Middle Neolithic to

Plate 2:  Sha Po’s modern semi-urbanised setting: house arrowed with black, white, and green roof canopy marks approximate centre of backbeach, while the plateau site is marked by the low tree-covered ridge extending from the upper middle of the photo to the right-hand side of view. Yung Shue Wan’s coastal market and bay are to the left, while Hong Kong Island is in the right background.

6   Piecing Together Sha Po

World War II, but with regionally important remains of Bronze Age and Six Dynasties–Tang date. But what makes the area doubly special is that overlooking the backbeach site from its low plateau to the north is a second, fascinatingly different, Bronze Age site at Sha Po New Village, which boasts traces of settlement and a craft workshop in which fine polished quartz earrings were made.10 Moreover, the archaeological evidence suggested that at different stages in Sha Po’s human story local communities had made very different uses of the backbeach, plateau, and surrounding landscape. This raised the possibility of reconstructing the changing lifeways and contrasting social landscapes of Sha Po’s communities over their six millennia history. We now continue by introducing the study region’s recent socio-economic history, geographical context, and landscape character.

Sha Po’s Local Context Yung Shue Wan’s recent socio-economic history The earliest historical mention of Lamma Island occurs in a gazetteer of 1464,11 where it is named ‘Pak Lo Mountain’ (泊潦山) and on a late sixteenth-century map of the Guangdong coast where it is marked as ‘Pok Liu’ (博寮).12 The name ‘Lamma Island’ is itself an Anglicisation of the modern Cantonese name (南丫島) (literally, ‘southern forked island’) which, like Sha Po, is a good example of the Chinese’s love of place names reflecting the local physical environment.

Map 2:  Late sixteenth-century map of the Guangdong coast: (1) Pok Liu (Lamma Island), (2) Cheung Chau, (3) Tai Ho Shan (Lantau Island), (4) Tuen Mun, (5) Kap Shui Mun, (6) Tsim Sha Tsui, (7) Lei Yue Mun, (8) Wong Nai Chung, (9) Po Toi, and (10) Hong Kong. Source: Guo Fei, ‘Coastal Map of Kwang Tung’, in Yue Da Ji. Place and publisher unknown, c. 1598. Reproduced courtesy of the University of Hong Kong Libraries.

Plate 3:  1960s’ overview of Yung Shue Wan and Sha Po Old Village with terraced plateau in background. Photo by Lee Kwan Wing (李坤榮).

Plate 4:  1960s’ view of Yung Shue Wan fishermen’s pier and waterfront. Photo by Lee Kwan Wing (李坤榮).

8   Piecing Together Sha Po

The first records mentioning settlement and landholding in the hinterlands of Yung Shue Wan date back to the Qing dynasty, but in turn refer to an imperial land grant dating back to the Ming, when Yiu Cho of Nantou (南頭姚祖)—just across Deep Bay to the north-west of modern Hong Kong—took possession of the island. His descendents formed the Yiu Yi Yin Tong ancestral trust through which they acted as landlords to Lamma’s rice-farming and fishing communities.13 Among Lamma’s main resident clans, the earliest were the Chaus of Wang Long (橫塱)—three brothers originally of Hong Kong Wai (香港圍)—closely followed by the Chans of Tai Wan, both of whom settled and began farming in the earlier eighteenth century as tenants of the Yius.14 Their fuller story is presented in Chapter 7’s discussion of the later historical archaeology of the study region, but here it will suffice to say that Sha Po Old Village was probably a late eighteenth- to early nineteenth-century ‘offshoot’ of earlier Chau clan settlements originating at Wang Long. As we will see in later chapters, fishing has been an important economic activity centred on Yung Shue Wan’s sheltered sandy bay from the present day back into prehistory. The more recent fishing community’s Tin Hau temple was most likely erected in the later eighteenth or early nineteenth century and provided a focus for later developments.15 With the temple in place the bay’s pivotal position between the fishing and farming communities encouraged the establishment of a small coastal market at Yung Shue Wan and the settlement’s present form evolved from those early beginnings. Today, the old and new settlements of Sha Po have been largely subsumed within the built-up village landscape of the larger ‘commuter settlement’ of Yung Shue Wan, which is now just a 25-minute fast ferry ride from the metropolitan bustle of Hong Kong’s central business district. Forty years ago, though, the Yung Shue Wan area was still essentially rural in character, but had evolved into a ‘satellite market gardening suburb’ following the decline of the traditional rice-farming economy.16 Much of that agricultural landscape still survived until as recently as twenty-five years ago, when the patches of vegetable fields still under cultivation today at Yung Shue Long (榕樹塱) and behind Tai Wan To (大灣肚)17 were mirrored in extensive plots stretching from Sha Po all the way back to Wang Long (see further discussion in Chapter 3).

A socio-topographically defined study area Our study area is centred upon the two Sha Po sites on the backbeach and plateau but it also embraces their wider landscape setting, which includes the Yung Shue Long valley to the north and the Wang Long valley to the south and east (see Map 1). That broader study area corresponds with the Yung Shue Wan Site of Archaeological Interest, which like many others was focused on known archaeological sites but was delimited by the sea and the topographic interface between lowlands and surrounding mountains. In truth, though, past human activities transcended those interfaces, for example through the widespread use of hillsides for cultivation terraces, fuel gathering, and burial grounds. Moreover, while local topography clearly influenced past patterns of human inhabitation and land use, the hills dividing separate ‘watershed communities’ were—by later historical times at least—also criss-crossed by trackways that interconnected such neighbouring communities and gave them access to local and more distant markets. By adopting a landscape approach, we acknowledge that throughout Sha Po’s human history local communities inhabited, engaged with, and experienced their local domain as part of a larger social landscape encompassing Lamma Island and its inshore waters, while at different periods in history that might also extend to the wider Hong Kong–Pearl River Delta coastal zone, and—perhaps through travel, trade, and exchange—out into the rest of Lingnan and beyond. These notions of wider connections and bigger social landscapes are then used to guide and frame interpretive narratives in later chapters.



Sha Po Tsuen: Hong Kong Archaeology in Microcosm   9

Map 3:  Map of YSWSAI study area showing: (1) backbeach with Sha Po Old Village highlighted, (2) plateau, (3) lagoon (former), (4) Yung Shue Wan (YSW) market and fishermen’s pier, (5) Wang Long, (6) Ko Long villages (both in Wang Long Valley), (7) Tin Hau Temple, (8) YSW Main Street, (9) YSW Back Street, and (10) Yung Shue Long Valley. Source: SMO. 1:1000 Scale Topographic Map, Sheet Nos. 14-NE-10D and 14-NE-15B. Hong Kong: Lands Department, 2015. Reproduced with permission of the Director of Lands. © The Government of the Hong Kong SAR. Licence No. 59/2015.

The physical environment While backbeach sites are prominent topographic features on less developed coastlines around Hong Kong, the example at Sha Po Old Village is far less obvious, hidden as it is among village houses and narrow lanes behind Yung Shue Wan Back Street. Although masked by modern development, older maps, and physical traces on the ground allow the backbeach to be traced as a roughly crescent-shaped area of raised ground mirroring the curve of the pre-reclamation shoreline (Map 3 brown shading). It measures approximately 150 m north-east–south-west by 50 m north-west–south-east and has a maximum height of around 5.1 mPD.18 While reclamation works and development have obscured the natural topography of the Yung Shue Wan coastline, the seaward face of the backbeach is still visible in places. This is especially true where it slopes down towards the sea at the southern end of Back Street. Here at a distance of some 50 m seaward

10   Piecing Together Sha Po

or west of the highest central plateau of the backbeach, the ground is between 1.2 to 1.5 m lower, while to the rear of the backbeach there is a flat low-lying (3.7  mPD) area latterly used as vegetable fields, paddy fields before that, and earlier still it is thought to have contained a shallow—but perhaps periodically quite extensive—freshwater lagoon (blue shading on Map 3).19 An even lower area (2.4 mPD)— reportedly once flooded at high tides—formerly existed between the southern backbeach and the sand spit upon which the Tin Hau Temple and North Lamma Clinic now sit, but it is now filled in and developed. To the north of the backbeach the top (25 mPD) and gently sloping southern flank of the Sha Po New Village plateau—known locally as Yung Shue Ling (榕樹嶺) (Map 3 green shading)—have been much altered by housing development, while the foot of its southern side and western end have both been cut back to accommodate development. Before such changes, it seems likely that the plateau’s western end would have effectively divided Yung Shue Wan into a larger southern bay, with the backbeach, lagoon, and valley behind rising gently towards Wang Long, and a smaller northern bay leading into the Yung Shue Long valley. Beyond the southern end of the backbeach the land rises steeply towards the heavily wooded crest of Tai Shan (大山) at an elevation of 56 mPD. Those surface landforms are reflected in the underlying geology (see Map 4), such that in the main valley bottoms there are extensive deposits of stream-transported alluvium (Qa and Qat) fronted by an arc of sandy beach deposits (Qb) marking the modern shoreline, behind which to the south is the Sha Po backbeach. The valley sides are bordered by bands of sandy clay and boulders—so-called debris flow deposits (Qd)—eroded and washed down from the surrounding low granite hills (gm), in particular during the summer rainy season. Interestingly, the Sha Po backbeach is one of a group known to archaeologists that are not yet officially recorded on geological maps by the Geotechnical Engineering Office.20 In the past, carefully managed fung shui woodlands21 were grown on slopes behind rice-farming villages as a more eco-friendly—and attractive—way of reducing landslide risk than the concrete used so widely today. While housing development has had a significant impact on the appearance of the Yung Shue Wan area since the 1970s, the same period has also seen an equally dramatic reforestation through plantations and post-agricultural scrub and woodland regeneration on hillsides that were stripped for firewood during the dark days of World War II and its immediate aftermath.22 Crucially, the study area is well watered by a number of small but perennial streams that were essential for human survival in all periods. The local drainage pattern also made the Yung Shue Long and Wang Long valleys ideal for wet rice cultivation.

A Regionally Important Site Although a number of Hong Kong multi-period backbeach sites have seen small-scale publication as summary reports—mostly in the Journal of the Hong Kong Archaeological Society (JHKAS)—just two such sites have so far been more substantially published: Sham Wan (深灣) and Sha Ha (沙下).23 However, as explained below, neither is really comparable with Sha Po, which has a quite different landscape context, qualitatively different material remains, and has seen an entirely different history of archaeological work. With respect to Sha Po’s archaeological highlights, there is the clearest evidence in the region for Bronze Age metallurgy—the melting and casting of bronze axes—occurring on the southern backbeach. Contemporary with that on the plateau we have traces of post-built structures—perhaps stilt-houses— seemingly used by craftspeople skilled in the manufacture of fine polished quartz rings. A programme of thermoluminescence testing commissioned by this project has shown that Sha Po’s kiln-based salt-lime industry began in the Six Dynasties period but continued at least until the end of the Tang.24 The dates

Map 4:  Geological map of Yung Shue Wan. Source: Geotechnical Engineering Office. HGM20 Series 1:20,000 Scale Map of, Solid and Superficial Geology Sheet 14 Cheung Chau (first edition). Hong Kong: Geotechnical Control Office, 1995. Reproduced with permission of the Director of Lands. © The Government of the Hong Kong SAR. Licence No. 59/2015.

12   Piecing Together Sha Po

seem to show a number of kilns in contemporary operation, while associated finds—including a locally unique moulded guan (官) brick—seemingly suggest imperial management of the industry. The rare discovery of two Six Dynasties burials with surviving skeletal remains—both apparently adult females—were found in different parts of the backbeach but both were buried with their feet pointing towards the same distant peaks on Lantau Island. Lastly, Sha Po’s Qing dynasty remains served to highlight the interpretive richness of more recent periods and the need for an expansion of historical archaeological research in Hong Kong. Fundamentally, then, the nature of the evidence clearly encouraged a presentation framed in terms of a multi-period social landscape,25 which is the first time this has been attempted in Hong Kong. Moreover, in methodological terms Sha Po’s history of intense but piecemeal development, and the patchwork of archaeological investigations of different scales, types, and methodologies that it produced, is fundamentally different from the aforementioned published backbeach sites at Sha Ha and Sham Wan, which had seen little or no prior development or archaeological work before they were investigated as relatively well resourced and, in Sha Ha’s case, large-area excavations. In fact, if we include Father Finn’s pre-war discoveries on the plateau,26 archaeological work at Sha Po has occurred intermittently over nine decades, but with a particular intensity since the early 1970s. As a result, the timing, location, and even the size and shape of many of the areas investigated reflected the nature of each particular impact rather than any form of overarching research design.27 Unfortunately, despite the discovery of exciting findings at Sha Po, the commercial context of many excavations meant that the time and money necessary for in-depth analysis and publication were usually lacking. Thus, although Sha Po was evidently a very important site, its prevailing context of archaeological investigation had prevented a wider recognition of the fact. Abroad there has recently been a greater focus on using the results of development-funded excavations—unpublished ‘grey literature’ reports—in archaeological research.28 Sha Po provided an excellent opportunity to demonstrate that, despite the imperfections of its rather piecemeal investigative approach, there was also a potentially rich research dividend to be derived from development-funded archaeological work in Hong Kong. This is especially the case in areas such as Sha Po where rich archaeological sites coincide with areas undergoing expansive village development. Overall, then, there was a compelling argument in favour of researching and writing a book that drew all the strands of evidence together in one overarching reanalysis and interpretation of the Yung Shue Wan area in general and the Sha Po sites in particular. If many passing through Sha Po are unaware of the history beneath their feet, they can easily be forgiven not only on grounds of limited publication, but also because the ancient landscape—whose most recent features remained in use and visible until at least the 1970s or 1980s—has in three or four decades been so dramatically transformed as to render it unintelligible to the eyes of all except older locals and inquisitive archaeologists. This book is our attempt to shed light on Sha Po’s hidden past and in so doing reanimate the lifeways and landscapes of its past inhabitants. The next section describes how that process unfolds over the course of the following eight chapters.

Piecing Together Sha Po We begin in Chapter 2 by presenting the twists and turns of Sha Po’s ‘site biography’, or how over a period of eight decades our present understanding of its archaeological treasures gradually came to light. It is a story that reflects Hong Kong archaeology as a whole, in that there were the discoveries made by pre-war pioneers,29 significant contributions by the Hong Kong Archaeological Society and the AMO and, most recently, a series of important finds made by archaeologists working in the commercial sector. During that



Sha Po Tsuen: Hong Kong Archaeology in Microcosm   13

story of discovery, fieldwork progressed from poorly recorded ‘antiquarian collecting’, through more formalised research digging, into the present era of AMO-licensed excavations working to agreed research designs. Having set out ‘how we know what we know’, we then move on in Chapter 3 to establish a fuller environmental context for the study region introduced above. We begin by exploring the making of Hong Kong’s present cultural landscape and then work back through its earlier forms, firstly into the age of rice farming where the long-term sustainable management of that particular socio-economic lifeway created highly distinctive cultural landscapes stretching back from the Qing dynasty to as early as the Northern Song in some areas. Then from the Tang dynasty moving backwards, we enter an era where, on the face of it, human impacts beyond the intensively industrialised backbeach areas seem to have been relatively slight. That said, the coastal focus seemingly exhibited by early historical populations was even more intensively expressed in prehistory when, once sea-levels had stabilised at more or less their present position, the resource-rich landscape of the New Territories and Pearl River estuary coastline and offshore archipelagos then took shape. In Chapter 4 we begin our chronological journey through Sha Po’s human story in the earlier Middle Neolithic, providing the necessary archaeological background and context by referencing discoveries made across the wider Hong Kong–Pearl River Delta region (a format also employed in Chapters 5–7 inclusive). The backbeach evidences a major break in activity until the Later Neolithic and we suggest that the patterning of activities is suggestive of a relatively low intensity usage of the site by a small-scale community of fisher-hunter-foragers. Artefactual evidence is also used to suggest that by the end of the Neolithic the Sha Po community, like others across the region, was exhibiting features attributable to a rise in social complexity, which probably reflected both internal change and the intensification of contacts with agropastoralist groups to the north of the Pearl River Delta. Chapter 5 examines one of Sha Po’s most fascinating and important periods of cultural development, the Bronze Age, a period during which the local community was making wider and more specialised use of the coastal landscape. On the plateau there was some form of stilt-house settlement associated with the specialised manufacture of fine quartz rings, while on the backbeach we have the region’s best evidence for non-ferrous metallurgy in the form of in situ bronze casting. The evidence for craft specialisation tells us that society was undergoing change and could perhaps support the work of artisans through some form of surplus production of food. Moreover, access to more advanced technology and exotic materials are both indications of a widening of external contacts, trade, and exchange, while a heightened interest in personal ornamentation and display points towards greater competition and the emergence of social hierarchies. In Chapter 6 we explore the contrasting evidence for activity spanning the Han, Six Dynasties–Tang, and Song–Yuan periods at Sha Po. Our study of the structural remains and artefactual evidence associated with Sha Po’s Six Dynasties–Tang kiln-based coastal industry is supported by the results of a programme of thermoluminescence dating of kiln remains. Collectively, the evidence suggests that Sha Po was a planned and imperially controlled kiln complex directed towards the production of salt, with lime as a process-related by-product. In a pattern typical across Hong Kong, the industry’s post-abandonment phase is associated with Northern Song and some Southern Song–Yuan ceramics. The final chronological discussion in Chapter 7 addresses the Ming and Qing dynasties, which at Sha Po could not be more different in that the former is virtually absent, whereas archaeological remains from the latter period are abundant and provide fascinating insights into the lives of local people. Moreover, those material remains can also be interpreted with reference to a particularly rich historical and anthropological resource resulting from documentary research and interviews with village elders between

14   Piecing Together Sha Po

the 1950s and 1980s. Recent historical research is a rapidly expanding field in archaeology, but sadly neglected in Hong Kong, and in this chapter we attempt to highlight its potential for the creation of more humanistic narratives and detailed interpretations than are possible in earlier periods. Chapter 8 provides an opportunity to draw all the strands of evidence together within an overarching synthetic analysis of patterns of human activity through time, which are then interpreted in terms of the development, use, and past experience of Sha Po’s multi-period cultural landscape. The shifting patterns of human activity during the 6,500-year span of our study also permit the changing backbeach landform to be modelled as it expanded westward through time. The main text is rounded off in Chapter 9 with a series of conclusions, which reflect on the value and significance of Sha Po’s archaeological resource in a local and regional context. We also offer some reflections on lessons learnt during the research that led to the production of this book and make some recommendations and predictions concerning future management of the Sha Po sites and their wider landscape. The final substantive component of the book is our Catalogue of Selected Finds, which, as the name suggests, presents photographs, drawings, and descriptions of around forty of Sha Po’s most interesting artefacts, most of which have not been seen in print before.

Conclusions Without giving too much away, we hope in this chapter to have conveyed a sense of Sha Po’s archaeological significance and revealed sufficient glimpses of its fascinating story to have whetted the appetite for the journey of exploration that lies ahead. That journey begins in earnest in Chapter 2, with our review of the process of investigation and discovery that provided the raw materials for this book. So we now move on to explore how we came to know so very much about the people and social landscapes of ancient Sha Po.

Notes 1. Place names within Hong Kong are given Cantonese transliterations in accordance with A Gazetteer of Place Names in Hong Kong, Kowloon and the New Territories, 2nd ed. (Hong Kong: Colonial Secretariat, 1969) with their Chinese characters when first used. Place names in China are presented in pinyin, again with their Chinese characters upon first appearance. 2. Strangely, the English spelling of the bathing beach ends in ‘Yeh’ while the AMO archaeological site is spelled ‘Ye’. We therefore use ‘Yeh’ when referring to the beach and ‘Ye’ for the archaeological site. 3. The Yung Shue Wan Site of Archaeological Interest actually includes a number of discrete archaeological ‘hotspots’, such as the Sha Po Old Village backbeach site and Sha Po New Village plateau site, set within a broader ‘buffer zone’. This approach is used by the AMO for most, if not all, of the 208 SAIs listed on their website in September 2014, at http://www.amo.gov.hk/form/list_archaeolog_site_eng.pdf. 4. Backbeach is now one of the two preferred terms (the other being backshore deposits) used by scholars for the raised sand bodies commonly found behind exposed sandy beaches in Hong Kong and other coastlines subject to the periodic effects of severe storms such as the summer wet season typhoons experienced locally, which can mobilise huge quantities of sand. Translation of village name is from William Meacham, ‘Sha Po Tsuen’, JHKAS XIII (1993a): 33. 5. Shellshear, Heanley, Schofield, Finn, and Chen Kung-che. 6. William Meacham, ed., Sham Wan, Lamma Island: An Archaeological Site Study, Hong Kong Archaeological Society Journal Monograph III (Hong Kong: Hong Kong Archaeological Society, 1978); William Meacham, ‘Lo So Shing’, JHKAS VII (1979): 16–26; ‘Sha Po Tsuen’, JHKAS XIII (1993): 33–54; Nigel Spry, ‘Sha Po Tsuen’, JHKAS XII (1990): 7–28. 7. Tang Chung, A Journey into Hong Kong’s Archaeological Past (Hong Kong: Regional Council, 1991). 8. Lung Kwu Tan near Tuen Mun and Tung Wan on Cheung Chau—which, strictly speaking, is actually a tombolo—are good examples of this type. 9. AAL, ‘Archaeological Rescue Excavations at Sha Po Tsuen’ (unpublished excavation report, 2011a); AAL, ‘Archaeological Watching Brief at Sha Po Tsuen’ (unpublished watching brief report, 2011b). In Hong Kong, rescue excavation involves a set-piece archaeological



Sha Po Tsuen: Hong Kong Archaeology in Microcosm   15

excavation in the pre-construction phase of development projects, whereas watching briefs require an archaeologist to monitor the engineering contractors’ groundworks during the construction phase. 10. This plateau workshop was first identified in the 1930s by Father Daniel Finn, SJ, who labelled the site ‘YSW’—for Yung Shue Wan—in his series of papers on Lamma Island’s archaeology published in the journal Hong Kong Naturalist. 11. Patrick H. Hase, ‘Some Notes on the History of Lamma, Especially Yung Shue Wan’ (unpublished paper, 2002), 5. Hase cites the Dongguang County Gazetteer of 1464 for this early reference to Lamma. 12. Map held on microform by Special Collection, Hong Kong University Library: Guo Fei, ‘Coastal Map of Kwang Tung’, in Yue Da Ji (place and publisher unknown, c. 1598). 13. Hase, ‘Yung Shue Wan’, 6. The original Chinese documents referring to the Yiu Yi Yin Tong were translated into English but sadly only the translations now survive. The name in Chinese characters therefore remains unknown. 14. Hase, ‘Yung Shue Wan’, 11–13. 15. Hase, ‘Yung Shue Wan’, 40. 16. Hase, ‘Yung Shue Wan’, 43. 17. Tai Wan To is today known in English as ‘Powerstation Beach’ due to the massive Hong Kong Electric power plant that now overlooks it. 18. Survey and Mapping Office, Explanatory Notes on Geodetic Datums in Hong Kong (Hong Kong: Lands Department, 1995). Hong Kong heights on land are expressed in metres above Principal Datum or ‘mPD’, where Principal Datum is 1.23 m below mean sea-level. 19. Previous testing revealed fine sediments suggestive of the former existence of the lagoon. Au Ka-fat, ‘Report on the Archaeological Survey on the Project of Small House Development at Yung Shue Wan Archaeological Site Lamma Island’ (unpublished report, 2001a). 20. J. C. F. Wong and R. Shaw, High-Level Coastal Deposits in Hong Kong, GEO Report No. 243 (Hong Kong: Geotechnical Engineering Office, 2009). As Wong and Shaw discuss, there was previously some speculation concerning the possibility that Hong Kong’s backbeaches may have formed during a period when ancient sea-levels were higher than today’s. But there is little archaeological evidence to support this idea and sea-levels seem to have stabilised at more or less modern levels by the middle Neolithic (around 6,000 years ago). The terms ‘raised beach’ or ‘sandbar’ are both inappropriate here, as the former relates to beaches formed by previously higher sea-levels, while the latter should be reserved for offshore sand bodies, typically formed at the mouth of estuaries. 21. Chapter 3 contains further discussion of the importance of fung shui in village placement and landscape management. 22. The deep practical and spiritual significance of fung shui woodlands for local farming communities was perhaps most clearly demonstrated by the fact that many survived World War II intact within an otherwise deforested landscape. 23. Sham Wan in South Lamma was excavated over several seasons and unearthed some very significant archaeological remains, but also made the crucial discovery that the earliest cultural horizons may lie buried beneath sterile deposits of windblown sand (William Meacham, ed., Sham Wan, Lamma Island, 1978). Sha Ha in Sai Kung is an example of the post-1997 pattern of major excavations being done as collaborations between the Antiquities and Monuments Office (AMO) and a range of institutions from the People’s Republic of China (AMO, The Ancient Culture of Hong Kong: Archaeological Discoveries in Sha Ha, Sai Kung [Hong Kong: AMO], 2005). 24. In this book we use the label ‘kiln’ for the ubiquitous fired clay industrial structures that characterise the Six Dynasties–Tang period in Hong Kong. We do, however, recognise that kilns usually have fully enclosed firing chambers, while the earlier historical examples in Hong Kong are open-topped. 25. In recent decades in European and North American archaeology, landscape has become an important conceptual framework used for the investigation, analysis, interpretation, and presentation of the human past. The idea of social landscapes has become particularly popular because the physical characteristics of landscapes and the values and meanings people attach to them are all communally or socially defined; hence the term ‘social landscape’ (Lynn Meskell and Robert W. Preucel, eds., A Companion to Social Archaeology [Oxford: Blackwell], 2004). 26. Daniel J. Finn, ‘Archaeological Finds on Lamma Island near Hong Kong, Part XIII’, Hong Kong Naturalist 7 (3–4) (1936c): 257–68. Finn was in fact one of several pre-war archaeological pioneers who visited the area including Shellshear, Schofield, and Chen Kungche (see further discussion in Chapter 2). 27. Mostly pipe trenches or Small Village House developments. 28. For example, one of the authors’ PhD research: Mick Atha, ‘Late Iron Age Regionality and Early Roman Trajectories (100 BC–AD 200): A Landscape Perspective from Eastern Yorkshire’ (PhD thesis, University of York, 2008); Mick Atha and Steve Roskams, ‘Premedieval Transitions at Wharram Percy’, in Wharram: A Study of Settlement on the Yorkshire Wolds, XIII, ed. Stuart Wrathmell (York: University of York Press, 2012), 63–82. 29. Solomon M. Bard, ‘Archaeology in Hong Kong: A Review of Achievement’, in Conference on Archaeology in Southeast Asia, ed. Yeung Chun-tong and Li Wai-ling (Hong Kong: University Museum and Art Gallery, the University of Hong Kong, 1995), 383–96. The ‘Pioneers’ was one of three groups mentioned by Bard, first head of the AMO, when discussing the development of archaeological work in Hong Kong.

2 How We Know about Ancient Sha Po

Introduction The Sha Po Old Village backbeach and New Village plateau sites both lie within the larger Yung Shue Wan Site of Archaeological Interest. The initial pre-war identification of the plateau site occurred as a result of agricultural activity and erosion, whereas more recent archaeological discoveries on the backbeach and plateau alike have resulted from investigations associated with a mixture of Small House developments, government infrastructure projects, and occasional research digs. The forty-eight areas1 investigated at Sha Po since the early 1970s have produced regionally significant discoveries but, in common with many other Hong Kong archaeological sites, most remain unpublished and unknown to the general public. Moreover, when viewed individually the full meaning and significance of such discoveries can easily be overlooked. One aim of this book, therefore, is to highlight the significant research value of collectively analysing old archives, unpublished reports, and published site summaries, especially when such findings are then presented through the lens of a multi-period social landscape. The narrative below follows the progress of fieldwork and discovery in chronological order: from the earliest finds made by pioneers of the pre-war era, to the 1960s–1990s salvage-research fieldwork of the Hong Kong Archaeological Society, right up to the most recent discoveries made by professional archaeologists within a development-funded ‘impact assessment’ context.

The Father of Lamma Archaeology The story of archaeological fieldwork in Sha Po can be traced all the way back to the early 1930s and the activities of one remarkable scholar—the eminent Classicist, linguist, geographer, and archaeologist Father Daniel Finn.2 Finn was an Irish Jesuit priest and lecturer in geography at the University of Hong Kong, and it was there that he came into contact with local archaeological pioneers Professor J. L. Shellshear of the Department of Anatomy and his associate Dr. C. M. Heanley, who together produced the first publication on Hong Kong archaeology.3 Finn was unique among the early fieldworkers in that he had studied classical archaeology first at Royal University in Ireland and then subsequently at Oxford. The others were, nevertheless, prominent scholars in their own fields and all had scientific training of some sort or another. In the first of his thirteen articles on Lamma Island archaeology, Finn records how the professor, before going on leave in 1932, ‘asked me to interest myself in the [archaeological] observations’.4 Shortly after that conversation, Finn relates finding sherds of prehistoric pottery, a bronze weapon fragment, and a stone spearhead in sand being unloaded from junks onto a quayside in Aberdeen. Enquiries revealed that the sand had come from Tai Wan on the western shore of Lamma Island, a site first discovered by Heanley and already known to be archaeologically rich due to his initial

Map 5:  Sha Po study area showing areas previously investigated by excavation (light grey) and monitored by watching brief (dark grey). The approximate extent of the backbeach is shown in brown outline. 1: 1972 Bronze Age pot find spot; 2–4: HKAS 1972 trenches ‘A’, ‘B’, and ‘C’ respectively; 5–7: HKAS 1988 trenches ‘D’, ‘E’, and ‘F’ respectively; 8–11: HKAS 1989 trenches ‘G-GT’, ‘H’, ‘X’, and ‘CC’ respectively; 12–13: HKAS 1995 trenches ‘J’ and ‘K’ respectively; 14–15: AMO 1997 trenches 1 and 2 respectively; 16: Au 2000–2001; 17: Li 2001; 18–24: AAL 2002 units (trenches) 2–7 and 9 respectively; 25: AMO 2004; 26–27: ERM 2005 trenches ‘T1’ and ‘T2’ respectively; 28: AMO 2010; 29: Peacock and Nixon (First Territory-Wide Survey); 30: Au 2001; 31–37: AAL 2008–10 trenches AA1, AA2, AA4, AA5, AA3, AA6, and AA7 respectively; 38–47: AAL 2008–10 watching brief areas A–J respectively; 48: AMO 2012. Source: SMO. 1:1000 Scale Topographic Map, Sheet Nos. 14-NE-10D and 14-NE-15B. Hong Kong: Lands Department, 2015. Reproduced with permission of the Director of Lands. © The Government of the Hong Kong SAR. Licence No. 59/2015.

18   Piecing Together Sha Po

discoveries and those he subsequently made with others.5 Although Finn’s thirteen articles published in The Hong Kong Naturalist were predominantly focused on his excavations at Tai Wan, where he identified an important Bronze Age site in the process of being destroyed by sand quarrying, he also discussed findings from two other important sites on the west coast of North Lamma: Hung Shing Ye to the south and Yung Shue Wan to the north. In Finn’s day the main Sha Po backbeach site was then unknown to archaeologists as it still lay hidden from view beneath the fields and houses of the Old Village. Only with the intensification of housing development, which occurred from the 1970s onwards, did the full significance of the site become clear. In contrast, by the 1930s a combination of widespread agricultural terracing and associated erosion on the New Village plateau had brought significant quantities of prehistoric artefacts to the surface. In addition to the remains visible on the plateau itself, examples of similar materials were also noted in erosion deposits on its slopes and at their foot.6 Finn’s ‘YSW’ site thus encompassed the plateau and surrounding slopes to the south, west, and north, but he also mentions finding similar materials on the northern flank of Yung Shue Long valley as well. There are, however, no records of Finn ever excavating on the plateau and it seems likely that everything he mentioned was surface collected from sediments disturbed by agriculture and/or redeposited by erosion. The main features of his ‘YSW’ site were a predominance of coarse and hard geometric ceramics—the latter having a distinctive grey fabric dotted with black inclusions—‘cups, glazed and unglazed, some with “trade-marks”’, plus evidence for a stone ornament workshop ‘notably of quartz rings’, numbers of crude pebble picks, a few bronze objects, and a unique hard pottery ‘horse’ figurine.7 Chen Kung-che also reported collecting quartz artefacts and debitage on the terraced plateau, which further supported the workshop hypothesis.8 It is thus quite clear that, even before the Pacific War of 1941–45 Lamma Island was already a prominent location on Hong Kong’s ‘archaeological map’, but many of Sha Po’s secrets were yet to be revealed.

Plate 5:  Father Finn (left) working at Tai Wan in the 1930s. © Irish Jesuit Archives, Dublin, reproduced with permission.



How We Know about Ancient Sha Po   19

Post-war Research and Salvage Work Prelude to legislation The war had a profound impact on Hong Kong and its people: local Chinese and foreigners similarly endured tremendous hardships and suffering during the occupation. Not surprisingly, therefore, in the years following liberation the city’s inhabitants were more focused on rebuilding their lives and looking to the future, rather than considering archaeology and the past. Nonetheless, as early as 1947 Professor S. G. Davis of Hong Kong University was leading groups of friends and students on weekend trips to Lamma Island where they conducted minor surveys and excavations, although artefact collection rather than systematic recording seems to have been their main focus.9 In 1956 a more formal group of archaeological enthusiasts, comprising scholars in a variety of departments at the University and members of the public, founded the University Archaeological Team, which continued revisiting sites identified by the pre-war pioneers, as well as searching for new ones.10 As a known archaeological ‘hotspot’, Lamma Island continued to be a focus of interest and a 1959 visit to the Yung Shue Long part of Finn’s ‘YSW’ site is recorded in a handwritten notebook in the archives of the Hong Kong Archaeological Society. The Society was formed in 1967 and in November of that same year carried out trial excavations on the western slopes of the New Village plateau but found only evidence for quite recent agricultural activity and a few sherds of disturbed prehistoric pottery.11 Since the early 1970s a significant proportion of archaeological investigations conducted at both Sha Po sites have occurred as a result of Small Village House developments,12 which are the most common type of small-scale construction projects regularly occurring in Hong Kong. Indeed, the first clues as to the archaeological richness of the backbeach site resulted in 1972 from foundation trenching for a new house at Sha Po Old Village. There, the local contractor unearthed an almost complete Bronze Age pot [1],13 which suggested that the backbeach and plateau were both in use during that period. That chance find was followed up by three small trenches excavated by the society [2–4], which yielded evidence for stratified deposits with green glazed historical pottery and kiln debris—probably Six Dynasties–Tang

Figure 1:  1972 hard geometric pot

20   Piecing Together Sha Po

Plate 6:  Sha Po Old Village: photo montage of 1972 HKAS and Museum of History excavation. ©  Hong Kong Archaeological Society, reproduced with permission.

in date—above layers with Bronze Age and Later Neolithic pottery.14 Although small-scale, these first formal excavations clearly suggested that the backbeach site mirrored the known pattern of multi-period activity at similar sites around Hong Kong’s coastline.

Sha Po emerges: The Society and the AMO The establishment of the Antiquities and Monuments Office (AMO) in 1976 marked the beginning of a new era of more formalised heritage management practice in Hong Kong. Thereafter, anyone wishing to carry out archaeological surveys or excavations was required to have a government licence, and fieldwork in general became more rigorously controlled.15 In the early 1980s the AMO commissioned an archaeological survey of the entire territory, which included the excavation of a small trench on the plateau [29]. In line with the negative result of the Society’s earlier trenches, the excavators could find no trace of



How We Know about Ancient Sha Po   21

in situ Bronze Age deposits, only disturbed prehistoric material including a surface find of half a stone mould for casting bronze axes.16 The site therefore seemed to have been significantly degraded since Finn’s day; however, as we will see below, more extensive recent excavations on the plateau have confirmed that undisturbed early to mid-Bronze-Age layers and features do still survive there. The Society eventually returned to the backbeach in 1988 and investigated the area to the south-west of the 1972 trenches [5–7]. The three further trenches yielded large quantities of fired clay kiln debris associated with distinctive Six Dynasties–Tang pottery and some Song material—but no kiln structures. Underlying prehistoric material was concentrated in the two more easterly trenches and a range of Later Neolithic to Bronze Age materials, consistent with the 1972 results, was found.17 The material patterning already suggested that the main focus of prehistoric activity was probably down the landward (eastern) side of the backbeach while, based on the volumes of kiln debris being encountered, the early historical kilns noted elsewhere around Hong Kong’s coastline were almost certainly awaiting discovery at Sha Po. The Society returned to the Old Village the following year to conduct research excavations in previously untested areas of the backbeach. Of the four trenches, two were located to the north of the 1988 area [8 and 9], one to the south [11], and another around 80 m to the north-east [10]. Trench G-GT [8] was dominated by Six Dynasties–Tang material in blackened soils, including apparently in situ kiln furniture, a fired clay ‘floor’, and much shell and coral mixed with charcoal and lime.18 Here again was striking proof of early historical industrial activity, albeit still without any intact kiln structures. Close to the water table were a number of Bronze Age sherds and ‘water rolled’ coarse pottery. Approximately 15 m further north, Trench H [9] revealed ‘a floor of large kiln blocks and large coral pieces’,19 which reinforced the impression of widespread early historical industrial activity down the seaward (western) side of the backbeach. Some 50 m north-east of the latter excavation, Trench X [10] investigated the centralnorthern part of the backbeach in an area that remains largely untested to this day. More kiln debris was uncovered mixed with shells, mammal bone, and lime in association with Tang and Bronze Age pottery. Interestingly, beneath that there were indications of midden deposits (ancient rubbish dumps) containing Bronze Age pottery with burnt (mammal?) bones and fish bones, with some Later Neolithic pottery underneath.20 The fourth, and by far the largest (6 × 6 m), of the trenches excavated in 1989—Trench CC—[11] was at the southern end of the backbeach and proved very productive with well-stratified, multi-period archaeological remains including some of regional importance. A series of cultural deposits was noted, sloping down from south to north and seemingly reflecting the seaward profile of the prehistoric backbeach. The upper deposits contained the by then familiar mix of kiln debris, lime, and Tang pottery with some Song material, but beneath that was clear evidence for an early historical midden deposit with mammal and fish bones and Six Dynasties green glazed celadon. Below the midden, a thin layer of pumice overlay the rare find of a well-preserved Jin dynasty burial of an adult female with hairpin and two finger-rings. Given Hong Kong’s destructive acidic soils, such preservation of organic remains only occurs where overlying deposits contain shells, coral, or lime—as was the case here. The possibility of in situ Bronze Age deposits within the backbeach, which was first raised by the complete pot found in 1972, was finally realised in Trench CC when an horizon rich in material of that period, including three complete stem cups, four sandstone moulds—two a matching pair—for casting bronze axes, possible crucible fragments, and coarse pottery sherds splashed with bronze were found.21 Here at last was definitive proof of bronze casting on a Hong Kong archaeological site, although the activity had first been suggested by Finn in the early 1930s following his discovery of a mould half at Tai Wan and bronze splashes on stones at Hung Shing Ye.22 A fuller understanding of the scale and intensity of prehistoric activity on the backbeach resulted from two excavations conducted in 1994 by the Society on behalf of the AMO on the landward side

22   Piecing Together Sha Po

of the backbeach [12 and 13].23 Trench J [12] revealed a Later Neolithic–Bronze Age layer very rich in ceramics, beneath which was the first evidence for a Middle Neolithic presence on site. Trench K [13], which was located immediately south of Trench CC, yielded several complete Bronze Age pots from a layer that was clearly the continuation of that identified in the adjacent trench. The Han–Six Dynasties midden noted in Trench CC also continued south across Trench K. The upper layers in Trenches J and K both produced the familiar pattern of Tang dynasty industrial activity, though still with no evidence of actual kilns. The final investigation of the ‘pre-impact assessment’ era occurred in the Old Village’s western block, when two trenches were excavated during the redevelopment of a house [14 and 15]. Past building work had mixed Tang pottery and kiln debris with Later Neolithic and Bronze Age material, but a Six Dynasties burial was identified cutting into Bronze Age deposits and the rare discovery of one half of a sandstone mould for a pair of fish-hooks was also made.24

The ‘Impact Assessment Era’: House Plots and Pipe Slots The implementation in 1998 of the Environmental Impact Assessment Ordinance25 marked the beginning of commercial or development-funded archaeology in Hong Kong. However, the AMO continued to oversee all pre-development archaeological investigations for Small Village House developments in sensitive archaeological sites such as Sha Po. A key effect of the new legislation at Sha Po was the extension of impact mitigation—archaeological excavations and monitoring works—to the long pipe trench alignments of infrastructure projects, which opened up new areas for investigation that had previously remained untested. At Sha Po, this invaluable change has shown that while utility installations can cause

Plate 7: Yung Shue Wan Back Street: photo montage of the 2000–2001 AMO excavation. ©  Antiquities and Monuments Office, reproduced with permission.



How We Know about Ancient Sha Po   23

significant damage to archaeological deposits, important remains can nevertheless survive below or in between. The first fieldwork in Sha Po of the ‘new era’ was an AMO investigation in 2000–2001 [16] at the northern end of the backbeach prior to the construction of a row of new village houses. An L-shaped block of three trenches (65 m2) produced the first in situ early historical kiln structure, as well as deposits dating to the Song, Six Dynasties–Tang, and Bronze Age. The very worn condition of much Bronze Age pottery suggested that most, if not all, had been redeposited from the plateau/slopes above, seemingly by multiple episodes of downslope erosion. In the early 2000s, a planned programme of utilities improvements in Sha Po Old Village prompted a series of AMO-commissioned investigations along footpaths in the area. First, in 2001, a single small trench [17] produced a familiar pattern of somewhat mixed upper layers with kiln debris, Song–Ming and Six

Plate 8:  Photo montage of the 2002 investigations (clockwise from top left): Unit 6 excavation looking towards Sha Po Old Village; Kiln K3 structure, kiln debris, and lime fragments in Unit 7.3; prehistoric stone arm-ring in Unit 6.1; Kiln K2 wall in Unit 7.1; prehistoric coarseware sherd in Unit 6.3.

24   Piecing Together Sha Po

Dynasties–Tang material, followed by layers with Bronze Age and then Later Neolithic artefacts.26 That isolated trench was followed in 2002 by a much more extensive programme of testing within footpaths across the backbeach [18–24]. Despite the narrowness of most of the corridors involved and the high levels of recent historical-modern disturbance, a number of important discoveries were made, most notably, two Six Dynasties–Tang kiln structures in Unit 7 [23].27 At lower levels in Units 4–7 and 9 [20–23 and 24 respectively], though, were apparently undisturbed Later Neolithic layers reflecting the aforementioned rich Neolithic deposit found in 1994 [12]. Similarly, the Bronze Age activity noted that year was also reflected in contiguous areas: to the west of the same property in Unit 6 were two possible burials, while a further Bronze Age focus was noted to the north-east in Units 6 and 9 [22b and 24]. A few Middle Neolithic red painted pottery sherds found in the same unit [22] should perhaps be seen as outliers of the deposit noted in 1994. One final discovery of interest was a large deposit of later Qing pottery that had seemingly been tipped down the rear face of the backbeach [24].28 Another opportunity to assess the modern status of Finn’s plateau site arose when a housing development was planned to the north of 22–23 Sha Po New Village [30], which was immediately upslope of the small test pit excavated in the early 1980s [29]. A series of thirteen test pits were excavated within the 570 m2 site, giving a total excavation area of 68.5 m2.29 Recent agricultural activity was found to have truncated prehistoric deposits along the rear and front edge of each terrace, but in between was an in situ cultural layer with frequent Bronze Age pottery and stone artefacts.30 Finn’s site therefore clearly still existed and, moreover, the discovery of fifteen post-holes at the base of the cultural layer suggested it may have been a rare Bronze Age settlement site, possibly with stilt-houses. That same year another plot just to the west was also earmarked for housing development [25] and an L-shaped pattern of three trenches totalling 56 m2 in area was excavated in advance of construction. A further forty-nine post-holes were found—again at the base of the Bronze Age cultural horizon—therefore suggesting a settlement extending some distance across the south-facing flank of the plateau. A sizeable assemblage of stone artefacts, representing all stages of ornament manufacture, strongly supported Finn’s earlier claim for the existence of a stone ornament workshop.31 The pottery comprised types commonly found on the backbeach and thus indicated that the plateau and backbeach sites were probably in contemporary use, but perhaps reserved for different activities (of which more in Chapter 5). In October to November 2004 a pair of trenches was excavated close to what was thought to be the former seaward face of the early historical backbeach [26 and 27].32 The uppermost of two cultural layers was dated to the Song dynasty and produced a finds assemblage including green glazed celadon pottery, part of an iron hoe blade (cha), a fragment of a rare Shang dynasty stone ritual object or zhang, and three fragments of human skull suggesting the disturbance of a burial of unknown date.33 The second cultural layer spanned the Sui to Tang dynasties and included good examples of pottery of the former period. Interestingly, the Song layer produced spreads of shells, mammal, and fish bones suggestive of midden deposits, while similar materials were found in rubbish pits cutting into the Sui–Tang layer. A few Later Neolithic and Bronze Age pottery sherds were found admixed with historical material but in situ prehistoric deposits were not identified.34 Then between 2008 and 2010 a series of seven rescue excavations (AA1–AA7) were conducted across the backbeach in advance of a major scheme of mains sewerage and water main installation [31–37 respectively].35 In addition, approximately 390 m of pipe trench—subdivided into ten blocks (coded A–J)— were subjected to archaeological monitoring (watching brief )36 during construction phase groundworks [38–47 respectively]. Despite widespread modern disturbance, a number of important discoveries were made, most notably as follows: in rescue excavations AA5D and AA5G, which bisected an area rich in previous findings of kiln debris, two Six Dynasties–Tang kiln structures (K4 and K5 respectively) were



How We Know about Ancient Sha Po   25

unearthed [34], while in watching brief Area ‘B’ [39], just south of kiln K1, another Six Dynasties–Tang kiln (K6) was found with a fired clay ‘working floor’ (K7). The survival of the two structures under Yung Shue Wan Back Street was quite surprising given that utilities disturbance extended in places to a depth of 1.2 m below surface. Associated with the kilns were Six Dynasties–Tang cultural layers, which were sealed by kiln post-abandonment spreads containing a mixture of Tang, Northern Song, and Southern Song–Yuan pottery.37 Another late Qing ceramics dump, similar to that noted in 2003 [24], was encountered a few metres further to the north in trench AA6 [36]. The main foci of Bronze Age activity were in trench AA3 [35] and adjoining watching brief Area ‘A’ [38]—both perhaps at the outer edge of the substantial hill-washed deposit noted to the north-west [16]—and in trench AA5A-F [34] near the important deposits found in 1989 and 1994. Later Neolithic activity was widely evidenced but had particular concentrations down the eastern side of the backbeach in AA1, AA2, AA4, and AA6 [31, 32, 33, and 36 respectively]. With the exception of AA6, Bronze Age deposits also broadly mirror this patterning,

Plate 9:  Photo montage of the 2008–10 investigations (clockwise from top left): Area ‘B’ watching brief, Kiln K4 in trench AA5D, excavation at AA5A, adze rough-out and Bronze Age coarse geometric pot both at AA5B, archaeologist examining findings during watching brief Area ‘B’

26   Piecing Together Sha Po

thus further confirming that the backbeach originated to the east and then advanced to the west with the gradual accumulation of sand. The two most recent investigations occurred ahead of housing development in 2010 on the plateau [28] and in 2012 at the very heart of the prehistoric backbeach [48]. The 2010 excavation revealed a cultural layer and further post-holes indicating a clear continuation of the Bronze Age settlement site found in 2001 and 2003–4.38 The 2012 investigation revealed a 1.8 m deep sequence of backbeach deposits with traces of earlier historical kiln debris mixed with later historical pottery above a Bronze Age layer with pits cutting into an underlying Later Neolithic horizon.39

Learning from Sha Po: Reflections on Past Practice Sha Po’s historiography Sha Po is one small, but important, part of Hong Kong’s archaeologically rich island-coastal environment. Since the earliest times, human subsistence, settlement, and associated activities have been focused around the coastline and areas immediately inland. The Sha Po experience, however, offers insights into a number of important questions concerning different aspects of archaeological heritage and its management within the region. We saw above how archaeological sites and materials on Lamma came to the attention of early pioneers such as Finn due to a combination of agricultural activity and sand quarrying, as well as natural processes of erosion. From the earliest days, then, impacts of one form or another were destroying and, at the same time, revealing archaeological treasures. The main innovation today is that such impacts are generally more effectively managed using archaeological mitigation measures. More recent discoveries have thus resulted from archaeological investigations carried out in advance of housing developments or as part of programmes of infrastructural improvement. Not surprisingly, since the 1930s the legislative, methodological, and intellectual frameworks governing archaeological work in Hong Kong have changed dramatically. Constants do however exist; development has always impacted upon buried archaeology in the territory and archaeologists—whether amateurs, professionals, or government officers—have had to respond. The above historiography of Sha Po fieldwork offers many excellent examples of Hong Kong archaeology in action. Investigations have occurred at Sha Po in all stages of Hong Kong’s eighty-plus years of archaeological history: pioneering amateurs, local societies, and professional archaeologists have all played their part—whether conducting ‘antiquarian collecting’, research or rescue excavations, or monitoring pipe trenches during watching briefs—and all have contributed to the Sha Po story. It is thus a story which, on so many levels, serves to represent Hong Kong archaeology in microcosm.

Problems and potentials The Sha Po sites, like many others in Hong Kong, have less than perfect histories of investigation, management, and publication, but the same could be said of archaeological sites in many places the world over. However, rather than dwelling on past imperfections which, given Hong Kong’s development-driven economy, could certainly have been far worse if archaeology had not achieved the prominence it did during the later colonial and post-colonial periods, we should instead embrace the possibilities presented by the data we do have. We therefore hope to demonstrate through this book that in spite of—or rather as a result of—Hong Kong’s intensity of development, the territory has a great deal to offer archaeologically and Sha Po provides one of the most fascinating chapters in the story.



How We Know about Ancient Sha Po   27

One of the main objectives of this project was to attempt to demonstrate that Hong Kong’s site archives, despite often being fragmentary and inconsistent in nature, were nonetheless a rich and significant resource that, if used creatively, might yield a significant research dividend. Sha Po, with its chequered history of multiple excavators, a myriad of excavation and recording methodologies, varying degrees of disturbance, and very patchy reporting and publication exemplifies this state of affairs. But to take disturbance as an example, during the 2008–10 utilities-related project, there were relatively few well-preserved pockets of stratified deposits beneath footpaths; however, the findings within these long corridors helped to ‘tie together’ discrete patches of better preserved archaeology recorded in locations excavated ahead of housing development. Thus by mitigating the impacts of government infrastructure projects at Sha Po, a series of supporting data sets now exist that allow something approaching a more holistic, multi-period ‘landscape’ study of human activity on the backbeach and plateau to be attempted. That approach will be applied in Part II and, in particular, Part III of the book, but before exploring cultural developments in Hong Kong and Sha Po, we must first understand the environmental context within which they occurred. Therefore in the final chapter of Part I we explore the development of the Hong Kong and Sha Po landscape, beginning in the built-up present and then peeling back the layers to eventually reveal the pristine environment encountered by Sha Po’s Middle Neolithic fisher-hunterforagers at around 4500 BCE.

Notes 1. Map 5 shows the location plan of all forty-eight areas investigated—either research digs, rescue excavations, or watching briefs—and when discussed below the relevant area number, and context/layer if applicable, will appear in square brackets in text (e.g., [44] or [44:01]). The reader can then refer to the figure as a means of ‘navigating’ their way horizontally across the site and vertically through the multi-period landscape. 2. Geoffrey A. C. Herklots, ‘Obituary. The Rev. Father Daniel J. Finn, S.J.’, The Hong Kong Naturalist 7 (3–4) (1936): 269–70. 3. C. M. Heanley and J. L. Shellshear, ‘A Contribution to the Prehistory of Hong Kong and the New Territories’, in Praehistorica Asiae Orientalis (Hanoi: s.n., 1932), 63–76. 4. Daniel J. Finn, ‘Archaeological Finds on Lamma Island near Hong Kong, Part I’, The Hong Kong Naturalist 3 (3–4) (1932): 226. 5. Ibid.; Walter Schofield, An Archaeological Site at Shek Pik (Hong Kong: Hong Kong Archaeological Society, 1972): 162–63. 6. Daniel J. Finn, ‘Archaeological Finds on Lamma Island near Hong Kong, Part XIII’, The Hong Kong Naturalist 7 (3–4) (1936c): 257–58. 7. Finn, ‘Lamma Island, Part XIII’, 258–61. 8. Chen Kung-che, ‘Archaeological Excavations in Hong Kong’, Acta Archaeologica Sinica 4 (1957): 4. 9. William Meacham, The Archaeology of Hong Kong (Hong Kong: Hong Kong University Press, 2009), 34. 10. Solomon M. Bard, ‘Archaeology in Hong Kong: A Review of Achievement’, in Conference on Archaeology in Southeast Asia, ed. Yeung Chun-tong and Li Wai-ling (Hong Kong: University Museum and Art Gallery, the University of Hong Kong, 1995), 384. 11. James Watt, ‘Excavations at Yung Shu Wan, November, 1967’, JHKAS I (1968): 36. 12. The Small House Policy was enacted in 1972 and affords every male ‘indigenous villager’ from a ‘recognised village’ a grant to build one small house on the village’s land (‘a male person at least 18 years old who is descended through the male line from a resident in 1898 of a recognised village, an entitlement to one concessionary grant during his lifetime to build one small house’), at http://www. landsd.gov.hk/en/legco/house.htm. 13. Ho Ching-hin, ‘Sha Po Tsuen’, JHKAS IV (1973): 19–20. 14. The Neolithic in Hong Kong is conventionally recognised to have early and late stages to both its Middle Neolithic and Late Neolithic periods. Only a small amount of early stage Middle Neolithic material was found at Sha Po but, in contrast, large quantities of late stage Late Neolithic material have been uncovered. Therefore, in the interests of simplicity the late stage Late Neolithic is shortened throughout this chapter to ‘Later Neolithic’. 15. Tracey Lie-Dan Lu, ‘The Management of Cultural Heritage in Hong Kong’, Hong Kong Institute of Asia-Pacific Studies, Occasional Paper No. 137 (2003): 11–12. 16. Brian A. V. Peacock and Taryn J. P. Nixon, ‘Report of the Hong Kong Archaeological Survey, Volumes I–III’ (unpublished report prepared for the Antiquities and Monuments Office, 1986), 116. 17. Nigel Spry, ‘Sha Po Tsuen’, JHKAS XII (1990): 7–28. 18. William Meacham, ‘Sha Po Tsuen’, JHKAS XIII (1993a): 36. 19. Meacham, ‘Sha Po Tsuen’, 38. 20. Ibid.

28   Piecing Together Sha Po 21. Meacham, ‘Sha Po Tsuen’, 40. 22. Finn, ‘Lamma Island, Part V’, 123. 23. The archives for these two sites remain inaccessible due to copyright issues. Our discussion of the findings is therefore based upon the brief summaries of results available for study at the Hong Kong Archaeological Society and AMO Discovery Centre. 24. AMO, ‘Sha Po Old Village, Lamma Island, 1997’ (unpublished excavation report, 1997). 25. The Environmental Impact Assessment Ordinance (EIAO 1998) heralded a new ‘commercial’ era in Hong Kong archaeology whereby developers—either private or government departments—were expected to allow time and provide funds for archaeological investigations ahead of developments where archaeological resources were known or likely to be impacted. A number of archaeological consultancies were established at that time or soon afterwards to take advantage of the commercial opportunities provided by the new law, and such companies competitively tender for impact assessment work. 26. Field Archaeological Co., ‘Footpath Excavation at Sha Po Tsuen, Lamma Island’ (unpublished excavation report, 2001). 27. Archaeological Assessments Ltd., ‘Sha Po Tsuen Rescue Excavation’ (unpublished excavation report, 2003). 28. Ibid. Whether this deposit was the result of a gradually accumulating rubbish tip or a major clearance and dumping event is unclear. 29. Au Ka-fat and Raymond Lee, ‘Archaeological Investigation at Yung Shue Wan, Lamma Island, 2000–2001’ (unpublished report, 2001). 30. Ibid. 31. Finn, ‘Lamma Island, Part XII’, 259; C. J. Liu, ‘Excavation on the Ancient Site at Sha Po Tsuen, Lamma Island, Hong Kong’, Kaogu 6 (2007): 14–15. 32. ERM, ‘Rescue Excavation at Small House Lot No. 1575s.B. in D.D. 3 at Yung Shue Wan, Lamma Island’ (unpublished excavation report, 2005). 33. Ibid. 34. Ibid. 35. AAL, ‘Rescue Excavations at Sha Po Tsuen’. 36. AAL, ‘Archaeological Watching Brief at Sha Po Tsuen’. 37. Ibid. 38. AMO (personal communication). 39. AMO (personal communication).

3 Social Landscapes and Ancient Environments

Introduction Archaeology uses ancient material remains to reconstruct the lifeways of past people, but our primary methods of data gathering—surveys and excavations—occur in the setting of the present cultural landscape, which is a complex mix of modern and ancient features. Indeed, the surface remains of the most recent chapter of our Sha Po story—the village houses and abandoned field systems of late Qing to early twentieth-century rice-farming communities—still exist, albeit masked by a combination of scrub-forest regeneration and modern development. However, the bulk of Qing dynasty archaeological remains discussed in this book, like those of all earlier periods, lie buried below ground and only came to light as a result of excavation. Those buried remains reflect human agency or social action—past events and activities—that in the course of the following six chapters provide the basis for a series of narratives relating how, over a period of some six millennia, local communities inhabited, exploited, shaped, and were in turn shaped by Sha Po’s physical environment, thereby transforming the same geographical space (our study area) into a number of fundamentally different places or social landscapes.1 However, if we wish to understand Sha Po from a social landscape perspective, we first need to have some appreciation of the ancient environment with which past people interacted on a day-to-day basis. Sha Po, like most areas of Hong Kong, has seen little research conducted in the field of environmental archaeology,2 which aims to illuminate the subsistence economies and environment of past peoples. Environmental data are also often an integral part of landscape-focused research projects, but the archaeological data available to this project resulted from many separate excavations focused primarily upon identifying the periods represented by layers, features, and artefacts unearthed in each location, rather than the broader patterns of human-environment interaction they reflected. Nevertheless, when the available data are collated by period and then examined in the spatial and social context of the entire Sha Po study area—as well as the wider regional setting—enough clues exist to allow some important features of the ancient environment to be sketched out. Furthermore, one of the writers has long experience of approaching archaeological research from a landscape perspective, which suggested that much could be gained by analysing, interpreting, and presenting the results of Sha Po’s diverse and varied investigations through the unifying lens of a multi-period social landscape. Below we offer a brief overview of Hong Kong’s premodern physical environment, working backwards from the landscapes of rice farmers, to those of Six Dynasties–Tang coastal industrialists, and then prehistoric fisher-hunter-foragers. But first we begin with an introduction to landscape and the benefits of considering the material remains and associated activities of past human communities from a social landscape perspective.

30   Piecing Together Sha Po

A Social Landscape Approach to the Human Past Wherever we live in the world, we inhabit, interact with, and experience the physical environment not as a meaningless, spatial backdrop to our daily lives, but rather as a series of interconnected ‘meaning-full’ places with rich social significance and associated cultural values. Such places, which reflect geographically and socio-historically specific interactions between environmental processes and human agency, are often termed ‘cultural landscapes’.3 But here we must make an important distinction between the notions of ‘cultural’ and ‘social’ landscapes on the basis that the same cultural landscape might be perceived and experienced in a variety of ways and carry very different social meanings for different groups of people. In that sense, we embrace the popular European Landscape Convention definition of landscape as ‘an area, as perceived by people, whose character is the result of the action and interaction of natural and/ or human factors’.4 Moreover, we also embrace a ‘social landscape’ approach as it more accurately reflects the human context within which the archaeological remains we discuss and interpret in this book were created, used, and understood by past social groups—communities—inhabiting Sha Po. Therefore, in each period, the social landscape is both a physical reflection of human-environment interactions, and the embodiment of socio-political, economic, and ideological-religious aspects of communities inhabiting Sha Po at different times in the past. Of course, while the modern ensemble of village houses, shops, restaurants, streets, and so on can be walked around and experienced as a ‘seamless’ social landscape, the remains uncovered in excavations are but small fragments of previously much more extensive—and equally seamless—landscapes. But those material fragments from antiquity require ‘piecing together’ archaeologically if the ‘evidence for particular social practices’5 that they embody is to be more holistically understood. The Sha Po study area should thus be understood as containing a series of contrasting social landscapes created by different groups at different times and for different reasons. Some of those reasons were practical, for example, relating to economic necessity or socio-political organisation, but others such as the placement and orientation of human burials (even entire rice-farming settlements) seemingly also drew on beliefs of a more ‘intangible’ nature. In the remainder of this book, then, we aim to interpret the archaeological evidence through the lens of a social landscape perspective, but also in the knowledge that the people creating those landscapes almost certainly did so within a very different ideological-religious framework from that of the authors and many other readers of this book. We begin our review of the ancient environment with a summary of the key features of the New Territories’ traditional agricultural way of life and the physical and social landscapes it sustained in our study area and across much of premodern Hong Kong.

The Social Landscape of Later Historical Sha Po The landscapes of Hong Kong’s former agricultural communities were a clear reflection of their fundamental reliance on wet rice cultivation, supported by a variety of other crops grown in dry terraced fields. Underpinning that economic ‘model’ were a series of deeply held beliefs within farming communities, which centred on the Chinese notion that human existence is inextricably bound up with nature6 and communities’ relationships with the local environment were shaped by a complex fusion of ancestor worship, animism, and geomancy (fung shui).7 Thus the people, their economy, belief system, and the physical environment combined to create a highly sustainable way of life and a social landscape that was a powerful reminder of who they were and where they came from. As anthropologist Rubie Watson observed:



Social Landscapes and Ancient Environments   31 [A]mong the people of rural Hong Kong geomancy is deeply intertwined with the ancestor cult and can be considered integral to many local religious practices. Geomancy not only allows humans to comprehend and take advantage of the forces of nature that surround them, but also guides the creation and maintenance of landscapes.8

In reality, the layout of field systems was primarily a matter of common sense use of the natural topography and drainage patterns, whereas the positioning of villages, temples, ancestral halls, graves, and shrines were all matters requiring consultation with the local geomancer or fung shui master. The longterm sustainability of local rice-farming communities relied upon careful landscape management, which reflected a deep knowledge of the environment, its challenges, and potentials. In a mountainous region prone to typhoons and landslides, communities managed environmental risk by growing a mixture of double crop rice in their larger low-lying paddy fields—the bulk of which were often owned communally by ancestral trusts9—while smaller, privately owned, terraced fields would, if well-watered, be used for single crop rice10 and/or crops including sugar cane, sweet potatoes, or peanuts.11

Plate 10: Typical New Territories rice-farming village setting taken by Hedda Morrison in 1946–47. ©  HarvardYenching Library, reproduced with permission.

32   Piecing Together Sha Po

Although every village’s fung shui landscape was a unique response to local environmental factors, one can, nonetheless, distinguish recurring patterns throughout the less urbanised parts of the New Territories that reflect a depth and consistency of beliefs regarding the auspicious placement of villages and their inhabitants. In the ‘classic’ example shown below, the compact two-row village nestles below a low hill with its back protected by its species-rich fung shui woodland. In front of the village is the flat area used for drying rice, while below are the community’s paddy fields stretching across a flat valley floor served by a network of carefully maintained stream-fed water channels. The outer perimeter of villages is invariably marked by armchair-shaped earth god shrines, which define the area under the deity’s protection, while the ‘master’ shrine guarding the village’s main entrance is often situated directly beneath an ancient banyan ‘spirit’ tree. A fine example grows beside the Tin Hau temple in Yung Shue Wan. Last but certainly not least, the ancestral burial ground will be sited out of view on an auspiciously orientated hillside not far from the village. The Sha Po study area is a good example of the adaptation of that ‘standard’ rice-farming template in an island locale with neither prime agricultural land nor ‘ideal’ fung shui settings. Sha Po Old Village was itself sited to take advantage of the slight ‘hummock’ of the backbeach, which was too sandy to be of great value agriculturally. However, the valley to the east was well supplied by small perennial streams and, with appropriate drainage and water management, the adjoining low-lying former lagoon was therefore perfect for wet rice agriculture. In fact, most—if not all of the low-lying fields between Sha Po and Wang Long—would have originally been rice paddies laid out surrounding their respective settlements, while all accessible hillsides such as the New Village plateau were terraced—often with boulder facings to limit erosion—and used for growing crops such as sweet potatoes, papaya, maize, and peanuts. A similar pattern of terracing is visible on the flank of Tai Shan and on both sides of the Yung Shue Long valley. The main areas not under cultivation were wooded slopes directly behind villages, the villages themselves, or ancestral burial grounds on nearby hills. Trees would have been far less in evidence than they are today and besides individual spirit trees, fung shui groves, and stands of fruit trees in the immediate environs of villages—it seems likely that all lowlands and most slopes near settlements would have been cleared and under cultivation. Even though Sha Po is in an unusually exposed location with no protective hill or woodland nearby, it nevertheless sits in a quite typical slightly raised position above what would have been a well-watered valley and paddy fields. By the time Sha Po was built, Yung Shue Wan was already expanding north around the bay from the Tin Hau temple, which may itself have afforded some geomantic protection to the village behind. The original settlements of Wang Long, Ko Long, and Yung Shue Long occupy more sheltered positions, each being slightly raised above the valley floor overlooking their former stream-fed paddy fields, but only Yung Shue Long has its back to a wooded hillside in what might be considered a more typical fung shui setting. A clear impression of the socio-economic and landscape transformation wrought over the last fifty years is provided by the 1960s’ photographs in Chapter 1 and the 1968 map shown below, in which almost all available land was then still under cultivation (marked ‘C’ on the map). In contrast to our good understanding of the results of later historical rice farmers’ interactions with the physical environment, the more ancient landscapes and environments of earlier historical and prehistoric communities lie hidden from view, but can be reconstructed—in overview at least—using data provided by archaeologists, engineers, and earth scientists.

Map 6:  1968 1:1200 scale map of Yung Shue Wan. Source: Crown Land and Survey Office (CLSO). Yung Shue Wan, Lamma Island. 1:1200 Scale Topographic Map Sheet C-229-SW-C / C-246-NW-A. Hong Kong: CLSO, 1968. Reproduced with permission of the Director of Lands. © The Government of the Hong Kong SAR. Licence No. 59/2015.

34   Piecing Together Sha Po

The Many Attractions of Sha Po’s Ancient Environment Beyond the more regional focus of broad environmental data summarised below, the majority of our specific evidence for Sha Po’s ancient—pre-rice-farming—environment and physical landscape has come from archaeological excavations, which form the basis for the Middle Neolithic to later historical landscape reconstructions presented in Chapter 8. The general character of available resources—foodstuffs and raw materials—and their occurrence on archaeological sites in Hong Kong have already been discussed in print, but we will examine the Sha Po evidence for such resources, in the context of wider landscape change, in the following four chapters.12 Here, though, we will restrict ourselves to some general observations concerning the environmental character of our study area prior to its wholesale transformation by rice farmers, more specifically as it would have appeared when encountered by Sha Po’s first inhabitants in the Middle Neolithic. A combination of ancient sea-level data, the modelling of subsea topography, and engineering cores drilled into the seabed, suggest that by the time the climate started to warm at the end of the last Ice Age—between 11,000 and 12,000 years ago—Hong Kong’s mountainous New Territories and islands were then upland areas overlooking a vast coastal plain drained by large meandering rivers, which ran through valleys blanketed by dense subtropical broadleaf forest.13 Over the next five millennia sea-levels rose and flooded the plain, its river valleys, and forests, eventually stabilising at more or less modern levels by around 5000 bce.14 With the basic structure of the archipelagic landscape in place, a combination of oceanic and climatic processes, in particular relating to heavy summer rains and typhoons, got to work creating the prominent tombolo formations and the backbeaches that from around 4500 bce proved so attractive to Sha Po’s Middle Neolithic inhabitants. It seems likely that behind the coast, the aforementioned forests would almost certainly have covered the valleys, flanks, and summits of all but the highest mountains in the region.15 In places at that time the coastal environment of our wider study region was dramatically different from the one we see today. For example, the Pearl River Delta was then a wide shallow bay studded with islands, while in areas such as Sai Kung and Deep Bay the coastline was further inland than today and there were numbers of small islands offshore. Eventually millennia of river-borne sediment and storm-blown sand would radically alter the coastline, in particular around the Pearl River mouth and Deep Bay,16 but when our Sha Po story begins, the archaeological and environmental evidence indicates that Hong Kong and the Pearl River Estuary formed one large region inhabited by culturally similar groups of mobile fisher-hunter-foragers. The environment they inhabited must have been superabundant in marine, shore, and land-based foodstuffs, while the forests and coastal hills provided a myriad of raw material resources. While we have no physical trace of ocean-going vessels, the apparent mobility of such coastal-maritime groups suggests that they had and made regular use of some form of boats, most likely dug-out canoes, which may or may not have been fitted with outriggers.17 Arriving by boat at Yung Shue Wan, one would have entered a sheltered sandy bay surrounded by densely forested hillsides, while in the centre was the wooded Sha Po plateau, which then extended further into the bay, effectively dividing it in two. To the left was a smaller beach at the mouth of the Yung Shue Long valley, from which drained one perennial stream, while to the right, behind the larger beach, was the gently rounded hump of the Sha Po backbeach. From its top Sha Po’s first people would have gazed inland over a shallow freshwater lagoon, with the densely wooded Wang Long valley and hills beyond, while to the south a second stream flowed from the wetland through the southern backbeach and out into the bay. In sum, for maritime-focused fisher-hunter-foragers, Sha Po provided an appealing mixture of attributes: a relatively sheltered bay and backbeach with access to coastal resources and fishing grounds, a freshwater lagoon fed by perennial streams, and a forested hinterland full of plant foods, raw materials, and wild game.



Social Landscapes and Ancient Environments   35

Conclusion In Part I we have introduced our study region, discussed the biography of investigation and discovery in terms of Hong Kong’s history of archaeological fieldwork, and then rounded off the process of contextualising Sha Po with an overview of the character and trajectory of environmental and landscape change from the modern era back into prehistory. Next, in Part II, we begin our chronological discussion of Sha Po’s human story, beginning in the pristine setting of Hong Kong’s newly formed archipelagic landscape.

Notes 1. The key ideas of the social landscape concept and some good examples of their archaeological application can be found, for example, in Gabriel Cooney, ‘Social Landscapes in Irish Prehistory’, The Archaeology and Anthropology of Landscape, ed. P. J. Ucko and R. Layton (London: Routledge, 1999), 46–64; Wendy Ashmore, ‘Social Archaeologies of Landscape’, in A Companion to Social Archaeology, ed. Lynn Meskell and Robert W. Preucel (Oxford: Blackwell, 2004), 255–71; Lynn Meskell and Robert W. Preucel, eds., A Companion to Social Archaeology (2004). 2. John Evans and Terry P. O’Connor, Environmental Archaeology: Principles and Methods (Stroud: Sutton, 1999). The Sha Ha study is a notable exception here (Tracey Lie-Dan Lu, Z. Zhao, and Z. Zheng, ‘The Prehistoric and Historic Environments, Vegetations and Subsistence Strategies at Sha Ha, Sai Kung’, in The Ancient Culture of Hong Kong: Archaeological Discoveries in Sha Ha, Sai Kung, ed. AMO [Hong Kong: AMO, 2005], 57–64). 3. Mechtild Rössler, ‘World Heritage Cultural Landscapes: A UNESCO Flagship Programme 1992–2006’, Landscape Research 31 (4) (2006): 334. Perhaps the most quoted definition of cultural landscape was that coined by Carl O. Sauer, who said: ‘The cultural landscape is fashioned from a natural landscape by a cultural group. Culture is the agent, the natural area is the medium, the cultural landscape the result.’ Carl O. Sauer, ‘The Morphology of Landscape’, in Land and Life: A Selection of the Writings of Carl Ortwin Sauer, ed. J. Leighly (Berkeley: University of California Press, 1965[1925]), 315–50. 4. Council of Europe, European Landscape Convention (Florence: Council of Europe, 2000). 5. John C. Barrett, ‘Fields of Discourse: Reconstituting a Social Archaeology’, Critique of Anthropology 7(3) (1988): 6. 6. Mick Atha, ‘A Neglected Heritage: Towards a Fuller Appreciation of the Landscapes and Lifeways of Hong Kong’s Rice Farming Past’, Asian Anthropology 11 (1) (2012): 132–33. 7. ‘Animism’: the idea that natural objects, most often boulders and trees in local belief systems, can house spirits that have intentions and desires which can impact upon the world of the living. 8. Rubie S. Watson, ‘Fengshui, Landscape and History in Rural Hong Kong’, Symbols Spring (2007): 3. 9. Patrick H. Hase and Lee Man-yip, ‘Sheung Wo Hang Village, Hong Kong: A Village Shaped by Fengshui’, in Chinese Landscapes: The Village as Place, ed. R. G. Knapp (Honolulu: University of Hawai‘i Press, 1992), 81. 10. Hase, personal communication. 11. Jack M. Potter, Capitalism and the Chinese Peasant: Social and Economic Change in a Hong Kong Village (Berkeley and Los Angeles: University of California Press, 1968), 57. 12. See Tracey L. D. Lu, ‘Natural Resources and Subsistence Strategies in Prehistoric Hong Kong’, Kaogu 6 (2007a): 36–45. 13. Edward Stokes, Hong Kong’s Wild Places (Hong Kong: Oxford University Press, 1995), 8; Fyfe et al., The Quaternary Geology of Hong Kong (Hong Kong: Geotechnical Engineering Department, 2000), 29. 14. Wong and Shaw, High-Level Coastal Deposits, 12–13. In truth, when the earth sciences, archaeological, and other data are compared, there is some uncertainty surrounding the timing and height at which sea-levels stabilised, but our own observations at Sha Po and Meacham’s research at Sham Wan (cited in Wong and Shaw) suggest that on Lamma Island at least the situation we describe probably applied. 15. Stokes, Wild Places, 8. 16. Many of the small near-coast islands that formed an early focus of human activity were eventually landlocked by sedimentary processes. Careful consideration of the ancient environmental context is therefore crucial to the interpretation of prehistoric coastal sites across the Hong Kong–Pearl River Delta region. 17. Stephen Davies (personal communication) noted that there is no evidence yet available for outrigger use in the prehistory of the Lingnan coast. He suggests that an alternative may have been the temporary creation of ‘double canoes’ when longer open-ocean voyages were planned. We are grateful for his insights on this and other matters involving boats.

Part II Sha Po’s Human Narrative

4 Sha Po’s First People: Neolithic Fisher-Hunter-Foragers

Introduction Sha Po is one of many backbeach sites dotted around the coastline of Hong Kong’s mainland New Territories and, in particular, islands which, by around 4500 BCE, evidence the activities of bands of boat-using1 hunter-fisher-foragers who then inhabited the area.2 At that same time, in the Yangzi Valley there was a fully developed agricultural Neolithic3 with often large villages inhabited by wet-rice cultivators who kept, among other animals, domesticated pigs, dogs, and fowl.4 By roughly 5000 BCE, hunterforagers in inland Guangdong had, as Drewett commented, ‘adopted “Neolithic” attributes of polished stone tools, pottery, and a formal burial rite with grave goods’.5 By 2400 BCE, the region’s first horticulturalists were well established in northern Guangdong at sites such as Shixia, where rice cultivation and processing, and corded and impressed geometric pottery were recorded.6 Based on the palaeoenvironmental and archaeological evidence, it seems clear that communities in Hong Kong were in fact part of a larger maritime-focused socio-economic region encompassing the Pearl River Delta—then in fact a shallow, island-studded bay—and adjoining areas of the South China coast.7 Fisher-hunter-foragers there inhabited a rich coastal environment with abundant marine and terrestrial resources—including many carbohydrate-rich wild plant resources—that supported a diversified and highly successful subsistence economy, which may explain why rice farming seemingly did not develop in Hong Kong before its introduction by immigrants to the region from the eleventh century ce onwards.8 The character and sequence of Neolithic developments in the Pearl River Delta–Hong Kong region have already been widely discussed elsewhere,9 so to avoid repeating such work we will present only an overview of cultural developments during the entire era. A more detailed discussion of the socio-political, economic, and ideological implications of changes in material culture will be presented in relation to the two main periods of Neolithic activity represented at Sha Po.

Neolithic Materiality and Lifeways in Hong Kong General chronology and patterning of human activity Middle Neolithic people seem to have been particularly drawn to islands—some with tombolo formations—although the primary focus of human activity throughout prehistory was the backbeaches, in particular on islands but also around the mainland coast of the New Territories.10 But from the beginning of the Late Neolithic a continued interest in backbeach areas can be contrasted with a new focus on prominent coastal headlands that provided opportunities for surveillance and perhaps defence, and may be an indication of increasing competition in society (a subject we will discuss further below in relation to artefactual evidence).

40   Piecing Together Sha Po

Based primarily upon changes in the kinds of pottery—and other distinctive artefacts—people used and radiocarbon dates on charcoal from layers in which they were found, the Neolithic in Hong Kong can be broken down into four cultural phases.11 The earliest securely dated sites are from the earlier Middle Neolithic (4500–3500 BCE),12 which is characterised by finds of fine white basins and red-painted bowls and cups with perforated pedestals—‘consumption’ wares used for eating and drinking—round-bottomed cord-impressed coarse cooking pots, and fired clay ‘vessel legs’ probably from ding tripod cooking pots.13 By the later Middle Neolithic period (3500–2700 BCE) consumption wares took the form of white pedestal bowls with perforations, incised lines, and impressed circles. In addition to corded decoration, cooking and storage wares now also had incised wavy lines and shell-like impressions.14 Hac Sa Wan in Macau was the first site in the Hong Kong–Pearl River Delta region where both Middle Neolithic types occurred and their relative dating could therefore be established.15 The relative phasing has since been further elaborated following the discovery of far richer, well-stratified deposits at Xiantouling in Shenzhen.16 The Late Neolithic is also divided into an earlier (2700–2400 BCE) and later stage (2400–1500 BCE). Sites of the former period are relatively few in number and are marked by the use of particularly large coarse cooking pots and relatively hard fine pottery with distinctive high-necked and foot-ringed forms—some with applied strips of clay and incised lines—with a range of stamped geometric decorations including ‘basketry’, concentric circles, leaf vein (or herringbone), and zigzag.17 The later stage, which has been identified at over thirty sites in the region, has fine pottery that is generally softer and chalky in texture and includes carinated and spouted forms with more elaborate geometric decoration.18 In recent years, based on archaeological discoveries in southern China, many mainland scholars now consider Hong Kong’s later stage Late Neolithic (2400–1500 BCE) to be part of the ‘Early Bronze Age’,19 even though bronze use in Hong Kong occurs certainly no earlier than 1200  BCE, and then only in association with hard geometric pottery.20 In the interests of consistency and clarity, we will follow the AMO’s 2007 paper and use the conventional Hong Kong periodic divisions of later stage Late Neolithic (2400–1500  BCE) and Bronze Age (1500–500  BCE). However, we do acknowledge that such neat boundaries are more of an archaeological creation than a cultural reality, and there is something of a transitional phase between the two periods, which is widely evidenced at Sha Po.

Continuity and change in the craft-domestic sphere Despite the above detailed changes in pottery use, the basic pattern of fine clay ‘consumption’ wares and corded coarse ‘cooking’ wares—designed for their toughness and resistance to the thermal shock of direct heating over fires—continued in use throughout the Neolithic. There is similar continuity in the use of thick-walled coarse clay cylinders known as pot-stands,21 which are thought to have been used to support round-bottomed coarse pots—presumably during cooking—but oddly enough most show few signs of sooting, but it may have worn off their smooth surface over time.22 A range of simple utilitarian pebble-based tools was also used right through the Neolithic period, including unmodified pebble pounders (or hammers)—with flattened ends caused by use—and a myriad of crudely chipped choppers, scrapers, and pointed picks, which probably saw diverse application in food gathering and processing, stone working, and other areas of craft activity. Some scholars have labelled the more pointed variety ‘oyster chisels or picks’, but we prefer to use the more neutral term ‘pebble tools’ to reflect the general uncertainty concerning the precise uses to which they were put.23 Sandstone or siltstone polisher-whetstones were also ever-present items used to shape and polish adzes made from hard volcanic rocks, which were the ubiquitous tree-felling and wood-working tool of prehistoric Hong Kong.



Sha Po’s First People: Neolithic Fisher-Hunter-Foragers   41

The adzes themselves were rectangular and trapezoidal in form and some had shoulders, while the introduction of more advanced stepped forms seems to have been a Late Neolithic development. Sharp-edged flake tools struck from fine-grained rocks such as dacite and rhyolite (volcanic lava) can be interpreted as general blades used for detailed cutting and processing tasks (i.e., dismembering and stripping carcasses, and cutting up plant foods). That general composition of stone ‘toolkit’ continued throughout prehistory, but with improvements in design and changes in emphasis through time. Although the sharpening and finishing of stone tool and weapon rough-outs seems to have often occurred within the general domestic context, major stone tool and ornament workshops are also known.24 Discoveries in Hong Kong and Delta sites of grooved stone bark-cloth beaters reflect the Middle Neolithic development of a clothing technology based on the processing of bark from mulberry and other suitable trees.25 By the Late Neolithic, though, bark-cloth had been replaced by spun fibre technology using stone and pottery spindle whorls, which may have been adopted through contact, trade, and exchange with agropastoralists in inland Guangdong, and provide a clear indication that twine, woven fabrics, and cordage could then be made.

Fishing, hunting, and foraging: Subsistence activities Waisted pebbles—usually interpreted as fishing net-weights—are another category of artefact occurring on sites throughout the Neolithic, while occasional larger versions have usually been interpreted as ‘boat-anchors’.26 Turning to hunting and warfare, a few leaf-shaped projectile points—arrowheads or spearheads—are known from Middle Neolithic sites, but from the earlier Late Neolithic their numbers and quality of manufacture greatly increased.27 By the end of the Late Neolithic, refinements in projectile point design such as the introduction of simple fluting—which was further developed in the Bronze Age—may partly be a consequence of craft specialisation, but they also created more efficient weapons in what was probably an increasingly competitive society. Fishing, hunting, and foraging are all recorded in the few shell middens identified in Neolithic Hong Kong. Later Middle Neolithic shell middens at Sham Wan on Lamma Island allowed the preservation of several wild boar and deer teeth, and bones from fish including headgrunt, ray, and shark,28 the last two being some indication that the community had already mastered the art of offshore fishing. An intensification of fishing activity is suggested by the increased numbers of waisted pebble net-weights found at some later stage Late Neolithic sites. However, the remains of fish are rare, as their survival relies upon the existence of shell middens such as those found at Sham Wan and Po Yue Wan.29 Interestingly, the same two species of fish dominated assemblages from both sites—the headgrunt (Pomadsys hasta) and marine catfish (Arius leiotetocephalus)—which may indicate targeted seasonal exploitation of both species when they came inshore en masse to spawn.30 Vertebrate remains also included the bones of deer, dolphins, pigs, sea turtle, bird, and dog, of which the latter may be Hong Kong’s only confirmed Neolithic domesticated animal. Po Yue Wan also produced the rare discoveries of bone projectile points and a locally unique bone awl.31 Further organic artefacts were unearthed at Tung Wan Tsai North, where perforated fish vertebra beads, shell ornaments, and scrapers all occurred as grave goods.32 However, Hong Kong’s coastal middens are few and relatively small-scale, especially when compared with the huge shell mounds of more permanent villages of the Pearl River Delta’s Hedang Culture (3000–1900 BCE).33 Our understanding of the importance of plant foods in the Neolithic coastal economy was greatly enhanced by a study of starch and phytolith34 remains on stone tools from the Middle-Late Neolithic

42   Piecing Together Sha Po

(c. 3000–2000  BCE) backbeach site at Xincun, Taishan.35 Starch residues revealed the use of palms, bamboo, bananas, freshwater roots and tubers, whereas the phytolith data showed a massive dominance of palms and very low incidence of rice. The only concerted effort to recover plant remains in Hong Kong occurred at the Sha Ha site in Sai Kung, where the AMO mounted a programme of soil sampling and analysis aimed at understanding past ‘subsistence strategies . . . environments and natural resources’.36 The results included high pollen counts of grassland species, phytoliths of domesticated gourd, and pollen/ phytolith evidence for trees of the oak and palm families, among which the latter is now even more significant in light of the Xincun, Taishan evidence for their targeted exploitation. Rice, the key cereal crop of the inland South China Neolithic, was identified by phytoliths and two partial charred grains—one from a layer dating to the final stage of the Late Neolithic and the other Bronze Age to Warring States in date—but the grains were too fragmentary to be identified as either wild or domesticated. Somewhat more definitive evidence has recently been recovered at the Guye shell-mound site in the Gaoming region of the Pearl River Delta, where twenty or more grains of domesticated rice were dated to 2000 BCE.37 However, based on the evidence from backbeach sites at Sha Ha and Xincun, Taishan, prehistoric rice cultivation at present seems unlikely in Hong Kong, although there was perhaps localised woodland clearance for slash-and-burn cultivation of some plants such as gourd, and probable cyclical harvesting of a range of starch-rich plant foods such as sago palm, freshwater roots and tubers, and also wild rice.

Neolithic homes: Stilt-houses or ground-level huts? Although no prehistoric structural timbers have ever been found in Hong Kong, excavations at Tai Wan in 1996 produced evidence for two earlier Middle Neolithic houses marked by stone-packed post-holes cut into clayey decomposed granite on the lower hillside behind the backbeach. The houses were reported to have internal, clay-lined hearths (fireplaces), which should indicate that they were ground-level, not stilt-house, structures.38 Similar locations in later Middle Neolithic sites at Fu Tei Wan and Kwo Lo Wan on Chek Lap Kok and Sai Wan on Cheung Chau all had post-holes suggestive of timber structures, while the use of packing stones at Fu Tei Wan indicates structures—like those at Tai Wan—of some permanence, whereas for simple temporary shelters one might expect to find little more than clusters of much smaller stake-holes. The massive excavations at Sha Ha in Sai Kung revealed a Late Neolithic occupation with evidence for a stilt-house settlement, while two unparalleled large ‘long-house-type’ ground-level structures were excavated near the ancient shoreline of Deep Bay at Ha Pak Nai.39 The rocky headland sites at Sha Lo Wan and Pa Tau Ku both produced quite rich assemblages of material and remains interpreted in terms of domestic, craft, and funerary activity. The former site is interesting for its apparent zonation of activities within enclosures defined by rock-cut post-holes, while the latter included the possible remains of up to twenty houses identified from post-holes and pits within square or circular features delimited by boulders.40

Contrasting patterns of mortuary behaviour Middle Neolithic human burials are rare but were thought to have been identified at Chung Hom Wan based on the burial of complete pots,41 while somewhat later examples were inferred at Kwo Lo Wan, Fu Tei Wan, and Sha Ha from ‘pits’ containing complete pots, adzes, and stone rings. However, most ‘grave’ pits were too small even for flexed burials and probably represent secondary burials,42 or maybe even some form of non-funerary ‘structured deposition’.43 In contrast at Sham Wan, the presence of overlying shell



Sha Po’s First People: Neolithic Fisher-Hunter-Foragers   43

middens meant that two sets of skulls and long-bones survived to reveal the position of burials, while many fragments of burnt human skull confirmed that multiple cremations had occurred.44 Based on the consistent pattern of grave goods we can suggest that Middle Neolithic society was perhaps relatively egalitarian. In contrast, in the earlier part of the Late Neolithic the inclusion of exotic, ceremonial objects in some burials—such as multiple large yue blades found at Yung Long and Fu Tei Au—may indicate a growing interest in ritual activity and material markers of status. They also suggest connections with groups in the Pearl River Delta and hint at contact, trade, and exchange—presumably via major rivers—with inland horticulturalists that made wide use of such objects, initially as jades imported through the Nanling Mountains from Liangzhu Culture farmers of the Middle Yangzi,45 but later also as locally produced copies. An opportunity to ‘meet’ the people themselves occurred in 1997 at Tung Wan Tsai North on Ma Wan Island, where a cemetery comprising nineteen burials with grave goods and dating from around the end of the Late Neolithic was discovered.46 Fifteen graves contained human remains, of which seven—in shell-rich sand—had almost complete skeletons. The better-preserved graves comprised seven primary and three secondary burials and included men, women, and children, which indicates the site’s use by an entire small community.47 Some adults’ incisors (front teeth) had been intentionally removed—presumably for ‘cosmetic’ reasons—which is a cultural trait noted in ‘Hedang Culture’ sites in the Pearl River Delta48 and in prehistoric communities bordering the South China Sea as far west as the Gulf of Siam.49 The inclusion of marine shells as grave goods at Tung Wan Tsai also occurs throughout the latter geographical range. Analysis of surviving teeth revealed the ‘shovel-shaped’ incisors so characteristic of Asian Mongoloid populations.50

A growing interest in ornamentation and display Middle Neolithic personal ornaments are comparatively few in number and mainly comprised small and relatively crudely made split rings (earrings) worked from quartz and other crystalline rocks. But during the Late Neolithic local communities seem to have developed a growing taste for white and clear crystal earrings, which see a dramatic improvement in quality that seems to coincide with the development of workshop areas that herald later accomplishments in polished disc, bangle, and earring working during the Bronze Age. By the later stage Late Neolithic new ‘exotic’ items also appear in the form of flanged ‘T’-section bangles and ge dagger-axe blades in a cream-yellow rock,51 which was too soft to be of practical use and the ‘weapons’ may have in fact fulfilled some aspects of the ceremonial role previously played by large yue axes mentioned above. Moreover, when viewed together with the shiny personal ornaments, they indicate that personal display associated with status and ceremony was maybe becoming a more important feature of local life, which in turn suggests the emergence of social hierarchies operating within a more competitive socio-political ‘landscape’.

Discussion The contrast in scale and character between the Hedang-type shell mound sites and many much smaller sites in Hong Kong, added to the relative seasonal abundance of food fish during winter to early summer, has led some scholars to propose that many backbeach sites were merely seasonal fishing camps within the territories of larger groups based more permanently in settlements at larger shell mound sites in the Pearl River Delta.52 The evidence for Later Neolithic rice cultivation at the large permanent shell mound

44   Piecing Together Sha Po

settlement at Guye epitomises the contrast between sites in the Delta and around Hong Kong’s coastlines and islands. On the other hand, some have argued that evidence such as the pair of timber-built rectangular ‘long-houses’ with rammed-earth floors found overlooking Deep Bay at Ha Pak Nai,53 and the Tung Wan Tsai North cemetery, show that some Hong Kong sites were more permanently occupied.54 Similarly, one could argue that the nature and scale of activities identified at Pak Mong also surpassed what one might expect at a seasonal camp.55 In reality, though, our chronological resolution for different periods or phases of activity at many Hong Kong prehistoric sites is often no better than several hundred years. For example, at Tung Wan Tsai North the cemetery may simply reflect the repeated use of the site by one or more small kin-based groups, while the presence of children in itself does not provide definitive proof of a more permanent occupation.56 In general, apart from the few extensively excavated sites such as Sha Ha, Yung Long, and some of the aforementioned headlands, many sites have seen quite small-scale and biased sampling—for example, focused only on backbeach areas—which cannot possibly hope to illuminate the full range of human activities occurring across the wider ancient coastal landscape. We should not expect to find settlement evidence in the loose sands of backbeach sites, but experience suggests it may survive in untested clayey lower hill slope areas nearby. Equally, in some cases where adjacent bays have contemporary prehistoric sites, such as Sha Po and Tai Wan, one might easily be a functionally distinct satellite or subordinate of the other. On balance, then, our understanding of patterns of landscape exploitation in Neolithic Hong Kong is far from complete. That said, we have clear evidence for settlement sites and burials in some lower hill slope areas and domestic, craft, and mortuary activity on backbeaches. Those sites were used by mobile, maritime-focused fisher-hunter-foragers who by the Late Neolithic seem to have gathered wild starchrich plants, cultivated others such as gourd, and kept dogs and possibly pigs. We would suggest that the Pearl River Delta and coastal Hong Kong formed one culturally integrated region, with larger and more permanent settlements in the Delta, and a range of longer-term and more intermittently used sites around the islands and coast. Although rice farming may have developed at some permanent settlements well within the Delta, the mobility essential to the successful exploitation of Hong Kong’s archipelagic environment, added to the availability of a diversity of food resources including starchy wild plants, almost guaranteed the very late adoption of wet rice agriculture at the coast. We now use that regional background to situate and contextualise the evidence for Neolithic people at the relatively compact backbeach site at Sha Po Old Village.

Neolithic Sha Po As explored in the previous chapter, the first people to make use of the Yung Shue Wan area of Lamma Island would have encountered a pristine landscape rich in food and raw material resources. Arriving by boat one would have seen a curving sandy bay bordered by low wooded hills with the gentle rise of the backbeach centre-right, the promontory of the Sha Po plateau in the centre, and small streams entering the sea from the Yung Shue Long Valley to the left and Wang Long Valley to the right. Walking inland on to the top of the backbeach one would have looked out across a shallow freshwater lagoon—probably somewhat brackish at high tide—bordered by dense woodland on three sides. The majority of diagnostically Neolithic material at Sha Po is of later stage Late Neolithic type (2400– 1500 BCE)—hereafter referred to as ‘Later Neolithic’—although given the highly fragmented nature of much of the pottery, it is possible that some earlier stage Late Neolithic material may also be present but could not be isolated. In many areas of the backbeach Later Neolithic pottery was found in layers

Map 7: Neolithic–Bronze Age physical landscape. Source: CLSO. Yung Shue Wan, Lamma Island. 1:1000 Scale Topographic Map Sheets 14-NE-10D and 14-NE-15B. Hong Kong: CLSO, 1981a and 1981b respectively. Reproduced with permission of the Director of Lands. © The Government of the Hong Kong SAR. Licence No. 59/2015.

46   Piecing Together Sha Po

also containing Bronze Age material.57 While in some instances this mixing is perhaps indicative of real continuity between periods, it may also reflect the difficulty during excavation of discriminating between different loose, sandy cultural layers, which in the more intensively reworked areas of the backbeach were subject to earlier historical or more modern disturbance.58 In contrast, some better-preserved areas of the backbeach revealed discrete earlier Middle Neolithic59 and more extensive and well-defined Later Neolithic60 and Bronze Age61 cultural horizons. The latter, together with the fascinating contemporary site on the plateau above, are reserved for discussion in Chapter 5. Below, Neolithic Sha Po is explored in chronological order beginning with the first mobile bands of hunter-fisher-foragers in the earlier Middle Neolithic (4500–3500 BCE) and then moving on to discuss the evidence for coastal communities of the Later Neolithic (c. 2400–1500 BCE). A series of thematic headings are used to guide discussion of the material remains, for example, in terms of domestic, craft, and subsistence activities within the communities concerned. In the interests of consistency and clarity those same headings are also used, where possible, in Chapter 5, which then provides a basis for comparison of the lifeways of Later Neolithic and Bronze Age communities at Sha Po.

Sha Po’s first people: Earlier Middle Neolithic (4500–3500 BCE) The activities of Hong Kong’s first coastal fisher-hunter-forager bands can be identified by their distinctive red-painted fine pottery, fine-corded cooking wares, and associated stone tools, all of which were found in small quantities near the eastern edge of the central backbeach immediately south of Sha Po Old Village. In 2002 several sherds of red-painted pottery and fine-corded coarseware jar were found [22a:05], while further painted pottery was unearthed just to the south in 2009 [32:L3]. These discoveries may respectively mark the western and southern edges of a localised Middle Neolithic deposit first encountered in 1994 [12:L3], which produced further characteristic pottery, three general-purpose pebble tools, two adze rough-outs, and a coarse pottery bowl.62 The findings, although limited, illuminate a fairly typical cross-section of life at the time. The painted pottery was the ‘tableware’ of the day, used for drinking and consuming food, which was cooked over open fires using the coarseware jars and bowls. The pebble tools were used throughout prehistory and probably fulfilled multiple roles as hammers, coarse choppers, and general-purpose food-processing implements—hence their longevity. Many prehistoric sites (see below) produce sandstone and siltstone blocks with smooth, dished surfaces (whetstones) that were used to shape and sharpen stone adzes, spearheads, and arrowheads. Rough-outs of adzes and other forms of tools are common finds throughout Hong Kong prehistory and tell us that people sourced, transported, and perhaps traded such objects unfinished—probably to reduce the chance of breaking in transit what were valuable goods—and they were then finished by the end-user.63 Adzes—a key indicator of the South China coastal Neolithic—are essentially an asymmetrical axe, with one flat face and one rounded or stepped. They were designed thus specifically for hafting (mounting on a wooden handle) with the blade horizontal, not vertical like an axe. When in use the curved face would be on top while the flat one faced the user.64 Further elaboration on their significance is provided in our ‘Later Neolithic’ discussion below. Sha Po, like a number of smaller coastal sites of this period, has an insubstantial and patchy cultural deposit with no recognisable features, which gives a general impression of intermittent use by mobile boat-using people. One can speculate that such groups were probably small in number and kinbased—perhaps comprising two or three extended families—and maybe no more than, say, fifteen to twenty-five members. At the Tai Wan site, which is in the next bay south from Sha Po, more extensive

Figure 2:  Middle Neolithic stone tools: (a) and (b) pebble tools, (c) and (d) adze rough-outs [all 12:L3]

48   Piecing Together Sha Po

Figure 3:  Hafted adze reconstruction

excavation produced relatively abundant artefactual evidence of the period, plus the rare remains of postbuilt structures on clay-rich slope deposits above the backbeach.65 A third Lamma site at Lo So Shing also yielded earlier Middle Neolithic materials. At Sha Po, although excavation was more piecemeal than at Tai Wan, sufficient areas of the eastern (earlier) part of the backbeach have nevertheless been sampled to argue that it really was much less intensively used in the earlier Middle Neolithic and, with Lo So Shing, was perhaps one of several coastal sites within the territory of a group whose main camp on Lamma was at Tai Wan. Surprisingly, Sha Po, Lo So Shing, and Tai Wan all appear to have a long break in human activity between the earlier Middle Neolithic and the Later Neolithic—a gap of well over a millennium—which is rather puzzling. Later-stage Middle Neolithic activity was recorded on Lamma but only at the southernmost site of Sham Wan, whereas the earlier stage of the Late Neolithic—exemplified by the site of Yung Long overlooking Deep Bay—has yet to be identified on the island. We now move on to examine the large volumes of evidence for human activity at Sha Po during the Later Neolithic.

The Later Neolithic (2400–1500 BCE) Although we have clear evidence for the inhabitation of Hong Kong’s rich archipelagic landscape throughout prehistory, it was not until the second half of the third millennium BCE that we again find clear evidence for a human presence at Sha Po. Sha Po is actually one of a myriad of sites evidencing Later Neolithic activity, which is far more widespread than that in earlier periods. There is also a significant degree of continuity—perhaps continual occupation in many instances—on into the Bronze Age. Although there is a relative abundance of material available when compared to the Middle Neolithic, interpreting the lifeway of Sha Po’s Later Neolithic community similarly relies heavily upon the patterning of artefactual evidence—pottery and lithics—as there are few obvious features to guide our reconstructions of past human activity. We begin our discussion of Later Neolithic Sha Po by examining the general patterning of evidence across the backbeach.



Sha Po’s First People: Neolithic Fisher-Hunter-Foragers   49

General patterning of activity Later Neolithic activity at Sha Po was found in excavations all the way up the eastern side of the backbeach from close to its southern tip, north-east through the Old Village, and then following the curve of the supposed inland edge of the backbeach north to the foot of the New Village plateau. The northern part of the backbeach has avoided significant development and therefore remains relatively poorly understood; however, the one small trench excavated in the middle of that area in 1989 [Square ‘X’:11] produced Later Neolithic and Bronze Age materials, and thus provides a ‘bridge’ between the Old Village and plateau’s foot, where a few potsherds dated to the terminal Neolithic or Early Bronze Age were probably redeposited by erosion from the plateau above. The main focus of domestic, craft, and possible funerary activity was on the highest central-eastern area of the backbeach where, several millennia later, Sha Po Old Village was built. One can but wonder at the treasures encountered by villagers when they prepared the site for development around the turn of the nineteenth century, but our task now is to explore the discoveries made by archaeologists in more recent times, beginning with those associated with the quest for food. Food from forest, sea, and shore A major frustration of Hong Kong archaeology is that we know that early settlers must have encountered and exploited an incredibly rich environment full of food and other material resources, but due to very poor organic preservation we only ever find a fraction of the things they must have used.66 Unfortunately, most of the food remains came from ‘transitional’ layers spanning the Later Neolithic to Bronze Age, but we discuss those deposits here and leave the discussion of ‘pure’ Bronze Age midden remains for Chapter 5. In contrast to Sha Po’s evidence for small-scale middens of earlier historical date (see Chapter 6), there was only a scattering of shellfish, fish, and mammal remains in Sha Po’s Later Neolithic and transitional deposits, but some nonetheless of potential interest. The skull fragments and teeth of a young pig were found in 1989 towards the southern end of the backbeach [11:05] among ‘transitional phase’ pottery (i.e., latest Neolithic to Early Bronze Age). It is unclear whether this animal was wild or domesticated, but based on its immaturity the excavator presumed it was the latter.67 Further immature pig skull and jaw fragments, together with sea turtle and large deer teeth and long-bone fragments, were found in similar mixed deposits about 20 m to the north-east [34:506–507], while at the north end of the Old Village the jawbone of another pig was also recovered together with a few shells and fishbone fragments [20:03–04]. In addition, a scattering of shellfish remains was noted throughout the backbeach in prehistoric layers disturbed by later activity, but much of the material may reflect the activities of Qing dynasty villagers or earlier historical kiln workers rather than the prehistoric inhabitants. Overall, there is a striking contrast between the dearth of food remains observed, and the huge volumes of potsherds from coarse corded cooking pots. There is also a range of stone artefacts that can be related to the acquisition and processing of foods and it is to them that we turn next. Hunting, fishing, foraging, and food processing toolkits? Although multiple uses of stone artefacts are often possible—indeed likely—many commonly found types can sensibly be linked to subsistence activities. For example, one could easily interpret the three fragments of ground-stone68 points found at Sha Po—two probably spearheads [20:03; 34:507] and the other an arrowhead [22a:05]—as evidence for the warlike mentality of local people, but they were also eminently suited to the task of hunting a variety of prey animals such as wild boar and deer. Waisted pebbles, routinely interpreted as fishing net-weights, were somewhat surprisingly found only in very

50   Piecing Together Sha Po

Figure 4:  Later Neolithic artefacts relating to fishing, hunting, foraging, and food processing: (a) spearhead [34:507], (b) pounder [22a:04], (c) stone flake [21:03], (d) waisted pebble ‘net-weight’ [11:L7], (e) pebble tool [22a:04], (f ) anvil stone [11:L5]

small numbers on the Sha Po backbeach, and just one came from a Later Neolithic context [11:L7]. Around thirty ‘general-purpose’ pebble tools in a variety of pointed (pick) and edged (chopper-scraper) forms were found in two main concentrations in the south-eastern [11:5–7] and central-eastern part of the backbeach [22a]. Dislodging larger shellfish from rocks was maybe one use, but they were probably tools of much wider application and could, for example, also have been useful for basic disarticulation of animal carcasses and breaking of marrow-rich long-bones. In contrast, sharp-edged struck flakes of hard, fine-grained volcanic rocks such as dacite and rhyolite, which were found in small numbers across the backbeach, were the personal knives of their day and would make ideal butchery tools for the finer processing of fish, meat, and no doubt plant foods as well. Pebble grinders, pounders, and flat or disc-shaped anvil stones, while perhaps linked to stone-working, would also be highly effective for processing plants such as palms, nuts, roots, and tubers, as well as wild grass seeds such as rice.69



Sha Po’s First People: Neolithic Fisher-Hunter-Foragers   51

Storing, cooking, and consuming food Pottery, as ever, is our main source of evidence for human activity and thousands of sherds were recovered from Later Neolithic horizons up the eastern side of the backbeach. Coarse corded, round-bottomed cooking wares were by far the largest category of pottery fragments recovered. Surviving rims were thicksectioned with everted, triangular, and hooked forms typical of the period, which were robust and ideal for moving and carrying. Mixed in with the corded wares in some areas was a significant minority of coarse geometric cooking pots with everted rims, which might indicate continuity into the earlier Bronze Age. The fine wares were characteristically soft and chalky in nature and often survived only as worn fragments; however, sufficient pieces of a fine carinated jar with concentric circles were found to suggest it was near-identical to an example from Tai Wan.70 The generally poor preservation of Neolithic pottery may explain why there were no identifiable fragments of soft pottery bowls and cups of the types found occasionally at other sites.71 There may also be cultural factors at work as low-fired chalky wares were

Figure 5:  Later Neolithic domestic artefacts: (a) to (d) coarse corded/incised cooking pots: (a) and (b) [14:04], (c) [34:520], (d) [34:510; CSF8], (e) soft geometric storage jar [34:507], (f ) fire grate fragment [14:04], (g) and (h) potstands, respectively from [32:204; CSF9] and [34:507]

52   Piecing Together Sha Po

Plate 11:  Later Neolithic corded cooking pot

relatively fragile and ill-suited to a mobile, boat-based lifeway when compared with tough, robust coarse cooking pots. One possibility, therefore, is that bamboo or wooden cups and bowls may have been the preferred option when groups were on the move. No trace of in situ burning was noted on the backbeach, but occasional fragments of pot-stands and clay fire-grates—as found within fire-pits or hearths at the Yung Long site—were recovered together, in particular from the area immediately south of Sha Po Old Village [22a].72 The latter finds and sheer volume of coarse corded round-bottomed cooking pot fragments on the backbeach together suggest that food preparation and cooking must have occurred there. However, in prehistoric backbeach sites the loose sediments, long timescales involved, and post-depositional disturbance factors often permit the identification and zoning of activities only in the broadest periodic terms—e.g., Later Neolithic—even though we know, or at least strongly suspect, that what we are seeing is the sum of many discrete episodes of human behaviour. If winter to early summer was one of the main periods of fishing activity in the region, then one would imagine some form of temporary shelters would have been desirable to keep out the cold north wind while processing and cooking food or carrying out craft activities on the exposed backbeach. However, as is typical of backbeach sites, the loose sand preserved no trace of stake- or post-holes,73 but the former use and importance of wood can be deduced from other evidence, such as the stone tools used to cut and work it.



Sha Po’s First People: Neolithic Fisher-Hunter-Foragers   53

Have adze will travel: The importance of woodworking At Sha Po at least twelve complete adzes—two in the form of rough-outs—and several fragments were found in Later Neolithic deposits spanning the southern end of the backbeach [13:L4; 11:5–7], the centre-east activity focus [12:L2], and northern end of the Old Village [14–15: 04–05]. Rectangular, trapezoidal, and stepped forms were all represented and a wide range of sizes were noted (5 cm to 23 cm long), which tells us that woodworking was fundamental to our maritime people’s economy. The stepped adze was a Later Neolithic design innovation with a thicker, longer, and more steeply angled cutting face, which must have made it a more powerful tool in skilled hands. Outside of the very largest example, which may have served a specialised, perhaps ceremonial, purpose,74 the larger adzes would be suitable for harvesting stakes, felling mature trees, and shaping artefacts big and small. Some of the Sha Po adzes are also very small (c. 5 cm long) and can only have been useful in finer, more detailed work such as the trimming of arrow and spear shafts, shaping of paddles, adze shafts, and other small artefacts. No house remains were found in Neolithic Sha Po, but post-holes elsewhere tell us that timber buildings were in use from the Middle Neolithic onwards. Similarly, while no trace of prehistoric boats has

Figure 6:  Later Neolithic stone adzes for wood cutting and shaping: (a) adze rough-out [34:510; CSF], (b) stepped adze [14:04; CSF], trapezoidal adzes, (c) [22a:01], and (d) [14:05; CSF]

54   Piecing Together Sha Po

been found in Hong Kong, the patterning of sites across the Hong Kong archipelago indicates that dug-out canoes—probably without outriggers but maybe sometimes used in pairs—were the key to a mobile and diversified coastal-maritime lifeway built around hunting, fishing, and foraging.75 While such gaps in the ‘organic record’ are a frustration, the survival of finished adzes and other stone artefacts is also complemented by a fascinating range of rough-outs, part-finished, and broken objects, together with tools relating to their manufacture.

Tool, weapon, and ornament manufacture In contrast to the plateau’s exciting remains of a Bronze Age stone ornament workshop (see Chapter 5), the Later Neolithic backbeach seems to evidence a pattern of less specialised stone-working activity.76 The general distribution of adzes is reflected in that of fine sandstone or siltstone whetstones (sharpening and polishing stones), some with dished surfaces seemingly reflecting the shape of adzes, others perhaps used for shaping and polishing rings, knives, and points. To the south of the Old Village, though, there seems to have been a particular focus of hard grey slate-schist working,77 as evidenced by several whetstones, large knife and adze rough-outs (see, for example, Figure 6a), several pieces of knife debitage (waste flakes), broken arm-rings (bangles) and cores, and a number of general-purpose pebble tools. A stone bead and an unusual finely polished tube, which was drilled from both ends but had broken during manufacture, were also found here. Although disturbed by recent activity, Later Neolithic deposits at the northern end of the Old Village produced interesting finds in the form of a rotary ring-polishing stone, which was below a layer containing a quartz ring fragment and possible quartz rough-outs.78 Several other quartz disc rough-outs were also noted near the landward face of the backbeach around 30 m to the south-east. It is interesting to note that there was only one fine sandstone saw blade found on the backbeach. Such saws are strongly associated with split earring manufacture and were a relatively common find in Bronze Age deposits on the plateau and at its foot. Spinning a yarn As mentioned above, by the Late Neolithic bark-cloth had been rendered obsolete by a spun fibre technology which, with the probable dearth of suitable animal hair, most likely used plant fibres such as those from hemp.79 Fired clay spindle whorls—perforated discs—which when mounted on long sticks—or spindles—provided the momentum for hand-spinning of fibres, are regular discoveries in Later Neolithic sites. Four were unearthed on the Sha Po backbeach, of which three were in undisturbed contexts [20:03; 11:L6–7; 12:L2], while the fourth was found with Later Neolithic materials but in a disturbed context [14:5]. Modern archaeologists are careful to avoid gender stereotyping when interpreting artefacts, but well-preserved burials from the Hedang site and others in the Pearl River Delta have shown that female burials were often accompanied by spindle whorls.80 It is therefore interesting to note that one of the above spindle whorls [20:03] was found within an irregular patch of soil staining that was thought by the excavators to be the possible remains of a grave.81 In addition to their gender associations, spindle whorls also tend to be ‘naturally’ associated with the weaving of fabric—which by the local Bronze Age was definitely occurring82—but there were many possible uses beyond clothing including containers or even sails. Moreover, spun fibre might also be used for fishing line and nets, or when several strands were twisted together to make cordage, may also have served as rigging for sailing canoes, anchor ropes, or carrying handles for pots—the list is endless.83 The versatility of spun fibre technology is reflected in its longevity and breadth of application in the modern world, but it must have been a truly life-changing innovation for Neolithic communities on the South China coast.

Figure 7:  Later Neolithic tool, weapon, and ornament manufacture: (a) whetstone-polisher [23:03], (b) knife roughout [22a:04], (c) stone saw [36:609], (d) ring core [34:507], (e) quartz ring fragment [20:03], (f ) rotary ring polishing stone [15:09; CSF25]

Figure 8:  Later Neolithic spindle whorls: (a) stone spindle whorl [20:03], (b) pottery spindle whorl [14:05]

56   Piecing Together Sha Po

Discussion Archaeological visibility or changing human behaviour? In the first half of the chapter, we presented the evidence for the activities of mobile maritime-coastal groups spanning roughly 3,500 years of the Neolithic and reaching to all corners of the Hong Kong archipelago. We then showed that in contrast to that general patterning—and like many other backbeach sites in the region—Sha Po has a less continuous story of human occupation, which in this instance involves an apparent hiatus in activity of almost a thousand years. The question then is whether that apparent gap in human activity is real (behavioural) or illusory (due to poor archaeological visibility). Sha Po has been sampled more widely than many backbeach sites and no definite later Middle Neolithic or earlier Late Neolithic remains have been found. It is possible, of course, that if activity in those periods was focused in the north-central part of the backbeach, where virtually no testing has occurred, the remains might yet await discovery. A case in point being the localised earlier Middle Neolithic deposit at Sha Po, which with a different pattern of development and testing could itself easily have been missed. Moreover, given the highly fragmented and worn condition of many prehistoric potsherds at Sha Po, it is also possible that occasional pieces of earlier Late Neolithic pottery might have been overlooked. What we can say with some certainty is that Later Neolithic Sha Po was far busier than at any earlier stage, and became busier still moving into the Bronze Age. Here we need to remember that the Sha Po Later Neolithic remains discussed above span a period of around a thousand years, or say forty to fifty generations. As such, the volumes of pottery, stone tools, and, in particular, shellfish and other food remains suggest a very low intensity, probably intermittent, use of the backbeach by small-scale, highly mobile groups.

Socio-economic considerations One of the most striking aspects of Hong Kong’s prehistoric backbeach sites is just how few shell middens there are; indeed, of the few identified, most are not really worthy of that name, so limited is the scale of the deposit. At East and Southeast Asian Neolithic residential sites such as Kidosaku in Japan and Nagsabaran in the Philippines, where shellfish were targeted as a regular food resource, one sees as a result the formation of massive shell-middens, which at the latter site created a prominent landscape feature.84 Similarly, the Neolithic ‘shell-mound’ sites of the Pearl River Delta, such as Hedang, Jinlansi, and Luoshanzui,85 earned their label from the low hills formed by deep deposits of shells which, incidentally, also preserved all manner of other organic remains. Given the clear cultural connections between groups moving around the coast and islands and those dwelling in the Delta, we can say with some confidence that if shellfish were such a desirable food resource in the Delta, the same probably applied across the larger socio-economic region. Shellfish, it seems, were just not really targeted by people visiting the coastal areas and islands of Hong Kong.86 Which then raises the questions, why were they going there and what were they doing? Here we should remember an old archaeological adage: absence of evidence is not necessarily evidence of absence. In other words, the absence of subsistence remains in prehistoric Hong Kong should not be taken to indicate that subsistence was not the main reason people were there, in all probability it was. Poor preservation of organic materials is probably a key to our understanding here. The relatively impressive survivals of organic remains at Sham Wan and Po Yue Wan occurred due to the presence of what were actually quite modest deposits of shells.87 Our subsistence data in Hong Kong are therefore massively biased towards sites where shells were collected, when in reality it was those very sites that were unusual



Sha Po’s First People: Neolithic Fisher-Hunter-Foragers   57

because such activity occurred. Based on the surviving evidence, shellfish were not the reason people were there; fish, marine and terrestrial mammals, and reptiles almost certainly were. And depending upon the season, we might envisage the targeting of different combinations of those resources in different parts of the study region. Many coastal sites may therefore originally have contained the bones of a variety of marine and terrestrial animals, but the majority are now zooarchaeologically blank due to a lack of bone-preserving shells. Here, we should emphasise how recent research has demonstrated that shellfish, however large their midden accumulations might be, were likely to be ‘a crisis or supplementary resource’88 beyond the starchy plant and protein-rich vertebrate mainstays of coastal prehistoric subsistence. Therefore, in general, during the Later Neolithic we would argue that many Hong Kong coastal sites— in particular those like Sha Po on smaller backbeach sites on islands—were probably used intermittently and for varying durations of residence. Probably they were used as bases for inshore and pelagic fishing, but also as locations from which terrestrial and marine reptiles and mammals were undoubtedly hunted. In addition, while scientific proof of the Xincun, Taishan variety is still lacking in Hong Kong, we nevertheless strongly suspect that, as an essential source of carbohydrate, starchy plants were also collected and processed at Sha Po and other contemporary sites throughout the region. The people visiting Sha Po were probably organised in kin-based, extended family groups, perhaps just two or three canoe loads in number, but they might also have identified themselves as part of larger communities with more permanent bases, for example, as seen at Tai Wan, Ha Pak Nai, perhaps Sha Ha, and at various shell mound sites in the Delta. In contrast to the wider evidence for subsistence continuity from the Middle Neolithic, our Later Neolithic community at Sha Po had adopted spun fibre technology and geometric pottery, could make sophisticated stone tools and weapons, and had perfected the manufacture of a wider and more refined range of personal ornaments. These are indicators, we would argue, of a growing interest in competition and display, which are both facets of increasing social complexity. Some changes were perhaps local innovations, but there was no doubt also influence through contact, trade, and exchange—either directly or more likely through intermediaries—with more complex agropastoralist societies to the north. In Chapter 5, we explore how the intensification of such contacts gave local communities access to new technologies, raw materials, objects, and ideas, which contributed significantly to the transformation of Bronze Age society.

Notes 1. Although we have no physical evidence for boats, the fact that so many sites are around the coast and on islands large and small logically suggests that from the earliest times prehistoric people of the Hong Kong–Pearl River Delta routinely used them. Moreover, we also have evidence for open-ocean fishing of large pelagic fish species and dolphins. While simple dug-out canoes would have sufficed for coastal, estuarine, and riverine travel, for longer journeys over the open sea we would suggest that the increased stability of double dug-out canoes may have been preferred. General ideas on probable lack of outrigger use and possible use of double dug-out canoes as an alternative were provided by Stephen Davies (personal communication). See also note 73 in this chapter for wider clues regarding boat use by prehistoric peoples of the South China Sea, Indian Ocean, and Western Pacific. 2. While our Sha Po story and that of many other local sites begin in the Middle Neolithic (around 4500 BCE), there are a few sites of potentially earlier date in the region. Seemingly, the earliest site of all is a large stone tool quarry and workshop found at Wong Tei Tung, Sai Kung, which produced by far the most impressive chipped stone tool assemblage yet found in Hong Kong. Initial scientific dating of sediments gave dates as early as the Upper Palaeolithic, far earlier than any other local site (S. S. Zhang and Ng Wai-hung, eds., The Important Discovery of Hong Kong Archaeology: Wong Tei Tung Palaeolithic Site [Hong Kong: China Review Academic Publishers], 2006). However, later retesting on behalf of the AMO by labs at Hong Kong and Oxford universities produced earliest dates of around 5700 BCE, which is still the earliest date for any prehistoric site in Hong Kong. Retesting report: Tracey Lie-Dan Lu, ‘Report on the Date of the Wong Tei Tung Archaeological Assemblage’ (unpublished AMO Archaeological Research Report, 2007), at http://www.lcsd.gov.hk/CE/Museum/Monument/form/Report on the Date of the Wong Tei Tung Archaeological Assemblage. pdf (accessed 11 February 2014).

58   Piecing Together Sha Po 3. The term ‘Neolithic’ in its classical sense refers to the period when farming began and is associated in China with villages and cemeteries, the domestication of wild cereals such as rice and millet and raising of animals such as dogs, pigs, and chickens. It is also a period materially marked by the use of ground-stone tools—rather than chipped stone technology—and the widespread use of pottery. The Hong Kong coastal Neolithic evidences ground-stone tools and pottery but has as yet only debatable evidence for rice cultivation and animal husbandry. 4. Peter L. Drewett, Neolithic Sha Lo Wan, Antiquities and Monuments Office Occasional Paper No. 2 (London: Archetype, 1995), 51. 5. Ibid. 6. Zhu Feisu, Peng Ruce, and Liu Chengde, ‘On Geometric Pottery from Shixia, Maba’, Wenwu Jikan III (1981): 225–33. 7. Qiao Xiaoqin, ‘The Maritime Expansion of Yue Culture and the Related Archaeological Evidences’, in Collected Essays on the Culture of the Ancient Yue People in South China, ed. Chau Hing-wah (Hong Kong: Urban Council, 1993), 34–39; Tang Chung, ‘Notes on Geomorphology of the Tai Wan Culture in the Circum-Pearl River Delta’, Tropical Geography 17 (2) (1997b): 179–83; Tang Chung, ‘A Study of Maritime Diffusion of Prehistorical Mongoloid’, Southeast Culture (11) (2000): 6–22. 8. Some indications of prehistoric manipulation of certain local plants were found at Sha Ha (AMO 2005), but the first convincing evidence for the establishment of rice-farming communities in Hong Kong still dates to the Northern Song. 9. Chau Hing-wah, Collected Essays on the Culture of the Ancient Yue People in South China (1993: 40–55); William Meacham, The Archaeology of Hong Kong (Hong Kong: Hong Kong University Press, 2009), 86–109; Shang Zhitan, Collected Essays on Hong Kong Archaeology (Beijing: Wenwu Publishing, 2000), 1–53. 10. Due mainly to natural siltation processes, which were active since the coastline stabilised around 5000–4000 BCE, but accelerated following the late prehistoric emergence of farming in inland Guangdong, such mainland coastal areas were in some instances quite different back in the Neolithic. The modern coastline of Deep Bay, for example, is perhaps 1 km west of its line in the Neolithic, when shallow coastal waters surrounded a number of small islands, which silted up and are now landlocked. Of course, in the last one hundred years many coastal areas of Hong Kong have also seen massive change due to reclamation and fill, such that many ancient coastlines are today far inland. 11. In this research, we follow the general characteristics and dating of the major phases of Hong Kong prehistory as set out in the paper that introduced a Hong Kong special section in China’s premier archaeological journal Kaogu (AMO, ‘Recent Archaeological Discoveries and Studies in Hong Kong’, Kaogu 6 [2007]: 3–9). 12. Sites include: Tai Wan, Lamma Island: CUHK 1996 excavation discussed in Shang, Collected Essays, 230–31; Chung Hom Wan, Hong Kong Island: Solomon M. Bard, ‘Chung Hom Wan’, JHKAS VI (1975): 9–25; Yung Long, Tuen Mun: William Meacham, ‘Middle and Late Neolithic at “Yung Long South”’, in Conference on Archaeology in Southeast Asia, ed. C. T. Yeung and W. L. Li (Hong Kong: University Museum and Art Gallery, the University of Hong Kong, 1995), 467–78. NB: There is an ‘Early Neolithic’ in inland Guangdong and elsewhere, but the term is not yet applicable to the Hong Kong–Pearl River Delta region. 13. Kevin T. W. Sun, ‘Research on the Prehistoric Remains in the Pearl River Delta’ (MA diss., Peking University, 2002), 11. Ding are a common feature of Neolithic assemblages in northern Guangdong and the Middle Yangzi but very rare in Hong Kong where potstands and round-bottomed jars seem to have fulfilled the same role. 14. Summary based on: Sun, ‘Pearl River Delta’, 12. Sites include: Sham Wan: Meacham, Sham Wan; Sha Ha: AMO, Sha Ha, Sai Kung; Kwo Lo Wan and Fu Tei Wan: William Meacham, ed., Archaeological investigations on Chek Lap Kok Island, Hong Kong Archaeological Society Journal Monograph IV (Hong Kong: Hong Kong Archaeological Society, 1994), 23–44 and 129–54; Shek Pik: Au Ka-fat and Tang Chung, ‘A Preliminary Excavation Report on Tung Wan Beach Site in Hong Kong’, in Treatises in the Commemoration of the 30th Anniversary of the Discovery of the Maba Human Cranium, ed. Guangdong Provincial Museum and Museum of the Qujiang County (Beijing: Cultural Relics Publishing House, 1988), 208–18. 15. Hac Sa Wan, Macau: William Meacham, ‘Hac Sa Wan’, JHKAS XI (1986): 106–8. 16. Pang Quanmin et al., ‘Preliminary Excavation Report on the Xiantouling Sand Dune Site in Dapeng, Shenzhen City’, in Archaeological Discovery and Research of Shenzhen, ed. Shenzhen Museum (Beijing: Wenwu Publishing, 1994), 28–41. Li Hairong and Li Junxiaong, ‘Xiantouling Site in Shenzhen’, in Major Archaeological Discoveries in China 2006 (Beijing: Wenwu Publishing, 2006), 34–37; Li Hairong and Li Junxiong, ‘A Late Neolithic Site in Xiantouling, Shenzhen, Guangdong’, Kaogu 7 (2007): 9–15; Shenzhen Municipal Institute of Archaeology, ed., Xiantouling Site in Shenzhen: Report on the Excavation in 2006 (Beijing: Cultural Relics Press, 2013). 17. Summary based on: Sun, ‘Pearl River Delta’, 13; sites include: Yung Long: Meacham, ‘Yung Long South’, 445–66; Sha Ha: AMO, Sha Ha, Sai Kung; Sha Lo Wan: Drewett, Sha Lo Wan. 18. Sites include: Tung Wan Tsai North, Ma Wan: Chau Hing-wah, Wu Yaoli, and Li Langlin, ‘A Preliminary Report on the Excavation of a Prehistoric Site in Tung Wan Tsai North, Ma Wan, Hong Kong’, Kaogu 6 (1999): 1–17; Sham Wan: Meacham, Sham Wan; Sha Ha, Sai Kung: AMO, Sha Ha, Sai Kung; Po Yue Wan, Cheung Chau: James R. Crawford, ‘Po Yue Wan’, JHKAS XI (1986): 64–79; Sha Chau off Tuen Mun: R. J. Frost, ‘Sha Chau’, JHKAS VI (1975): 37–50. 19. Based on archaeological findings, mainland scholars now in general agree that the Bronze Age in Guangdong was divided into three major phases: the first evidenced a range of characteristics in different regions, followed by the more widely distributed ‘double-F pottery phase’ (Hong Kong’s Bronze Age), and then finally the ‘asterisk pottery phase’. However, no consensus has been reached regarding the date ranges for these phases (You Xiaolei, ‘The Preliminary Study on Henglingshan Cemetery and the Relative Problems’ [MA diss., Shandong University, 2011]). 20. Meacham, ‘Sha Po Tsuen’, 44–46. 21. For example, see Chau, Ancient Yue, Plates 45, 62, 102, and 138.



Sha Po’s First People: Neolithic Fisher-Hunter-Foragers   59

22. At Yung Long, many incised pot-stands were found in small clay-lined fire-pits in close association with clay fire-grate fragments and charcoal; however, such clear contextual associations are the exception rather than the rule. 23. The term ‘oyster pick’ was first used in the 1961 report on the Neolithic shell midden site in Dongxing, where the authors observed the modern use of oyster picking tools near the site (Mo Zhi and Chen Zhiliang, ‘A Neolithic Shell Midden Site in Dongxing, Guangdong’, Kaogu 12 [1961]: 644–49, 688). 24. For example, a Late Neolithic workshop at Sha Ha (AMO 2005) produced adzes and pebble picks, while the large-scale manufacture of stone ornaments was noted at Pak Mong (Wong Wan-cheung, ‘Discussion of the Slit Ring Manufacturing Technology in the Pearl River Delta: A Case Study of Pak Mong Site, Hong Kong’ [MA Diss., Xiamen University, 2009]). 25. Tang Chung, Forget Me Not – The Historical Roots of Hong Kong (II): Origins of Clothes – Barkcloth (Hong Kong: Centre for Chinese Archaeology and Art, 2011). 26. For example, see examples in Chau, Ancient Yue, 205–7. 27. The Yung Long site in particular produced a large number of this date. 28. William Chan, ‘Identification and Analysis of Excavated Fish Remains’, in Sham Wan, Lamma Island: An Archaeological Site Study, Journal Monograph III, ed. William Meacham (Hong Kong: HKAS, 1978): 248–57. 29. For example, Sham Wan, Lamma Island: Meacham, Sham Wan; Po Yue Wan, Cheung Chau: Crawford, ‘Po Yue Wan’. Coastal middens (ancient rubbish dumps) often contain shells, which are alkaline, that can neutralise Hong Kong’s acid soil conditions thus permitting the survival of any other organic materials such fish, reptile, and mammal bones. 30. Chan, ‘Fish Remains’, 256. A note of caution: both species also have particularly robust crania (skulls), which no doubt survive far better than other species, or indeed other more fragile elements of the same fish. The apparent importance of the two species may therefore be exaggerated. Either way, they were caught in significant numbers and made a good contribution to diet. 31. Chau, Ancient Yue, 219; Crawford, ‘Po Yue Wan’, 69–72. We are reminded of the paucity of our organic survivals in Hong Kong by well-preserved Chinese sites such as Hemudu (5000–3000 BCE), which produced evidence for the exploitation of a vast range of domesticated and wild animal and plant food resources, and abundant wooden and bone artefacts that far outnumbered those made of stone (Kwang-chih Chang, The Archaeology of Ancient China [New Haven and London: Yale University Press, 1986], 208–16). 32. Chau et al., ‘Tung Wan Tsai North’, 12–13. 33. The superb preservation created by the Hedang shell mound permitted the survival of a wide range of organic remains including seventy-seven human burials, ‘many bones of mammals and aquatic animals’ plus a ‘vigorous bone industry, which produced hairpins, needles, awls and weaving shuttles’. Charles F. W. Higham, The Bronze Age of Southeast Asia (Hong Kong: Cambridge University Press, 1996), 84; Guangdong Provincial Museum and Foshan Museum, eds., Hedang Site in Foshan: Excavation Report from 1977 Winter to 1978 Summer (Guangdong: Guangdong People’s Publishing House, 2006), 105–8, 135–36. 34. Phytoliths are microscopic silica bodies found in plants, which can be diagnostic to species and survive for millennia in the ground even in Hong Kong’s acidic soil conditions. They can therefore reveal clues regarding past vegetation surrounding archaeological sites and the types of plants used or eaten there by people. 35. Xincun, Taishan: X. Yang et al., ‘Sago-Type Palms Were an Important Plant Food Prior to Rice in Southern Subtropical China’, PLoS One 8 (5): e63148, at http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0063148 (accessed 21 August 2014). 36. Tracey Lie-Dan Lu, Z. Zhao, and Z. Zheng, ‘The Prehistoric and Historic Environments, Vegetations and Subsistence Strategies at Sha Ha, Sai Kung’, in The Ancient Culture of Hong Kong: Archaeological Discoveries in Sha Ha, Sai Kung, ed. AMO (Hong Kong: AMO, 2005), 57–64. 37. Z. H. Liu et al., ‘Research on the Excavated Ancient Cultivated Rice Species Discovered in Guangdong’, Guangdong Agricultural Sciences (9) (2010): 52–64. 38. According to Shang (Collected Essays, 230–31), two contemporary quite well-preserved Middle Neolithic house foundations (F1 in Area VII and F2 in Area VIII) were identified in the 1996 excavation carried out by the CUHK team. F1 was oval-shaped, about 35 m2 in size; F2 was rectangular in size, about 48.2 m2 in size. Large quantities of pottery and stone finds were recorded inside and around the two residential features; Tang Chung, ‘Archaeology and the Origins of Hong Kong’, New Asia Life Monthly 32 (6) (2005): 6–8. 39. Ha Pak Nai: Au Ka-fat and Mo Zhi, ‘Working Paper on the Ng Ka Yuen, Ha Pak Nai Sandbar Site Excavation’, JHKAS XIV (1998): 5–40. Building F1 measured 12.5 m N-S by 8.5 m E-W, which is very large compared with any other proposed prehistoric structures in Hong Kong, and had a west-facing porched doorway looking out to sea (and to benefit from the afternoon sun and avoid cold easterly and north-easterly winds). 40. Sha Lo Wan: Drewett, Sha Lo Wan; Pa Tau Ku: Tang Chung, ‘Preliminary Report and Submission of the Pa Tau Kwu Site Excavation’ (unpublished excavation report, 1992); Tang Chung, ‘Investigation Report on the 1994 Pa Tau Kwu Areas I and II Excavation’ (unpublished report, 1994); Tang Chung, ‘New Discoveries of Hong Kong’s Ancient History’, Historical Research 3 (1997a): 32–52. 41. Bard, ‘Chung Hom Wan’. 42. Primary burials are where a person was interred and then remained undisturbed, whereas secondary burials involve the exhumation of primary burials and their reburial, often with the bones stacked together, in much smaller pits. Interestingly, the practice of secondary burial still persists among indigenous rural communities such as those on Lamma Island. 43. Human burial seems the most likely option here, although some of the ‘burials’ previously identified from ‘grave goods’ alone might be objects placed as offerings to ‘earth gods’ or ancestors, which would fall under the generic archaeological category of ‘structured

60   Piecing Together Sha Po deposition’ (i.e., things buried on purpose for reasons unknown, but usually assumed to have ‘ritual’ meaning). Ritual, by the way, is a catch-all word used by archaeologists to label all features and finds they cannot explain by other means! 44. F. P. Lisowski, ‘Human Remains and Burials at Sham Wan’, in Sham Wan, Lamma Island: An Archaeological Site Study, ed. William Meacham (Hong Kong: Hong Kong Archaeological Society, 1978), 232–46. 45. Zhang Chi and Hung Hsiao-Chun, ‘The Neolithic of Southern China: Origin, Development and Dispersal’, Asian Perspectives 47 (2) (2008): 299–329. 46. Here, it must be acknowledged that these Ma Wan Phase II burials are considered by some scholars of Guangdong archaeology to be ‘Early Bronze Age’ in date, e.g., Bu Gong, ‘Archaeological Observations on the Development of Lingnan Culture’, Journal of Historical Anthropology 3 (2) 2005: 11. 47. Chau et al., ‘Tung Wan Tsai North’. 48. For example, sites at Hedang, Youyugang, Jinglanshi, and Xiankezhou (Sun, ‘Pearl River Delta’, 42). 49. Charles Higham and Tracey Lie-Dan Lu, ‘The Origins and Dispersal of Rice Cultivation’, Antiquity 72 (1998): 874. The practice may have been a (deeply unpleasant) rite of passage into adulthood. 50. Interestingly, the shape of the teeth tested actually fell somewhere between the forms labelled as ‘sinadonty’ (typical of NE Asian populations) and ‘sundadonty’ (typical of SE Asian and Pacific groups). 51. Both types of new object were shaped from an unusual soft, cream-yellow coloured rock—white on fracture—which analysis conducted at Hanoi in the 1930s suggested was a kind of amphibole rock known as tremolite. Interestingly, Finn records seeing many examples of rings in this same material in a private collection derived from Thanh Hoá in the Red River Delta area of Vietnam (Finn, ‘Lamma Island, Part VIII’, 52–53). The absence of cores from such bangles on Hong Kong sites may indicate the importation of finished objects, while the presence of many similar bangles from sites in the Red River Delta area of north-east Vietnam suggests one possible source. Such bangles seem to have continued in use into the Bronze Age as evidenced by an example found in a grave of that date at Shek Pik (see Chapter 5). 52. Shang Zhitan and Chen Shilong, ‘Characteristics and Problems of the Prehistoric Dune Sites near the Mouth of the Zhujiang River’, Wenwu 11 (1990): 51; Steven Wai Hung Ng, ‘The Spatial Pattern of Prehistoric Sites around Estuary of Pearl River’, JHKAS XIV (1998): 60–61. 53. Au Ka-fat and Mo Zhi, ‘Ha Pak Nai’. 54. Chau et al., ‘Tung Wan Tsai North’, 15. 55. Tang Chung, Shang Zhitan, and Wong Wan-cheung, ‘Preliminary Report on the Excavation of Pak Mong Site on Lantau Island’, Kaogu 6 (1997): 57. 56. Indeed, the assumption that the presence of children on site argues against a seasonal use of the site is perhaps questionable when the coastal societies involved most likely made habitual use of boats and pursued a highly diversified subsistence strategy to which men, women, and children no doubt contributed. Moreover, women and children did not have to walk—as might be the case for inland hunter-gatherers employing all-male groups for seasonal hunting expeditions—so, if boat-using groups did move to different sites at different times of the year, they could easily have done so as extended kin-based groups, setting up camp in different sites, potentially for months at a stretch, in order to exploit particular resources through hunting, fishing, and foraging. 57. Spry, ‘Sha Po Tsuen’, 7–28; HKAS 1994 Square J Layer 2 (information from unpublished archive); Archaeological Assessments Ltd. (AAL), ‘Sha Po Tsuen Rescue Excavation’ (unpublished excavation report 2003); AAL, ‘DSD Contract No. DC/2007/18 Yung Shue Wan and Sok Kwu Wan Village Sewerage, Stage 1 Works: Archaeological Rescue Excavations at Sha Po Tsuen’ (unpublished excavation report, 2011a); AAL, ‘DSD Contract No. DC/2007/18 Yung Shue Wan and Sok Kwu Wan Village Sewerage, Stage 1 Works: Archaeological Watching Brief at Sha Po Tsuen’ (unpublished watching brief report, 2011b). 58. Due to their loose sandy sediments and complex site formation processes, which comprise both human reworking and natural processes, backbeach sites are notoriously difficult to excavate in terms of clear stratigraphic units. Boundaries between strata are invariably ‘fuzzy’, feature edges are extremely difficult to define, and dense, smooth objects such as historical kiln furniture often occur in prehistoric horizons due to post-depositional downward ‘migration’. 59. Middle Neolithic: HKAS (1994) Trench ‘J’, Layer 3 (information from unpublished archive). 60. Later Neolithic: Meacham, ‘Sha Po Tsuen’, 33–54 – Square CC, Layers 6 and 7; AAL, ‘Sha Po Tsuen’, some areas. 61. Bronze Age: Meacham, ‘Sha Po Tsuen’, Square CC Layer 4; HKAS 1994 Square K Layer 3; AMO, ‘Sha Po Old Village’ (unpublished excavation report, 2001). 62. HKAS 1994 Trench ‘J’, Layer 3; AAL, ‘Sha Po Tsuen’. A ‘rough-out’ is a roughly shaped stone tool, which was seemingly the preferred condition that such objects were transported from the stone quarry site to the final user. 63. In the interests of efficiency (i.e., not transporting more fragile finished objects but also not moving lots of excess stone), stone blocks would be roughly shaped at source and moved on to end-users. Insufficient study of the provenance, distribution, and consumption of stone tools has thus far been conducted to allow a meaningful assessment of their socio-economic significance in prehistoric Hong Kong. We therefore do not yet know whether stone was mostly sourced by the end-users themselves, was traded down-the-line between groups, or involved quarrying specialists and middlemen. 64. The normal mode of use on large pieces of wood, such as a house timber or dug-out canoe, would involve the user standing with one foot to either side of the log, bending forward and using a swing similar to that used for chopping firewood with an axe. Removed chips of wood would fly behind the user between their legs, and the rounded upper surface would leave a series of ‘scooped’ marks on the worked wood (observed on timber from waterlogged sites). 65. Shang, Collected Essays, 230–31, 274, 285.



Sha Po’s First People: Neolithic Fisher-Hunter-Foragers   61

66. Lu, ‘Natural Resources’, 36–45. Lu quite rightly argued that the problem of differential preservation in favour of non-organic remains rather than organic materials had also been compounded by the use of inadequate recording and sampling in many previous excavations. This will hopefully change soon. 67. Meacham, ‘Sha Po Tsuen’, 47, 48, and 53. Young pigs do not necessarily indicate that they were domesticates raised on site. Young wild boar would no doubt be targeted by hunters as they would be easier to hunt and kill than their larger, more wary, and much more dangerous parents. Full zooarchaeological study of prehistoric animal remains from Hong Kong sites is desperately needed. 68. Grey fine-grained rock such as shale, slate, or schist was invariably used for such points. Useful analyses of ground-stone tool-weapon raw material types and sources are provided in the Sham Wan and Sha Ha reports. 69. Here again, the important evidence for the prehistoric exploitation of starchy plants from the Xincun site in Taishan, Guangdong is particularly notable. 70. See Chau, Ancient Yue (Plate 130), 210. 71. For example, as illustrated in Chau (Ancient Yue, Plates 129, 134, and 135) stem-cups from Lo So Shing, Lamma and Xiantouling, Shenzhen, and bowl from Lung Kwu Tan, Tuen Mun. Some have suggested that there is a higher proportion of consumption wares on the larger ‘base camp’ shell mound sites in the Delta, which was taken as an indication of more permanent settlement there, but in his 1998 paper Steven Wai-hung Ng argues that the data are not particularly clear-cut. 72. The Yung Long fire-pits were only found on that site: whether this was an accident of preservation or reflects some specialised local activity is unclear, but the more widespread presence of pot-stands and fire-grates at many sites does, we feel, support their connection with an equally commonplace activity, and cooking seems most likely. That said, a recent publication on salt-working in China has speculated that the Yung Long evidence may reflect a special focus there on the production of salt from seawater (Li Shui Cheng, ‘Ten Years of Salt Archaeology in China’ , Kaogu Xue Yanjiu 9 [2012]: 377–78). 73. But see the dramatic evidence in Chapter 5 for a timber-built settlement of some sort on the Bronze Age plateau. 74. The 23 cm long trapezoidal adze rough-out was made from a hard slate or schist, but would probably be of little practical use for wood-working. It may therefore have been used as a spade or some other form of digging tool, or alternatively might have served some ceremonial purpose or other. See Catalogue of Selected Finds No. 34. 75. The spread of Neolithic boat-using peoples is evidenced elsewhere in the region. Sites at Andarayan and Nagsabaran in northern Luzon respectively had domesticated rice and pigs with Taiwanese nephrite (jade) objects and pottery styles at around 2000 BCE (Philip J. Piper et al., ‘A 4000-Year-Old Introduction of Domestic Pigs into the Philippine Archipelago: Implications for Understanding Routes of Human Migration through Island Southeast Asia and Wallacea’, Antiquity 83 (2009): 687–95. Also, rice cultivators arrived with Philippines-style pottery in the far-flung Marianas by 1500 BCE; an ocean voyage of 2300 km (Hsiao-chun Hung et al., ‘The First Settlement of Remote Oceania: The Philippines to the Marianas’, Antiquity 85 [2011]: 909–26). See also Dorian Fuller’s fascinating discussion of early boat technology in Asia (Dorian Q. Fuller et al., ‘Across the Indian Ocean: The Prehistoric Movement of Plants and Animals’, Antiquity 85 [2011]: 544–58). 76. By ‘less specialised’, we mean remains evidencing relatively low-intensity activities seemingly carried out as part of the daily routine of food preparation, consumption, rubbish disposal, and tool and weapon finishing and sharpening. Specialised working produces denser scatters of materials diagnostic of a particular type of activity, which occur in zones dedicated to that activity (i.e., in stone workshop areas often dedicated to the production of certain types of tools, weapons or ornaments). 77. ‘Workshop’ covers south end of area 22a, west of area 32, and north-east end of area 34. See Catalogue of Selected Finds Nos. 30, 33a, and 34. 78. AMO, Sha Po Old Village. See Catalogue of Selected Finds No. 25. 79. See Chapter 6 for discussion of use of hemp in Bronze Age woven fabric as was found at Kwo Lo Wan, Chek Lap Kok preserved in metallic corrosion products (Chau, Ancient Yue, 231). Another textile fragment, in this case Han, was similarly preserved by metallic corrosion at So Kwun Wat. 80. J. J. Yang, ‘Issues on Burial Customs of Shang and Zhou Dynasties in the Lingnan Area’, Jianghan Kaogu 2 (2010): 71–81. Evidence from the Central Plains shows that spindle whorls were in most cases female grave goods, and possibly of ‘older female’ (aged over 40). However, there are also examples of spindle whorls being buried with males (D. Wang, ‘Analysis of Spindle-Whorl: From Neolithic Age to Bronze Age in Shandong Province’ [unpublished MA thesis, Department of Archaeology and Museology, Shandong University, 2009], 83). 81. AAL, ‘Sha Po Tsuen’, 35. In the loose sand of backbeach sites, it can be notoriously difficult to pick up burial pits in plan and, when no bone survives, soils stains and suspected grave goods can be all archaeologists have to go on. 82. Meacham, Chek Lap Kok, 176 and 184–85. 83. While direct evidence for prehistoric double dug-out or outrigger sailing canoes is lacking, the sheer distances covered by Neolithic peoples as they spread out across the South China Sea, Indian and Pacific Oceans indicate that they were not paddling simple dug-out canoes on such long, open ocean voyages (see again Fuller et al., ‘Indian Ocean’). Double-hulled dug-out canoes or outriggers created stability for open-water travel, which may have been sufficient for inshore journeys between the islands, mainland, and delta, but sails provided the technological leap that led to the colonisation of the Pacific from the Late Neolithic onwards. 84. For example, at Nagsabaran in northern Luzon in the Philippines, Later Neolithic shell gathering created a midden 3  m deep × 100 m wide × 600 m long; the Jomon Culture site at Kidosaku in Tokyo Bay, Japan, although only comprising three simultaneously occupied houses, which were rebuilt several times, approximately 30–45 tons of clams—the most popular shellfish on site—were gathered with an estimated total number of around three million (H. Koike, ‘Prehistoric Hunting Pressure and Palaeobiomass: An Environmental Reconstruction and Archaeozoological Analysis of a Jomon Shellmound Area’, in Prehistoric Hunter-Gatherers in

62   Piecing Together Sha Po Japan: New Research Methods, ed. T. Akazawa and C. M. Aikens [Tokyo: University of Tokyo, 1986], 27–53), as cited in Colin Renfrew and Paul Bahn, Archaeology: Theories, Methods and Practice (London: Thames and Hudson, 2012), 294. 85. Guangdong Provincial Museum and Foshan City Museum, eds., Excavation Report on the 1977–1978 Hedang Site in Foshan (Guangzhou: Guangdong People’s Publishing House, 2006); Mo Zhi, ‘New Findings on Guangdong Archaeological Investigations and Excavations’, Kaogu 12 (1961): 666–68; Qiu Licheng and Liu Chengji, ‘Preliminary Findings of Cuntou Site in Dongguang’, Journal of Guangdong Provincial Museum 2 (1991):70–73; Li Yan, ‘The Second Season of Excavation of the Cuntou Site at Dongguang, Guangdong’, Wenwu 9 (2000): 25–34. 86. The high alkalinity of shells means that, even in Hong Kong’s acidic ground conditions, if shells were dumped in prehistory, they should still survive today. It is possible, of course, that some ancient shell middens might have been quarried away for use in early historical lime production. No such evidence has yet been recognised or recorded, but it could easily be missed in the challenging excavation conditions of backbeach sites. 87. Although as noted by Meacham, the preservation of the Jin dynasty burial and other organic materials in Square CC also owed much to the presence of lime in overlying Tang dynasty deposits (Meacham, ‘Sha Po Tsuen’, 46 and 53). 88. It has been calculated that a single large (red) deer carcass was calorifically equivalent to around 52,000 oysters or more than 150,000 cockles (Renfrew and Bahn, Archaeology, 293).

5 Bronze Age: Technology, Trade, and Increasing Social Complexity

Bronze Age Life in Hong Kong and the Pearl River Delta Introduction The Bronze Age (c. 1500–500 bce)1 is a widely evidenced and particularly exciting period in the Hong Kong–Pearl River Delta region’s human story, as it embodies a series of dramatic changes in material culture and technology that coincide with an apparent expansion of trade and exchange networks. Such socio-economic developments are reflected in a range of exotic ‘prestige goods’2 such as extremely rare jade yazhang and bronze halberds and daggers, which were imported and sometimes buried with prominent individuals.3 Collectively, the material evidence suggests the emergence of increasing competition within a society that was becoming more complex and hierarchical in nature. In contrast to those suggestions of socio-political change, there are clear indications of continuity—perhaps with some intensification—of the diversified fishing, hunting, and foraging economic lifeway. Strangely, in the later Bronze Age—from around 700 BCE until the Western Han (c. 100 BCE)—there appears to be a significant drop in activity across Hong Kong, although the degree to which this change is a cultural or archaeological phenomenon is open to debate (see further discussion below).4 The Bronze Age is also a period during which Chinese historical documents refer to Lingnan, and with it Hong Kong and the Pearl River Delta, as the homeland of the Southern Yue people.5 Bronze Age sites in Hong Kong include Shek Pik and Hai Dei Wan on Lantau Island, Kwo Lo Wan on Chek Lap Kok, Sha Ha in Sai Kung, Tung Wan Tsai (South) on Ma Wan, and Sham Wan and Tai Wan on Lamma Island. Backbeach sites remained the main focus of human activity, but the Late Neolithic headland and tombolo sites were, for some reason, abandoned in favour of new sites located on lower hillsides and plateaux such as Man Kok Tsui on Lantau Island, So Kwun Wat near Tuen Mun, and, of course, Sha Po.6 The Bronze Age is also notable as the first period in which we find some traces of human activity inland, for example on terraced alluvium in the Lam Tsuen Valley.7 Given the general abundance of Bronze Age remains across Hong Kong, it is particularly surprising that there is even less evidence for settlements or houses than in earlier periods, which makes the post-built structures on the Sha Po plateau all the more significant (see below).

Pyrotechnology, trade, and craft specialisation Although there seems to be significant socio-economic continuity from the Later Neolithic into the earlier Bronze Age—hence our discussion of a transitional phase—by the Bronze Age proper, at around 1000 BCE or somewhat earlier, a quite new and distinctive material ‘signature’ had emerged. Improved pyrotechnology—the control and manipulation of fire—was a key advance for both metals and pottery

64   Piecing Together Sha Po

production. The transference of technological knowhow, either by itinerant artisans or via trade and exchange-related information flow, meant that by around 1000 BCE the casting of simple bronze objects such as axes, arrowheads, and fish-hooks was evidenced in Hong Kong by the finished objects, sandstone ‘bivalve’ moulds, and, very occasionally, finds of splashed molten bronze from the actual casting process.8 No raw materials or furnace structures have been found to suggest local smelting of copper or tin, nor have any ingots of the two metals thus far been discovered. The artisans responsible for casting finished bronzes may therefore have relied primarily upon the remelting of scrap bronze, which could be done ‘on demand’ using small coarse pots as crucibles (see evidence from Sha Po below). Having said that, no material clues survive to show the method by which they achieved a sustained temperature of 950oC in order to melt and pour the bronze into moulds. Moreover, there is absolutely no evidence in Hong Kong for the casting of larger, thin cross-section, more technically challenging items such as bronze swords, daggers, and halberds, which therefore appear to be imported, exotic items acquired—perhaps via intermediaries—through trade with more technologically advanced societies to the north. It should be acknowledged that some scholars see such weapons as markers of more violent interactions between the forces of central Chinese states and warlike Yue ‘barbarians’, who presumably carried off such weapons as ‘spoils of war’.9 Either way, further distribution of such ‘prestige goods’ may then have occurred via elite-level exchange networks operating between different groups inhabiting the coastal-riverine area of Lingnan.10 Without detailed provenancing studies and analysis of regional distributions of artefacts from their point of production to consumption (use) it is impossible to say how things got from A to B, but down-the-line trade seems maybe the most likely mechanism, in particular where exotic ‘prestige’ goods are concerned.11 More advanced kiln technology allowed specialist pottery manufacturers, such as those at Meihuadun and Yingang in Boluo County on Guangdong’s East River (Dongjiang), to produce high-fired stoneware storage jars decorated with double-F and other stamped geometric motifs,12 while mastery of fast wheel pottery throwing led to the production of fully wheel-made fine stem cups or gui (簋), some with green glaze and resembling proto-porcelain in terms of hardness and quality. The term ‘proto-porcelain’ (原始瓷) is commonly used by Chinese scholars to describe fine porcelainlike glazed pottery produced during the Shang and Zhou dynasties.13 Although clearly an advancement on earthenware, such ceramics were still technologically inferior to the higher-fired porcelains produced after the Eastern Han period.14 While it must be acknowledged that some scholars have challenged the scientific validity of the concept, definition, and use of this term,15 its widespread acceptance among archaeologists in Guangdong justifies its use in our research. Most significantly—in both socio-economic and political terms—such proto-porcelains produced at the Meihuadun and Yingang kilns in Bolou County in inland Guangdong were not only discovered at large inland sites such as Shixia (Phase 4)16 and Henglingshan,17 but also appeared throughout Hong Kong, including on outlying islands such as Lamma. A significant proportion of the hard pottery has marks inscribed inside the hollow pedestal of stem cups or inside the rim of storage vessels, which was a feature widely evidenced on pottery grave goods at the large Bronze Age cemetery at Henglingshan in Boluo County, Guangdong.18 The great variety of such marks probably argues against their being identifiers of individual potters. We therefore use the term ‘potter’s marks’ in the sense that they were applied when pots were ‘leather hard’ during the manufacturing process (by potters), but are probably not the ‘signatures’ of individuals. No such pottery workshops have been found in Hong Kong and much of the hard pottery and proto-porcelain found here was probably imported down-river from Meihuadun, Yingang or other specialist producers. For functional reasons, cooking pots continued to be round-bottomed relatively thick-walled coarseware jars and basins, but now with everted flaring and/or dished rim forms, and geometric rather than corded



Bronze Age: Technology, Trade, and Increasing Social Complexity   65

decoration, although there is significant overlap between the two types.19 The apparent scarcity of Late Bronze Age (or ‘Early Iron Age’) activity in Hong Kong (c. 500–200 BCE) is most notably evidenced by the very rare finds of so-called ‘asterisk pattern’ or ‘rice character’ pottery typical of this late period, which is a comparatively common discovery in inland Lingnan.20 The Meihuadun pottery centre also produced—in the same fabric as the stoneware jars—a series of animal figurines mostly interpreted as oxen, but also dogs and sheep,21 which are found in small numbers on local sites, with notable clusters at Man Kok Tsui on Lantau Island and in North Lamma, in particular at Sha Po–Tai Wan. The figurines have prompted some to interpret them as evidence for Bronze Age animal domestication and farming in Hong Kong,22 but we would question that literal connection. As well as providing food, plants also continued to be harvested for their fibres, which were spun and woven, as was remarkably proven by the survival of a tiny 0.8 cm by 0.7 cm scrap of woven hemp fabric found preserved in surface corrosion on a bronze spearhead in a grave at Kwo Lo Wan.23 The preservative qualities of bronze corrosion products also led to the survival of a piece of ‘woven straw matting’ beneath a fine bronze axe in another grave at Hai Dei Wan.24 While the latter was probably used to lay out the deceased and grave goods, the relatively complex weave and spun fabric of the Kwo Lo Wan piece could mean it was part of the persons’ clothing, but might also be a burial sheet. With respect to lithic technology, there was continuity in the use of utilitarian tools such as the versatile pebble picks, stepped and shouldered adzes, pounders, anvils, grinding pebbles, and whetstones. But there was also a noticeable increase in the numbers, styles and quality of polished stone personal ornaments, such as split earrings, disc ornaments,25 and arm-rings, which at sites such as Sha Ha, Man Kok Tsui, and Sha Po were being produced in what seem quite well-organised workshops with remains also of complete and broken finished objects, rough-outs and waste materials (debitage), as well as fine sandstone saws, whetstones, and rotary polishers used in the manufacturing process.26 Although bronze was available, its supply was sufficiently restricted—either by genuine scarcity or perhaps by socio-political factors—to ensure the continued use of ground-stone projectile points. But here again we see technological design improvement expressed in developed fluted forms, which perhaps reflect a growing interest in martial display, as exemplified by fine sets from Shek Pik and Sham Wan, which in the latter case range in size from 6.7 cm to 16.5 cm long.27 The high quality of execution and use of standardised forms can also perhaps be interpreted as the work of a few skilled artisans, rather than the more variable, but nonetheless functional, work of many hands, which is suggested by earlier types.

Subsistence continuity but some intensification? The subsistence economy continued to be founded upon fishing, hunting, and foraging, although our understanding of the true importance of each activity is once again significantly hindered by issues of archaeological visibility and poor organic preservation. In common with earlier periods in Hong Kong prehistory we have no clear indications of the kinds of things being cooked in the coarse cooking pots, but we can imagine a range of foodstuffs including starchy plants, shellfish, and the meat of marine and terrestrial vertebrates. More definitive answers will hopefully come from future scientific analyses of any food residues that might survive.28 As shown at Xincun in Taishan, Guangdong starchy plants were no doubt an essential component of diet but, as in the Neolithic, to date we have no solid evidence from a Hong Kong site to support that belief. At Sham Wan the targeting of two fish species—the headgrunt and marine catfish—seems to have continued from the Later Neolithic into the Bronze Age, when some of the recorded deer and pig remains were no doubt also deposited. Further limited middens were noted at Sha Po (see below) and at Tung Wan Tsai South on Ma Wan, where 47 species of shellfish were

66   Piecing Together Sha Po

recorded together with occasional remains of fish, barking deer, and a larger species of deer.29 Although food remains are poorly represented in this period, there is good artefactual support for a continuing, perhaps intensified, interest in fishing as indicated by numerous finds of bronze fish-hooks and waisted pebble net-weights. While, as discussed in Chapter 9, the aforementioned projectile points were as wellsuited to hunting as they were to warfare.

Social hierarchies? Prestige goods and mortuary display Bronze Age cemeteries on the scale of the Later Neolithic–Early Bronze Age example at Tung Wan Tsai North are unknown in Hong Kong, as in general is the preservation of human skeletal remains. The one notable exception is Shek Pik, the rich Bronze Age site on Lantau Island, where in 1938 six very poorly preserved skeletons were identified by Schofield.30 Interestingly, the graves contained neither hard pottery nor bronzes, but skeletons 4 and 6 were accompanied by comparatively rich assemblages of grave goods, including a tremolite T-flanged arm-ring, sets of grey fluted arrowheads, a shale dagger, clam shells, shark’s teeth, and bones of fish and mammals.31 In contrast to Shek Pik, many of the sites listed above produced only occasional burials that were identified by clusters of pottery, stone, and, very occasionally, bronze grave goods, around which it was sometimes also possible to define a grave pit. At Kwo Lo Wan on Chek Lap Kok, eight graves were found including Grave 1 with a pair and Grave 2 with two pairs of bivalve sandstone axe moulds ‘in casting position’32 (i.e., standing upright as if awaiting the molten bronze). Interestingly, Grave 2 was also the richest on site with two pots, a set of ten quartz earrings of different sizes and a bronze spearhead. The use of bivalve moulds as grave goods may have socio-political significance, because it is likely that those controlling the means of bronze object production (e.g., mould carvers and/or owners and bronze casters) would be individuals of somewhat elevated status. Moreover, early metallurgy may well also have been considered an almost ‘magical’ process, transforming patinated scrap bronze or ingots into gleaming golden objects, which were no doubt greatly prized. Man Kok Tsui is another burial ground of note where, on a small hillock, multiple graves were inferred from finds of twelve complete pots, adzes, and stone rings, although bronzes were rare.33 At the remarkable late Shang to Spring and Autumn period (1100–476 BCE) cemetery discovered at Henglingshan in Boluo County, which contained 302 burials, pottery grave goods also far outweighed those in bronze, which included three ritual bronzes comprising two bells and a ding tripod vessel.34 The general scarcity of bronze objects in society was thus reflected in funerary contexts and only Shek Pik and Tai Wan, in particular, have produced significant numbers of bronze artefacts, albeit with few definitively from graves. The most remarkable evidence for emerging social differentiation was perhaps that contained in grave M6 at Tai Wan, which contained a collection of imported jadeite objects comprising an extremely rare yazhang (sceptre) and a number of tube and perforated plaque-like ornaments that mark out the deceased person as someone with power and status well above the ordinary. The yazhang is said to have been used by pre-Qin Chinese military leaders to direct troops in battle, but it is extremely unlikely that it retained that same cultural meaning and function within Hong Kong’s coastal communities.35 What we can say, however, is that together with imported bronze halberds, swords, and daggers—all also found at Tai Wan—the yazhang is part of the category of exotic trade goods, which were also no doubt signifiers of socio-political prestige. As such, they highlight Tai Wan on Lamma Island, and maybe Shek Pik on Lantau Island, as important communities containing individuals or groups with relatively high status within Bronze Age Hong Kong and particularly good—most likely elite level—connections that extended across large areas of coastal-riverine Lingnan.



Bronze Age: Technology, Trade, and Increasing Social Complexity   67

Emergence of an ideological-religious symbolic language Last but by no means least, the Bronze Age is also characterised by the remarkable rock carvings (petroglyphs), for example at Po Toi and Big Wave Bay,36 which are thought to be reflections of the religious and ideological world of our maritime-focused people, but also showcase their art and sense of the aesthetic. Interpretations of the carvings, which include abstract and more figurative representations of animals and humans, have been many and varied. Fascinatingly, though, the carvings have many stylistic parallels in motifs stamped on pottery and cast into bronze objects found locally, as well as others occurring much further afield. Meacham draws some useful parallels between designs that occur on rock art, local Bronze Age pottery and bronze weapons, Shang–Zhou bronzes, and drums made by the Dong Son culture of south-west China and northern Vietnam.37 The only excavated evidence for the carvings’ Bronze Age date was unearthed by Hong Kong’s lone Chinese amateur archaeologist of the pre-war era, Chen Kung-che (陳公哲), who found human burials, complete pots, bronze objects, and a fish-hook mould directly below the Shek Pik rock carving.38 However, the existence over a significant geographical area of a common symbolic repertoire on rock art, ‘prestige goods’ weaponry, and utilitarian pottery suggests that the ideology underlying such symbols was deeply embedded in religious, martial, and domestic spheres of society. It also indicates that the emergent social complexity identified during the Later Neolithic had reached a much greater level of maturity and Hong Kong–Pearl River Delta communities were becoming integrated, presumably through trade and exchange, into larger socio-political formations. Our task now is to explore how Sha Po, with its fascinating juxtaposition of plateau and backbeach sites—both with rare and significant findings—adds to our understanding of this dynamic period in Hong Kong’s human story.

Bronze Age Sha Po: A Community Experiencing Significant Socio-political Change The general pattern of activity The general shape of the Bronze Age coastal landscape appears to have changed little from that shown in Map 7 in Chapter 4, although there was perhaps some expansion of the backbeach westward towards the sea. Socio-economically, though, the Bronze Age constitutes one of the most dramatic and busy periods in the Sha Po story, when both the backbeach and plateau to the north were seemingly in contemporary use, but for quite different purposes. Across the south-facing slope of the plateau was a settlement, probably of stilt-houses, whose inhabitants included skilled stone-workers focused on the production of fine quartz ornaments. While down below, against a typical background patterning of domestic-craft activity, there was perhaps the region’s best evidence for the activities of bronze metallurgists, who were melting and casting axe-heads at the southern end of the backbeach. Also, on the central backbeach, there are suggestions of four, perhaps more, Bronze Age burials. When considered as a whole, we appear to have a dynamic social landscape, evidencing a wide range of general subsistence-domestic, funerary, and specialised craft activities with broad zones for each. However, as ever in prehistoric coastal archaeology we must be mindful of the fact that our material evidence is an ‘aggregated overview’ of a myriad of discrete, perhaps often-repeated, events whose pattern and distributions changed through time. This is especially true on heavily reworked backbeach sites such as Sha Po, where the identification of fine subphases of Bronze Age activity is not easily achieved. There is, nonetheless, much of interest and we begin our discussion on the higher ground of the Sha Po plateau.

68   Piecing Together Sha Po

Plateau Settlement and Workshop Introduction The Sha Po plateau is first discussed in print by Father Daniel Finn, who labelled the site ‘Yung Shue Wan’, or ‘YSW’ for short, in his series of articles published in the Hong Kong Naturalist.39 His knowledge of the site was based entirely on surface finds, which had been disturbed by the large-scale terracing and ongoing cultivation activity of local farmers. However, even without excavation it was already evident from fragments of bronze and frequent geometric stamped pottery that the site was Bronze Age in date. Moreover, the large quantities of quartz rough-outs and rings, stone knives and polishers occurring on the plateau, its slopes, and at its base indicated that stone ornament manufacture had been a major activity in that period. However, it was not until the beginning of the twenty-first century that the existence of a settlement was suggested by discoveries of post-holes in three separate excavations on the south-facing flank of the plateau.

The plateau domestic scene The recent development of the south-facing flank of the plateau prompted archaeological excavations in three adjacent plots within an area some 40 m east–west by 20 m north–south, which all revealed a similar pattern of Bronze Age remains: namely, a single cultural horizon with a fine collection of ceramics of the period, from which post-holes had been cut into the underlying subsoil. The cultural horizon contained a mix of geometric stamped pottery comprising hard storage jars and coarse cooking pots (fu)—but also some continuity of corded cooking wares—and a number of fine wheel-made stem cups, including several green glazed examples. In the 1930s Finn also recorded finding green glazed stem cups on the plateau, together with several very unusual sherds of glazed double-F pottery, a few green glazed examples of which were also found in 2004.40 In line with aforementioned findings at the Meihuadun and Yingang kiln sites and Henglingshan cemetery, several stem cups and hard geometric jars from the plateau and backbeach bore inscribed marks respectively inside the hollow pedestal or on the inside of the rim.41 Much of the hard pottery had a grey fabric with black spheroidal inclusions—a type noted by Finn as being particularly common at Sha Po—which seems to be a product of the Meihuadun and/or Yingang kilns, where it is appositely described as having ‘black sesame dots’.42 The green glazed proto-porcelain stem cups are also probably products of the same kilns. Also present were a few pot-stands, presumably relating to cooking, while the range of stone artefacts was notably different in composition from that observed in the backbeach. On the plateau there were no crude pebble choppers and picks, just a single trapezoidal adze and one waisted net-weight. A pebble pounder or grinder and three pitted anvil stones may have been used for stone-working, but the Xincun, Taishan evidence suggests they could also be used in food processing. A collection of more specialised grinding, cutting, and polishing implements seemingly connected to stone ring manufacture is discussed further below. The dearth of bronze objects recently found may partly be explained by an anecdote related by Finn in 1936, in which he records being handed a piece of bronze spearhead found on the plateau, but then goes on to mention reports of ‘old bits of bronze’ being previously bought from villagers by passing pedlars.43 Perhaps the most unusual class of artefacts recovered from the Sha Po plateau is the three pottery animal figurines: a ‘horse’ torso found in the 1930s, an ox’s head found in 1995 (both of which are made of the same hard grey ‘Meihuadun-type’ pottery), and another ‘ox’ made of a coarse reddish-brown earthenware found during the AMO’s excavations in 2004.44



Bronze Age: Technology, Trade, and Increasing Social Complexity   69

The animal figurines: What do they mean? The discovery at the Meihuadun kiln of double-F pots and animal figurines in the same hard pottery confirmed the shared origin suggested by Finn in 1936.45 Indeed, his ‘horse’ torso and the ox head found in 1995 are remarkably similar in two important respects: they are apparently made from the same hard pottery and both were covered in small ‘pin-prick and circle’ markings, thought by Finn to represent the animal’s hide. With its wide mouth and broken horns, the head does indeed resemble an ox, although it could feasibly be that of a young water buffalo. Despite missing its head and legs, Finn’s figurine, he argued, had a body shape that was most like that of Han representations of horses, while a series of deep perforations along its flanks and chest appeared to coincide with points ‘for the attachment of horse’s trappings .  .  . and apparently that for the typical Chinese chariot and not merely for riding’.46 While acknowledging that horses and chariots were never important to the riverine-coastal communities of the South,47 Finn nevertheless made convincing connections between the figurine, its hard pottery fabric, and the strong stylistic connection of the double-F motif with late Shang objects from the Central Plains.48 He thus implied that the presence of representations of northern domesticated animals on Hong Kong sites should only be taken as an indication of northern cultural influences on Bronze Age farming communities far from the coast in inland Guangdong. The whereabouts of the ‘horse’ torso are unknown, so we will probably never be able to offer it up to the 1995 ox head to see whether they fit, but the possibility is tantalising. Could Finn’s figurine actually be an ox, and the perforations on the torso reflect the attachment points for oxen pulling a humble cart? The last figurine to be discovered on the Sha Po plateau is strange on several counts. Unlike the others, it is made from partly ‘burnished’ reddish-brown earthenware—not hard pottery—and, despite earlier identification as an ox,49 is almost certainly not a representation of a domestic animal (unless Bronze Age children kept small furry pets) because from many angles it most closely resembles the squatting posture

Figure 9:  Father Finn’s animal figurine body found in the 1930s and head found in 1995

70   Piecing Together Sha Po

Plate 12:  Bronze Age pottery animal figurine from the AMO 2004 excavation. © Antiquities and Monuments Office, reproduced with permission.

of the house mouse (Mus musculus cataneus). We therefore feel that the latter may be a local creation, whereas the other two are no doubt imports from kilns in inland Guangdong. So we come to the following questions: Why were they made? What was their purpose and what do they mean? To deal with the most recent discovery first, if we are correct in our identification, then we would suggest that mice as commensal animals would be regularly seen,50 and possibly even eaten, but the figurine might best be interpreted as a child’s toy. In that context, the ‘burnishing’ may simply be a consequence of intensive use of a relatively low-fired, much-loved plaything. The ox head and ‘horse’ torso are more problematic as the bones of neither animal are evidenced in securely dated, sealed archaeological deposits dating earlier than later historical times. Certainly, by the Bronze Age we can be fairly sure that hard pottery and figurines were being imported from areas inland where rice farming and animal husbandry were being practised. But to interpret imported domestic animal figurines as literal evidence for the use of those animals in a totally different environmental and socio-economic context (i.e., a resourcerich coastal periphery inhabited by hunter-fisher-foragers) is just not supported by the archaeological evidence.51 Moreover, if the figurines are such literal representations, one wonders why in Hong Kong there are no pigs or dogs, when these were the most widely occurring early domesticates in China and South East Asia?52 So to return to our earlier questions, we would interpret the imported figurines as one



Bronze Age: Technology, Trade, and Increasing Social Complexity   71

of three things: either potters’ ‘samples’ demonstrating the quality and fineness of their work, or objects originally produced for use in some form of ritual related to household or agricultural fertility, or as we suggested for the Sha Po ‘house mouse’, they were simply the toys of Bronze Age children. Whatever the case, we should expect that their context of use, value, and meaning could easily have changed between the site of their production in Boluo County in Guangdong and their eventual home—almost 200 km distant by water—at Sha Po and other contemporary sites at the coast.53

Suggestions of a stilt-house settlement Perhaps the most dramatic discovery on the plateau was a scattering of over seventy post-holes (see Figure 10), spread across all three of the larger excavation areas [25, 28, and 30].54 The post-holes were identified cutting into the decayed granite subsoil following removal of the Bronze Age cultural horizon. The smaller examples are perhaps better described as stake-holes, and it was those that were the shallowest features, whereas most of the larger examples (>25 cm diameter) were 25 cm or more deep, although only a few contained possible packing stones.55 As the post-holes were filled with the same soil as the Bronze Age layer above, it was impossible to tell whether they were cut from the top of that layer, during its formation, or at the beginning of Bronze Age activity on the plateau. But the absence of floor surfaces compacted by the trampling of feet, hearths (fire-places), or telltale inside-outside patterning of finds, suggests that any houses or structures were not built and occupied at ground-level, but were instead on stilts.56 Drawing on that evidence, and no doubt also the notion of seasonal use of sites such as Sha Po, there are those who would argue that the post-holes simply record the presence of temporary, easily dismantled structures, rather than more solidly built stilt-houses.57 But in two out of three excavation areas on the Sha Po plateau [25 and 30] that view is countered by the lack of intercutting or ‘double’ post-holes and absence of post-removal ‘scars’,58 which might be expected if regular dismantling or multiple phases of rebuilding had occurred on more or less the same spot during repeated seasonal occupations occurring over a number of years, even multiple generations. In contrast, though, the 2010 excavation has several paired, overlapping, and multiple post-holes, which in that area do perhaps suggest rebuilding on the same spot. On balance, the evidence is somewhat equivocal. However, even if residence was only seasonal, the community would still be there for months at a time and some form of weatherproof shelter would presumably be required. Overall, the plateau was apparently not intensively used, but each structure may have lasted a generation or so, and was then either refurbished in situ or rebuilt on an adjacent plot. Either way, the focus of activity may have gradually ‘migrated’ across the plateau through time creating the pattern observed. In the published report the excavators speculated that one cluster of ten post-holes may have formed a 3 m by 2 m subrectangular stilt-house. This structure seems very small for a domestic dwelling but might be big enough for sleeping and shelter if most of the community’s activities occurred outdoors.59 Overall, for the reasons outlined above, we would support the interpretation that the Sha Po plateau had some form of ‘settlement’ involving stilt-house structures, but with little prospect of larger area excavations in future a fuller understanding of its character may remain elusive. However, when it comes to the question of craft activities on the plateau, we have much more solid and substantial evidence upon which to base our interpretations.

72   Piecing Together Sha Po

Figure 10:  Composite plan of Bronze Age post-holes found on the plateau (dashed outline in 2004 excavation indicates oval structure identified by excavator). Original plans © Antiquities and Monuments Office and reproduced with AMO permission.

Craft specialisation and the importance of display The material evidence suggests that the Sha Po Plateau was used by a group of skilled artisans operating what can best be described as a stone ornament workshop. When Finn visited the site in the 1930s he described finding disc-shaped quartz rough-outs and cores scattered across the plateau, down its slopes, and around its base.60 Recent findings on the plateau reflect Finn’s observation in that quartz rough-outs were by far the most common residue of ring manufacture, followed by cores, and then broken polished rings.61 Very occasional grey schist rough-outs and broken rings were found, but the overwhelming focus was clearly on the manufacture of ornamental rings made from white or clear quartz. The stone working toolkit comprised pebble polisher-pounders, pitted anvils, fine sandstone whetstones, and saws, but rotary polishers like the one found on the backbeach (see below) were surprisingly absent.62 By drawing on Finn’s detailed analysis of the materials, as well as the work of later scholars,63 we can offer an overview of the earring manufacturing process. First, a chunk of quartz was chipped into a rough disc somewhat larger than the finished ornament to create the rough-out.64 Whether that process occurred at Sha Po or elsewhere is unclear, but the apparent lack of larger quartz chunks or scatters of small pieces of waste (debitage) on site suggests that quartz was maybe partially processed at the quarry—most likely on Lamma—and arrived at Sha Po as rough-outs.65 The rough-out’s surfaces were then ground down to the required thickness probably using pebble polishers, anvils, and whetstones. Next the central hole was bored from the back (flat) face of the disc using some form of hollow rotary drill, which Finn suggested



Bronze Age: Technology, Trade, and Increasing Social Complexity   73

need be nothing more complex than a length of bamboo of the desired size, with fine sand and water creating the necessary abrasive medium. Thereafter the rings’ inside and outside diameters were polished and, if an earring was the desired end product, the slot was carefully sawn through from one or both sides to complete the ornament.66 A final overall polish and an ornament of the highest quality was produced using what seems, to our modern eyes at least, very basic technology but made highly effective by skilled hands and an intimate knowledge of the raw materials and processes involved. The plateau workshop’s focus on a particular category of stone ornament—clear and white quartz earrings—plus the evident refinement of forms and quality of finishing, arguably reflect several related changes in Bronze Age society. Firstly, they probably indicate the development of a greater degree of craft

Plate 13:  Artefactual evidence for a white quartz earring workshop on the Sha Po plateau: (a) rough-out, (b) polished blank, (c) finished ring (half ), (d) polishing stone, and (e) sandstone ‘saw’. © Antiquities and Monuments Office, reproduced with permission.

74   Piecing Together Sha Po

specialisation, which in turn suggests that the artisans involved had to be supported by the wider community, perhaps through the redistribution of surplus food resources. That points towards the emergence of more centralised authority in local communities, and individuals—probably also lineages—with the power to organise and command. The increasing use of showy jewellery also suggests a growing concern with personal display and the use of appropriate material markers of status and identity in what was becoming a more competitive society.

A Myriad of Activities on the Bronze Age Backbeach Introduction Interestingly, the ceramics evidence from the Sha Po plateau and backbeach—which in both areas comprises the same range of cooking, storage, and consumption wares—clearly suggests that both areas were in contemporary use, but the activities and use of space in the two zones seem otherwise to have been quite different. The plateau’s focus on settlement and fine craft activity can be contrasted with the backbeach evidence for bronze casting, stone tool finishing-repair, human burials, and food preparation, consumption, and waste disposal. Finn’s 1930s’ observation that there had been significant downslope erosion of materials67 from the plateau settlement and workshop has been more recently confirmed by investigations at the northern end of the backbeach [16, 35, and 38]. There two distinct layers were noted containing Early–Middle Bronze Age ceramics with quartz ring cores, rough-outs, and stone saws which, in 2000 in particular, were identical in composition to the plateau assemblages but the coarser pottery had been heavily weathered during its journey downhill.68

Figure 11:  Bronze Age fishing and hunting gear: (a) barbed bronze arrowhead [22a:02], (b) double fish-hook casting mould [15:07; CSF26], (c) grey schist arrowhead [22a:02]



Bronze Age: Technology, Trade, and Increasing Social Complexity   75

Suggestions of subsistence continuity and hunting The earlier patchy evidence for small-scale shellfish exploitation continues into the Bronze Age on the southern backbeach and a range of species from rocky shore (mainly turban shells, whelks, and oysters) and some sandy shore (clams) were identified there, together with occasional fish bones from head grunt and catfish [11:L4; 34:E-F536].69 Also present were very occasional reptile and mammal bones including those of dolphin, sea turtle, large deer, crocodile, a possible porcupine, and, most surprisingly, the tooth of a rhinoceros.70 Other evidence for subsistence was found in a small isolated excavation on the northeastern backbeach, which reportedly produced ‘some burnt bones and a few fish bones (head grunt)’ [10:L5].71 Again the above evidence hardly warrants the term ‘midden’, but it does give us some clues regarding the range of non-plant resources being exploited at Sha Po. In addition, the notion (at present untested) that starchy plant foods were probably a very important dietary staple for Hong Kong’s prehistoric populations can also be applied to the Bronze Age inhabitants of Sha Po.72 Once again hunting can be inferred from seven projectile points found in three excavations on the south-eastern backbeach, of which five—two spearheads and three arrowheads—were made of grey schist [11:L4; 22a:02; 48:C3],73 while two cast bronze arrowheads had almost identical barbed and tanged forms [11:L4 and 22a:02].74 Besides the few fish bones, fishing was indicated by discoveries of waisted pebble net-weights [11:L4], a stone ring-shaped net-weight [34:E502],75 and the rare find in 1997 of half a sandstone mould for a pair of slender bronze fish-hooks [15:07].76

Domestic and craft activity areas While we have no indications that people actually lived on the backbeach—indeed, this seems unlikely, given its exposed position and the plateau’s evidence for possible settlement—the south-eastern end evidences quite intensive domestic and craft activity. For example, excavations in 2002 alone produced over 50 kg of pottery associated with fire-grate and pot-stand fragments and over 60 stone artefacts [22a and 22b]. A rare ‘asterisk’ pattern hard pottery sherd of supposed Late Bronze Age date was also found within the assemblage [22b:01].77 The fire-grates, pot-stands, and pottery—which is dominated by geometricstamped coarse cooking pots plus hard storage jars and a smaller number of stem cups—collectively indicate that food was being stored, cooked, and presumably consumed on the backbeach although, as noted above, food residues themselves are typically scarce. Many contexts also contain a substantial assemblage of corded coarse pottery, whose use overlapped with geometric cooking wares during the Early Bronze Age. The stone artefacts from that same general area provide an interesting contrast to the assemblage composition from the plateau, which was dominated by tools, residues, and artefacts related to ornament manufacture. Whereas on the central-eastern to southern backbeach the assemblage was similar to that discussed under ‘Later Neolithic’ in Chapter 4, in that it comprised—in descending order—polisherwhetstone fragments, pebble tools, pounders and anvils, finished artefacts, and rough-outs. The modest size of the assemblage and rarity of waste material and part-finished objects suggest that this was not a ‘workshop’ as such; instead, the remains probably reflect activity occurring within a daily domestic context: rough-outs being worked into their final form, while finished tools and weapons such as adzes and projectile points were probably being repaired and sharpened in readiness for use. As is typical throughout Hong Kong, the Sha Po stoneworkers selected the same type of grey laminated schist—often

76   Piecing Together Sha Po

Figure 12:  Bronze Age pottery: (a) coarse geometric cooking pot [34:520], (b) hard pottery stem cup with inscribed ‘potter’s mark’ inside pedestal foot [11:L4], (c) hard pottery ‘double-F’ storage jar [38:103], (d) hard pottery lid [16:T2L5; CSF12], (e) [12:L3], and (f ) [13:L3] both hard pottery storage jars with net and lozenge decoration and ‘potter’s marks’ inside rim

banded yellow ochre or dull orange—for the manufacture of the larger type of polished arm-rings, knives, and projectile points. A key consideration was clearly the stone’s natural propensity to fracture into large flat sheets that were then ideal for either purpose. As with our discussion in Chapter 4, we would suggest that many pebble tools, in addition to their likely use in stone tool, weapon, and ornament manufacture, were probably also involved in a variety of food-processing activities.

Burials behind the beach? The possibility of Bronze Age graves on the south-eastern backbeach was first raised in 1972 by the discovery of an intact ‘double-F’ jar, whose survival had presumably resulted from its intentional burial in



Bronze Age: Technology, Trade, and Increasing Social Complexity   77

Figure 13:  Bronze Age stone artefacts: (a) stone knife rough-out [22a:02], (b) pebble tool [23:02], (c) polishing stone [22a:01], (d) stone flake [22a:01], (e) stone pounder [22a:03]

antiquity, although no associated feature was identified. In the acidic sediments of Hong Kong sites, which can completely destroy organic materials such as human bone, a localised soil stain, and discrete cluster of artefacts may be all that survives to indicate the former presence of a burial. Moreover, in the loose sand of backbeach sites it may be very difficult to identify the opening of a grave pits, but the presence of a ‘structured deposit’78 (i.e., one that was intentionally placed within a feature and did not result from discard or accidental loss) can be a decisive clue for archaeologists. We will therefore discuss all possible burials, but only use the label ‘grave’ when we deem the weight of circumstantial evidence warrants it. In 2002, when a cache of four grey schist knife rough-outs was discovered with Bronze Age coarse pottery directly above an elongated oval soil stain measuring 2.4 m by 1.1 m in the surface of the underlying Neolithic horizon, a Bronze Age burial was inferred [Grave 1—22a:02].79 Grave 1 was orientated with its long axis on a bearing of approximately 140 degrees (roughly north-west–south-east), with the broader (head?) end pointing inland to the south-east. Fascinatingly, this apparent orientation is broadly paralleled by that of two earlier historical graves (Graves 2 and 3) described in Chapter 6, which contained skeletal remains of individuals buried ‘facing the sea’. 80 Two other possible burials in the same area of the backbeach were suggested by further discoveries of discrete artefacts clusters, one also in 2002 and the other in 2009. The first ‘grave’ assemblage comprised a small hard geometric jar, a stem cup, and a large rock oyster shell,81 together with a number of unidentifiable bone fragments. The pottery gives a date firmly in the Bronze Age and similar vessels were recorded as grave goods at Henglingshan cemetery [22a:02/03].82 The approximate location of the other possible

78   Piecing Together Sha Po

Limit of excavation Bronze Age geometric pot (in layer 02) stone

stone

Cluster of four knife rough-outs (in BA layer 02)

Dark soil stain in surface of Later Neolithic layer 03

Conjectured extent of Grave 1 in BA layer 02 0

1m

Figure 14:  Plan of Grave 1

grave was marked by finds of a small globular jar and shallow bowl, both of which were undecorated and made from the same smooth, fine clay [32:204/205].83 The bowl had probably been slightly disturbed by more recent activity but they were found one above the other in different but contiguous contexts. Based on the mixture of soft and coarse geometric pottery found in association, a date in the Early Bronze Age seems most likely here. If all three examples discussed above are indeed burials, then their positioning in an area evidencing a wide range of other domestic and craft activities suggests that patterns of usage on the south-eastern backbeach may well have shifted during the Bronze Age, but over the millennia such subphases have become impossible to detect archaeologically.



Bronze Age: Technology, Trade, and Increasing Social Complexity   79

Sha Po’s backbeach metallurgists As early as the 1930s Finn argued that bronze casting had taken place at Tai Wan, where he recovered one half of a bivalve axe mould,84 and at Hung Shing Yeh, where he found ‘definite proof of local casting in the shape of the metal droppings or outpourings apparently from the vicinity of the mouth of the mould’.85 Against that North Lamma background, the surface find in 1985 of one half of a bivalve axe mould on the Sha Po plateau, which occurred during the First Territory Wide Survey, is noteworthy.86 However, it is on the backbeach below that we find Hong Kong’s most compelling evidence for bronze casting; there near the southern end of the backbeach, the society unearthed a pair of sandstone bivalve axe moulds in intimate association with coarse potsherds splashed with bronze casting slag and possible fragments of square crucibles [11:L4]. Laboratory analysis of the slag showed that it contained copper, tin, and lead, which would only be present in the liquid bronze alloy when it was prepared for casting.87 Nearby were two further half moulds for making different axes, thus confirming that at least three different axe types could have been cast by the metallurgists working on the southern backbeach. In addition, half of a small, fine sandstone double fish-hook mould was discovered on the central backbeach in sediment disturbed by a Southern Dynasties burial [15:07].88 In terms of the process itself, we would suggest that some form of forced draught technology—bamboo blowing tubes (tuyères) being more likely perhaps than bellows—would be required to heat charcoal to achieve the sustained 950oC required to liquefy bronze (1083oC if copper and tin ingots were used). But some form of hollow clay-lined hearth or pit would surely be necessary to retain the heat? We have no such features in the excavated areas, although they may exist nearby; however, the physical evidence described above seems to suggest an ingenious but simple technical solution. We wonder whether the coarse pottery found with casting slag inside was in effect part of a portable ‘bowl furnace’. If a coarse bowl or basin was half-filled with burning charcoal and either a crucible or smaller coarse bowl containing

Plate 14:  Bronze Age axe and fish-hook moulds. © Antiquities and Monuments Office, reproduced with permission.

80   Piecing Together Sha Po

Figure 15:  Speculative reconstruction of bronze casting at Sha Po (based on objects and residues)

the ingots or scrap bronze was placed on top, this would then provide a suitable heat-retentive enclosed environment within which tuyères could be used to achieve the temperature necessary to make bronze. No structural residue would result, but any slag skimmed off the molten metal would finish up inside the coarse pottery vessel, while any localised traces of heating and charcoal would probably be erased by time and later activity. While this may not be the answer, we would argue that the Sha Po findings are certainly consistent with such a methodology. Experimental testing of the above idea should ideally be attempted to see if the process is indeed practicable. Finally, the use of the southern tip of the backbeach for metallurgy was probably a deliberate choice in order to keep noxious fumes and fire-risk as far away as was possible from the plateau settlement.

Conclusions What we hope the above presentation of evidence and discussions have demonstrated is that Sha Po, while reflecting many typical aspects of the period, is in several ways a unique and regionally significant Bronze Age site. There is the extremely rare juxtaposition of plateau and backbeach sites in contemporary occupation, but each records different uses: the former with residential and domestic activity alongside fine ornament manufacture, and the latter with domestic, funerary, and stone working activity, together with striking evidence for bronze metallurgy. The Sha Po area can thus quite legitimately be described as a Bronze Age social landscape, which no doubt also lay within a larger North Lamma territory including the seemingly higher status, bronze-rich site at Tai Wan.



Bronze Age: Technology, Trade, and Increasing Social Complexity   81

At a more detailed level, the plateau has good evidence for a number of post-built structures which, in the absence of evidence for ground-level activities, can be interpreted as stilt-houses or shelters of some form. The lack of intercutting post-holes in two out of the three larger excavations is interesting, as it does not suggest the regular rebuilding—seasonal or otherwise—of structures on the same spot, whereas this may indeed be inferred by the multiple intercutting post-holes in the 2010 excavation (see Figure 10). Whether this means that some structures were used longer-term by a group of limited size (and were therefore not regularly rebuilt), or that site usage was very sporadic and varied in intensity through time, is hard to establish. However, the case for the operation of a quite specialised workshop producing clear and white quartz rings is compelling and offers further support for the notion of an increasingly competitive and complex society on North Lamma. The degree of product specialisation on the plateau is interesting in that it seems to indicate production well beyond the needs of the Sha Po community, which in turn suggests the presence of a group of artisans serving wider ‘markets’, perhaps bartering their goods in return for subsistence support, and maybe answering to a polity with political reach that transcended kinship-based allegiances. Of undoubtedly the greatest archaeological importance are the fascinating remains of ancient metallurgy in the form of bronze casting on the southern backbeach, which is unparalleled anywhere within the region. The artisan’s ability to transform ingots or, more likely, dull scrap bronze into shining golden objects must have seemed almost ‘magical’ and such individuals may therefore have been afforded special status.89 Many questions still remain regarding the nature of the process and whether, for instance, scrap bronze or ingots provided the ‘raw’ material, a forced draught was produced by tuyères or bellows, and clay hearths or our suggested pottery ‘bowl furnace’ were used. Whatever the case, bronze axes, fishhooks, and barbed arrowheads may all have been cast at Sha Po; however, this is certainly not bronze casting on anything like an industrial or mass-produced scale. Instead, we should perhaps be thinking in terms of the activities of a few specialist artisans, either moving between communities trading their highvalue products and services or maybe, given the increasing evidence for social differentiation in Bronze Age society, serving the needs of particular communities under one tribal leader or chief. The above evidence for the emergence of a more complex and probably stratified Bronze Age society in Hong Kong makes the apparent dearth of Late Bronze Age activity all the more surprising. However, we wonder whether the scarcity of asterisk-pattern pottery and the occasional ‘late’ radiocarbon dates from ‘Middle’ Bronze Age deposits might actually be related. Although geographically remote from the ‘war zones’ of the Warring States Period, perhaps there was nevertheless sufficient political turmoil inland to cause a breakdown of long-standing supply networks, which limited access to asterisk-pattern pottery and iron implements, and led to the conservation and retention of older styles of hard geometric pottery and bronzes at the coast? Here, we must keep the changes occurring in our coastal society in perspective; although imported prestige goods, the development of craft specialisation, and rich burials all point towards increasing social complexity, such socio-political change may have relied heavily upon thriving external trade and exchange networks. Alternatively, the evidence for increasing competition, weaponry, and display could be indications of a society where intergroup warfare was becoming endemic, which in turn caused a breakdown of supply networks from within. The truth at present eludes us, but suffice it to say that nothing approaching the apparent breadth and intensity of human activity evidenced during the Bronze Age occurred again at Sha Po until the Six Dynasties–Tang period. In Chapter 5, we explore that early historical world, beginning in what appears to have been a much quieter Sha Po landscape during the Han dynasty.

82   Piecing Together Sha Po

Notes 1. Strictly speaking, the Bronze Age in Hong Kong continues until the beginning of the Qin dynasty in 221 bce, but that date has little meaning in local archaeology. We prefer to use the 1500–500 bce date range as this seems a good compromise based on the physical evidence and all available scientific dates (see notes 4 and 15 below). 2. The use of ‘prestige goods’ is seen as a key marker of increasing social complexity and competition in societies and is often associated with a move towards what we would commonly term ‘chiefdoms’. Prestige goods can be defined as: ‘a limited range of exchange goods to which a society ascribes high status or value’ (Renfrew and Bahn, Archaeology, 582). 3. A rich burial with jade yazhang and other objects at Tai Wan, Lamma Island was found within a ‘double-F’ (Bronze Age) cultural horizon; Tang Chung, A Journey into Hong Kong’s Archaeological Past (Hong Kong: Regional Council, 1991), 87–89, although a radiocarbon date potentially associated with the grave gave a date of c. 1700 bce (Meacham, Archaeology, 113). The Bronze Age community at this important site also had access to imported bronze daggers and halberds (dagger-axes). 4. This phenomenon is identified on the basis of seventeen radiocarbon dates, which suggest a gap in activity between the ‘end’ of ‘Bronze Age’ cultural activity around 700 bce and the few Western Han sites beginning around 100 bce; Meacham, Archaeology, 130–35. 5. Shang Zhitan, ‘Preface’, in Collected Essays on the Culture of the Ancient Yue People of South China, ed. Chau Hing-wah (Hong Kong: Urban Council, 1993), 10–11. 6. Hai Dei Wan: Bernard Williams, ‘Hai Dei Wan’, JHKAS VIII (1980): 27–51; Kwo Lo Wan: Meacham, Chek Lap Kok, 155–86; Man Kok Tsui: S. G. Davis and Mary Tregear, ‘Man Kok Tsui Archaeological Site 30, Lantau Island, Hong Kong’, Asian Perspectives 4 (1960): 183–212; Sha Ha: AMO, Sha Ha; Sham Wan: Meacham, Sham Wan; Shek Pik: Chen, ‘Archaeological Excavations’, 5–6; Walter Schofield, An Archaeological Site at Shek Pik (1975); So Kwun Wat (AMO 2010); Tai Wan: Finn, ‘Lamma Island, Parts I–XIII’; Au Ka-fat et al., ‘Preliminary Report on the Excavation of Tai Wan Site, Lamma Island’, in Cultures of South China and Neighboring Regions: Essays in Honor of Professor Cheng Te-k’un on the Occasion of the Sixtieth Anniversary of His Academic Career, ed., Tang Chung (Hong Kong: Chinese University Press, 1994), 195–208; Tung Wan Tsai South: Pamela R. Rogers et al., ‘Tung Wan Tsai: A Bronze Age and Han Period Coastal Site’ (Hong Kong: AMO, 1995). 7. Lam Tsuen Valley: Steven W. H. Ng et al., ‘Archaeology Investigation at Lam Tsuen Valley, Hong Kong’, JHKAS XV (2002): 33–47. 8. Bronze splashes at Hung Shing Yeh: Finn, ‘Lamma Island, Part V’, 123. Bivalve moulds and bronzes at Kwo Lo Wan: Meacham, Chek Lap Kok, 161; Shek Pik: Chen, ‘Archaeological Excavations’, 5–6; Schofield, Shek Pik, 66. 9. Tang Chung, ‘Study of the Bronze Dagger with Bow-Shaped Guard and Human Mask Design Discovered at Shek Pik of Lantau Island in Hong Kong’, in Collected Essays on the Culture of the Ancient Yue People in South China, ed. Chau Hing-wah (1993), 101. 10. ‘Elite-level trade and exchange’ is something we expect to see when societies become more complex, competitive, and stratified and a few more powerful or wealthy individuals seem to have preferential access to certain, often exotic or rare objects or materials. Other indicators of such socio-political changes might be the emergence of craft specialisation (i.e., metallurgy, stone-working, or pottery production) and an increasing interest in personal ornamentation and display, which often carries over into mortuary rituals. 11. Provenancing is the study of raw materials and artefacts, usually through laboratory analysis, to identify the item’s chemical ‘fingerprint’ and geographical origin. Once that is known, it is then possible to plot how such materials or artefacts were distributed across a region and where they were consumed (used). Patterns and mechanisms of trade and exchange can then be deduced. 12. Meihuadun and Yingang are the classic Guangdong kiln sites producing many types of high-fired stonewares found on Hong Kong sites. The latter kiln also produced some types of animal figurine found locally. Yingang: Gu Yunquan, Li Ziwen, and Deng Hongwen, ‘Excavation of the Yingang Site at Boluo in Guangdong’, Wenwu 7 (1998): 17–30; Meihuadun: Liu Chengji and Yang Shaoxiang, ‘Excavation on the Kiln Site at Meihuadun, Yuanzhou, Boluo, Guangdong’, Kaogu 7 (1998): 28–43. 13. The Chinese Ceramics Society, ed., Zhongguo tao ci shi (Beijing: Cultural Relics Publishing House, 1982), 77. Based on scientific analysis, some scholars further refined their definition to include pottery with a ‘“china-stoned” [kaolin?] body, glazed and having relatively lower water absorption rate’, Luo Hongjie and Li Jiazhi, ‘Preliminary Discussion on the Definition of Proto-porcelain’, Kaogu 7 (1998): 71. 14. Ye Hongming and Cao Heming, ‘Views on the Origin of Chinese Porcelain’, Wenwu 10 (1978): 85–86; Sun Tianjian, ‘The Invention of Proto-porcelain and Its Significance as a Milestone’, China Ceramics 39 (3) (2003): 60–61. 15. Wang Changsui, Li Wenjing, and Chen Yue, ‘The Re-exploration of the Concept of “Proto-porcelain” and the Origin of the Celadon’, Kaogu 9 (2014): 91. 16. Zhu Feisu, ‘Excavation and Phasing of the Shixia Site’, A Collection of Studies on Archaeology 10 (2013): 599. 17. Guangdong Provincial Institute of Cultural Relics and Archaeology (GPICRA), Henglingshan in Boluo County: Excavation Report of Shang and Zhou Period Cemetery in 2000 (Beijing: GPICRA, 2005). 18. The 302 Bronze Age graves in the Henglingshan cemetery included 588 complete pots of which 426 bore inscribed marks using 236 different symbols. The monograph includes a discussion of such symbols and offers a range of possible interpretations including personal names, markers of individual potters or owners or (the excavator’s choice) different teams of pottery artisans working at the kilns. GPICRA, Henglingshan. 19. Two key characteristics for ceramic cooking pots used directly over open fires are embodied in the heavily tempered, thick-walled, and round-bottomed jars and basins used throughout Hong Kong prehistory: toughness and resistance to thermal shock. They worked so well that they did not change in terms of their round-bodied form and coarse tempered fabric until earlier historical times (e.g., Han-to-Tang coarse cooking pots have different surface decoration and finishes, but the forms are little different).



Bronze Age: Technology, Trade, and Increasing Social Complexity   83

20. Meacham uses the ‘asterisk’ pottery evidence to argue for a real gap in Late Bronze Age activity, and as a basis for dismissing as ‘outliers’ two ‘late’ radiocarbon dates (780–380 bce and 368 bce–24 ce) produced by deposits containing typical Middle Bronze Age deposits; Meacham, Archaeology of Hong Kong, 130–35. 21. ‘Oxen’ at Man Kok Tsui and Hai Dei Wan: Peter Y. K. Lam, ed., Archaeological Finds from Pre-Qin Sites in Guangdong (Hong Kong: CUHK Art Gallery, 1984), 258–59; and at Sha Po Tsuen: Meacham, Archaeology, 111. ‘Dog’ at Hai Dei Wan: Chau, Ancient Yue, 233. ‘Sheep’ at Tai Wan: ibid. 22. For example, Chen Wen, ‘Bronze Age of Hong Kong and Its Periodisation’, Huaxia Kaogu 3 (2002): 77; Meacham, Archaeology, 111. 23. Meacham, Chek Lap Kok, 176, 184–85. 24. Meacham, Archaeology, 121. 25. These could be finished items, for example, to be worn on a necklace, but could also be earrings processed up to their final sawing stage of manufacture. 26. For example, Davis and Tregear, ‘Man Kok Tsui’; AMO, Sha Ha. 27. Shek Pik: Schofield, Shek Pik, Plate CXIX; Sham Wan: Chau, Ancient Yue, 232–33. 28. In China and many other parts of the world, residue analysis has been highly successful in identifying a wide range of foodstuffs preserved, for example, as fatty acids (lipids) or starch grains in pottery vessels (see examples in Renfrew and Bahn, Archaeology, 271–73; in particular 9,000-year-old ‘rice wine’ from the Early Neolithic village of Jiahu in Henan). 29. Rogers et al., ‘Tung Wan Tsai’, 56–65. 30. Schofield, Shek Pik, 33. 31. While Shek Pik is unquestionably an important Bronze Age site with all the diagnostic artefacts of that period, the six burials may somewhat pre-date that main phase of occupation and fall into the aforementioned transitional stage spanning the traditional boundaries between Neolithic and Bronze Age. 32. Meacham, Chek Lap Kok, 156, 179–80. 33. Meacham, Archaeology, 118. 34. GPICRA, Henglingshan, 518–19. 35. Tang, Archaeological Past, 87–89. 36. Eleven undisputed Bronze Age rock carvings are known and all are afforded the highest level of legal protection available in Hong Kong as Declared Monuments. 37. William Meacham, Rock Carvings in Hong Kong (Hong Kong: William Meacham, 2009b), 106–17. 38. Chen, ‘Archaeological Excavations’, 5–6. 39. In 1958 Finn’s thirteen articles—focused mainly on the Tai Wan site, but also Hung Shing Yeh and Yung Shue Wan—were collected together in a volume Archaeological Finds on Lamma Island near Hong Kong, 1933–1936, which was edited by Father T. F. Ryan and published by the University of Hong Kong. In this book, we prefer to use the original thirteen articles as our source. 40. Finn, ‘Lamma Island, Part X’, 256–62; ‘Lamma Island, Part XIII’, 259; C. J. Liu, ‘Excavation on the Ancient Site at Sha Po Tsuen, Lamma Island, Hong Kong’, Kaogu 6 (2007): 10–29. While green glaze quite regularly occurs on stem cups and much more rarely on jars, it was almost certainly something achieved by design during the firing process; in contrast, the much rarer ‘brown glaze’ reported by Finn on double-F storage jar sherds may in fact be examples of so-called ‘kiln glost’ caused accidentally when fuel ash dripped from the kiln ceiling onto the upper body of pots during firing. Margaret Medley, The Chinese Potter (Oxford: Phaidon, 1989), 40–41; Peter Y. K. Lam (personal communication). 41. Stem cups: e.g., HKAS 1989 excavation archive (C.16.20, C.16.21, and C.16.22), AAL 2002 excavation archive (2002.028.00235), and AMO 2004 excavation archive (SF135). Jars: e.g., HKAS 1994 excavation archive (C.16.95), and AMO 2004 excavation archive (SF115 and SF118). 42. Liu, ‘Sha Po Tsuen’, 27; Liu and Yang, ‘Meihuadun Kiln Site’, 28–43; Gu et al., ‘Yingang’, 17–30. 43. Finn, ‘Lamma Island, Part XIII’, 259. 44. 1930s ‘horse’: Finn, ‘Lamma Island, Part XI’, 54–58; 1995 ‘Ox’: Meacham, Archaeology of Hong Kong, 111. The latter was coded ‘NLS’ in the HKAS archive and is therefore presumed to come from the top of the plateau near North Lamma School; 2004 ‘Ox’: Liu, ‘Sha Po Tsuen’, 19. 45. Finn, ‘Lamma Island, Part XIII’, 259. 46. Finn, ‘Lamma Island, Part XI’, 54–58. 47. Finn, ‘Lamma Island, Part XI’, 58. 48. See also T. Y. Jao, ‘Prehistoric Remains in South China and the Culture of Yin-hsu’, Ta-lu Tsa-chih 8 (1954) as cited in Meacham, Archaeology, 115. 49. Liu, ‘Sha Po Tsuen’, 19. 50. Commensals are animals, such as house mice and rats, which live with another species (people) but neither is dependent on the other for survival. Harper Collins, Collins English Dictionary (London: Harper Collins, 1994). 51. Here, we are also reminded of the hard pottery lamb found in Bronze Age deposits in the next bay south from Sha Po at Tai Wan. Sheep are not recorded in any site locally and were probably never kept as a domesticated animal in the study region. 52. Greger Larsen et al., ‘Patterns of East Asian Pig Domestication, Migration, and Turnover Revealed by Modern and Ancient DNA’, PNAS 107 (17) (2010): 7686–91. One figurine discovered at Hai Dei Wan has been described as a dog (see Chau, Ancient Yue, 233), but the head much more closely resembles that of a water buffalo or an ox as suggested in an earlier publication (Lam, Finds from Pre-Qin Sites [1984]: 258–59).

84   Piecing Together Sha Po 53. Most long-distance trade and exchange in the ancient world, in particular in heavy bulky items such as pottery, used water transportation, in this instance down the East River (Dongjiang). The distance was estimated using the measurement feature on Google Earth, which incidentally gave an ‘as-the-crow-flies’ distance of around 100 km. 54. Site [25]: AMO, ‘Excavations on the Ancient Site at Sha Po Tsuen, Lamma Island, Hong Kong’ (2007b). 55. The post-holes were suboval to circular in plan and most varied in size from c. 12 to 25 cm ‘diameter’, with one or two others being much larger at between 50 to 70 cm ‘diameter’, while depths ranged from 8 to 54 cm. 56. The use of stilt-houses in Hong Kong prehistory is another aspect of largely unproven ‘conventional wisdom’, where scholars have suggested they ‘must have’ existed and the probability has become reality as a result of the retelling (e.g., in Shang Zhitan and Ho Ching-hin’s prefaces to Collected Essays on the Culture of the Ancient Yue People of South China, edited by Chau Hing-wah [Hong Kong: Urban Council, 1993], 10–11). The Sha Po plateau is nevertheless a distinct possibility and we present it in this book as just that (a possible stilt-house settlement) but Sha Po, like many other sites, just does not have enough evidence to be certain, which is extremely frustrating. 57. Meacham, Archaeology, 115–16. 58. On well-preserved sites, this activity can be identified by elongated surface outlines of post-holes caused by posts being pulled over and then dragged out to be reused elsewhere; however, we are unaware of any instance where this has been identified (or recognised) locally. 59. Experience elsewhere in Hong Kong (e.g., Sha Ha) suggests that even with relatively large area exposures the task of defining individual structures can be difficult, and more so when isolated smaller-scale excavations are employed. However, the Sha Po excavation areas were determined by the nature of the development (Small House), so we and the excavators had to work with the data before us. 60. Here we must acknowledge a scholarly debt to Father Daniel Finn, whose remarkably detailed research into the manufacture of North Lamma rings and the techniques and processes that produced them is still full of useful insights almost eight decades after his death. Finn, ‘Lamma Island, Part VIII’, 40–47; ‘Part XIII’, 259. 61. Liu, ‘Sha Po Tsuen’. 62. We wonder whether such rotary polishers were more effectively deployed in the manufacture of ornaments made from the somewhat softer grey schist, which might then account for their absence on the plateau. This hypothesis requires further research, perhaps through examination of stone residues on the surface of rotary polishing stones. 63. Tang Chung agreed with Finn about the use of bamboo drills and also emphasised the meaning of ring-polishing stones as one of the drilling tools. Tang Chung and Cheng Wai-ming, Archaeological Investigation at Hac Sa, Coloane, Macau (Hong Kong: Chinese University Press, 1996); Tang Chung, ‘Prehistoric Tubular Boring Technique in East Asia’, in Prehistoric Jade Carving Techniques, ed. Qian Xianhe and Fang Jianneng, 145–56 (Taipei: National Taiwan Museum, 2003); Emily H. Y. Chan, ‘More Than Meets the Eyes: Slotted Rings and Social Complexity in Bronze Age Hong Kong’ (unpublished final-year project, Department of Anthropology, CUHK, 2012). Y. T. Xiao, The Pre-Qin Island Inhabitants of the South China Sea: A Study on Seashore Sand Dune Sites (Beijing: Cultural Relics Publishing House, 2004), 136–40. Wan-cheung Wong, ‘Discussion of the Slit Ring Manufacturing Technology in the Pearl River Delta: A Case Study of Pak Mong Site, Hong Kong’ (MA diss., Xiamen University, 2009). 64. In finds from Pak Mong there is evidence for the basic shaping of quartz rough-outs by cutting (sawing in our terminology), in addition to the more widespread use of chipping (Wong, ‘Pak Mong’, 26). 65. Recent discoveries by the authors at Pak Kok Tsui at the northern tip of Lamma suggest one possible source of good quality white quartz in the island. Mick Atha, ‘Survey-cum-Excavation at Pak Kok Tsui, Lamma Island’, JHKAS XVII (forthcoming). 66. Finn sensibly argued that many of the so-called thin ‘knives’ found on workshop sites were actually ‘saws’ used, probably with fine sand and water, to cut slots in the earrings. Although the blades have no teeth, we feel that ‘saw’ more accurately reflects the function rather than knife, which suggests the action of a sharp cutting edge. 67. Finn, ‘Lamma Island, Part XIII’, 258. 68. Observations based on study of AMO site archives and artefacts. A similar pattern was also observed in Trench AA3 [35: 306 and 307] excavated in 2009; AAL, Sha Po Tsuen, 30–31. 69. Respectively: Meacham, ‘Sha Po Tsuen’, 47; AAL, Sha Po Tsuen. 70. Ibid. Apart from the rhinoceros, the vertebrate remains are fairly typical of prehistoric sites where such evidence survives, but the former is a rare and unusual find. We would highlight the fact that the animal remains in Bronze Age layer 536 were directly beneath much richer midden deposits of apparent early historical date, and therefore might relate to that later activity (AAL, Sha Po Tsuen, 45). 71. Meacham, ‘Sha Po Tsuen’, 38; There is no information as to whether the bones were mammal or otherwise, but their burning may be an indication of cooking, or if human perhaps cremation. 72. Future starch and phytolith analysis of stone tools in Hong Kong will hopefully provide some more solid evidence for the range of plants being exploited throughout prehistory. 73. One spearhead found in 1989 [11:L4]: Meacham, ‘Sha Po Tsuen’, 43; an arrowhead in 2002 [22a:02]: AAL, Sha Po Tsuen, 26; a spearhead and two arrowheads in 2012 [48:C3]: AMO, unpublished archives. 74. Another example of the same type was found in Bronze Age deposits at Sham Wan (Meacham, Sham Wan, 227). 75. A similar ring-shaped net-weight was found at Man Kok Tsui: Lam, Finds from Pre-Qin Sites, 256–57. 76. The only other example was found by Chen Kung-che at Shek Pik. 77. This is similar to examples found in Lung Kwu Sheung Tan: William Meacham, ‘Report on Salvage Excavations at Lung Kwu Sheung Tan 1990’, JHKAS XIII (1993b): 23, Figure 13; Lam Tsuen: Ng et al., ‘Lam Tsuen’, 42; and Ngau Tam Mei: Au Ka-fat and Mo Zhi,



Bronze Age: Technology, Trade, and Increasing Social Complexity   85

‘Archaeological Investigation Report on the Main Drainage Channels for Ngau Tam Mei Project, Phase II’, unpublished report, 1999. Such bold and closely packed asterisk patterns are a distinctive type not seen beyond the early Western Han (Zhu Hairen, ‘Research on Han Archaeology in Hong Kong’, in Archaeology of Western Han Southern Yue Kingdom and Han Culture, ed., Chinese Academy of Social Science and Institute of Cultural Relics and Archaeology of Guangzhou (Beijing: Science Publishing, 2010), 47. 78. ‘Structured deposition’ is a term coined by J. D. Hill (1995) in his seminal study Ritual and Rubbish in the Iron Age of Wessex, BAR British Series Vol. 242 (Oxford: Tempus Reparatum). 79. AAL, Sha Po Tsuen, 36; see also Catalogue of Selected Finds No. 28a–d. 80. This contrasts sharply with Shek Pik where the six individuals were buried with their heads towards the sea. 81. Shells have been recorded elsewhere in Hong Kong as items of grave goods: Schofield, Shek Pik, Plate CXVIII; Chau, ‘North Tungwantsai’, 12–13. And also in the Pearl River Delta: Li Ziwen, ‘Preliminary Report on the Excavation of the Shell Midden Site at Yinzhou, Sanshui City, Guangdong’, Kaogu 6 (2000): 27. Indeed, shells appear as grave goods in prehistoric sites from Hong Kong all the way west to peninsular South East Asia (Charles Higham, personal communication). 82. AAL, Sha Po Tsuen, 36; see also Catalogue of Selected Finds Nos. 11a and 11b. A very similar stem cup (M272.2:3) and several comparable small jars (M068.1, M110.4, and M111.4) were found in graves at Henglingshan cemetery in Boluo County, Guangdong; GPICRA, Henglingshan. 83. AAL, Sha Po Tsuen, 27–28; see also Catalogue of Selected Finds Nos. 10a and 10b. A very similar small jar was found at Shek Pik in 1937 in a burial with well-preserved ‘Skeleton VI’, whose grave goods also included a sandstone saw, nine particularly fine fluted projectile points, and two trapezoidal adzes, one of which was stepped; Schofield, Shek Pik, 40 and 71 and Plates CXIX and CXX Nos. 3 and 4. 84. Tai Wan, Lamma Island: Finn, ‘Lamma Island, Part III’, 151. 85. Hung Shing Yeh: Finn, ‘Lamma Island, Part V’, 123. 86. Brian A. V. Peacock and Taryn J. P. Nixon, The Hong Kong Archaeological Survey (Hong Kong: AMO, 1986), 116. 87. Meacham, ‘Sha Po Tsuen’, 40, 48–53. 88. AMO, Sha Po Old Village, 45. See also Catalogue of Selected Finds No. 26. 89. The discovery of two pairs of casting moulds in a comparatively rich grave (Burial 2) at Kwo Lo Wan on Chek Lap Kok might mark it out as that of a bronze artisan-metallurgist. Meacham, Chek Lap Kok, 162.

6 On the Edge of Empire: Han–Yuan Sha Po

Introduction In this chapter, we present the evidence for earlier historical activity at Sha Po spanning roughly the second to thirteenth centuries CE. We review the broad historical-archaeological themes shaping this period and provide an updated analysis based on the excavated detail from Sha Po. The short-lived dynasty made famous by Qin Shi Huang and his Terracotta Army, left no trace in Hong Kong,1 and our exploration therefore begins in the Han dynasty, when our study region was absorbed by imperial China, albeit as a coastal periphery still inhabited by maritime-focused communities of Yue people.2 We then move on to examine the relatively abundant evidence for Six Dynasties–Tang activity, in particular associated with the important salt-lime industry that dominates historical sequences at many of the region’s backbeach sites. We open our Six Dynasties–Tang discussions, though, by reviewing the evidence for subsistence as provided by midden deposits, followed by an examination of the two Six Dynasties burials found on the backbeach. Thereafter, the main presentation focuses on the construction and use of Sha Po’s Six Dynasties–Tang kiln industry, followed by its abandonment and the area’s after-use in the Northern Song to Southern Song–Yuan periods.

Life at the Edge of Empire: Han Dynasty Introduction Hong Kong was notionally part of the Southern Yue Kingdom during the earlier Western Han and was absorbed by the Han upon their conquest of the region in 111 BCE.3 However, historical references to the Hong Kong region are extremely rare but one crucial example in the Han History (Han Shu) states that in Panyu Prefecture, which encompassed Guangzhou and the eastern side of the Pearl River Delta including what is now Hong Kong, ‘[t]here is . . . a Salt Official’, while another one looked after the area west of the Pearl River.4 In terms of archaeological remains, at the time of writing just six of seventeen sites have slight traces of Western Han date, while twelve have thin spreads of Eastern Han material.5 The range of objects recovered includes so-called ‘seal-stamp-over-net’ storage jars—the classic ‘type fossil’ of the local Eastern Han—and coarse round-bottomed cooking pots, occasional iron tools, copper alloy arrowheads, and coins.6 Almost all known Han sites are on coastal backbeach or lower hill slope sites in the same topographic zones favoured by later prehistoric groups, and a similar coastal lifeway is thus implied.7 That pattern of small-scale coastal communities suggested by the archaeological evidence is therefore seriously at odds with the affluence and power represented by the Lei Cheng Uk Eastern Han tomb, which was unearthed in 1955 in West Kowloon during site formation works for one of Hong Kong’s first public



On the Edge of Empire: Han–Yuan Sha Po   87

housing estates.8 Many similar tombs have been found in the environs of the major Han settlement at Guangzhou,9 whereas Lei Cheng Uk is as yet unique in Hong Kong. Given Hong Kong’s widespread transformation of ancient coastal landforms, if other tombs existed, in particular around the Kowloon peninsula, one might expect at least one or two to have been discovered, but this has not been the case. Han dynasty Hong Kong thus seems to indicate the continuity of a society comprising maritime-focused indigenous coastal communities living under the watchful eye of one or more high-status imperial officials—probably connected to salt production—presumably with administrative and military support of some form or another.

Sha Po’s enigmatic traces of Han activity The Han materials unearthed on the Sha Po backbeach, although including a small number of characteristic artefacts, were typically sparse and enigmatic in that no in situ layers or features were identifiable. The main focus of Han dynasty activity was on the central-southern backbeach, although the materials either occurred mixed into the surface of Bronze Age deposits or in later layers associated with industrial activity.

Continuity of backbeach domestic activity? The pottery assemblage is a typical mix of Han domestic wares, including hard cooking pots such as that found in 1994 [13:02], which had a distinctive everted, dished rim, and round bottom with bold cord decoration. According to Zhu, it resembles a pot found on Kau Sai Chau, which should be later Eastern Han in date.10 One side of a large cooking pot of similar form, but in a harder fabric, was found nearby in the surface of a Bronze Age deposit [34:F536].11 The dished rims have the appearance of a lid-seating, but no lids were found. Seal-stamp-over-net storage jars were represented by a flat-bottomed base with near-vertical walls found in pre-Tang deposits on the southern backbeach [11:03]12 and another small bodysherd recovered nearby from post-Tang layers [34:D529].13 Also, the entire side profile of a typical hook-rimmed storage jar with sloping shoulders and near-vertical walls—but no decoration—was found in 1995.14 The arrival of iron at the coast Associated with the aforementioned corded cooking pot was a three-winged, socketed copper alloy arrowhead whose form resembles others found at Tung Wan Tsai (South)15 and by Maglioni in Eastern Guangdong.16 Based on such inter-site parallels and intra-assemblage associations, the arrowhead is likely to be late Eastern Han in date. Nearby in the same layer was a socketed iron axe-head of a type in use from the Warring States to Qin–Han periods.17 Perhaps the most interesting artefacts of this period, though, were two iron hoe-heads (cha) found on the northern [16:T1L5]18 and west-central backbeach [26:T2L1],19 in contexts dated to the Sui–Tang period and Song dynasty respectively—although the latter also overlay Sui–Tang deposits. Two main cha types are known: one referred to as a ‘straight-lined cha’ and the other a ‘U-shaped cha’20—both Old Village examples are of the latter type. The northernmost cha is an important and quite remarkable survival given its thin-walled design, which is similar to another found in Dongpingcun, Sichuan dated to the Eastern Han,21 but unlike the latter the Sha Po example bore no inscription. Despite being thicker in cross-section, less than half of the second hoe-head survived,22 but an attempted reconstruction is shown below. Lastly, in 1994 a badly worn coin was discovered in the same layer as other Han materials and was therefore suspected to be a Wuzhu type [13:02].

Figure 16:  Han pottery: (a) hard pottery storage jar [34:536; CSF13], (b) rubbing of a c. 1.5 cm square seal stamp from a storage jar sherd [34:529], (c) corded hard pottery cooking pot [13:L2], (d) hard pottery storage jar (plain) [HKAS 1995: Square M, L3]

Plate 15:  Han corded hard pottery cooking pot. © Antiquities and Monuments Office, reproduced with permission.

Plate 16:  Han socketed iron axe. © Antiquities and Monuments Office, reproduced with permission.

Plate 17:  Iron hoe (cha)

90   Piecing Together Sha Po

Discussion The pottery assemblage is fairly typical of Hong Kong sites, such as Kau Sai Chau and Tung Wan Tsai South, in that it is small in size and comprises types best interpreted as utilitarian domestic wares that tended not to be used as burial objects. All the vessels found would be used for the cooking and storage, rather than consumption, of food. Interestingly, while iron cha are known to have been used as early as the Warring States period north of Lingnan,23 archaeological findings in Guangdong suggest that it was only during the Han dynasty that iron tools became more commonly available in the south.24 This trend is reflected, for example, by three fragments of iron ‘spades’ within the Han finds assemblage from Tung Wan Tsai South, which look suspiciously like socketed cha. Such cha were mounted on long wooden handles and are conventionally interpreted as farmers’ tools used for hoeing and other tasks such as cutting drainage channels.25 However, their discovery near Sui–Tang industrial deposits—most intimately in a layer pre-dating or contemporary with Kiln K1 (see below)—is intriguing and may indicate an association with the coastal kiln industry rather than agriculture, which seemingly arrived much later in Lamma Island.26 The discovery of ‘two large axes or hoes’27 associated with Tang kilns at Sham Wan Tsuen—one of which is almost certainly a cha, but ‘square-U’ in form—perhaps adds to this argument. The bulk of the above finds came from later (mostly Six Dynasties–Tang) contexts, which in the case of the single seal-stamped potsherd discovered in 1989 was taken to indicate the possibility of continued use of Han pottery ‘in the immediate post-Han period’.28 Another possibility is that Han period activity did occur, but it was so short-lived, temporary, and materially poor that any resulting ‘cultural horizon’ would at best be thin and discontinuous. When one factors in the loose, sandy layers of the backbeach cultural sequence—whose diffuse boundaries are notoriously difficult to identify in section, never mind when excavating in plan—such a slight Han deposit would be virtually impossible to discriminate archaeologically from thicker and more artefact-rich Six Dynasties–Tang deposits lying on top, or equally rich ‘Bronze Age’ deposits underneath. One might interpret the findings above as evidence for the local Yue people’s quite limited access to Han material culture, or as an indication of a very limited Han Chinese presence at Sha Po, perhaps in the form of a few administrators and military overseers of an early version of the Imperial Salt Monopoly, which is thought to have existed in Hong Kong at this time.29 In sum, the Sha Po findings reflect the more general pattern evidenced at many local sites, which seemingly indicate that Hong Kong was part of a wider coastal region experiencing increasing interaction with Han Chinese authority, but was certainly not an area experiencing mass immigration of inland farm populations.

Six Dynasties–Yuan Coastal Industry and Maritime Trade Historical background The Six Dynasties–Yuan period in coastal Lingnan is characterised historically in terms of the activities of the Imperial Salt and Pearl Monopolies and the rich maritime trade flowing through South China’s preeminent city-port on the site of modern day Guangzhou. During the Tang dynasty, Tuen Mun was a significant military base from which war junks patrolled the surrounding sea lanes. It was also the place where foreign merchant vessels were required to stop before travelling on upriver or further along the coast.30 In his in-depth study of salt-working in Hong Kong, Hase argues that despite the supposed abolition of the Imperial Salt Monopoly between 23 and 763  CE (Eastern Han to mid-Tang), third- and fourth-century references to a salt official in Dongguan—‘The Eastern District’ of Panyu Prefecture,



On the Edge of Empire: Han–Yuan Sha Po   91

which included Hong Kong—suggest that imperially controlled salt-working actually did continue on the Guangdong coast.31 A later record in Xin Tang Shu (New Tang History), which refers to salt-working in Dongguan in 758  CE, provides further evidence for the operation of the Imperial Salt Monopoly throughout the Six Dynasties–Tang period.32 During the Nanhan period (907–71 CE) the military base at Tuen Mun was improved and reinforced, Castle Peak became the kingdom’s Holy Mountain, saltworking was reorganised, and the Tai Po pearl industry reached its zenith.33 Song dynasty governments maintained strict control of salt production and supply as a key source of income for the imperial treasury. Early Southern Song historical records refer to a salt-field known as Kwun Fu Cheung (官富場), probably located in Kowloon Bay, which is thought to have been one of the biggest imperial salt production grounds in South China.34 An imperial garrison of 150 soldiers was established there by the Salt Official in order to suppress salt smuggling,35 probably throughout the entire territory of Hong Kong. The latter is suggested by a record of Kowloon troops being despatched to Lantau Island in 1200 CE by Qian Zhiwang (錢之望), the prefect of Guangdong, to put down a revolt and, as he put it, to ‘wipe out that nest of Yao’.36 The Joss House Bay rock inscription provides a more local historical source that records a visit to the area in 1274 CE of a ‘salt administration officer’.37 The Song dynasty ends with the famous and tragic tale of the Boy Emperors, within which locations traditionally associated with both Lantau Island and Kowloon Bay (Kwun Fu) are again mentioned.38 The former Sacred Hill (Sung Wong Toi) received its name through this association and the commemorative rock, which survives in the modern park of the same name, is believed to have been inscribed in the Yuan dynasty. The Song dynasty is also the era when, according to genealogical records, several clans first settled the best rice-farming land in the New Territories.39 The establishment and consolidation of major ricefarming clans in the New Territories can be related to a series of imperial grants of wasteland to prominent men, whose clansmen—and their tenants—then converted the land to arable.40 Thus in the twelfth century, the Kam Tin Valley was granted to the Tang clan and the middle Beas River Valley to the Haus.41 The availability of large areas of waste for such grants is significant as it suggests that those prime ricefarming lands had not by then been opened up for agriculture. So perhaps we can surmise that pre-Song populations continued to follow a pattern of cultivating and gathering starchy plants in the hinterlands of their coastal sites—perhaps using some form of slash-and-burn (swidden) agriculture—but were not settled farmers. Here it is interesting to note that the few Song to early Ming references to barbarian peoples in Guangdong Province record them using slash-and-burn agriculture.42 In 1318 CE a petition by Chang Wei-yen of Tai Po makes mention of ‘Yao’ and ‘Shan-lao’ hill-tribes, which is interesting in light of postwar aerial photographic43 and more recent LIDAR44 evidence for high-level terracing—remote from ricefarming villages and not the ‘herringbone’ pattern of tea terraces—on hillsides in the New Territories. Unfortunately, documentary evidence has little to offer us regarding Lamma’s settlement history during the Six Dynasties–Yuan periods, but Hase draws what may be a useful analogy with Lantau Island, about which early fourteenth-century Cantonese scholar Ng Loi said this: The [Yao] people live by fishing and salt making, they do not practise agriculture [i.e., rice farming]. In the Shiu Hing period [1131–1162 CE], the young and strong were removed and conscripted into the Navy. Since then the island has remained empty . . . Today the island has just a couple of hundred families involved with it. They dig yams and hunt deer, which every so often they take to the City to exchange for rice.45

Ng Loi recorded the lasting impact on Lantau of what Hase termed ‘Imperial punishment raids’, but we can only speculate whether the enforced depopulation of Lantau was, as Hase has tentatively suggested, also implemented on Lamma.46 But the reference to the Lantau people’s non-agricultural lifeway

92   Piecing Together Sha Po

is, nonetheless, very interesting and there are unquestionably far fewer materials of diagnostically Yuan to Ming date at Sha Po, on Lantau, and at many other sites across the region when compared with either preceding or later periods.

Six Dynasties–Yuan archaeological overview That historical background finds meaningful parallels in a Six Dynasties–Tang archaeological record that is dominated by a kiln-based industry, which must have been a very prominent feature of coastal landscapes at that time.47 While limited trenching at many sites has revealed perhaps one or two kilns, more extensively excavated sites have revealed what appear to be planned complexes of up to thirteen kiln structures arranged in rows, although the construction of Kiln 11 over Kiln 12 at Sham Wan Tsuen offers a rare example of multiple phases of industrial site use.48 The discoveries of shells and coral and mounds of solidified lime in some kilns,49 together with occasional structures interpreted as slaking pits,50 led to the industry’s understandable association with the manufacture of lime.51 However, there are some serious question marks over the kilns’ use primarily for the manufacture of lime. Firstly, there is the kilns’ open-topped, cylindrical design with external plastered floors, elongated surface stoke-channels, and patterns of radial supporting brackets, which collectively seem ill-suited to the relatively simple process of burning lime (see generic kiln design in Figure 17 below). Next there is the fired clay kiln furniture (see Plate 22) that comes in five main forms—flat bars (possibly fragmented radial supporting brackets), cylindrical bars, flat-ended props, squat props, and irregular spacers—which collectively seem unnecessary for lime burning, but are perhaps rather better suited to supporting containers of some sort. Lastly, although lime was unquestionably associated with the kilns, the reasons for the occurrence of two different types of lime residues in different contexts of discovery have also never been adequately explored. Doubts over the ‘lime burning hypothesis’ led local Chinese scholars to reconsider the archaeological remains in light of historical records, which mentioned a Han–Tang Imperial Salt Monopoly in our study region and, in Yuan–Ming times, discussed the use of large limecoated woven bamboo pans with brine boiling ‘stoves’, leading to an alternative interpretation focused on salt.52 Perhaps more compelling still is the evidence presented in a late Tang document Ling Biao Lu Yi (嶺表錄異) by Liu Xun, which records contemporary salt-working along the Lingnan coast.53 The process described involved the concentration of brine by leaching,54 followed by flame testing of salinity levels, and then boiling down of brine using bamboo basins coated with ‘oyster ash’ (lime?). Based on our own review of the evidence, we would support the idea that salt was the main purpose of the industry, while the lime found in some kilns was probably a process related by-product, but could of course also have had many other uses beyond the industry. For now we must leave it there, but we will develop our argument more fully as we work through our detailed examination of the Sha Po evidence in the passages below. While physical evidence of the Song salt industry remains elusive, many Six Dynasties–Tang kiln sites have post-abandonment spreads containing a mixture of late Tang, Northern Song, and Southern Song–Yuan pottery. Some continuity of site use is thus indicated, but it seems the salt industry itself was probably reorganised during the Song dynasty into fewer, larger centres using different technology and the kiln-based industry consequently died out. A major factor behind that reorganisation may have been the depletion of coral and shell sources used for the making of the lime by-product, but the historically attested difficulty of controlling salt smuggling across the myriad of backbeach kiln sites scattered around the coast may also have prompted a move towards a more centralised mode of production. Material evidence directly connecting Hong Kong with the vibrant Six Dynasties–Tang maritime trade of that time has mainly been limited to the presence in excavations of ceramics—many of them



On the Edge of Empire: Han–Yuan Sha Po   93

Figure 17:  Generic kiln design

produced in Guangdong,55 but also Zhejiang and Hunan—that were also exported as tradewares throughout South East Asia, and even further afield.56 The production of tradeware ceramics and their distribution along the ‘Maritime Silk Route’ dramatically expanded and intensified during the Northern Song to Yuan period. A review of Song dynasty sites in Hong Kong identified roughly forty locations that had produced tradeware ceramics and other materials of that date.57 In contrast to the widespread evidence for industrial activity and the consumption of tradeware ceramics, there has not been a single Six Dynasties–Tang settlement site found in Hong Kong,58 which has encouraged notions of a boat-based or stilt-house dwelling population that left no recognisable traces of land-based settlement.59 That lack of pre-Song permanent settlement also fits with the aforementioned existence of vast tracts of wasteland granted by imperial decree to prominent men in the Northern Song. However, in contrast to the widespread occurrence of Song–Yuan ceramics, settlement sites of the period, such as that found at Shek Kong, are relatively rare discoveries.60 But as we write the recent redevelopment of the To Kwa Wan–Kowloon Bay area is continuing to yield dramatic discoveries of settlement and cemetery remains, and vast quantities of tradeware ceramics, which strongly reflect the area’s supposed importance as a regional imperial power base as is suggested in historical texts.61 Turning to funerary activity, the recent discovery of what appears to be an extensive late Tang cemetery at San Tau in North Lantau62 has significantly augmented the previously very limited funerary evidence for Hong Kong’s Six Dynasties–Tang period. The grave goods recovered—in particular, fine bronze belt fittings, off-cuts of silver ingots, and large iron-bladed weapons—seem to suggest that the cemetery was

94   Piecing Together Sha Po

used by military and civilian outsiders connected to the maritime trade rather than an indigenous community, but the presence of kiln debris also indicates industrial activity in the area.63 Beyond San Tau, there are very few Six Dynasties–Tang burials known in Hong Kong,64 but two were discovered at Sha Po and are discussed below. Song–Yuan burials were similarly scarce until the exciting discoveries made recently at To Kwa Wan.65 Previous scientific dating of a small number of kilns gave a date range spanning the fifth to tenth centuries CE, which correlates well with the dates suggested by ceramics and coins found in association, although the bulk of that material is Tang dynasty in date. While it is extremely unwise to draw simplistic parallels between historical records and archaeological discoveries, we nevertheless think that in a salt-making region of supposed importance in the Han–Six Dynasties–Tang period, and undeniable prominence thereafter, the widespread and consistent archaeological evidence for a kiln-based coastal industry is much more likely to be connected to salt-working rather than lime burning, about which history is comparatively mute. That and other issues are addressed below in our discussion of the Sha Po evidence, which begins with subsistence, then covers funerary activity, and finally the industrial remains.

Life, death, and industry in Six Dynasties–Yuan Sha Po Introduction While the Han dynasty was a period of relative quiet on the backbeach, the following Six Dynasties to Tang period witnessed a dramatic increase in activity centred upon the coastal industry introduced above. At Sha Po the kiln industry spread out across a backbeach which, probably as a result of a number of major typhoons, had almost doubled in width in the millennium or more since the height of the Bronze Age. Virtually every area excavated on that expanded backbeach has produced quantities of kiln debris, but it was only within the last decade that several clusters of kiln structures came to light as a result of development-funded archaeology. Beyond the very obvious remains of kiln structures and associated industrial residues, the site has also produced significant artefactual evidence reflecting the daily lives of the kiln workers. In addition, midden deposits found at Sha Po offer insights into at least some aspects of early historical diet, while two burials provide an opportunity to get closer to the people themselves and their belief systems. Given the evidence for continuity between use and post-abandonment phases, we felt it made more sense to approach the Six Dynasties–Yuan period, not using strict periodic subdivisions, but rather through a series of thematic headings addressing interrelated spheres of cultural activity: focusing first on the evidence for past subsistence and diet, then mortuary behaviour, and lastly the main discussion covering the construction, use, and abandonment of the kiln industry. Finally, all the strands of evidence for this important chapter in Sha Po’s human story are drawn together in a closing synthetic discussion. Kiln workers’ meals? Six Dynasties–Tang subsistence A myriad of resources from land, coast, and sea Marine shells from a variety of coastal habitats were found with varying intensities in every area excavated on the Sha Po backbeach. Some doubtless entered the site through natural processes (e.g., storm deposits of sand containing shells), others may have been the raw material for lime production, but some were almost certainly collected by people as food and discarded with fish and animal bones in food-rich middens or in smaller rubbish dumps. We begin below by examining the character of the middens and

Map 8: Six Dynasties–Song physical landscape. Source: CLSO. Yung Shue Wan, Lamma Island. 1:1000 Scale Topographic Map Sheets 14-NE-10D and 14-NE-15B. Hong Kong: CLSO, 1981a and 1981b respectively. Reproduced with permission of the Director of Lands. © The Government of the Hong Kong SAR. Licence No. 59/2015.

96   Piecing Together Sha Po

the range of resources they contained, and then consider how such materials can be interpreted in terms of past food procurement strategies and diet. Excavations in 1989 [11:3a]66 revealed early historical midden deposits containing marine shells and the bones of fish (mainly headgrunt and catfish), dolphin, terrestrial mammals including bovids (water buffalo?), suids (wild boar?), and a carnivore (probably a civet)—some with signs of butchering (i.e., cut-marks).67 The deposit included an Eastern Han seal-stamped sherd within a larger assemblage of Six Dynasties–early Tang pottery. Similar food remains, but with few datable artefacts, were later encountered in a midden layer located immediately to the west [34:E-F537] that appeared to be earlier than Kilns K4 and K5, which were dated by thermoluminescence (TL) to the Jin–early Tang period. The undisturbed nature of the midden was confirmed by the existence of two large pieces of cracked but still articulated—connected—green turtle carapace with numerous other bones from the same species. Associated with them were marine shellfish (including whelks, oysters, clams, and mussels), fish bones (including those of dog shark, croaker, sea bream, and shark), and mammal bones of suids (wild boar?), unidentified rodents, and large cervids (deer).68 The most unusual bone was the rib of a dugong, a once common marine mammal of the Pearl River Delta and Hong Kong’s coastline but now—due to a combination of disturbance, pollution, and habitat loss—almost certainly extinct in local waters.69 Among thirteen shark vertebrae were seven carved and perforated examples, presumably for stringing on a necklace or bracelet. Four more such beads were found with mammal bones in a continuation of Han–Sui middens in excavations just to the south [13:L2]. The presence of fish vertebrae beads in the above Han to Sui–early Tang midden deposits seems to indicate the continuity of a cultural practice dated elsewhere in Hong Kong to the prehistoric period.70 However, the bulk of midden materials on the southern backbeach appeared to post-date Bronze Age deposits, and the aforementioned Han dynasty storage jar [34:F536] at their base presumably marks

Plate 18:  Early historical pig cranium, dugong rib, and photos of dugong. Dugong photos kindly provided by Amina Cesario and Agnese Mancini via Dave Baker at HKU.



On the Edge of Empire: Han–Yuan Sha Po   97

Plate 19:  In situ green turtle carapace in early historical midden deposits

the genesis of such activity. Following the abandonment of Kilns K4 and K5, Song dynasty activity then disturbed and incorporated some of the earlier midden materials—including four more pierced and carved large fish vertebra—within a kiln post-abandonment spread [34:E-F535],71 which was also noted in 1989 [11:L2]. Further subsistence remains were identified in excavations at the supposed western edge of the backbeach [26 and 27], which comprised shells, fish, and mammal bones in shell-rich midden spreads of Song date as well as several small Tang rubbish pits.72

Early historical continuity of prehistoric foodways The great breadth of food resources discussed above give us an excellent indication of the ecological richness of Hong Kong at that time, and strongly suggests that the diversified hunting, fishing, and foraging based foodways of local prehistoric groups persisted into earlier historical times. What it does not reflect, of course, is the no doubt very significant contribution to diet made by plant foods, in particular as a source of carbohydrate. As argued in preceding chapters, plants such as palms, yams, lotus root, bamboo shoots, and wild grasses, may have been starchy staple foods of local populations for millennia. Nevertheless, the Sha Po middens do tell us that local communities gathered a wide variety of shellfish from sandy and rocky shores, but in volumes that did not make a significant contribution to their diet. In other words, shellfish were relatively easy to collect and good to eat, but not a dietary staple. Fishing was apparently important, and some species such as headgrunt and catfish, which also dominate fishbone assemblages from prehistoric and early historical sites with shell middens, are abundant in inshore waters at spawning time and seem to have been especially targeted. The presence of shark and dolphin indicate offshore fishing probably using baited hooks, but perhaps also harpoons. Today, green turtles only rarely visit Hong Kong’s beaches—including Sham Wan on Lamma—but they were once abundant and no doubt represented a seasonal food captured when ashore after laying their eggs, which were themselves a

98   Piecing Together Sha Po

source of food. A potentially significant contribution to subsistence was likely to have been provided by hunted mammals such as wild boar—still regularly encountered by after dark dog-walkers in the wooded hinterlands of Yung Shue Wan—and large deer, which are both usually present in well-preserved prehistoric or historical middens. Wild boar and, in particular, large deer, are woodland animals and their presence in early historical middens indicates that Lamma’s environment was then still densely forested, and in general remained so until the late Ming or early Qing, although the Sha Po plateau may have remained open (see below). Overall, what the above midden evidence tells us is that the early historical community using Sha Po backbeach exhibited a significant degree of subsistence continuity with local prehistoric groups. This seems to suggest the continued presence in Han to Tang times of local maritime-focused groups rather than the mass immigration of Han Chinese people from outside.

Buried facing the sea: Two graves from the Six Dynasties Introduction We noted above the general scarcity of the settlements and cemeteries of communities involved in the widespread coastal industry. A combination of small-scale excavations, generally poor preservation, and a high incidence of later disturbance have meant that very few burials have been detected and fewer still have yielded human skeletal remains. However, in 1989 a Six Dynasties burial discovered on the Sha Po backbeach [Grave 2—11:L3c] was overlain by a spread of lime [11:L2] and shell-rich midden deposits [11:L3a], which had neutralised the acid soils thus allowing good preservation of the skeleton.73 A second Six Dynasties burial excavated in 1997 [Grave 3—15:7] also contained skeletal remains, but this time in poorer condition.74 The two Sha Po burials stand out from the handful of other Six Dynasties burials found previously in Hong Kong, which were mostly identified by grave goods alone, although one at Pui O produced fragmentary human remains.75 The almost complete skeleton in Grave 2 was carefully excavated and recorded before removal, and was later subjected to laboratory analysis. In contrast, Grave 3 was unfortunately excavated in difficult circumstances and only the upper third of the skeleton was removed with minimal recording in what amounted to a rapid salvage exercise. As a result, much contextual information and physical evidence was lost and no study of the skeletal material had occurred before our recent review of the fragmentary human remains, grave goods, and archives. We begin below with an examination of the graves’ location and character, then their orientation, followed by a discussion of the skeletal remains in terms of the age, sex, health, and racial characteristics of the interred individuals, and finally the grave goods are described and then used as a basis for dating. Grave location and character The two graves were positioned on the highest part of the backbeach close to its longitudinal centreline, with Grave 2 near the southern end and Grave 3 in the centre. These raised locations were probably chosen as they were considered to be beyond the risk of inundation or storm disturbance (but see discussion of Grave 2 sequence below). Defining the outline of graves cut into and backfilled with very similar sand can be notoriously difficult,76 and neither Grave 2 nor Grave 3 was recognised until the excavators encountered the skeletons. While the trench at Grave 2 was extended to allow full exposure in plan, at Grave 3 the burial extended under an adjoining building and was never fully defined. Both graves were recorded in section as very shallow features; indeed so shallow that some form of later truncation had almost certainly occurred.



On the Edge of Empire: Han–Yuan Sha Po   99

The recorded depth of Grave 2 is—as acknowledged in the report—odd to say the least, in that it barely disturbed underlying Middle Bronze Age deposits, and was sealed just 10 cm above the skeleton by an unbroken layer of pumice [11:3b].77 The skeleton’s articulated condition and lack of weathering or disturbance led the excavator to conclude that it was indeed a burial—and not a body washed ashore or laid out to decompose—but that ‘a grave of only 10–15 cm depth could have been dug . . . and after interment a small amount of sand was mounded over’.78 However, it seems highly unlikely that Grave 2 would have originally been so shallow, especially if the backbeach was in use at that time, as seems possible based on the Han–Six Dynasties pottery finds and early range of some thermoluminescence dates from kilns (see Table 2 below). Moreover, such a shallow grave would almost certainly have attracted the unwelcome attention of scavenging animals and might easily have been exposed by storm action. Based on the archaeological relationships, we would see Grave 2 being cut from the surface of an originally thicker layer of Han to Jin backbeach deposits [11:3d; 34:E-F536]. The grave upper fill and surrounding deposits were then truncated—either by human activity or perhaps more plausibly by storm action—to within 10 cm of the body. The pumice layer and more sand were subsequently washed and/ or blown back on top, and the midden deposits would then have accumulated on top during the later Six Dynasties–Tang period. While obviously speculative, we feel this is a reasonably coherent attempt to explain the strange sequence encountered at Grave 2. The original stratigraphy around Grave 3 is difficult to interpret due to later disturbance. Moreover, the skull and upper thorax of the skeleton were unearthed near the western corner of the trench and the grave clearly extended under an adjoining property to the south-west. As a result the grave was never fully exposed in plan, but we would suggest that the sequence should have comprised a Tang layer [15:03] sealing the grave, which was itself cut into the aforementioned Bronze Age layer. Whatever the case, significant post-interment truncation of the upper grave fill is again indicated, but is here less easily explainable due to the reduced level of contextual detail available.

Burials and the art of placement The orientations of the two Sha Po graves are strikingly similar in that the head in Grave 2 was reported to be pointing towards the east-south-east (a bearing of 120 degrees),79 whereas from the photographic and drawn records in the Grave 3 archive we would estimate the head direction to be somewhere between east-south-east and south-east (115–135 degrees). Both burials therefore have a general east-south-east orientation with the head pointing towards the hills and feet towards the sea. Interestingly, a late Jin–Southern Dynasties (400–600 CE) burial found in Pui O, Lantau Island in 1983 was also orientated with the head pointing inland, in this instance approximately to the north.80 The preliminary report of another, particularly rich, Jin dynasty burial at Pak Mong, also on Lantau Island, unfortunately included no information regarding grave orientation.81 The Sha Po and Pui O graves seem to indicate a concern with grave orientation relative to the local landscape, rather than an adherence to a particular compass bearing or direction. Fascinatingly, if the two individuals buried at Sha Po had sat upright they would have looked straight out to sea towards the prominent peak of Lin Fa Shan on Lantau Island, which has been described as the ‘anchor’ of the nearby Tin Hau temple’s main fung shui line.82 That orientation is also broadly mirrored by the most convincing of the three possible Bronze Age burials discussed in Chapter 5 (Grave 1), thereby suggesting continuity in grave orientation on the backbeach spanning well over a millennium.83 Here it must be remembered that placement in the landscape continues to be the paramount consideration for local farming and fishing communities when deciding where to bury their dead. Although

100   Piecing Together Sha Po

the myriad of Qing–modern graves on the hills surrounding Sha Po have a variety of alignments, they all nevertheless have two things in common: they have hills behind and look out over the sea. .

Who were they and what were they like? 84 (Age, sex, stature, health, and race) On the rare occasions where skeletal remains survive in Hong Kong, an entirely different dimension is added to the archaeological story in that we can approach individual people and learn something of their lives. The skull, pelvis, and teeth can collectively provide useful clues as to an individual’s sex, age, and racial group, long bones allow calculation of stature, and study of the skeleton as a whole may offer insights into the person’s health and nutrition when alive, and even the cause of death. The near complete skeleton in Grave 2 was identified as that of a Southern Mongoloid female aged between late twenties and early thirties. The racial type was no doubt determined based on skull and tooth morphology (shape), sex on the basis of skull shape, and the pelvis (which for obvious reasons is quite different in women), while age can primarily be determined by studying the ends of long bones and tooth eruption stage and wear. For the age range suggested here, the long bone ends would be fully fused and all molars, including wisdom teeth, would be present. She was estimated, from observation of the articulated skeleton, to have stood around 1.6 m tall and she had no obvious pathologies (signs of injury or disease) apart from minor periodontal disease, although her teeth were generally in good condition.85 The site photographs appear to show a depressed fracture of the right temporal bone (temple), but as this was not mentioned in the report one can only assume that it was dismissed as damage caused postdeposition or even during excavation. Otherwise, it could easily have been a possible cause of death. The site records showed that in Grave 3 the skull, mandible (jawbone), and five other skeletal elements were identified within the trench, while an undisclosed number of other bones were excavated by digging horizontally into the section. With few shells and no lime nearby, the skeletal remains were far more fragmentary and incomplete than in Grave 2 and had not been studied following excavation. Our recent examination of the physical archive revealed that very few bones now survive and most of the skull was unavailable for study. The surviving elements include the mandible—broken into two halves—which had a quite complete set of teeth in good condition with no signs of decay. The third molar (wisdom tooth) was present indicating the person had reached adulthood, while the limited degree of wear suggested an age range similar to the woman in Grave 2. The incisors were of the ‘shovel-shaped’ type characteristic of Mongoloid peoples, while the pointed chin and small size of the teeth are both features associated with females—males tend to have larger teeth and generally squarer jaws.86 Other surviving identifiable elements included the following: the first cervical vertebra, or atlas, upon which the skull sits; six thoracic vertebrae to which the ribs attach; fragments of a rib; a fragment of the left radius (lower arm bone); and a distal phalanx (finger-end). Long bones can be used to provide a good estimate of overall stature, but the radius was unfortunately too fragmentary to allow this. The only other Six Dynasties inhumation burial with human remains was one at Pui O, but poor preservation seems to have discouraged any assessment of the human remains. We therefore have no same period comparative data against which to further assess the people buried at Sha Po. Dating accompanying jewellery and grave goods When compared with other Six Dynasties graves found in Hong Kong, the two Sha Po burials seem quite sparsely furnished with grave goods, but whether this implies that they had a somewhat lower social status in life is, for a variety of reasons discussed below, not entirely clear. Most unusually for burials of this period, Grave 2 contained no pottery grave goods at all, but the woman was laid to rest wearing three personal ornaments comprising an almost 20 cm long, U-shaped



On the Edge of Empire: Han–Yuan Sha Po   101

silver hairpin

silver finger-rings

Figure 18:  Sketch of Jin skeleton in Grave 2 based on site photo

silver hairpin found just beyond the skull and two small silver wire rings found near her fingers at waist level.87 Very similar hairpins and rings were found in Jin dynasty (265–420 CE) graves in Guang­ dong, Jiangsu, and near Beijing,88 which also fits with the pre-Sui to Tang stratigraphic position of the grave. Early historical graves often have pottery grave goods by the head and/or feet of the deceased, but the lower half of Grave 3 remained unexcavated so further finds may yet await discovery. Near the skull, though, there was a poorly preserved necklace comprising bead-like fragments made of an unidentifiable material, but unfortunately the item was not available for photography or study.89 The only item of grave

102   Piecing Together Sha Po

Plate 20:  Silver hairpin from Grave 2

goods—rather than personal ornamentation—was a well-made green crackle glazed bowl, which was complete and in very good condition. Comparison with previously excavated examples clearly indicates a date in the Southern Dynasties (420–589 CE).90 Judging the relative status of our two Sha Po women is difficult given the small comparative data set; however, the other Six Dynasties burials mentioned above included larger numbers of pottery grave goods than at Sha Po, although in the case of Pui O these included kiln wasters of much poorer quality than Grave 3’s green glazed bowl.91 In contrast, the Jin dynasty burial at Pak Mong was particularly rich, having two four-lugged green glazed jars and three pairs of stacked green glazed bowls, together with a pair of iron scissors.92

Discussion Although assessment of the broader significance of our two burials may be difficult, Sha Po is certainly special in having two graves of that era with preserved human remains. But perhaps the most interesting aspect of the funerary evidence is the fact that both women were buried with the same general orientation. This may be an indication of shared geomantic-religious beliefs concerning auspicious placement of

Plate 21:  Southern Dynasties green glazed bowl from Grave 3



On the Edge of Empire: Han–Yuan Sha Po   103

the dead and seems unlikely to be accidental, especially when the Pui O individual was also placed with their feet to the sea, head to the hills. We also see some wider significance in the hairpin and rings worn by the woman in Grave 2, which found parallels in graves across Jin dynasty China. So although Hong Kong was a coastal periphery of the Jin Empire, local communities clearly had access to wider markets, presumably via maritime trade and exchange networks. Now we move on to deal with the activities of the living, in particular relating to their construction, operation, and eventual abandonment of the kiln-based industry evidenced across the length and breadth of the Sha Po backbeach.

Construction, use, and abandonment of the Sha Po kilns Introduction Some overall sense of the scale of Six Dynasties–Tang industrial activity at Sha Po can be gained from the fact that fired clay kiln debris—kiln structural fragments and kiln furniture—have been found in every archaeological excavation conducted within the backbeach (see examples in Plate 22). Most significantly, since 2000, a total of seven Six Dynasties–Tang industrial structures (K1–K7—numbered in order of discovery) have been identified, comprising six kilns and another fired clay structure interpreted by the writers as a ‘working floor’ of some sort. In order to avoid too much descriptive detail in text, we provide a summary of kiln dimensions and characteristics in Table 1, to which reference can be made while reading our discussions below. The seven structures were each surrounded by dense deposits of kiln debris—a pattern repeated across Hong Kong—and the existence of several other intense kiln debris spreads almost certainly indicates that several more kilns await discovery. In contrast to the linear arrangements of kilns revealed in open-area investigations such as those at Lo So Shing, Sham Wan Tsuen, and Yi Long, the many smaller interventions at Sha Po have so far revealed three discrete kiln groups of broadly Six Dynasties–Tang date. A key goal of this research was to try to establish whether the kilns formed a planned industrial complex, potentially under imperial control (i.e., multiple kilns in contemporary use) or, alternatively, the seven were the cumulative remains of a smallscale ‘cottage industry’. Eighteen thermoluminescence (TL) dating samples (S1–S18 inclusive) were therefore submitted for testing and the very interesting results are shown in Table 2 and incorporated in our discussions below. As a cross-check and ‘control’ for the TL results, samples S1 and S2 were taken from a group of squat props associated with charcoal previously dated by radiocarbon to 364–640 CE.93 Using the TL laboratory’s narrower ±10% estimated error, the resulting TL date ranges for S1 (593– 851 CE) and S2 (395–689 CE) are well correlated with the radiocarbon results, and we therefore use TL dates with a ±10% estimated error throughout our discussions below.94 In our following exploration of the industrial remains we work from north to south, examining the three kiln groups in terms of their character, date, and sequence of construction, use, and abandonment. Thereafter we review the other locations within the backbeach with intense industrial remains. Finally, the Sha Po industry is considered in terms of its wider regional context and contribution to our understanding of Six Dynasties–Tang kiln sites across Hong Kong. The northern kilns The northern group of three structures (K1, K6, and K7) was discovered in rescue excavations before house construction in 2000–2001 [16:K1], which was the first in situ kiln discovered at Sha Po after decades of finding only kiln debris, while the other two were identified a decade later during archaeological monitoring for sewer and water pipe installation [39:K6 and K7].

Plate 22:  Kiln furniture: (a) flat-ended props, (b) cylindrical bars, (c) flat bars, (d) spacer, (e) squat props

K1

K2

K3

K4

K4

K5

K6

K7

0.45 m 1.20 m Not excavated. Should survive. 2.00 m 0.30 m 1.20 m Yes. Radial, wellpreserved.

Not excavated.

Not excavated. Should face between south and NE. Not excavated. Should face between SW and north.

Yes. Surfaced with stone.

Interior Floor?

Oysters and coral in kiln base.

Lime/Shell Residues in Backfill?

Comments

Convex design quite unlike other more ‘trough-like’ features interpreted as ‘slaking pits’ or ‘clay-mixing troughs’.

Internal floor stones fragile due to heating to high temperature. TL testing of wall by Sun Yat-sen University gave a date of 677–786 CE. Yes. Truncated. Not excavated Not excavated. Unusual stepped upper wall design: but should But none 0.30 m thick above ext. floor and 0.45 m survive. visible in thick below. surface. Yes. Surfaced Yes. 8 kg flat Support brackets sealed under convex Yes. Wellpreserved, with fired lime. Many external floor. Unusually overlaps inner convex. clay/stone. shells. face of kiln wall. Floor had rectangular post-slot angled at 20o from vertical towards centre of kiln. No. Upper Yes. Surfaced None found. Kiln wall relined on two separate occastructure with fired sions following damage resulting from truncated by clay/stone. severe overheating and/or long-term use. utilities. Not visible Not excavated Not excavated. Only the stoke-channel and possibly in sample but should part of kiln wall identified in a narrow excavation. survive. pipe trench. No. Area much Not visible Yes. 0.7 kg flat Approximately 10% of backfill sample disturbed by in sample lime pieces. excavated. utilities. excavation.

Yes. Fragmentary.

Surface Exterior Floor?

2.00 m 0.7 5m 1.00 m Upper wall Not excavated. truncated. Should face NW–N or S–SE. Not Not Not Not Not Yes. excaexcaexcaexcaexcavated. Facing SSW. vated. vated. vated. vated. 1.75 m Not 0.40 m 1.00 m Yes. Radial. Not excavated. excaSix (all Should face vated. truncated). between SW and SE. Overall dimensions in plan of exposed subrectangular portion: 1.5 m wide by 2.40 m long. Long axis None noted. orientated roughly east–west. Wall along south side 0.0 8m high by 0.30 m wide by 1.30 m long.

Kiln 2 1.70 m

K3

1.50 m

Kiln 1 1.90 m

Yes. Facing east.

1.95 m

2.10 m 0.30 m 1.00 m Yes. Eight surviving. Radial?

StokeChannel and Orientation?

Inside Diameter Max. O/All Support Wall Height Brackets? Width Top Base

K2

New Old (site) K1 Y1

Kiln Ref.

Table 1:  Summary of kiln characteristics

Table 2:  Kiln structure and kiln debris thermoluminescence (TL) testing results Sample Provenance (original site code, area, trench, SPT No. archive ID); SPT trench IDs as used in Kiln Map 5 shown in square brackets [  ] Ref. S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18

SPT1989; Trench G; Layer 6 (1.0–1.2 m); B2726; [8:L6] SPT1989; Trench G; Layer 6 (1.0–1.2 m); B2726 [8:L6] SPT2000; Y1:18; B2456 [16:K1–18] SPT2000; B2456 [16:K1] SPT2002; Unit 7.1; Kiln 1; Co. 08; Bag 6 [23:K2–08] SPT2002; Unit 7.1; Kiln 1; Co. 08; Bag 6 [23:K2–08] SPT2002; Unit 7.3; Kiln 2; Co. 04; Spit 4; B2000 [23:K3–04] SPT2002; Unit 7.3; Kiln 2; Co. 04, Spit 4; B2000 [23:K3–04] LIYSWII2008; RE AA5D; Kiln K1; Co. 545; Bag 1 [34:K4–545] LIYSWII2008; RE AA5D; Kiln K1; Co. 526; Bag 15 [34:K4–526] LIYSWII2008; RE AA5G; Kiln K2; Co. 542; Bag 1 [34:K5–542] LIYSWII2008; RE AA5G; Kiln K2; Co. 542; Bag 1 [34:K5–542] LIYSWII2008; WB Area B; Kiln K3; Co. 106; Box 41 [36:K6–106] LIYSWII2008; WB Area B; Kiln K3; Co. 106; Box 41 [36:K6–106] LIYSWII2008; WB Area B; Structure K4; Co. 108; Box 91 [36:K7–108] LIYSWII2008; WB Area B; Structure K4; Co. 111; Box 94 [36:K7–111] LIYSWII2008; WB Area B; Structure 117; Box 105 [36:117] LIYSWII2008; WB Area B; Structure 117; Box 105 [36:117]

TL Age

N/A

Calendar Calendar Date Date* (with Estimated Error) ± 10%** ± 20%** 1290 722 CE 593–851 CE 464–980 CE

N/A

1470 542 CE

395–689 CE

248–836 CE

K1 K1 K2

1650 362 CE 1340 672 CE 1240 772 CE

197–527 CE 538–806 CE 648–896 CE

32–692 CE 404–940 CE 524–1020 CE

K2

1140 872 CE

758–986 CE

644–1100 CE

K3

1190 822 CE

703–941 CE

584–1060 CE

K3

1120 892 CE

780–1004 CE 668–1116 CE

K4

730

K4

1282 CE 1209– 1355 CE 1430 582 CE 439–725 CE

1136–1428 CE 296–868 CE

K5

1660 352 CE

186–518 CE

20–684 CE

K5

1570 442 CE

285–599 CE

128–756 CE

K6

1610 402 CE

241–563 CE

80–724 CE

K6

1130 882 CE

769–995 CE

656–1108 CE

K7 K7

2260 248 BCE 1538 474 CE

474–22 BCE 700 BCE– 204 CE 320–628 CE 166–782 CE

N/A

1532 480 CE

327–633 CE

174–786 CE

N/A

1430 582 CE

439–725 CE

296–868 CE

* The calendar date is calculated from the year of testing (in this case, 2012) and is produced by subtracting the TL age from the current year (i.e., for S1: 2012–1290 = 722 CE). ** The calculation of date ranges based on the testing laboratory’s estimated error is calculated based on the TL age not calendar date (i.e., for S1: 1290 ± 10% gives a TL age range of 1419–1161; then 2012–1419 = 593 CE and 2012–1161 = 851 CE); giving a calendar date range (±10% error) of 593–851 CE.



On the Edge of Empire: Han–Yuan Sha Po   107

Kiln K1 was inserted into a Jin to Tang dynasty layer [16:L5], which had been baked to an intense red surrounding the structure [16:L4]. Layer L5’s assemblage of Jin, Sui, and Tang materials potentially spanned the fourth to eighth centuries CE, and included the locally unique and important find of a fired clay ‘tomb brick’ with a raised motif in mirror image of the character ‘官’ (guan, meaning ‘government official’) moulded on one edge. There are a number of published parallels for this general type of moulded brick, all outside Hong Kong and mostly dating from the Jin to Southern Dynasties (fourth to sixth centuries CE).95 The presence on K1’s eastern side of a quite typical flat-bottomed, U-shaped ‘stoke-hole’—or perhaps more accurately a ‘stoke-channel’—and traces of an external floor to the north, were both indications that something close to the kiln’s full height had survived. This was also suggested by traces of eight fragmentary support brackets that seem to have been radially mounted in slots on the upper surface of the kiln wall (see K3 and K6 below for better preserved examples). The excavator also mentioned three ‘ventilation holes’ in the kiln wall but their precise location and arrangement were not recorded.96 The occurrence of mid-Tang pottery (c. eighth century CE) in pre-kiln layer L5 suggests that K1’s construction should have occurred then or later. It is therefore interesting to note that thermoluminescence testing of a sample of kiln wall during the original project gave a date range of 677–786 CE for the last firing, which broadly agrees with the date (538–806 CE) obtained from our testing of Sample S4 from the kiln backfill. In contrast, the thermoluminescence result from a large cross-shaped fired clay block (Sample S3)—supposedly found standing in situ within the backfilled kiln—was surprisingly early

Plate 23:  Guan ‘tomb’ brick

108   Piecing Together Sha Po

Plate 24:  Kiln K1. © Antiquities and Monuments Office, reproduced with permission.

at 197–527 CE. This suggests that the large block was not in K1 when it was last fired, but may have been dumped there during its backfilling. Beyond the dated thermoluminescence samples, the remainder of fired clay kiln furniture in K1’s backfill took the more typical form of cylindrical bars, flat-ended props, bricks, and undiagnostic lumps. Quite unusually, the backfill yielded just two other artefacts but both were exciting in their own way: the first was a Tang dynasty green glazed ceramic ink palette,97 and the other was a Northern Song Mingdao Yuanbao (明道元寶) coin with kai (楷) calligraphy script and a minting date of 1032–33 CE.98 The coin’s exact place of discovery within the backfill remains unrecorded, but if the coin was not intrusive it indicates that the kiln’s backfilling was not completed until the early eleventh century or later. Covering the backfilled kiln was an extensive post-abandonment layer [16:T1/L3] containing pottery of late Tang–Northern Song date mixed with Southern Song–Yuan material including celadon bowl fragments—some being Tong’an types—and smaller quantities of Quanzhou jar with moulded dragon motifs and Dehua qingbai bowl sherds all from Fujian. A particularly interesting find was a lugged shoulder sherd from a yellowish-brown glazed wine jar made in the Qishi kilns at Foshan, Guangdong in the Northern Song period, which bore a very clear stamp with characters reading ‘Leung family wine’ (梁 宅酒).99 One final item of particular interest was an unusual cream-coloured flat brick with impressed lattice pattern, of a type usually described as ‘tomb bricks’ by Chinese scholars and dated to the Han–Six Dynasties period.100 Ten metres to the south-west of K1, Kiln K6 was also surrounded by intensely heat-reddened deposits that produced just three prehistoric finds [39:109]. This further reinforced the notion that all or most of the Bronze Age material found near K1 had come down from the plateau above and did not extend far beyond its foot. The kiln was first recognised during archaeological monitoring of preparatory engineering works for a deep pipe trench, during which sheet-piling was driven ‘blind’ through the northern edge

Plate 25:  Northern Song coin and Tang ink palette from K1 backfill

Figure 19:  Lattice brick and stamped wine jar sherd



On the Edge of Empire: Han–Yuan Sha Po   111

of the structure. This unfortunate event nevertheless allowed recording of structural traces still adhering to the sheet-pile and excavation of the excised backfill.101 The remainder of the kiln was preserved in situ by diverting a new water-main up and over it, which meant that only limited excavation of the upper backfill was carried out. Modern disturbance had truncated deposits surrounding the kiln, but the structure’s survival to almost its full original height was indicated by a series of fired clay support brackets [39:106a] mounted radially in slots in the wall’s upper surface. These bars were clearly added after the main clay ‘cylinder’ of the walls had been built and all had broken off in line with the inner face of the kiln, one fragment being visible in the surface of the backfill. Such slots and rectangular-section clay brackets have many parallels in well-preserved kilns; for example in K3 below and other major sites in the region.102 In the absence of recognisable pre-kiln deposits surrounding the structure, an initial estimation of the structure’s period of use was attempted using materials recovered from the disused structure’s backfill, which contained an interesting assemblage of kiln debris, potsherds, and industrial residues [39:107]. The roughly 150 pieces of kiln debris comprised the usual mix of cylindrical bars, flat-ended props, bricks, and spacers. The ceramics consisted of Tang glazed and slipped jars, late Tang–Northern Song glazed jars and celadon, Song–Yuan material including green glazed celadon, Quanzhou dragon-motif jar sherds, and white qingbai glazed Dehua ware—the last two both from Fujian. The parallels with the postabandonment layer surrounding K1, in terms of both types and dates, are quite strong and suggested the kiln’s abandonment in the middle Tang at the earliest and then backfilling extending perhaps later than K1 into the Southern Song–Yuan period. Perhaps the most interesting component in the backfill was 697 grams of white, flat fragments of a soft chalky material identified as lime (Ca(OH)2). The better preserved fragments bore clear rectilinear impressions on one side, apparently from a container or structure made from wooden strips or perhaps coarse-woven split bamboo (see Plate 27). The lime fragments were thin, averaging just 1–2 cm in thickness.

Figure 20:  Plan of K6

112   Piecing Together Sha Po

Plate 26:  Kiln K6 showing radial bars

Two samples of K6’s wall fabric were sent for thermoluminescence dating and a pair of contrasting results was produced: sample S13 gave a date range of 241–563 CE, whereas sample S14 yielded 769–995 CE. On the face of it, the same kiln wall producing two quite distinct date ranges is rather odd; however, there are possible explanations. One is that the pre-Tang dates from S13 record an earlier—or even initial—heating of an exterior, basal part of the wall, whose ‘radioactive clock’ was then never ‘reset’ by later firings of the kiln, which in contrast did produce a later date for sample S14. Certainly K6’s structure had an orangey-red rather than pink colouration—suggesting moderate heating—and had a relatively localised halo of heat-reddened sand (compare evidence for intense heating at K4 below). Alternatively, given the absence of pre-Tang material found near K6, the early date of S13 may well be erroneous. However, we would argue against that based on the Six Dynasties materials and thermoluminescence dates at K1, as well as the pre-Tang thermoluminescence dates produced by the structure of K7 (see below). In contrast, the thermoluminescence date range from sample S14 matches that of mid-late Tang dynasty pottery found within kiln backfill 107, and both arguably reflect the final phase of Kiln K6’s use. Although the kiln ceased production during the eighth to tenth centuries, based on the Northern Song and Southern Song–Yuan materials also present, activity continued in the area until its backfilling was completed at the very earliest in the twelfth century. Working floor K7, found just 4  m east of K6, was a subrectangular fired clay structure with long axis orientated roughly east-west, and comprised a smooth-surfaced, slightly cambered (convex) floor [39:108] bordered by low walls [39:111 and 113]. It was built on top of, or perhaps slightly set into, the same backbeach deposits recorded at K6, but here their colouration lacked the intense heat-reddening of further west, thus suggesting a different use history. The western end of the structure had been truncated by modern disturbance and an unknown proportion extended north-east beyond the trench edge. Following investigation, the surviving structure was preserved in situ. The upper surface of the low wall

Figure 21:  Six Dynasties finds: (a) Jin hairpin [11: Grave 2; CSF6a], (b) Southern Dynasties–Sui lid [34:541; CSF16], (c) Southern Dynasties lotus bowl [34:505; CSF17], (d) Six Dynasties bowl [11:L3; CSF18], (e) Six Dynasties–Sui cup [11:L3], (f ) Sui–early Tang cup [16:T2L5; CSF19], (g) Six Dynasties jar [16:T2L5], (h) Southern Dynasties bowl from Grave 3 [15:07; CSF15], (i) Sui stem cup [26:L4; CSF20]

Figure 22:  Tang pottery. Cups: (a) [34:505] and (b) [35:312]; bowls: (c) [16:T1L5; CSF22] and (d) [16:T1L5]; lugged storage jars: (e) [34:538], (f ) [39:107], (g) [35:318], (h) [34:538], (j) [36: unstrat.], (k) [36:607]; Tang–Northern Song: (m) [16:T1L5], (i) jar [39:107], (l) ink palette [16:K1 backfill; CSF21], (n) spout (jar) [16:T1L5]



On the Edge of Empire: Han–Yuan Sha Po   115

Plate 27:  Flat slabs of lime with linear bamboo-wood impressions from K6 backfill

bordering the southern side [111] bore three subcircular impressions; much as if large cylindrical kiln props had been pressed into the surface when the clay was still wet. Similar impressions were found on blocks in a kiln structure dump [114] to the west and on another group of large blocks [117] stacked on edge immediately to the south of K7. Levels (height measurements) taken on the floor confirmed what was apparent to the naked eye: the surface had been intentionally cambered so that the centre of the floor was up to 10 cm higher than the edges. It was clear from the colour of the structure itself, as well as the layers surrounding it [115] and underlying it [119], that the intense heating evidenced at K6 was lacking here. Also there was an apparent difference between the darker red, softer, and more crumbly floor, and the pinkish, harder fabric of the surrounding walls. Overall, it can be surmised that whatever K7 was, it was not repeatedly fired to high temperatures—indeed, its open, low-walled design is clearly ill suited to such use—and it may therefore only have been fired once when first built to bake the structure, but what was its function? Conventional wisdom might see K7 interpreted as a pit for slaking the lime found in the backfill of K6; although, unlike other sites, no trace of lime was found on the structure. Furthermore, the structure’s low walls and cambered surface are quite different from the more trough-like appearance of other supposed slaking pits103 and clay-mixing troughs104 excavated elsewhere. Perhaps due to its slightly lower position relative to K6, modern disturbance had failed to remove two areas of post-abandonment deposits at K7: one being a 10 cm thick spread of material over its surface [110] and the other a somewhat thicker deposit at its periphery [115]. Layer 115 produced just two sherds from a slip-coated jar and a green crackle glazed bowl—both Tang in date—together with over a

Figure 23:  Plan of working floor K7

Plate 28:  Working floor K7



On the Edge of Empire: Han–Yuan Sha Po   117

hundred items of kiln debris including bars, spacers, bricks, and props. Layer 110 yielded a much larger assemblage of kiln debris and a single datable potsherd from a Tang slip-coated jar. The ceramics indicated that K7 maybe went out of use somewhere in the mid-late Tang dynasty, but what did the thermoluminescence dating reveal? Samples S15 from the floor [108] and S16 from the wall [111] were tested along with two samples (S17 and S18) from the collection of apparently reused wall blocks [117] unearthed nearby. Three of the dates provided very convincing evidence that blocks 117—S17 (327–633 CE) and S18 (439–725 CE)— were indeed reused from K7 wall 111, from which sample S16 gave a date range of 320–628 CE. Floor 108, though, produced a surprisingly early date range of 474–22 BCE—which seems anomalous when compared with the wall results—but an answer can readily be found by returning briefly to the physics underlying thermoluminescence dating. The very early date produced by the floor (S15) almost certainly indicates that when it was fired to harden the clay a low temperature was achieved—hence its soft, red fabric. The clay’s naturally accumulated radioactive signal was therefore not reset to zero and, in thermoluminescence terms, the floor therefore appears much older than it really is. In contrast, the walls’ pinkish colouration looked higher fired to the naked eye, and was confirmed to be so by the quite consistent thermoluminescence dating results it produced. Significantly, the scientific dating results suggest that the Sha Po kiln industry—at the northern end of the backbeach at least—commenced production during the Six Dynasties but continued into the later Tang dynasty. That early origin is mirrored in thermoluminescence dates obtained from kiln bar samples at Lo So Shing and radiocarbon (14C) dates from charcoal at Yi Long.105 Although it was not possible to establish a physical or stratigraphic link between the post-abandonment deposits at K6 and K7—the former having been removed by modern disturbance outside the kiln—one can infer that K7 was abandoned during the Tang dynasty, while K6’s backfilling continued into the Southern Song or even perhaps the Yuan dynasty. Whereas K1, lying a few metres to the north, was apparently backfilled in the eleventh century or somewhat later. Whether this means that K1 and K6 both continued in use later than K7 is impossible to say but, based on the time span indicated by thermoluminescence dating of the structures and ceramics-coinage dating of post-abandonment deposits, it seems reasonable to suggest that all three probably overlapped in date and were in contemporary use.

The south central kilns The two south central kilns [23:K2 and K3] were discovered some 10 m apart during rescue excavations in 2002.106 Kiln K2 had been cut in half by the construction of a septic tank, but electricity cabling had caused only superficial damage at K3. The pre-kiln deposits contained very frequent kiln debris, occasional Tang pottery, and just a single Six Dynasties potsherd. Due to contamination, work at Kiln K2 was limited to defining the structure. Photographic records show few surface finds in the backfill, and while there is certainly kiln furniture present, no pottery or lime is evident—whereas much lime was visible in the surface of K3’s backfill (of which more below). As in the environs of K1, the kiln’s use had created a deep zone of pinkish heat-altered sediments [23/ K2:04/07], while under the external floor were shallower reddened bands [23/K2:03] indicative of less intense heating. Two sherds of a green crackle glazed bowl from pre-kiln context 04 suggest that the kiln was constructed during the Tang dynasty. Despite limited excavation and modern truncation, it was possible to show that the extant half of the kiln survived to its full original height with preserved floors [23/K2:08] attached to the outside of the upper walls. However, when compared with K3 (see below) and other well preserved kilns, the design of K2 is very unusual, perhaps unique. Interestingly, the excavators noted that above the level of the external

Figure 24:  Plan and south-east-facing section K2

Plate 29:  Kiln K2



On the Edge of Empire: Han–Yuan Sha Po   119

floors, the walls’ exterior changed from the uneven finish produced by slab- or coil-building within a pit, to two courses of fairly uniform and well-finished bricks. This indicates that the external floors were constructed on the contemporary ground surface at the level of the kiln’s original wall top, and the brick built upper part was a later addition. Without excavation it was impossible to establish if support brackets were present, but certainly none were visible at the surface. Nor was there any trace of the kiln’s stoke-channel, which must therefore have been orientated somewhere between south and north-east. Several large sherds of a pink-slipped coarse basin with dished, everted rim were found between the kiln and septic tank [23/K2:02];107 in what was probably disturbed kiln backfill. The type is diagnostically Tang and Guangdong-made, and an almost identical pot was recently found at San Tau.108 The pottery from pre-kiln layer 04 and post-kiln deposit 02 together provide a clear Tang timeframe for the kiln’s construction and use, but was that reflected in the thermoluminescence results? Two samples (S5 and S6) from the kiln’s external floors [23/K2:08] were tested to provide a date range for its last firing. The samples yielded quite consistent thermoluminescence results, which in turn produced calendar date ranges of 648–896 CE (S5) and 758–986 CE (S6)—broadly spanning the mid-late Tang dynasty. Taken as a whole, the dating evidence suggests that the kiln was constructed, used, and abandoned within the Tang dynasty. Around 10 m north-east of K2, Kiln K3 was surrounded by a ‘halo’ of intensely heat-reddened sediments and appeared to survive to its full original height. Removal of the kiln’s backfill [23/K3:03–07 inclusive] revealed a 2.6 m diameter structure with slightly incurving walls [23/K3:10]. In the base of the emptied kiln was a thin, compacted clay surface [23/K3:08] over a hard ‘brownish rock-like layer’ [23/ K3:09],109 which was almost certainly the kiln floor made of fired clay and/or stone slabs. Surrounding the structure at the contemporary ground surface was a smooth-surfaced, slightly domed fired clay floor [23/K3:11] that—in a unique variation of the type—extended to the inner face of the kiln wall and thus overlapped the outer ends of a series of radial, fired clay support brackets, which extended vertically down the inner face of the kiln.110 Again no evidence for a stoke-channel was identifiable in the narrow excavation, but it was most likely orientated somewhere between south-west and north. Another singular feature of this kiln was a 7 cm square hole in the external floor, which had probably been formed around a wooden post set in the thick clay before the floor was fired hard. Located on the north-eastern side of the kiln, the hole was angled at around 20 degrees from vertical towards the centre of the kiln. Unfortunately, the opposite side of the kiln was beyond the trench edge and there was therefore no possibility of checking whether a corresponding hole existed there. But what might its purpose be? The fact that it was intentionally moulded into the structure and angled towards the centre of the kiln is suggestive; for example, it might have supported some form of removable wooden beam or ‘A’-frame used during loading and unloading the kiln. The backfill contained abundant kiln furniture and collapsed kiln structure together with occasional Tang dynasty pottery. The most interesting finding, however, was over 8 kg of thin, flat lime fragments, many of which had rectilinear impressions on one face, presumably from a flat-bottomed wood or split bamboo container. Two kiln wall bricks from backfill layer 04 (samples S7 and S8) were submitted for thermoluminescence dating and produced very similar date ranges of 703–941 CE and 780–1004 CE respectively, which also had a good correlation with the dates from K2’s external floor. The lack of Six Dynasties material in pre-kiln deposits around K2 and K3, exclusively Tang pottery in both post-abandonment backfills, and thermoluminescence dates collectively indicate that the south central pair of kilns were constructed and used during the Tang dynasty and abandoned in the later Tang or Nanhan (907–71 CE) period.

Figure 25:  Plan and south-east-facing section K3



On the Edge of Empire: Han–Yuan Sha Po   121

Plate 30:  Kiln K3: overview (square hole arrowed), grate-bars, and square hole (detail)

The southern kilns The final two kilns [34: Kilns K4 and K5]111 were discovered in 2009, cut into the southern backbeach sequence of Han to Six Dynasties middens discussed above. An unusually broad zone of heat-reddened sand surrounding Kiln K4 gave advanced warning of its existence. Although the kiln survived to around 1 m high, recent utility trenching and Qing–modern agriculture had truncated the walls and there was no trace of radial support brackets or external floors in this instance, and the stoke-channel was not exposed in the narrow trench. The kiln backfill [34:523, 524, 526, and 545] was excavated to reveal walls with a pronounced incurvate form that meant a 2 m internal diameter at the floor level was reduced to just 1.5 m at the wall top. A quite typical local kiln floor was constructed from flat stones and fired clay slabs [546] set into clay bed [547] and surfaced with fire-blackened clay [527]. Interestingly, the unusually intense heating noted in surrounding sediments was dramatically reflected in the kiln structure, which had an equally unusual use history. While the vast majority of kilns evidence only one phase of build and use, K4 is a rare—perhaps unique—example that was relined during its lifetime not once but twice. The whitish-pink colouration and crumbly fabric of the original wall [522.1] suggests that the first relining [522.2]—itself very high-fired—occurred after an episode of particularly intense overheating, perhaps due to exceptionally high winds during the passage of a storm. In contrast, the dull red hues of the second relining [522.3] suggest firing to a far lower

Figure 26:  Plan of K4

Plate 31:  Kiln K4: overview, floor, and relined wall



On the Edge of Empire: Han–Yuan Sha Po   123

temperature. Relining is a very rare feature of such kilns,112 so why the operators went to the trouble of doing so twice in K4 is somewhat puzzling. The backfilled structure yielded a huge quantity of fired clay kiln furniture, bricks, and structural debris. The bulk of the furniture was in the primary fills—presumably the collapsed remains of the final firing’s grate-bars and props—while the wall bricks and fabric had tumbled in later. Evidence for K4’s date of use and abandonment came in three forms: pottery in the kiln backfill suggested a date of abandonment during the Tang dynasty; thermoluminescence samples from the lower backfills (S9 and S10) respectively produced date ranges of 1209–1355 CE and 439–725 CE; while kiln post-abandonment deposits to the south-west of K4 [529] contained pottery of late Tang–Northern Song and Southern Song–Yuan date. The date of thermoluminescence sample S9 corresponds with that of pottery from postabandonment layer 529, while the data range of S10 suggests that K4 was perhaps last used no later than the early Tang. On the basis that sample S9 was recovered from the black ashy deposits [545] in the very base of the kiln, which we presume to relate to the last firing, we propose that K4’s repeated relining may reflect its extreme longevity of use and it may indeed have been first fired in the Six Dynasties–early Tang (S10), but was perhaps last used as late as the Southern Song or even Yuan dynasty. That argument is challenged somewhat by the presence of only Tang pottery in the backfill, and K4 is certainly the only kiln at Sha Po that has suggestions of use beyond the Tang dynasty. Roughly 11 m south-west of K4 were the remains of Kiln K5, which was cut into the continuation of early historical midden remains noted above [34:542]. Within the narrow pipe trench only the southfacing stoke-channel [539] was exposed in plan, while the exterior of the kiln body [549] was apparently visible in section. Oddly, in this case there was no heat-reddening of sediments surrounding the kiln. The stoke-channel’s U-shaped form in cross-section was typical of other excavated examples in that it comprised two parallel rows of upright stone slabs bordering a flat floor made of fired clay blocks. Covering the structure was a post-abandonment layer [541] containing kiln debris, Tang dynasty storage jar and celadon bowl sherds, and a Southern Dynasties–Sui crackle glazed pottery lid.113 Two pieces of kiln furniture recovered from layer 542 beside K5 were thermoluminescence tested and gave dates of 186–518 CE (S11) and 285–599 CE (S12)—both firmly pre-Tang. This discussion of the southern pair of kilns would not be complete without mentioning the layer of kiln furniture, bricks, and flat pieces of lime found with Tang dynasty pottery in the north-western part of the Society’s adjacent excavations conducted in 1989 [11:L2],114 which seems very likely to be associated with K4 and/or K5. Underlying the kiln debris layer was the pre-Tang midden discussed above, which contained Six Dynasties’ celadon bowl sherds. When considered in overview, there is a generally early date range for the ‘construction and use’ phase of industrial activity at the southern end of the backbeach. The pre-Tang activity indicated by Southern Dynasties to Sui pottery was similarly reflected in the early thermoluminescence dates from three out of four samples from K4 and K5. However, we must acknowledge that the thermoluminescence dates from samples outside K5—in contrast to the others from kiln backfill or actual structures—can only tentatively be connected to its date of last use. Nevertheless, there is a strong impression that Kilns K4 and K5 were probably both early constructions and then were in contemporary use from the Six Dynasties period into the Tang. However, it seems possible that K4 continued to operate well beyond the end of the Tang dynasty, perhaps even into the Southern Song–Yuan, which may explain its exceptionally heat-damaged appearance and unique double relining. Before moving on to collectively assess and interpret the above industrial evidence, we must first discuss a number of other locations on the backbeach where intense industrial deposits may indicate the positions of further, as yet undiscovered, kilns.

124   Piecing Together Sha Po

Plate 32:  Kiln K5

Other industrial ‘hotspots’ Beyond the seven structures and their dense spreads of kiln debris were several other discrete concentrations of kiln-related deposits that stood out against the general ‘background’ scatter of such material across the backbeach (see Map 13 in Chapter 8). Most notable among these was the discovery in 1989 [8:L3] of an apparently in situ cluster of squat, flat-ended kiln bars surrounded by charcoal, and flat slabs of lime bearing ‘mat impressions’, which it was assumed had fallen ‘into a woven container upon firing or . . . [were] . . . placed therein while still hot and unslaked’.115 Radiocarbon dating of the charcoal produced a calibrated date of 364–640 CE, which as previously noted compares particularly well with one of two thermoluminescence dates obtained from the squat kiln bars themselves (395–689 CE). This cluster of squat bars may represent a separate, perhaps post-kiln, stage in the industrial process, but large quantities of kiln furniture and structural fragments found in association, added to ‘a floor of large kiln blocks and large coral pieces’ discovered nearby [9:L5],116 indicate that a kiln is probably close at hand. Another trench excavated that same year [10:L2–4] was reported to have unearthed ‘much kiln debris . . . and pieces of lime’,117 which again strongly suggests a kiln in the vicinity. While a year earlier, one of three small trenches excavated by Spry [6:18] produced a large volume of structural debris that had resulted from, it was argued, ‘the demolition of a kiln’.118 The general locations of four further probable kilns are suggested by intense scatters of kiln furniture and structural debris in association with Six Dynasties and/or Tang pottery as follows: in the north [19] and centre [26/27] of the western side of the backbeach, in the geographical centre of the backbeach [15/20], and also to the south-east [32]. The significance of the above patterning of ‘possible’ kiln structures is assessed with that of confirmed structures in Chapter 8’s discussion of cultural landscape developments through time. However, our next challenge is to collectively review and interpret the Sha Po industrial remains.



On the Edge of Empire: Han–Yuan Sha Po   125

Situating the Song–Yuan activity As is typical of local backbeach sites, the Sha Po kiln complex provided clear evidence for an ‘after-use’ or post-abandonment phase of activity that extended from the later Tang–early Northern Song into the Southern Song–Yuan period. Kiln K4 has the potential to be a rare example of a kiln whose use extended well into the Song dynasty, but this remains a tentative interpretation based on limited dating evidence. Many of the Southern Song tablewares—mostly green glazed celadon, white qingbai, and some brown glazed temmoku types—continued unchanged into the Yuan dynasty and can only be dated as twelfth to fourteenth century, but there is little or no material of diagnostically Yuan date.119 As a whole, the tradeware ceramics found across the backbeach were dominated by tablewares and storage vessels produced by kilns in Fujian (e.g., Tongan, Nanan, Dehua, and Quanzhou) or Guangdong Provinces (e.g., Qishi).120 Cooking wares are apparently absent, but may in some instances have been mistaken for later ‘village wares’. Continuity of use of the Sha Po backbeach beyond the Tang dynasty is thus clearly indicated, but there is only one location at the northern end of the backbeach where a reasonably clear but thin Southern Song–Yuan cultural horizon was recorded [35:305]. Also ‘Tang–Song’—probably better described as ‘Tang–Yuan’—pottery was also reported as disturbed finds in Qing and twentieth-century topsoil on the Sha Po plateau [e.g., 25:L3]. The discovery of Tang–Yuan ceramics on the Sha Po plateau is intriguing, as it suggests that the area was then cleared of vegetation and in use for purposes unknown. Here we are reminded of Ng Loi’s reference to the economy of Song dynasty salt-makers and fisher-folk of Lantau, in which he mentioned their

Figure 27:  Song–Yuan pottery: (a) celadon bowl [34:517], (b) celadon dish [39:107], (c) fragments of Quanzhou storage jar one with dragon decoration [39:107], (d) basin [34:519], (e) lugged basin [31:105]—the latter is perhaps Nanhan (Five Dynasties) to Northern Song

126   Piecing Together Sha Po

digging for yams—perhaps a staple source of carbohydrate—backed up by rice, which was obtained from ‘the City’ in exchange for hunted deer. Whether yams were a cultivated or wild plant food is unclear, but it is possible that the plateau may have witnessed some form of cultivation during the Tang to Southern Song–Yuan period. One final and highly enigmatic feature, which could be a disturbed burial, was found during archaeological monitoring in an area otherwise characterised by late Qing deposits [43:103]. A localised deposit of silty brownish sand was found to contain almost half of a small Southern Song–Yuan—possibly even early Ming—celadon bowl associated with three diagnostically human bones comprising a patella, tarsal, and phalanx, as well as two rib fragments and vertebra centrum, which may also be human. If really a burial, it must have originally been located surprisingly close to the top of the active shoreline.

Discussion: Putting the Sha Po kilns in their socio-economic context Some unusual features of the Sha Po industry As Table 1 shows, in many respects the character of the Sha Po kilns and residues reflect the published corpus of evidence; however, several discoveries are worthy of particular mention. For instance, the carefully smoothed, cambered surface of ‘working floor’ K7 was very unlikely to have been created by accident and its design would be suitable for various types of processing, perhaps requiring the drainage of excess liquid away to the edges—but to what end? The structure was extremely clean and no lime, clay, or other residue was noted during excavation. Previous discoveries of ‘bathtub’shaped fired clay features were interpreted either as clay mixing troughs or lime slaking pits, but K7 is unlike any of these. So what other processes evidenced or inferred above might require a very smooth and clean surface cambered to aid the removal of excess water? One possibility is that K7 was indeed a ‘working floor’ used to roll out cylindrical bars and shape bricks, brackets, and other elements of kiln furniture or structure. The floor might also be suitable for the final finishing and packing of salt prior to its onward transportation. Some water would be needed to maintain clay plasticity in the former process, while evaporated salt might still hold some water following boiling. Whatever the case, it is somewhat surprising that K7 is the only example of its type in Hong Kong, and it may therefore be a local solution to a practical need addressed via other means elsewhere. The double relining of K4 represents a unique investment of effort in a local kiln structure, which incidentally would have significantly reduced the kiln’s capacity, so why go to so much trouble? It is possible that the Sha Po backbeach was by that time fully occupied by teams of kiln operators and there was no free space to build anew; therefore, short of abandonment, repair was the only available option. The dating evidence also suggests that extreme longevity of use may be another possible reason. The excellent preservation of K3 revealed important details of the kiln’s upper structure, in particular the unique arrangement of radial brackets, whereby they were sealed beneath the peripheral surface floor rather than being constructed on top of it or cemented into it as seems to have been the case elsewhere. Also unusual was the strange square ‘post-hole’ moulded into the peripheral surface floor at K3, which may have supported some form of removable timber beam or ‘A’-frame, but why have such features never been reported before? Perhaps K3’s uniqueness is more apparent than real in that most excavated kilns are in a much poorer state than it and, elsewhere, radial brackets and external floors are usually either missing or badly truncated? Finally, the large cross-shaped block found in K1 is unique in Hong Kong, but its thermoluminescence date is much earlier than the structure itself, which places a question mark over its supposed use during the kiln’s final firing, but it does nevertheless look like some form of central supporting structure.



On the Edge of Empire: Han–Yuan Sha Po   127

Ruminating on residues As noted above, some scholars have been content to see the presence of lime within the kilns as irrefutable evidence for their exclusive use as ‘lime kilns’, but it is not quite that simple. Firstly, some kilns (such as K4 at Sha Po) have no trace of lime whatsoever, while in those that do the lime is found in two quite distinct forms, each with different contexts of deposition, which logically relate to different stages in the industrial processes carried out on sites. The first involves the discovery of a solid mass of lime mixed with ash in the very base of a kiln; for example as found in Sham Wan Tsuen kilns 10, 11, and 12. But so far none of this type has been found at Sha Po. Such lime-ash masses do not appear to have been dumped, but instead seem to be an in situ deposit marking the end of a particular kiln’s use phase. Some have suggested that such findings record the intentional in-kiln slaking of a charge of quicklime,121 but accidental slaking due to the occurrence of a sudden rainstorm before the kiln could be emptied is perhaps a more likely scenario. This seems more reasonable when one considers that quicklime is an easily shovelled powder—even with radial brackets cluttering the kiln’s mouth—whereas slaked lime is a noxious, hot paste that eventually sets as a solid mass, which in either state would be more awkward to mix and dig out. The solid masses are certainly very different in character from the second situation, in which many thin, flattened fragments of lime with woven wood/bamboo impressions are found within a kiln’s backfill (e.g., Sha Po K3 and K6) or, much more rarely, in open work areas.122 In the latter case the excavator suggested that ‘several layers of mat impressions’ had formed on unslaked lime (quicklime) following its collection as a thin layer within a ‘woven container’, but this mechanism for the formation of impressions seems very unlikely in the case of quicklime powder. On the other hand, the very flat profile and sharp-edged, parallel impressions on lime from K6 suggest to the writers that the lime must have been liquefied, and therefore slaked, before it was poured into or, more likely, firmly plastered onto the woven surface of a quite rigid container before it then cooled and set hard. Only thus could such sharp-edged, clear impressions of the container have been formed. It is worth remembering here that in more recent times lime was shipped from the point of manufacture in unslaked form and then slaking would occur immediately before it was used.123 This would certainly be the best option if the lime was to be used, for example, for caulking boats, tanning leather or in construction works. Significantly perhaps, Hong Kong’s Six Dynasties–Tang coastal kiln complexes sometimes have features identified as slaking pits, which could indicate that the lime processed there was probably for local and immediate use, rather than onward shipment in slaked form. We would argue that the different patterns of lime occurrence and other evidence discussed above are in fact highly informative when considered as a whole. Although modern limekilns are invariably enclosed structures, the Six Dynasties–Tang kilns could nonetheless be used to burn lime—albeit inefficiently—although their grate design and associated furniture seem unnecessarily complicated for that role,124 but contrastingly useful when placed to support and level-up large ‘pans’ of brine during boiling. That said, the large solidified masses of lime and ash found in some kilns clearly confirm that they were at least occasionally used to make batches of quicklime. Fired clay pits with lime residues similarly suggest that lime was slaked on some kiln sites although, as argued above, K7 was poorly suited for that purpose and may have had other uses. The occurrence within kilns of thin, flat pieces of lime with wood/bamboo impressions suggests that slaked lime was poured into or plastered onto the surface of flat-bottomed, rigid containers. If such containers had been used to collect and perhaps transport the slaked lime, then why were they always filled to such a shallow depth? This simply does not make sense when there were apparently sizeable batches of lime being burnt in some kilns. Why burn large batches but ship it out in thin sheets of slaked lime?

128   Piecing Together Sha Po

Given the physical evidence, the most logical explanation for the consistent occurrence of wood/ bamboo impressions on thin, flat fragments of lime—never on thicker slabs—is that coarsely woven, flat, and quite rigid containers were plastered with freshly slaked lime as part of the industrial process. This can most reasonably be interpreted in terms of the water- and fire-proofing of large woven bamboo pans with lime in preparation for their direct heating on top of the purpose-built supporting brackets in order to boil down concentrated brine.125 The smooth exterior floors and wide extent of radial brackets—reaching far beyond the outer walls of kilns—make perfect sense if the goal was to evenly distribute heat under the base of boiling pans that may have measured more than 3  m in diameter. The wide diameter but shallow depth of heat-reddened sediment outside some kilns—for example at Sha Po K1—is a pattern one might expect if hot air was funnelled radially between the exterior floor and underside of the boiling pan. The furniture, comprising bars, props, and spacers, is suitable for levelling-up and stabilising the presumably quite brittle lime-coated pans during the boiling process.126 Moreover, the long stoke-channel would be absolutely essential if fuel had to be regularly fed into the kiln when a large pan was positioned on top. The pans no doubt frequently wore out, became burnt, or needed recoating with lime, hence the many flat lime fragments discovered. Also the very rare discovery of intact kiln furniture similarly attests to its disposability and regular breakage in use. In sum, the thin fragments of lime, broken kiln furniture, structural design, and patterning of heating all support the kilns’ use for boiling brine to make salt.

Cottage industry or imperially controlled monopoly? A key goal of our research was to establish whether Sha Po’s early historical kilns represented the cumulative efforts of generations of small-scale indigenous producers—perhaps operating just one or two kilns at any given time—or, alternatively, were built en masse as a planned industrial complex under imperial control. To further that goal, we commissioned a programme of thermoluminescence testing of kiln debris and structural remains associated with the seven structures, and then compared and contrasted those results with the artefact dating and stratigraphic evidence. That revealed some clear and culturally significant patterning in the data. Firstly, the thermoluminescence dates for working floor K7, which we maintain was fired only once when first built to bake the structure, suggest that our industry began operation during the Six Dynasties–Sui period. The five structures in the northern and southern kiln groups also seem to have been constructed and in use from the pre-Tang era, and then the northern three were abandoned in the mid-late Tang, while K5 remains uncertain and K4 may have continued into the Song dynasty. The central kilns produced remarkably similar results and could actually have been built as a pair in the mid-late Tang dynasty, but both were subsequently abandoned in the latest Tang to Nanhan period. We therefore have two spatially discrete clusters to south and north that may be of broadly contemporary build and use, while the northern three and K4 were seemingly still in use by the time the relatively late central pair was constructed. But is that really evidence for planning and the controlling hand of non-indigenous authority? We think it probably is, as there are additional artefactual clues, in particular associated with the northernmost kiln K1, which further support the case. We refer, of course, to the lattice pattern and guan character bricks. We can find no parallels for the latter anywhere else in Hong Kong, but very similar lattice bricks were found in Southern Dynasties (i.e., immediately pre-Tang) tombs at Baoan, Shenzhen.127 Even more interestingly, two of the Baoan tombs also included green glazed calligraphy dishes similar in type to the one recovered from K1’s backfill. What is perhaps most notable, therefore, is that of the many Six Dynasties–Tang sites excavated in Hong Kong only Sha Po has produced such objects, and they therefore collectively represent a unique and important



On the Edge of Empire: Han–Yuan Sha Po   129

assemblage reflecting social status, imperial officialdom, and literacy. Why the two bricks were brought to Sha Po remains a mystery. They are clearly not items used by ordinary kiln workers, but neither were they found in a context clearly connected to imperial authority, such as a high-status tomb. They nevertheless have such connotations, and must be an indication that someone associated with the backbeach industry had official connections and kept records of the operation. That person would most logically be some form of administrator—probably a salt official—whose ‘circuit’ involved visits or perhaps even longer stays while he checked on the salt-working community at Sha Po.

Conclusions With its thin scatter of Han remains, Six Dynasties–Tang kilns, and Northern Song–Yuan after-use, Sha Po reflects any number of earlier historical sites around the coastline of Hong Kong. However, ‘the devil is in the detail’ as they say, and it was through a process of collective analysis and synthetic review that a number of particular—in some cases singular—aspects of the Sha Po evidence raised interesting interpretive possibilities. We have, for example, iron hoes (cha) of supposed Han date and agricultural application turning up in Six Dynasties–Tang industrial contexts. Cultivation of the Sha Po backbeach is a Qing and twentiethcentury phenomenon, and we strongly suspect that those hoes are more likely to have been used for mixing clay or lime rather than tilling soil. Although no salt survives as proof, we feel we have presented above a clearly and logically argued case for the existence of a widespread and long-lived kiln-based salt industry spanning the Six Dynasties–Tang period, possibly later in some cases. Lime was certainly produced in some kilns—although this is not evidenced yet at Sha Po—but mainly as a process-related by-product rather than a primary output of the industry. Salt was the ‘white gold’ of China’s imperial treasury and, based on the thermoluminescence and other dating evidence, Sha Po was one of a myriad of coastal kiln complexes that were planned, built, and operated under imperial control. Perhaps depletion of coral reserves in local waters eventually rendered the lime-hungry boiling pan method unsustainable. Moreover, the historically attested endemic salt smuggling on Lantau may also have occurred on Lamma in fishing communities where salt was essential for preserving their catch. One can imagine that the local community would have fought to maintain their traditional lifeway and economy while also attempting to satisfy the demands of their imperial overlords. The eventual Song reorganisation of the salt industry around larger, more centralised, and easily policed centres such as Kwun Fu (Kowloon Bay) probably had practical, economic, and political motivations, but either way by the Nanhan period Sha Po’s kiln-working industry, if not already completely abandoned, was certainly in terminal decline. Maritime-focused fishing communities no doubt continued to use the area, but in Chapter 7 we complete our journey through Sha Po’s archaeological story with an examination of the Ming–Qing period, when the former rice-farming communities now resident on Lamma settled the island and began creating the landscape of villages, paddy fields, and terraces that can still be recognised today if one knows where to look.

Notes 1. The Qin conquered Lingnan in 214 BCE and subsequently set up three counties in the region: Nanhai, Guilin, and Xiang. Some archaeological evidence dated to the Qin period was identified in Guangdong, particularly in Guangzhou—all with military connotations—including a shipyard (Mai Yinghao, Li Jin, and Chen Weihan, ‘Summaries on the Two seasons of Excavations on the Qin

130   Piecing Together Sha Po Dynasty Shipyard Site’, in Three Major Archaeological Findings of the Qin and Han Periods in Guangzhou, ed. Cultural Council of Guangzhou City, 5–42 [Guangzhou: Guangzhou Publishing, 1999]), wall ruins (Zhang and Huang, Southern Yue Kingdom, 38), and the graves of two possible Qin soldiers (Mai Yinghao, ‘Preliminary Report on the Excavation of a Qin Tomb in Luogang, East of Guangzhou’, Kaogu 8 [1962]: 404–6); and the recently identified late Qin city site (Shang Jie, ‘The Discovery of the Mount Shixiong Ancient City Site in Wuhua, Guangdong Province and its Initial Research’, South-East Culture 1 [2013]: 27–29). 2. In around 204 BCE—a few years prior to the final fall of the Qin Empire—the lieutenant of Nanhai founded the Southern Yue Kingdom and declared himself king (Zhang Rongfang and Huang Maiozhang, History of Southern Yue Kingdom [second edition], ed. Lingnan Wenku Editorial Committee and Guangdong Association of Chinese Culture [Guangzhou: Guangdong People’s Publishing, 1995]). Historical documents record that the Southern Yue Kingdom became a subject-state of the Han Court in 196 BCE, after which products such as lychee, cane sugar, mackerel, silver pheasants, and honey candles were offered to the Han Court as tribute in return for raw iron, iron agricultural tools, and livestock such as horses, sheep, and cattle. That relationship continued until the Han conquered the region in 111 BCE (ibid.). 3. Zhang and Huang, Southern Yue Kingdom, 418–22. 4. Patrick H. Hase, ‘Salt in Hong Kong’ (unpublished paper, 2003), 12. 5. The only local finding of early Western Han date, which coincides with the Southern Yue Kingdom, was at Pak Mong where a sealstamped urn, tripod pot, pottery box, and iron hoe and axe were found (Tang et al., ‘Pak Mong’, 61–62). 6. A hoard of sixty Ban Liang and Wu Zhu coins at So Kwun Wat (AMO website) associated with one of two burials, also preserved bamboo matting and woven hemp fabric. 7. Zhu, Research on Han Archaeology in Hong Kong, 43–46; sites of particular note are: Tung Wan Tsai South (Rogers et al., ‘Tung Wan Tsai’); Kau Sai Chau (William Meacham, ‘Report on an Archaeological Survey and Excavation of Northern Kau Sai Chau’ [on behalf of the Hong Kong Archaeological Society, December 1993–February 1994], unpublished report, 1995a); and most recently Yim Tin Tsai (Mick Atha, ‘Survey-cum-Excavation on Yim Tin Tsai, Sai Kung’ JHKAS XVII [forthcoming]). 8. Hong Kong Museum of History, Lei Cheng Uk Han Tomb (Hong Kong: Hong Kong Museum of History, 2005). The entrance to the tomb was unearthed during site formation works for one of Hong Kong’s first public housing estates. 9. Similar to Type VI late Eastern Han tombs found in Guangzhou. CPAM of Guangzhou, the Municipal Museum of Guangzhou, and Institute of Archaeology Chinese Academy of Social Science, eds., Excavations of the Han Tombs at Guangzhou (Beijing: Cultural Relics Publishing House, 1981), 368–71. 10. Zhu, ‘Han Archaeology in Hong Kong’, 48. 11. See Catalogue of Selected Finds No. 13. 12. Meacham, ‘Sha Po Tsuen’, 50. 13. AAL, ‘Rescue Excavations at Sha Po Tsuen’, 42. 14. Details from HKAS archives of 1994 excavations, which also mentions this pot found in ‘Square M’ in 1995, but the location of the latter trench was not included in the archives. 15. Rogers et al., ‘Tung Wan Tsai’, 136. 16. Rafael Maglioni, Archaeological Discovery in Eastern Kwantung, Journal Monograph II (Hong Kong: HKAS, 1975), 91–92. 17. Zhu, ‘Han Archaeology in Hong Kong’, 55. 18. AMO 2000–2001 archive; see also Catalogue of Selected Finds No. 5a. 19. ERM, ‘Yung Shue Wan’, 14; see also Catalogue of Selected Finds No. 5b. 20. Dong Shouxian, ‘Research of Iron Tools of the Han Dynasty’ (MA diss., Zhengzhou University, 2010). 21. Liu Shixu and Zhang Zhengning, ‘Investigation of a Han Dynasty Bronze Smelting Site in Dongpingcun, Xichang, Sichuan’, Kaogu 12 (1990): 1073; Dong, ‘Iron Tools’, 2010. 22. ERM, ‘Yung Shue Wan’, 14. 23. Li Jinghua, ‘Ancient Iron Agricultural Tools in Henan’, Agricultural Archaeology 1 (1985): 57. 24. Yang Shiting, ‘A Few Questions on the Early Iron Tools in Guangdong’, Kaogu 2 (1977): 204. 25. Yang Kuan, Development History of China’s Ancient Smelting Technology (Shanghai: Shanghai People’s Publishing, 1982), 297; Dong, ‘Iron Tools’, 17. 26. Although see discussion under ‘Post-industrial Background Spreads’. 27. Meacham, Chek Lap Kok, 189. 28. Meacham, ‘Sha Po Tsuen’, 46. 29. Lo Hsiang-lin, Hong Kong and Its External Communications Before 1842 (Hong Kong: Institute of Chinese Culture, 1963), 5. 30. Lo, Hong Kong Before 1842. 31. Hase, ‘Salt’, 14 (notes 28 and 29). 32. Ibid., 14; Au Ka-fat and Tung Po-ming, ‘Ancient Salt Industry in Hong Kong and Shenzhen Region’, JHKAS XIV (1998). 33. Lo, Hong Kong Before 1842, 23 and 38–41; Hase, ‘Salt’, 16. 34. James C. Y. Watt, ‘A Brief Report on Sung-Type Pottery Finds in Hong Kong’, JRASHKB 11 (1971): 142. 35. Lo, Hong Kong Before 1842. 36. Hase, ‘Salt’, 23. 37. AMO website (accessed 15 October 2014). 38. Lo, Hong Kong Before 1842; Jen Yu-Wen, ‘The Travelling Palace of Southern Sung in Kowloon’, JRASHKB 7 (1967): 21–38. 39. Hugh D. R. Baker, ‘The Five Great Clans of the New Territories’, JRASHKB 6 (1966): 26.



On the Edge of Empire: Han–Yuan Sha Po   131

40. Michael J. E. Palmer, ‘The Surface-Subsoil Form of Divided Ownership in Late Imperial China: Some Examples from the New Territories of Hong Kong’, Modern Asian Studies 21 (1) (1987): 14. 41. Patrick H. Hase, Custom, Land and Livelihood in Rural South China: The Traditional Land Law of Hong Kong’s New Territories, 1750–1950 (Hong Kong: Hong Kong University Press, 2013), 31–32. 42. Hase, ‘Salt’, 18 43. K. A. Styles and M. H. Law, ‘Some Observations about Man-Made Features on Natural Terrain in Hong Kong’, QJEGH 45 (2012): 131–38. 44. Alice C. S. Lai et al., ‘The Territory-Wide Airborne Light Detection and Ranging [LIDAR] Survey for the Hong Kong Special Administrative Region’; Robert J. Sas Jr. et al., ‘Detection of Old Agricultural Terraces in Steep, Vegetated Terrain Using Airborne LIDAR: Case Studies from Hong Kong’, both papers presented at the 33rd Asian Conference on Remote Sensing, November 2012. LIDAR is an airborne remote sensing technique of particular value when archaeological remains lie hidden beneath dense tree or scrub cover. We would hope and expect in future to see its more widespread use in Hong Kong archaeology. 45. As cited in Hase, ‘Yung Shue Wan’, 3. 46. Ibid., 5. 47. The kilns were first noted by C. M. Heanley in an article in 1935, in which he discussed finding ‘cone furnaces’ on several sandy beaches on Lantau and Lamma Island. His label had nothing to do with the shape of the kilns, but instead related to his observation that some kiln furniture resembled Egyptian cone seals! C. M. Heanley, ‘Fields of Hong Kong’, Hong Kong Naturalist 6 (2–3) (1935): 235–36. More recently, Li recorded 59 kiln sites but more have since been discovered. Li Long-lam, ‘Studies on 5–10 Centuries’ Brine-kilns in Hong Kong’, JHKAS XVI (2009): 94–103. 48. Kiln complexes: Hugh Cameron, ‘Lo So Shing’, JHKAS VII (1979b): 125–29; Hugh Cameron, ‘Discussion of Tang Lime Kilns at Sham Wan Tsuen’, in Archaeological investigations on Chek Lap Kok Island, ed. William Meacham (Hong Kong: HKAS, 1994), 223–30; Hugh Cameron, ‘Yi Long’, JHKAS VII (1979a): 118–24. 49. Lime can be produced by burning the calcium carbonate in shells and coral (CaCO3) to make quicklime (CaO), which could then be converted to lime (Ca(OH)2) by the addition of water (slaking). Lime masses: Cameron, ‘Lo So Shing’, 128–29; Cameron, ‘Sham Wan Tsuen’. 50. Slaking pits: Cameron, ‘Lo So Shing’, 128–29; William Meacham, ‘Pui O’, JHKAS X (1984): 62. 51. For example, Meacham, Archaeology of Hong Kong, 151–52. Here we should also mention that, based on the excavation of a fairly typical ‘lime’ kiln at Siu Lam, one local archaeologist has also argued in favour of pottery production. Liu Mao, ‘The Study of the Siu Lam Kiln Site in Hong Kong’ (unpublished report, 2002). However, the kiln design, furniture, residues and, above all, lack of misfired pottery wasters (usually present in large numbers at kiln sites) all argue against this interpretation. 52. The salt idea has been explored by several Chinese scholars since the late 1990s, e.g., Au Ka-fat and Tung Po-ming, ‘Ancient Salt Industry in Hong Kong and Shenzhen Region’, JHKAS XIV (1998): 81–87; Li, ‘Brine-Kilns in Hong Kong’, 94–103. All drew upon the historical sources, in particular illustrations in the Yuan dynasty Ao Bo Tu (Chen, Chun, ‘Salt Production’, in Ao Bo Tu, Volume 2 [Taipei: Taiwan Commercial Press, 1984], Plate 42, and Ming dynasty Heavenly Creations [Song Yingxing, ‘Salt Production’, in The Heavenly Creations, Volume 1 [Shanghai: Guoxue zhenglishe: Shijie shuju yinxing, 1936], 99–125). 53. Liu Xun, Ling Biao Lu Yi (嶺表錄異) as cited in Ji Chengming, ‘Analysis on Several Research Materials Available on Tang Dynasty Salt Production Techniques’, Salt Industry History Research (1) (1993): 44–46. 54. In antiquity leaching of salt trapped in intertidal mud was the most widely used method of salt-working along the Lingnan coast (including Hong Kong), as it involved relatively simple technology and did not require the construction expense and space needed for solar evaporation-type salt-pans (Hase, ‘Salt’, 57). 55. Meacham, Chek Lap Kok, 202; Mick Atha, ‘A Military and Civilian Cemetery of the Mid to Late Tang Maritime Trade? Ground Penetrating Radar (GPR) Surveys and Excavations at San Tau, North Lantau, Hong Kong’, in Collected Essays of the International Conference on Historical Imprints of Lingnan: Major Archaeological Discoveries of Guangdong, Hong Kong and Macao, ed. Susanna L. K. Siu et al. (Hong Kong: Hong Kong Museum of History), 176–217. 56. Regina Krahl et al., eds. Shipwrecked: Tang Treasures and Monsoon Winds (Washington D.C.: Smithsonian Institution). 57. Sharon W. Y. Wong, ‘A Preliminary Study of the Distribution and Consumption of Ceramics in Hong Kong during the Song–Yuan Period’, Bulletin of the IPPA 26 (2006): 140–46. 58. Occasional Tang materials have been excavated in the inland New Territories: for example, pottery at Tin Tak Kung, Pat Heung (now part of Tai Kong Po Site of Archaeological Interest) and at Tai Hom Tsuen, and a complete burial jar at Shek Kong. However, the scale of excavation and findings were too limited to understand the true character of Tang activity on site, and further research is therefore needed. 59. Meacham, Archaeology of Hong Kong, 154–56. 60. Shek Kong: AECOM, ‘Consultancy Agreement No. C8016, Environment Team Consultancy for Express Rail Link, Final Archaeological Rescue Excavation Report’ (unpublished excavation report, 2014). Other significant Song settlement remains, including house foundations, yards, and drains, were unearthed at Mong Tseng Wai (Mo Zhi and Li Ziwen, ‘Archaeological Excavations on Hok Chau, Mong Tseng Wai’, JHKAS XV [2002]: 1–17). As discussed in Chapter 9, the probability of detecting even quite extensive settlement remains is significantly reduced by local evaluation methodologies, which sample minute percentages of impacted areas using auger holes and small test pits, rather than geophysics and overlapping patterns of long, thin trenches. 61. ERM, ‘Archaeological Survey-cum-Excavation for Sacred Hill (North), Interim Report’ (unpublished report, 2014). 62. Mick Atha, ‘Excavations at San Tau’.

132   Piecing Together Sha Po 63. Ibid. 64. Six Dynasties burials: Pak Mong—Tang et al., ‘Pak Mong’, 61–62; Pui O—Meacham, ‘Pui O’, 64–66. Tang burials: at Lung Kwu Chau an isolated cluster of 8 bowls and 3 small dishes found in 1974 could be a burial (William Meacham, ‘Tung Kwu: Phase 4’, JHKAS VI [1975]: 58–59). 65. Song burials: Sham Wan Tsuen—Hugh Cameron and B. V. Williams, ‘Sham Wan Tsuen: Chek Lap Kok’, JHKAS X (1984): 21–24; Yuen Leng Tsai—HKIA, ‘Working Paper on the Rescue Excavation at Yuen Leng Tsai’ (unpublished excavation report, 2003); To Kwa Wan—ERM, Sacred Hill (North), 16–18. 66. The original report (Meacham, ‘Sha Po Tsuen’) groups this midden deposit, a pumice layer, and a Jin dynasty burial all within Layer 3, which is problematic when trying to discuss the site sequence in single context, stratigraphic terms. To facilitate such discussion, we therefore subdivide Layer 3 as follows: 3a—midden deposit, 3b—pumice layer, 3c—Jin burial, and 3d—yellow layer into which grave was cut. For clarity and consistency, we also allocate a new number to the pumice found within the surface of our Layer 536 in 2009 and label it 536a. 67. Meacham, ‘Sha Po Tsuen’, 40–48; bovids are members of the cattle family including water buffalo, while suids are the pig family including domesticated pigs and wild boar. The report does not identify beyond the level of family, so we do not know whether these are wild or domesticated animals, but assume the former. 68. AAL, ‘Rescue Excavations at Sha Po Tsuen’; cervids are members of the deer family and large deer bones are a feature of prehistoric and early historical sites around Hong Kong’s coast, where conditions allow their preservation. Croakers are another economically important food fish: the yellow croaker (黃花魚) being particularly prized in modern times and consequently overfished to the point of population collapse in the South China Sea. 69. Oral histories record dugongs in Sai Kung district during the twentieth century but decline seems to have been sharp since World War II (http://forteanzoology.blogspot.hk/2013/09/muirheads-mysteries-hong-kong-dugong.html). According to the IUCN website, the dugong is listed as ‘vulnerable’ and is now ‘declining or extinct in at least a third of its range’ (including the Hong Kong–Pearl River Delta region), at http://www.iucnredlist.org/details/6909/0. 70. For example, Meacham, Sham Wan, 229. 71. AAL, ‘Rescue Excavations at Sha Po Tsuen’, 44–47. 72. ERM, ‘Yung Shue Wan’, 19. 73. Meacham, ‘Sha Po Tsuen’. 74. AMO, ‘Sha Po Old Village’. 75. Meacham, ‘Pui O’, 65. 76. During the writers’ excavation of a Tang cemetery on a backbeach at San Tau—in this case, with no surviving human remains—some graves were identifiable in plan, but only with repeated trowel cleaning and regular water misting. 77. Meacham, ‘Sha Po Tsuen’, 40. 78. Ibid., 47. 79. Ibid. 80. Meacham, ‘Pui O’, 65. 81. Tang et al., ‘Pak Mong’, 61–62. A further Six Dynasties burial was identified at Sham Wan by two lugged jars (kuan) with bowl covers associated with a fragment of human mandible, while skull fragments and Wu Zhu coins were found nearby, but grave orientation could not be determined. Meacham, Sham Wan, 94–96. 82. Hase, ‘Yung Shue Wan’, 40; Map 2. 83. That continuity offers a striking contrast with the four different orientations noted at San Tau during the Tang dynasty, which is one of several factors influencing the idea that it was a non-indigenous burial ground used by various ‘outsider’ groups. 84. This heading is ‘borrowed’ from the title of Chapter 11 of Renfrew and Bahn, Archaeology. 85. Meacham, ‘Sha Po Tsuen’, 47. Periodontal disease is caused by poor dental hygiene and results from calculus formation around the gum-line. Caries (tooth decay), on the other hand, attacks the cusps of teeth and relates to the regular consumption of sugar, which was evidently not a significant part of the Jin–Southern Dynasties diet of this woman. 86. William M. Bass, Human Osteology (Columbia: Missouri Archaeology Society, 1995), 86–87 and 297. 87. Meacham, ‘Sha Po Tsuen’, 40. Although not mentioned in the report, ordinary Six Dynasties–Tang hairpins tended to be made of copper alloy (bronze or perhaps leaded bronze for such a long, fine object), better quality ones were made of silver, while those of the elites were fashioned from gold. 88. Jiang Linhai and Zhang Jiuwen, ‘Excavation of Tombs 8, 9 and 10 at Xiangshan near Nanjing’, Wenwu 7 (2000): 10–12; Qiu Licheng, ‘An Eastern Jin Tomb at Pingshigang of Zhaoqing City, Guangdong’, in Huanan Archaeology, ed. Guangdong Provincial Institute of Cultural Relics and Archaeology and Guangzhou City Institute of Cultural Relics and Archaeology (Beijing: Wenwu Publishing), 259–60, M1:15. Wenwu 10 (1983): 68 as cited in Meacham, ‘Sha Po Tsuen’, 46. 89. Glass necklaces are not unknown in this period but they generally survive quite well. The poor preservation in this case perhaps indicates the use of an organic material. 90. The Sha Po bowl was similar to green crackle glazed bowls found in Tomb No. 14 at Nanhuasi, Qujiang (Peter Y. K. Lam, Archaeological Finds from the Jin to the Tang Periods in Guangdong [Hong Kong: CUHK Art Gallery, 1985]: 163 No. 48B) and Tomb No. 9 at Xihe, Shaoguan (ibid., 165). Both were dated to the Southern Dynasties (see Catalogue of Selected Finds No. 15). 91. Meacham, ‘Pui O’, 64–66. 92. Tang et al., ‘Pak Mong’, 61–62.



On the Edge of Empire: Han–Yuan Sha Po   133

93. Meacham, ‘Sha Po Tsuen’, 44. 94. The TL lab—P. L. Leung Ancient Ceramics Scientific Authentication Centre—quoted an estimated error range of ±10–20%, but the cross-check suggested that the narrower TL error range could reasonably be used in our discussions. Both error ranges are shown in Table 2. 95. Lam, Finds from the Jin, 65–67. See fuller details in Catalogue of Selected Finds No. 3. 96. AMO, ‘Rescue Excavation at Lot 1687, Back Street, Sha Po Tsuen’ (unpublished excavation report, LI22, 2001). 97. See fuller details in Catalogue of Selected Finds No. 21. 98. Zhang Yue, ‘A Preliminary Study on Song Dynasty Currency and Its Culture’ (unpublished MA diss., School of Ethnology and Social Science, Minszu University of China, 2006), 34–35. See also Catalogue of Selected Finds No. 7b. 99. Chen Zhiliang, ‘Investigation on Ancient Shiwan Kiln Sites in Guangdong’, Kaogu 3 (1978): 195–99; Sharon W. Y. Wong, ‘A Preliminary Study of the Ceramics of the Song–Yuan Period’, Kaogu 6 (2007). See also Catalogue of Selected Finds No. 23. 100. For example, Chen Chunli, ‘Preliminary Report on the Excavation of a Southern Dynasties Tomb at Meihuacun, Dongshan, Guangdong’, in Guangzhou Wenwu Kaogu Ji, ed. Institute of Cultural Relics and Archaeology of Guangzhou (Beijing: Wenwu Publishing, 1998), 189, 192; Yi Xibing and Ma Jianguo, ‘Preliminary Report on the Excavation of Southern Dynasties Tombs at Zhongxing Primary School, Taojin Road’, in Archaeological Discoveries and Research in Guangzhou Vol. I (Beijing: Wenwu Publishing, 2005), 135. See also Catalogue of Selected Finds No. 2. 101. AAL, ‘Archaeological Watching Brief at Sha Po Tsuen’ (unpublished report, 2011b), 15. 102. Cameron, ‘Yi Long’, 120–21; ‘Lo So Shing’, 127; ‘Sham Wan Tsuen’, 191. 103. Cameron, ‘Lo So Shing’, 128–29; Meacham, Archaeology of Hong Kong, 156. 104. Cameron, ‘Sham Wan Tsuen’, 224. 105. Cameron, ‘Lo So Shing’, 125; ‘Yi Long’, 120. 106. AAL, ‘Sha Po Tsuen Rescue Excavation’. 107. Ibid., 33. 108. Mick Atha, ‘Archaeological Investigation at San Tau, Lantau Island’ (unpublished report, 2012b), 147. 109. AAL, ‘Sha Po Tsuen Rescue Excavation’, 34. 110. Ibid. 111. AAL, ‘Rescue Excavations at Sha Po Tsuen’. 112. The only other clear example we could find in the published literature was Kiln 1 at Yi Long, Lantau Island. Cameron, ‘Yi Long’, 118. 113. See also Catalogue of Selected Finds No. 16. 114. Meacham, ‘Sha Po Tsuen’, 39—‘Square CC’. 115. Ibid., 47—‘Square G’. 116. Ibid., 38—‘Square H’. 117. Ibid., 38—‘Square X’. 118. Spry, ‘Sha Po Tsuen’, 17—Square ‘E’. 119. The ceramics categories used in the discussions below follow those used in our recent Tang–Song period excavations at San Tau, which followed Sharon Wong’s 2007 summary of Song–Yuan pottery from Hong Kong sites, and William Meacham’s 1994 Chek Lap Kok monograph. 120. AAL, ‘Rescue Excavations at Sha Po Tsuen’; ERM, ‘Yung Shue Wan’, 14. 121. Cameron and Williams, ‘Sham Wan Tsuen’, 19. 122. Meacham, ‘Sha Po Tsuen’, 47. 123. HKU Architecture Department, ‘The Lime Kiln Industry of Sai Kung’, JHKAS VII (1979): 139–40. Report of a survey of the Sai Kung area for the AMO, undertaken by students of the Architecture Department of Hong Kong University. 124. That said, it has been suggested that the large quantities of cylindrical kiln bars found in backfills were in fact used inside kilns to space layers of fuel (wood or charcoal) and lime raw materials (shells and coral) during lime burning. For example, Au, ‘Rescue Excavation at Lot 1687’. 125. In line with Au and Tung’s earlier research, Li cited post-Tang historical sources that talked of large basins measuring approximately 3.2 m diameter being coated with lime and then backed with clay for heating over flat-topped ‘salt stoves’. Au and Tung, ‘Ancient Salt Industry’, 84–85; Li, ‘Brine-Kilns’, 98. The latter process is clearly reflected in Ji Chengming’s account of Liu Xun’s Ling Biao Lu Yi (嶺表錄異) (in the supplementary notes on page 37 of the 1983 ‘Lu Xun version’), which similarly explains the use of lime-coated bamboo ‘pans’ in the final boiling stage, but also reveals details of the preceding brine collection-concentration process in use in late Tang dynasty Lingnan. 126. Some question marks, nonetheless, remain over the specific functions of different elements of furniture and further thought and perhaps experimental work is needed before a more complete interpretation is possible. 127. Wen Benheng and Rong Daxian, ‘Preliminary Report on the Excavation of a Southern Dynasties Tomb in Baoan, Shenzhen’, Wenwu 11 (1990): 39–43.

7 A Time of Great Change: Sha Po during the Ming–Colonial Era

Introduction The Ming and Qing dynasties in Hong Kong were a time of significant demographic change: some of this was locally produced as established farming lineages grew in size and spread out to create new communities but, from the late seventeenth century onwards, far more occurred as a consequence of immigration by outsiders. The early Qing dynasty, following the devastation of the Coastal Evacuation1 and typhoons which followed it, was a period of government-induced mass migration—in particular by Hakka peoples, but Punti also—into Xinan County’s relatively depopulated countryside. In many areas, though, such early Qing immigrants found the best low-lying, well-watered rice land already claimed and they were therefore left with little choice but to settle in ever more marginal and challenging locations, usually as tenants of the larger and earlier established clans. By the Ming and Qing dynasties the historical sources available—such as imperial land and taxation records, the Xinan County Gazetteers, as well as clan genealogies—offer more detailed insights into the lives of Hong Kong’s people. On Lamma, our earliest historical references to the establishment of farming communities can thus be found in official documents dating to the Qing dynasty, which in some instances refer back to imperial land allocations made during the Ming. Such accounts help construct a broad socio-historical framework within which the more detailed narratives suggested by Ming and, to a far greater degree, Qing archaeological discoveries can then be situated. Based on the character and distribution of archaeological remains from Sha Po Tsuen, it made sense to break our discussion down into sections covering the Ming and Qing–Colonial periods separately. But first we must explore north-west Lamma’s social history beginning with the Ming.

Social History of North-West Lamma’s Ming–Colonial Era Colonial officers and historical anthropologists In attempting to provide some socio-historical context for our discussions of Sha Po’s later historical archaeology, we are greatly aided by the anthropological and historical research of several generations of colonial officers—most notably Schofield, Coates, Hayes, and Hase—whose interviews, observations, and writings recorded many fascinating details of the New Territories’ traditional way of life before it was irretrievably altered and, ultimately, ceased to exist.2 Thus Patrick Hase’s research on Lamma’s social history drew upon a variety of unpublished sources including transcripts of interviews conducted with local elders by district officers Austin Coates and James Hayes in the 1950s and by Hase himself in the 1970 and the 1980s.3 We are also particularly fortunate that several Qing Certificates of Title referring to land claims and disputes on Lamma Island survived to be included in Hase’s recent study of the traditional



A Time of Great Change: Sha Po during the Ming–Colonial Era   135

land law of Hong Kong’s New Territories.4 These records offer fascinating insights into the relationships between resident farmers and absentee landlords, and socio-political and economic activities of Lamma communities, during the Qing dynasty and early decades of the lease.

Absentee landlords and rice-farming clans Between the Southern Song and Ming dynasties large tracts of uncultivated (untaxed) wasteland in the New Territories were granted by imperial decree to ‘worthy officials’, who then undertook to have peasant farmers convert the land to arable.5 It was under such circumstances that ‘Pok Liu Chow’ (舶寮洲) or Lamma Island was granted by imperial decree to Yiu Cho of Nantou, Guangdong, most probably during the Ming dynasty.6 The Yiu clan then established the Yiu Yi Yin Tong (Chinese characters unknown) ancestral trust to oversee their responsibilities with respect to the payment of imperial land tax and role as absentee landlords to the island’s farming and fishing communities.7 It thus seems likely that some farmland may have been opened up on Lamma during the later Ming dynasty, but any pre-existing land tax registrations would have been cancelled during the Coastal Evacuation. Interestingly, we have hardly any archaeological evidence for Ming activity at Sha Po—and what little there is probably dates to the earlier Ming—while all existing villages on the island seem to date from the early eighteenth century or later. According to Certificates of Title and other surviving historical documents issued by the Xinan County Magistrate in 1736, 1883, and 1889, most of the cultivated land, stake-net fishing stations, and hill-slopes on Lamma were then owned by the Yiu Yi Yin Tong.8 One branch of the Yiu clan actually moved from Nantou and settled at Luk Chau on Lamma around the end of the eighteenth century.9 Although the surviving documents date to the Qing dynasty, the Certificate of Title issued by the Xinan Magistrate in 1736 states that Lamma was ‘bequeathed to them [the Yiu clan elders] by their forefathers’ (presumably in the Ming), while one of 1883 refers to Yiu clan land registrations made ‘in the previous dynasty’ (i.e., clearly in the Ming).10 Given the widespread famine and loss resulting from the Coastal Evacuation, it is probable that few members of any late Ming communities that were established on Lamma, returned to reclaim their farmland. Against that historical backdrop, the Yiu Yi Yin Tong’s registrations of land ‘already under cultivation’ during the 1730s are interpreted by Hase as evidence for their attempts to renew claims broken during that major hiatus.11 It also suggests that the Tong already had tenants in residence and farming the land at that time. The earliest of the Tong’s various declarations of landholding refers to areas west of Wang Long, which is widely accepted as the founding village of Lamma’s oldest resident clan, the Chaus (周), who moved to Lamma from Heung Kong Wai (today’s Wong Chuk Hang San Wai). Hase estimates an initial date of settlement at Wang Long of around 1710 by three brothers, the younger two of whom later established satellite villages at Ko Long and Tai Yuen by around 1750.12 Sha Po Old Village, the place central to this book, was also a Chau clan settlement, which was established by villagers from Ko Long, and therefore probably dates from around 1780 to 1805.13 Perhaps two generations thereafter, between roughly 1850 and 1870, the Tsang clan also settled at Sha Po and rented most of their fields from the Chaus.14 The Chaus, like Lamma’s other major clans, the Chans and Ngs, were obliged to pay rent-charge to the Yiu Yi Yin Tong, despite the fact that large areas had been converted from waste to arable by the farmers themselves and were therefore not previously recorded in the County Land Tax Registry. Disputes between Lamma’s resident farmers and absentee landlords inevitably arose and, as noted above, were recorded in official documents, which show that, in most cases concerning landholding rights, judgements made by the Magistrate respected the Customary Land Law, and found in favour of the landlords.15 The Tong still dominated landholding on Lamma when Britain took possession of the New

136   Piecing Together Sha Po

Territories in 1899, but by 1904–5 all land had to be registered for Crown rent, which brought about a wholesale reconfiguration of pre-existing landlord-tenant relationships—but why did the change of governance have such a dramatic effect? The answer lies in a combination of the administrative frailty of the Qing government and opportunism on the part of landlords. Almost unbelievably, given the dramatic demographic change that had occurred, no imperial survey of New Territories landownership had taken place during the entire Qing dynasty, and landlords had therefore not felt obliged to register the often huge areas of new arable land opened up by their tenants. Thus by the late nineteenth century Lamma’s tei kwat (subsoil) landowners, the Yius, were levying a rent-charge from their tei pei (topsoil) tenants on an area of land somewhere between twenty and twenty-five times as large as that they had declared for government land tax in 1744.16 As a consequence, tenant farmers’ rents were generally low—as the landlords could easily afford them to be—but when the Hong Kong government introduced the Block Crown Lease, the old landlords such as the Yiu Yi Yin Tong lost control of those areas of land not previously registered for Land Tax at the County Yamen.17 Moreover, the government also determined that any existing permanent tenants of tax-registered lands could not be dispossessed by their former landlords nor could their status as permanent occupiers on fixed rents be altered. The government thus ‘instituted a major land revolution in the New Territories, and specifically on Lamma, by making the old tenants new Crown Lessees, paying rent only to the Government’.18 Almost overnight, therefore, the Yiu Yi Yin Tong was cut out and Lamma’s resident rice-farming clans took formal control of the land they and their ancestors had opened up and worked for generations.

The impact of piracy and policing on patterns of settlement Piracy was at various times along the South China coast a serious menace to shipping and coastal communities alike, in particular around Hong Kong’s bays, channels, and islands with their many opportunities for concealment and ambush.19 The eighteenth century, when most of north-west Lamma’s early settlement occurred, was a period of endemic piratical activity in Hong Kong waters. The infamous local pirate Cheung Po Tsai caused significant economic disruption in the late eighteenth and early nineteenth centuries, but was finally defeated by the Qing Navy at Tung Chung in 1810.20 While Cheung was perhaps the last well-known figure, piracy actually persisted locally until it was largely suppressed in the later nineteenth century as a result of regular patrolling by the British Royal Navy and Chinese Imperial Maritime Customs.21 Against that background the patterning of eighteenth-century settlement near Yung Shue Wan—itself an earlier nineteenth-century development—seems to indicate a quite understandable concern with security. It is probably no coincidence that the Chau clan’s early settlements at Wang Long, Ko Long, and Tai Yuen, and their somewhat later village at Yung Shue Long, were all established in sheltered hollows several hundred metres inland where they could not be observed from the coast.22 The Chaus’ initial avoidance of the area immediately inland of Yung Shue Wan may also in part be a reflection of the relatively poor soils there, but there were many other sites—more visible than those chosen—that were ignored. A similar desire for concealment is suggested by the Chan clan’s positioning of the old village at Tai Wan (probably founded between 1725 and 1730), while the Ng clan’s hilltop settlement at Tai Ping (established c. 1800) overlooks Chau lands in the Yung Shue Long valley, but is itself concealed. Somewhat further afield, but nonetheless worthy of mention, is the old village of Pak Kok—settled around 1730 by Chans unrelated to those at Tai Wan—which occupies a perfectly concealed position just inland from the later and more visible new village.23 Other factors beyond security and soil quality may also have played their part. For example, Tai Ping’s relatively remote location may simply reflect the fact that it was a later foundation

Map 9:  1905 map of Yung Shue Wan with enlarged map of late Qing to early twentieth-century landscape. Source: CLSO. Lamma Island, D.D.3 Sheet 2. 1:1980 Scale Map. Hong Kong: CLSO, 1905; CLSO. Yung Shue Wan, Lamma Island. 1:1000 Scale Topographic Map Sheets 14-NE-10D and 14-NE-15B. Hong Kong: CLSO, 1981a and 1981b respectively. Reproduced with permission of the Director of Lands. © The Government of the Hong Kong SAR. Licence No. 59/2015.

138   Piecing Together Sha Po

and all prime low-lying locations had by then already been claimed by other clans. We should also bear in mind that fung shui (or geomancy) would probably have played some part in determining the precise details of where new villages were built, and how they were arranged.24 Moreover, while on the subject of ‘wind and water’: water was obviously a key concern for wet rice farmers, and all the early low-lying villages mentioned above were actually in ‘classic’ slightly raised positions overlooking land fed by small streams.25 Primarily, though, the early Qing settlement pattern in the hinterlands of Yung Shue Wan reflects an economic focus on rice farming married to a need for domestic security. Improved security of sorts would arrive in the wake of the First Anglo-Chinese or Opium War (1839–42), which led to the British possession of Hong Kong (1841), and eventually its formal creation as a Crown colony in 1843. Soon after its establishment, the British entrepôt became the main gateway for opium transhipment into the Chinese mainland and elsewhere,26 much of it as registered cargoes, but significant quantities also as smuggled contraband. The dramatic increase in numbers of foreign and local merchant vessels operating in Hong Kong waters prompted the Qing Imperial Maritime Customs to establish a number of customs houses or posts—such as at Fat Tau Chau—for collecting tax on foreign trade and to prevent smuggling.27 Lamma Island, which remained Chinese territory until the New Territories’ lease in 1898, was an ideal base for surveillance of the important shipping lane of the East Lamma Channel. According to Hase,28 ‘as part of the “Customs Blockade” of Hong Kong’ the Qing government operated a small Maritime Customs Post between 1860 and 1898 near the Yung Shue Wan Tin Hau Temple. According to the Qing dynasty Guangdong Tu Shuo (Illustrations of Guangdong), the post was small with around ten soldiers stationed there.29 Interestingly, in the Tin Hau Temple is a wooden tablet dated 1877, the donation of a Chinese naval officer, which presumably indicates that Yung Shue Wan was a stopping place on the war junk patrol route south of Hong Kong.30 Protection of a different sort was provided by the Tin Hau Temple, which is probably the oldest structure in Yung Shue Wan, although the earliest dated object—a wooden tablet of 1876—records a restoration of the building, which was perhaps first erected in the late eighteenth or early nineteenth century.31 The temple affords geomantic protection to the fishermen’s moorings, the ancient flat-rock landing place and breaming ground in front (see further details below), as well as Sha Po Old Village to the rear. The temple’s foundation is thus thought to slightly pre-date the development of a small market at Yung Shue Wan, supposedly around 1810–30, near the flat-rock landing place and on the landward side of a beach-top footpath running around the bay just above the high-water mark. The first row of shops at Yung Shue Wan was built between Main and Back Streets just north of the Tin Hau Temple, while a second row of shops was built out over the beach in the later nineteenth century.32 Therefore the relatively late development of the coastal strip can be seen as a natural response by villagers to the much improved security resulting from the gradual suppression of piracy and later military presence in the area.

The later historical economy of Sha Po and environs Throughout the Qing dynasty, the Yung Shue Wan area had two distinct populations: one of rice farmers—some of whom also operated coastal stake-nets—and the other of Hoklo fisher-folk who lived either on their boats or in stilt-houses at O Tsai. It is reported that ‘about 40–50 trawlers’ anchored in the bay towards the end of the nineteenth century.33 As Lamma’s population grew and the subsistence economy came under ever-increasing pressure, conflicts arose between local farming and fishing communities and their Yiu clan landlords. The Yiu Yi Yin Tong’s two Certificates of Title taken out in 1883 and 1889 are, respectively, clear evidence that local farmers and fishermen were resisting the Tong’s demands for rent on coastal stake-net sites and hillsides used for agriculture and fuel gathering.34 The



A Time of Great Change: Sha Po during the Ming–Colonial Era   139

latter disputes are probably also an indication that landlords and resident clans alike both recognised the economic opportunities presented by the Hong Kong urban population’s insatiable demand for perishable foods and fuel, which Lamma was particularly well placed to supply via Aberdeen. By the middle of the nineteenth century a number of wholesale buyers of livestock, fish, fruit, charcoal, and fuel grass, among other things, had established themselves in Yung Shue Wan. Some of Lamma’s fuel-grass was used locally at the boat breaming facility located in front of the Tin Hau Temple. There, on beaches to either side of the creek, boats were upended and had their hulls breamed—burnt clean of barnacles and other detritus—before being caulked (waterproofed) with lime and then repainted. Oral history records that the Yiu Yi Yin Tong upset the local fishermen by claiming ownership of the beaches and thus the right to charge for their use. In this instance the Yius came up with an ingenious compromise by agreeing to donate the money from such rents to the upkeep of the temple—thereby simultaneously gaining ‘face’ while placating the fishermen who already supported the temple’s upkeep.35 Early twentieth-century annual reports by the district officer (south) show that Sha Po and nearby villages were heavily involved in poultry and pig husbandry, while cattle reared in South Lamma were driven and held in fattening pens specially constructed for the purpose by the Chau clan at Sha Po.36 All the above products, including live cattle, were shipped by boat to middlemen in Aberdeen for onward transportation to Hong Kong.

Discussion The preceding overview of Ming–Qing Sha Po, Yung Shue Wan, and north-west Lamma more generally, has sought to provide some socio-historical and political background for the archaeological discussions that follow. The clan genealogies, oral testimonies, historical documents and historic landscape together offer some useful, and in places fascinating, insights into the area’s traditional way of life since the Coastal Evacuation. They also give us a general chronological framework, with reference to which we might expect to find few, if any, traces of later Ming date, but have a much higher expectation of encountering later remains connected with the settlement and development of the Yung Shue Wan–Sha Po coastal area. We next move on to explore the later historical archaeology of that area, beginning with the rather scant evidence for Ming period activity, followed by the relative abundance of Qing and early twentiethcentury remains reflecting the life and times of Sha Po Old Village’s farming community and their near neighbours overlooking the bay.

Sha Po’s Later Historical Archaeology The missing Ming: Demographic shift in a time of change? Nothing approaching the intensity of human activity connected with the demise and post-abandonment phase of the coastal industry—between perhaps the tenth and thirteenth centuries—is seen again at Sha Po until the later Qing dynasty, in the form of nineteenth-century structures and rubbish dumps relating to the settlement of the Old Village. The historical suggestions that a substantial depopulation of Lantau, and perhaps also Lamma, occurred during the later Southern Song are therefore apparently reflected in the archaeological evidence. Whether that resulted on Lamma from punitive action by the Southern Song authorities, or was a consequence of an economic shift towards more centralised production in a few larger salt-fields is unclear—but the gap appears to be real.

140   Piecing Together Sha Po

Hardly any definitively Ming material was recovered from the Sha Po backbeach or plateau sites, and none from well-stratified cultural horizons clearly of that date. Some categories of pottery, such as the yellowish-brown or yellowish-green glazed stonewares found at Sha Po and elsewhere, were long-lived and can only be dated as ‘Southern Song–Ming’. But against the general absence of post–fourteenth century material, such types are probably more realistically identified as ‘Song’ rather than ‘Ming’. The few probable Ming artefacts that were recovered either came from Song spreads overlying Tang industrial horizons, in Qing deposits, or in contexts disturbed by modern activity. Two lid sherds, one bearing the character ‘酒’ (wine)—typical of Ming dark brown wine jars—were recovered from a late Qing dump deposit [36:605] behind the original housing block of the Old Village. A grey, lugged bodysherd of Thai manufacture and a probable sixteenth-century date was found in a disturbed context [39:102] near Kiln K6, while a paddle-stamped fifteenth- to sixteenth-century Bau Malay37 sherd was found in another mixed layer [32:204] in the south-eastern part of the backbeach. Several fragments of celadon bowl from a layer disturbed by modern foundations above Kiln K1 [16:T4/L3] bore quite fine incised floral and

Figure 28:  Ming pottery: (a) ‘bow-lugged’ jar [36:605], (b) lid [36:605], (c) lugged jar sherd probably Thai and sixteenth-century [39:102]



A Time of Great Change: Sha Po during the Ming–Colonial Era   141

Plate 33:  Ming pottery: lugged storage jars (above) and wine jar lid (below)

lotus leaf decoration—suggestive of Ming date—but the sherds were too small to allow a definitive identification and may be Southern Song–Yuan. For most periods covered by this book, finds assemblages are simply so large as to render full description impracticable; the fact that the artefacts listed above are the sum total of ‘definitive’ Ming material recovered from Sha Po rather emphasises the contrast. Taken as a whole, the patterning of Ming material on the backbeach is maybe indicative of little more than occasional visits by boat-dwelling fisher-folk, traders or war junk patrols, perhaps to collect drinking water or other natural resources. The presence of Thai and Bau Malay jars at Sha Po, which were also found at Penny’s Bay and in the Sha Tsui shipwreck, is worthy of note here as such tough, robust storage wares were probably popular with maritime-focused groups. The evidence from Sha Po, Yung Shue Wan, and Lamma more generally, clearly indicates that the fifteenth to seventeenth centuries were a relatively quiet period in the island’s history. However, from around 1730, the well-documented settlement of the island by rice-farming clans heralded a period of sustained economic and population growth that eventually led to the Chau clan’s establishment of a new settlement at Sha Po, thereby beginning the latest chapter in our archaeological story of life on the ancient backbeach.

142   Piecing Together Sha Po

Village growth and daily life in Qing–Colonial Sha Po Introduction As noted above, the historical records suggest that Sha Po Old Village was constructed on the backbeach sometime around 1800—perhaps soon after the erection of the Tin Hau Temple—while Yung Shue Wan began to spread along the coast a decade or two after that. According to the 1905 map of the Block Crown Lease Survey (Map 9), Sha Po then consisted of one main row of four houses orientated southwest–north-east and two smaller, similarly aligned, blocks of houses to the north-west. Besides the four main excavations on the Sha Po New Village plateau, the archaeological investigations providing the data for this research were conducted in an area encompassing the houses and fields of the Old Village and the southern end of the earliest, eastern part of Yung Shue Wan following the line of Back Street. Significant quantities of nineteenth and early twentieth-century material were, not unexpectedly, discovered in such areas, but with quite distinctive and culturally meaningful patterning. A general scattering of such material was encountered in disturbed near-surface deposits in every trench excavated, but there were also several more intense areas of apparent rubbish disposal, some of which were also related to apparent revetment of the rear face of the backbeach. Interestingly, the foundations of two old buildings were found close to the extant eastern block of houses of the Old Village and clearly relate to an earlier phase of the settlement. We now move on to explore the activities of Sha Po–Yung Shue Wan’s nineteenth to early twentiethcentury communities through the material remains they left behind. Dwelling: House foundations and yards When investigating the environs of old villages, perhaps the most obvious traces we might expect to encounter are the remains of earlier buildings and their outside surfaces—houses and yards. Excavations at the south-western end of Sha Po Old Village revealed the wall foundations of two separate structures— S1 at the south-east end of Trench 33 and S2 in the north-east half of Trench 22—both probably relating to an earlier phase of the settlement. Structure S1 survived as a short length of wall observed running south-west–north-east across the width of a narrow trench [33:409] excavated a few metres south of the main Old Village blocks. The wall’s construction of rounded granite boulders bonded with a sandy orange mortar is quite typical of

Plate 34:  Structures S1(right) and S2 (left)



A Time of Great Change: Sha Po during the Ming–Colonial Era   143

Plate 35:  Old and new houses at Sha Po Old Village

older buildings on Lamma and one might therefore expect it to be pre–World War II in date. Interpreting such a small fragment of a structure is difficult; however, the presence of plaster on the north-west face of the wall, which seems to overlap two layers of cement flooring, suggests that this was the inside of the building. Further confirmation was provided on the wall’s south-east side, where a rubble-rich deposit [33:403] had been dumped to level-up the rear slope of the backbeach for an area of granite paving [33:410]—presumably an outside yard—which abutted and was therefore later than the wall. The rubble deposits contained late nineteenth-century to pre-war porcelain with village ware, a 1950 King George VI 10-cent coin, and a near-intact wine jar of post-1946 manufacture (see below), which collectively suggested a middle twentieth-century date (or later) for the paving—the building itself may be somewhat older. The wall’s alignment matches that of the extant Old Village blocks and is perhaps the remains of a free-standing house or outbuilding added to the village during the earlier twentieth century. The second structure (S2) was defined by a single course of dressed grey granite blocks that defined the partial footprint of an apparently open-plan rectangular building with traces in section of a thin, rammed earth (chunnam) floor. The foundations were generally around a half metre wide, but included a locally expanded area in the middle of the eastern side, which probably marks the former position of a doorway. The excavated portion measured approximately 4 m wide by 6.5 m in plan, while the long axis was on precisely the same alignment as the present two-row housing block of the Old Village.38 The width of the building is within the range typical for traditional village houses, whose long, thin shape was largely determined by the length of pre-cut roof timbers supplied to local markets by producers near Guangzhou.39 The two rows of six houses forming the Old Village at Sha Po are similarly long and narrow, as are many

144   Piecing Together Sha Po

other houses in nearby villages and the older parts of Yung Shue Wan. While some still exist in single storey traditional form with their distinctive stacked tile roofs and pointed gable ends, many houses have been rebuilt using reinforced concrete, but still respecting an architectural template determined by the standardised roof timbers of yesteryear. Deposits associated with the foundations contained a variety of nineteenth-century domestic materials that provide a general timeframe for the house. We now move on to examine the wider evidence for rubbish deposits identified behind old housing blocks in Sha Po Old Village and Yung Shue Wan Back Street.

Traditional patterns of rubbish disposal and recycling It has been written, tongue in cheek, that ‘archaeology is rubbish’,40 because many of the things archaeologists find were accidentally broken and intentionally discarded by people in the past. Before World War II, villagers on Lamma, as in the rest of the New Territories, had local solutions for disposing of domestic rubbish that was, in the main, organic food remains or broken artefacts—bits of pottery, tile, and the like—they dumped it ‘out the back’ in the household or village rubbish tip or midden. Similarly, when a settlement’s footprint needed enlarging to accommodate a growing community, such domestic rubbish was often combined with building rubble and used as fill to expand the developable area. In addition to the aforementioned use of rubble fill for paved yards, two discrete nineteenth to early twentieth-century rubbish dumps with dense concentrations of finds were identified in excavations behind the oldest extant two-row block of Sha Po Old Village [36:605 and 24:01]. A further rubbish deposit of similar date was also encountered to the rear of early houses in Back Street, in two separate campaigns of fieldwork in 2002 and 2009 [respectively 18:01–02 and 42:103]. While the Back Street deposit survived as a roughly 40 cm deep spread several metres in extent, those behind the Old Village

Figure 29:  Late Qing rubbish dumping and granite revetment used to level up sloping rear profile of backbeach and expand developable area



A Time of Great Change: Sha Po during the Ming–Colonial Era   145

had a quite level upper surface but a distinct sloping interface with the Tang dynasty layer underneath, indicating that they had been tipped down the rear face of the backbeach. The northernmost example [36:605], in particular, provided clear evidence of rubbish dumping (see finds discussion below), which was also used as fill for the expansion of the settlement platform out over the fields to the rear. Also within this material was a number of roughly squared grey granite blocks [36:606]—measuring on average 20 × 25 × 30 cm—which when viewed in section seemed to be a revetment (retaining wall) for the dumped material, effectively converting it to fill. The revetment stopped in line with the north-east corner of the Old Village block and was clearly related to it. Following its construction, rubbish continued to be tipped over the revetment, eventually burying it and in-filling the area beyond. Two final observations concern the aforementioned granite paving [33:410] which, although midtwentieth century in date, is made from strikingly similar blocks to those used in the late nineteenth to early twentieth-century revetment. It is possible that an earlier revetment—similar to that found 40 m further north—was robbed out and relaid as the paved yard. The paving probably marks the southern limit of intense activity associated with the construction and inhabitation of Sha Po Old Village, as nothing similar was evidenced in excavations further to the south.

Have you eaten rice yet? Reconstructing domestic consumption patterns The range of finds in the various rubbish spreads and dumps, in particular the larger ones mentioned above, provides some interesting insights into various aspects of everyday life; here, we examine materials relating to that true Chinese passion: mealtime. Food processing and cooking are represented by sherds from rice grinding bowls, many stoneware storage jars and cooking pots with wall-mounted handles, plus a number of lids, which are all very typical

Plate 36:  Village children eating lunch in Sha Tin, probably late nineteenth to early twentieth century. Courtesy of FormAsia Books, Hong Kong.

Figure 30:  Qing to early twentieth-century ceramics: (a) Wun Yiu/Wuhua spouted wine/water jar [42:102–3; CSF24], (b) Wing Lee Wai Winery wine jar [33:403], (c) Wun Yiu bowl [36:605], (d) brown glazed dish [36:605], (e) lid [36:605], (f ) cooking pot [36:605], (g) basin [36:605], (h) oil lamp-stand [42:102], (i) candle-holder [36:605], (j) oil lamp-dish [42:102]

Figure 31:  Rubbings of Chinese characters on Qing ceramics (company or brand names): (a) ‘Wo Sang’ with money symbols [36:605], (b) ‘Sang Lee’ (‘profit making’) moulded on rim of storage jar [34:502], (c) ‘Yiu Thai’ brand name moulded on base of opium pot [36:605]



A Time of Great Change: Sha Po during the Ming–Colonial Era   147

Plate 37:  Typical late Qing to early twentieth-century ceramics: (a) brick and tile, (b) stonewares for rice grinding, cooking, and serving, (c) oil lamps, (d) Shiwan tiles, (e) net-weights, (f ) gaming pieces, and (g) porcelain tablewares

nineteenth- and early twentieth-century kitchenware. In addition, several chunks of thick-walled orange earthenware vessels with ‘castellated’ tops found in Back Street [40:102; 42:102; 43:102] and at the Old Village [23:02] were identified, based on parallels at Tung Lung Fort, as pieces of charcoal cooking stoves.41 The storage jars and cooking pots typically have a brown glossy glaze, which in the latter case is often on the inside only. Although they were mass-produced utilitarian wares, the cooking pots are actually skilfully made, sometimes with walls just 2–3 mm thick. Invariably, the flat bases are blackened by soot from their use over smoky, wood- or grass-burning stoves. Some of these village wares at Sha Po bore moulded inscriptions (see rubbings in Fig. 31), being either the name of the workshop that made the pots or the company that commissioned their production.

148   Piecing Together Sha Po

Compared to the kitchenwares, both the larger rubbish dumps and more localised spreads contained relatively larger quantities of broken ceramic spoons, cups, bowls, plates, and dishes—tablewares as seen in Plate 36—used for serving, eating, and drinking. These included the remains of flat-bottomed, brown glazed stoneware dishes—a myriad of which came from the northern midden [36:605]—which judging by their similar glaze and fabric were probably made at the same kilns as the cooking pots—somewhere in Guangdong.42 There were also many blue-and-white bowls and rather fewer plates and tea cups—mostly

Plate 38:  Qing blue-and-white porcelain: (a) bowls, (b) rice wine warmer, (c) ‘double happiness’ bowl, (d) dish



A Time of Great Change: Sha Po during the Ming–Colonial Era   149

Plate 39:  Qing teapot made at Yixing kilns

from Hong Kong’s very own porcelain kilns at Wun Yiu, Tai Po43—but also better quality blue-andwhite types from Fujian and Guangdong, and Jiangxi white qingbai wares, probably from Jingdezhen. A number of polychrome cup fragments, including two from Jiangxi, date to the early twentieth century. The blue-and-white porcelain was mostly decorated with a combination of ‘double happiness’, Buddhist inscriptions, bamboo, lotus, and floral motifs (Plate 38). The variety of ‘tea sets’ suggested by the many cups found was complemented by finds of teapot body, lid, and spout fragments in Back Street [18:01; 19:01], and numerous others in various locations around the Old Village [24:01] including an attractive pumpkin-shaped moulded and painted teapot made at Yixing, Jiangsu during the nineteenth century (Plate 39). While tea was no doubt an ever-present accompaniment to meals, rice wine was also a widely consumed beverage—presumably with dinner rather than lunch given its potency—and thus the discovery of part of a four-lugged, spouted wine jar among the Wun Yiu pottery from Back Street [42:102] is not surprising.44 Similar jars were unearthed during excavations of the eighteenth to early nineteenth-century fort at Tung Lung Chau and at the Wun Yiu kiln site itself.45 The upper body of a blue-and-white ‘sweet pea pattern’ rice wine warmer and pourer,46 with traces of a small delicate handle, was made in Fujian in the nineteenth century. A fine gourd-shaped, brown glazed wine jar (Fig. 30(b)), bearing the inscription ‘天津香港永利威酒莊’ (Tianjin Hong Kong Wing Lee Wai Winery), was found in near-complete condition in deposits associated with the granite paving outside structure S1 [33:403]. Wing Lee Wai Winery—which was first established in Nanhai, Guangdong in 1876 and moved to Hong Kong in 1903—remains one of the very few traditional old Chinese wineries still operating in Hong Kong. Such brown glazed gourd-shaped bottles were products of the Shiwan kiln in Guangdong between 1925 and 1949, a date range which correlates well with the ‘Tianjin’ company name that was adopted after 1946.47

150   Piecing Together Sha Po

Plate 40:  (a) ink-stone, (b) penholder, and (c) ink bottle

Literacy, leisure, and light: Other aspects of domestic life Circumstantial evidence for literacy was provided in the form of a brown glazed ceramic spouted ink bottle found in the Back Street rubbish spread [42:102] bearing the Shanghai-based ink manufacturer’s name around the shoulder ‘上海山打營業公司留香墨汗’ (ink produced by Shan Da Company of Shanghai) and the potter’s name ‘何源記’ (‘Ho Yuen Kee’) on the base. It is interesting to note that the most readily available published examples of similar bottles found elsewhere, including some with identical inscriptions referring to the same manufacturer, were unearthed in Chinese settlement sites in North America dated to the early twentieth century.48 Further evidence for literacy was found in disturbed deposits nearby, which produced a plain rectangular ink-stone of suspected Ming dynasty manufacture, while a much finer quality, but totally worn-out, shale ink-stone of probable Qing date was found in the northern midden [36:605].49 The latter bore a well-executed melon-and-leaf decoration on the upper surface and the Chinese character ‘有’ (‘have’) inscribed on the back. The inscription might relate to ownership and the owner’s name was perhaps inscribed on the missing portion that still awaits discovery in some future excavation. The final piece of writing-related material culture—interestingly found near the shale ink-stone—was the base of a cylindrical blue-and-white porcelain penholder from Fujian or Jiangxi and of nineteenth-century type.



A Time of Great Change: Sha Po during the Ming–Colonial Era   151

The playing of board games, and perhaps gambling, was suggested by the discovery of five probable gaming counters (Plate 37(f )), among which two came from Qing deposits at the south of the Old Village [23:03; 22:01], and two more from the nineteenth-century Back Street rubbish spread [18:01]. These roughly shaped ‘discs’ measured between 3.5 and 5 cm in diameter and were very similar to examples found at Tung Lung Fort, which had also been shaped from old ceramic tiles.50 Before the quite recent advent of mains electricity on Lamma, illumination for indoor gaming, calligraphy, and all other activities relied upon oil lamps and candles.51 Three fragments of ‘two-tier’ oil lamp-stands were found in the Sha Po northern rubbish dump [36:605]: a pearly green glazed middle dish and a brown glazed middle dish and stem were both nineteenth to early twentieth-century Shiwan types, whereas another brown glazed example was probably nineteenth century in date (Plate 37(e)).52 The main Back Street rubbish spread [42:102] contained the top of another Shiwan pearly green glazed lamp-stand found with a brown glazed candle-holder of Qing to early twentieth-century date. A second spread nearby [19:01] contained fragments of a further lamp-stand and candle-holder of similar date. Lastly, an oil lamp-stand was found in association with house foundation S2 [22:01] just south-west of the Old Village.

Drugs and medicines Two small glass medicine bottles were unearthed from the northern midden [36:605], one of which was bluish-green in colour with a distinctive ribbed cylindrical body, rounded shoulders, and a narrow upright neck, most of which was missing. On the base were the Chinese characters ‘博濟軒’ (‘Pok Chai Hin’), which literally means a hall for giving mercy and offering pecuniary aid to the general public. Names like this were popular with Chinese medicine makers and clinics, and Pok Chai Hin is probably therefore a company name. Many medicine bottles of this type were found in nineteenth to early twentieth-century Chinese immigrant settlement sites in North America; however, there they are often erroneously interpreted as ‘opium bottles’.53 As the historical records attest, once the British East India Company had established a steady flow of opium into China from around the middle of the eighteenth century, its consumption became a common leisure activity for some groups in Hong Kong Chinese society, while for others engaged in more physical work it also offered a welcome pain-killing effect. Although opium paste—so-called prepared opium— was normally smoked by Chinese people for both its pleasurable and analgesic effects, the latter could also be achieved by eating it. Its widespread use continued locally until as late as 1945 when it was finally prohibited by law.54 Against that later historical pattern of use, the discovery in nineteenth to early twentieth-century rubbish dumps at Sha Po Old Village [22:01; 23:04; 33:403; 34:505] and behind Back Street [18:01] of five small ceramic opium pots—plus three others in surface spreads [26–27:01]—is hardly unexpected. Also found in the Back Street dump was the perforated top of a typical Shiwan-made opium pipe bowl and one half of a larger ceramic opium pot.55 The number is small when compared with the 173 opium pots found during excavations in the grounds of the restored Chinese mansion at Tai Fu Tai, San Tin.56 The difference probably reflects the relative wealth of the two communities and maybe also the contrasting social context of the drug’s use. The small pots originally had lids and it has been noted that such containers were also sometimes used to store other types of medicinal ointment, although opium was most commonly their contents.57 One of the small cylindrical containers from Sha Po carried the inscription ‘耀泰’ (Yiu Thai), which possibly refers to a local brand (see rubbing in Figure 31). As Sinn has recently discussed, in the later nineteenth century throughout the Chinese diaspora, Hong Kong was considered to produce the finest quality prepared opium, with many branded types being available. But how should

152   Piecing Together Sha Po

Plate 41:  Opium paraphernalia and glass medicine bottles: (a) lid of opium pipe bowl, (b) prepared opium pots, (c) medicine bottles

we interpret the five small opium pots found at Sha Po? Interestingly, in over forty years of excavations in the Sha Po–Back Street study area there has been only one possible fragment of ceramic pipe-bowl recovered, but no remains of actual pipes or other definitively opium-related paraphernalia. In contrast, North American sites have produced a wide range of ceramic and metal pipe-bowls, as well as other objects clearly relating to opium smoking. However, it should be remembered here that overseas Chinese were generally wealthier than their stay-at-home countrymen, to whom they routinely sent back money from Gold Mountain.58 In the Sha Po–Back Street area those who could afford opium would probably be few in number and relatively occasional users, and they probably used the cheapest bamboo pipes, which do not survive archaeologically. So perhaps the Sha Po opium pots should be interpreted in terms of the drug’s occasional medicinal use as an analgesic by hard-working villagers to ease aches and pains resulting from agricultural work or fishing, or even just plain old age, rather than its recreational use.



A Time of Great Change: Sha Po during the Ming–Colonial Era   153

Plate 42:  Modern stilt-houses at O Tsai, Yung Shue Wan

Farmers and fishermen Staying with the theme of farming and fishing, although Sha Po was at heart a rice-farming settlement, fishing has been an important aspect of Lamma’s economy throughout its human occupation right up to the present day. During the nineteenth to early twentieth century, fishing at Yung Shue Wan took three different forms: offshore trawling by boat-based fisher-folk, inshore sampan fishing by a mix of Hoklo fishermen and Punti villagers, while the latter also operated the many stake-net fishing stations around the coast. Although the majority of fishing boats working out of Yung Shue Wan were no doubt crewed by fisher-folk living on their boats or in stilt-houses—as some still do today along the northern shore of the bay at O Tsai—a mid-twentieth-century account mentions that Sha Po was a ‘major source’ of ‘landsmen’ employed as extra hands by Hoklo fishermen.59 It is therefore interesting that in total well over one hundred barrel-shaped clay net-weights have been found in the study area encompassing the Old Village and Back Street.60 Over the last few decades of archaeological fieldwork at Sha Po, small numbers have occurred in near-surface deposits disturbed by development and such weights therefore appeared to be a later historical phenomenon—indeed Spry suggested that they were ‘a type still in use until recent times’.61 The recent discoveries of two discrete clusters of twenty-seven and forty-two such net-weights—respectively found in the rubbish dumps behind Back Street [18:01-02] and in a large-scale rubbish dump behind the Old Village [36:605]—could be quite confidently dated to the nineteenth to early twentieth century by pottery found in association. However, such basic functional objects are likely to have been used in similar form for centuries, at some point replacing the grooved soapstone weights of probable later prehistoric–Han date, which in turn replaced the ubiquitous waisted pebble net-weights that are such a feature of many local prehistoric sites.

154   Piecing Together Sha Po

Plate 43:  Ceramic net-weights

In the end, some affirmation for Spry’s suggestion above was provided by current Yung Shue Wan fishermen who said that lead net-weights only became common on Lamma by around the middle of the twentieth century.62 Common sense conservatism was clearly a guiding principle in the lives of Lamma’s traditional fishing and farming communities alike; however, within one generation of World War II—when our archaeological story notionally ends—the old way of life was disappearing fast as villagers pursued opportunities abroad and the nearby colonial metropolis exerted an ever greater socio-economic and political influence on its hinterlands. Lamma Island’s communities, and Yung Shue Wan’s in particular, were in flux, tenaciously retaining what aspects they could of their old New Territories value systems and way of life while, at the same time, embracing the opportunities presented by the island’s emergent touristic and commuter-based economy.

Reflections This fourth and final chapter of Part II has taken our archaeological story from the mysteriously quiet and seemingly empty coastal-backbeach landscape of the Ming dynasty on into a contrastingly vibrant period of busy socio-economic activity. Historical sources assisted in the creation of a specific socio-economic and political context for our discussions of the settlement and utilisation of Yung Shue Wan’s hinterlands and coast. The archaeological evidence helped illuminate perhaps 150 years of socio-economic activity



A Time of Great Change: Sha Po during the Ming–Colonial Era   155

and domestic life in and around Sha Po Old Village and Yung Shue Wan Back Street. The remains of earlier houses, yards, revetment walls, and rubbish dumps offered insights into the character and progress of settlement evolution over maybe a century or so. The latter also revealed how Sha Po villagers ‘recycled’ rubbish and rubble as fill behind revetment walls and under paved yards—thereby putting waste material to good use as the community grew and needed room to expand. In net-weight dumps, we found evidence for the mixing and interaction of fishing and farming communities, with economic expediency overcoming any ethnic-cultural barriers to ensure that boats were crewed and fish were caught, to the general benefit of both communities. Hearth and home were well reflected in a wide range of domestic objects relating, predominantly, to the preparation and consumption of food and drink, but also lighting, games, literacy, and medicines. We can imagine a glowing charcoal brazier, cooking in progress, the evening-dark interior of a traditional village house, dimly lit by oil lamps and candles, but sufficiently good for writing—of maybe clan histories or business transactions—with ink bottle and stone, penholder and brush all to hand, and a nearby dinner table set with teapot, porcelain cups and bowls, and brown glazed dishes steaming with food—rice wine also if funds allowed. In Hong Kong such intimate details of domestic life can really only be written about Qing or later settlements and few, if any, have ever been thus described in print. For much of local archaeological history our interpretations are by necessity—or more accurately, due to lack of material detail and understanding of socio-historical context—broad brush and vague. In contrast, the rich contextual detail of historical archaeology offers an opportunity to write fuller and more nuanced interpretations of everyday life through the collective reanimation of ordinary objects within the social context of their past use. What surprised us somewhat was how difficult it was to find published material against which we could compare our Sha Po discoveries. Besides the reports from excavations at Wun Yiu’s blue-and-white porcelain kilns and a few other unpublished reports from Qing sites—most notably Tung Lung Fort—no substantial site or assemblage of Qing domestic objects has ever been published in Hong Kong or nearby areas of mainland China.63 However, further research led us to overseas sources concerned with the archaeology of the Chinese diaspora to North American and Australia. As recently explored by Elizabeth Sinn, such communities mostly emigrated via Hong Kong during the nineteenth century, and were also then supported and provisioned through the port while living overseas.64 Many of our Sha Po objects had been shipped out to Chinese overseas communities and, to our surprise, these locally neglected artefacts were regarded as important research objects by archaeologists abroad. A study of inventory records from Kwong Tai Wo—a nineteenth-century Chinese store in California—showed that Wun Yiu ‘Double Happiness’ and ‘Bamboo’ pattern blue-and-white porcelain bowls were the cheapest types costing 2 to 5 cents each. While both patterns are common at Sha Po and in California, the emigrants seem to have also favoured celadon bowls and more expensive porcelains. Similarly, overseas Chinese demanded the top-end brand of opium e.g., ‘Lai Yuen’, while containers found at Sha Po are of a less famous brand.65 Several research papers and postgraduate theses have focused on these unglamorous daily objects, while a series of excavations have sought to investigate and understand life in early Chinese settlement sites.66 Moreover, a number of museums even include these objects in permanent exhibits dedicated to early Chinese immigrants. Thus things dismissed locally on account of their ‘ordinariness’ are studied overseas as valuable cultural markers of the Chinese diaspora. Historical archaeology is an area of research and writing in which Hong Kong, and indeed greater China, has rather lagged behind. Part of the problem is that many local scholars would regard materials dated later than 1800 to have little research value; indeed, the Antiquities and Monuments Ordinance— Hong Kong’s basic heritage legislation, which is itself something of an ancient relic—uses 1800 as the year before which a relic (artefact), site or structure made by people can legally be considered an

156   Piecing Together Sha Po

‘antiquity’. As we approach the twentieth anniversary of Hong Kong’s handover to China, perhaps we are also approaching a point of sufficient post-colonial remove to allow the materiality of people’s everyday lives in the early colony and the immediately preceding period to become an academically valid and politically acceptable subject for historical archaeological research. The archaeological resource is rich, the socio-historical context full and complex, and the potential for fascinating and alternative narratives is therefore correspondingly enormous.

Notes 1. The Coastal Evacuation was ordered in 1662 by the Qing emperor Kangxi and involved the forced depopulation of a vast swathe of the South China coast—probably including the entire area of modern Hong Kong—in order to deny support to Ming loyalists under Koxinga based in Taiwan. Perhaps as few as one-tenth of the estimated 16,000 people driven out returned to their land after the rescission; however, the government then actively encouraged immigration to repopulate the land. The origin of the Hakka migration into Hong Kong dates from this period. James Hayes, ‘The Hong Kong Region: Its Place in Traditional Chinese Historiography and Principal Events Since the Establishment of Hsin-An County in 1573’, JRASHKB 14 (1974): 118–19. 2. Walter Schofield, Austin Coates, James Hayes, and Patrick Hase all served as district officers in the British colonial administration: Schofield in 1922–38, Coates in 1949–56, Hayes in 1957–62, while Hase was a city and New Territories district officer from the 1970s to 1980s. Their unpublished notes, memoranda, and publications, in particular those in the latter category produced by Hayes and Hase, are a rich source of information in the English language on many aspects of the traditional way of life in Hong Kong’s New Territories, Lamma Island included. Without their interest and efforts, much of the detail of the old rural way of life would have been lost unrecorded and we therefore owe them a significant debt of gratitude. 3. Hase, ‘Yung Shue Wan’, 45. 4. Patrick H. Hase, Custom, Land and Livelihood in Rural South China (Hong Kong: Hong Kong University Press, 2013). 5. Ibid., 31. 6. Ibid., 32. 7. Ibid., 54–55. 8. Ibid., 206–14. 9. Hase, ‘Yung Shue Wan’, 15. 10. Hase, Custom, 207–9. 11. Ibid., 55. 12. Hase, ‘Yung Shue Wan’, 11. 13. Ibid.,12. 14. Ibid., 17. 15. Ibid., 206–14. 16. Ibid., 10. 17. Yamen: the official office and residence of Chinese bureaucrats and mandarins in Imperial China. 18. Hase, ‘Yung Shue Wan’, 25. 19. Hayes, ‘The Hong Kong Region’, 114. 20. Anthony K. K. Siu, ‘A Study of the Ch’ing Forts on Lantau Island’, JRASHKB 19 (1979): 195. 21. Hase, ‘Yung Shue Wan’, 18. In truth, pirate attacks continued well into the twentieth century, but then as sporadic, isolated incidents, rather than the large-scale piratical activities of old. 22. Ibid. 23. Hase, ‘Yung Shue Wan’, 14. 24. Atha, ‘A Neglected Heritage’, 140. 25. Ibid., 142–43. 26. Elizabeth Sinn, Pacific Crossing: California Gold, Chinese Migration, and the Making of Hong Kong (Hong Kong: Hong Kong University Press, 2013), 191. 27. AMO, ‘Preliminary Report on the 2004 Fat Tau Chau Archaeological Investigation’, Huaxia Kaogu 4 (2007c): 35–41, 55. 28. Hase, ‘Yung Shue Wan’, 19. 29. Mao and Rui, eds., Reprint of the late Qing [1862–74] Tongzhi Cui Wen Tang edition [original ed. Gui Wencan] (Yangzhou: Jiangsu Guangling guji keyinshe, 1993), 17. Relevant page on Yung Shue Wan is in Volume 13 under Xinan County chapter) at http://goo. gl/aYI20p. 30. Hase, ‘Yung Shue Wan’, 19. 31. Hase, ‘Yung Shue Wan’, 40. 32. Hase, ‘Yung Shue Wan’, 29. 33. Interview by James Hayes in 1966 of an 84-year-old villager. Cited in Hase, ‘Yung Shue Wan’, 46. 34. Hase, Custom, 213.



A Time of Great Change: Sha Po during the Ming–Colonial Era   157

35. Hase, Custom, 214. 36. Hase, ‘Yung Shue Wan’, 30–31. The cattle pens appear to be depicted on the 1905 survey map (see Maps 9 and 14). 37. Bau Malay is a type of South East Asian paddle stamped earthenware pottery named by Bill Solheim, which seems to have been a fairly common tradeware along the South China coast. 38. AAL, ‘Sha Po Tsuen Rescue Excavation’, 36. 39. Hase and Lee, ‘Sheung Wo Hang’, 85–86. 40. Tony Robinson and Mick Aston, Archaeology is Rubbish (London: Channel 4 Books, 2002). 41. Pamela R. Rogers, ‘Tung Lung Fort’ (unpublished excavation report, 1985), Fig. 35. Nine fragmentary examples were found at Tung Lung Fort. Such ‘clay stoves’ were also unearthed in an overseas Chinese settlement found in the Union Passenger Terminal Site in Los Angeles, USA. Roberta S. Greenwood, ‘Old Approaches and New Directions: Implications for Future Research’, in Hidden Heritage: Historical Archaeology of the Overseas Chinese, ed. Priscilla Wegars (Amityville: Baywood Publishing, 1993), 393–94. 42. The question of identifying the sources of so-called village wares, which occur in huge numbers on most Qing sites and are routinely considered difficult to date closely, is one that needs addressing in Hong Kong and South China archaeology. Certainly, the very glossy, translucent brown glaze of much Qing to mid-twentieth-century village ware is quite different to Ming dynasty and earlier types. At Sha Po, they can be dated to the nineteenth to first half of the twentieth century based on their associations with more closely datable ceramics. 43. The pale body and use of well-painted, recognisable designs of much of this material suggests a date before 1890, after which the Wun Yiu kaolin reserves were becoming worked out. As a result, the clay became increasingly red and the porcelain of poorer and poorer quality, which could only be sold very cheaply. Expense on decoration was therefore kept to the minimum and designs became ever more simple and abstract, eventually giving way to stamping as a cheaper alternative (Pat Hase, personal communication based on oral testimony of the last surviving potter in interviews conducted by James Hayes in the 1960s). 44. See Catalogue of Selected Finds No. 24. 45. Rogers, ‘Tung Lung Fort’, Figs. 6 and 7; Au et al., Wun Yiu, 69. 46. Comparanda found on a Canadian website (http://www.chinesecol.com/treasure4.html), which includes very similar objects, and refers to others in the Asian American Comparative Collection and in displays at the Barkerville Chinese Museum, British Columbia. 47. Information mainly based on an interview dated 22 May 2012 with the present elderly manager (Mr. Y. K. Ching 程貽楷) of the business based in Sheung Wan, Hong Kong. The winery specialised in producing traditional Chinese wines with brands such ‘Ng Ka Py’ and ‘Mui Kwe Lu’. Their wines have been exported overseas since 1925. According to the winery, they started using the name ‘Tianjin Hong Kong Wing Lee Wai’ following their takeover of Wing Fung Yuk in 1946. A bottle of ‘Ng Ka Py’ or ‘Mui Kwe Lu’ in 1946 would cost about HK$ 3.5 (Y. K. Ching, personal communication). ‘半邊雞,一壺永利威’ (‘half side of a chicken and a bottle of Wing Lee Wai’ was a well-known advertising slogan of Wing Lee Wai Winery post–World War II): Apple Daily (6 October 2012) (酒香情4:百年五加皮 不認輸). 48. ‘A North American Pioneer Chinese Virtual Museum’, at http://www.chinesecol.com/treasure10.html (accessed 14 August 2010). 49. See Catalogue of Selected Finds No. 38. 50. AAL, ‘Sha Po Tsuen Rescue Excavation’, 23. A number of similar gaming counters were recovered from Qing deposits excavated at Tung Lung Fort. See Rogers, ‘Tung Lung Fort’. 51. Enquiries with HK Electric revealed that their oldest live account on Lamma Island dated back to 1963 (Eric Chau, personal communication), but whether this date marked the beginning of mains electricity supplies to domestic homes there is not clear. A date after World War II seems likely though. 52. A 26.5 cm tall pearly green oil lamp-stand with a top dish decorated with three raised dots, a ridged body, and a flat-based bottom dish is shown in the Taiwan National Museum of History, where it is identified as a product of the Shiwan kilns in Guangdong. Very similar types of glossy brown glazed lamp-stands were unearthed during excavation of eighteenth- to early nineteenth-century deposits at Tung Lung Fort, where complete examples stood 17.8 cm tall and were wheel-made in two pieces and joined below the flange. Pamela R. Rogers, ‘Tung Lung Fort’ (1985), Fig. 34. 53. Priscilla Wegars, ed., ‘Asian American Comparative Collection: Artifact Illustrations’, at http://webpages.uidaho.edu/aacc/illus.htm (accessed 14 August 2010). 54. Ko Tim-keung, Hong Kong Now and Then (Hong Kong: Joint Publishing, 2005), 112. 55. Peter Lee, Opium Culture: The Art and Ritual of Chinese Tradition (Rochester: Park Street Press, 2006), 187–96. 56. Solomon M. Bard, ‘Report on Archaeological Survey at Tai Fu Tai, Phase II’ (unpublished report, 2001), 16. 57. Ibid., 16. Two such opium containers unearthed from Tai Fu Tai, San Tin were sent for residue analysis and opium traces were detected in one of them. J. K. Yang and V. R. Hellmann, ‘What’s in the Pot? An Emic Study of Chinese Brown Glazed Stoneware’, Proceedings for the Society for California Archaeology 11 (1998): 59–66. 58. ‘Gold Mountain’: the nineteenth-century Chinese name for California. 59. Hase, ‘Yung Shue Wan’, 35. 60. Such net-weights are low-fired and reddish in colour—suggesting that they were probably baked in an open fire—and the majority weigh about 15 g each. In use, they would have been strung along the bottom part of fishing nets, thus ensuring that they sank evenly in order to trap any fish in their path. Although ideally suited for inshore net-fishing from small boats, it is unclear whether such weights were also needed in the kind of shore- or sampan-based stake-net fishing seemingly common in Qing to early twentiethcentury Lamma. James Hayes, ‘Stakenet and Fishing Canoe: Hong Kong and Adjacent Islands in the 19th and 20th Centuries’, Proceedings of the Eight International Symposium on Asian Studies (1986): 573–98; Hase, ‘Yung Shue Wan’, 21.

158   Piecing Together Sha Po 61. Spry, ‘Sha Po Tsuen’, 9. 62. Interview with Lamma fisherman from Yung Shue Wan conducted in 2008. 63. Rogers, ‘Tung Lung Fort’; Au et al., Wun Yiu. 64. Sinn, Pacific Crossing. 65. Ruth A. Sando and David L. Felton, ‘Inventory Records of Ceramics and Opium from a Nineteenth Century Chinese Store in California’, in Hidden Heritage: Historical Archaeology of the Overseas Chinese, ed. Priscilla Wegars (Amityville: Baywood Publishing, 1993), 160–63. 66. Archaeological excavations at Chinese settlement sites revealed that ‘Double Happiness bowls occur in high frequencies on sites occupied prior to about 1870, but are scarce in later assemblages; Double Happiness appears to have been superseded by Bamboo style bowls after about 1870’. Ruth A. Sando and David L. Felton, ‘Inventory Records of Ceramics and Opium’, 160, at http://www.cas.umt.edu/anthro/anth495cim/documents/Merritt_Dissertation_Complete.pdf; http://www.cas.umt.edu/ anthro/anth495cim/ArtifactID.htm; http://www.uiweb.uidaho.edu/aacc/index.htm; http://www.scahome.org/publications/ proceedings/Proceedings.11Yang.pdf; http://www.wolfcreekarcheology.com/Stoneware.htm.

Part III Exploring the Multi-period Social Landscape

8 Reconstructing Sha Po’s Landscapes and Lifeways

Introduction We began in Part I by introducing our Sha Po study area in terms of its archaeological highlights and their contribution to the site’s local and regional significance. We outlined the history of archaeological discovery and then introduced the idea that a multi-period social landscape could be reconstructed from the myriad of material traces left by Sha Po’s succession of human communities. Our chronological discussion of Sha Po’s human story then unfolded in Part II’s four chapters, which examined the changing nature of archaeological remains through time as a means of illuminating many aspects of the ancient lifeways of past communities and their changing interactions with the local environment. The specific socio-historical conditions governing such interactions through time resulted in archaeological ‘signatures’ that reflect not only socio-political organisation and economic activity, but in some cases also ideological-religious beliefs as well. Our task here in Part III is to attempt to create a more holistic, overall understanding of those past lifeways by reconstructing a succession of social landscapes shaped, inhabited, and experienced by ancient communities at Sha Po. Such reconstructions, it must be acknowledged, reflect periods of human activity of unknown continuity and perhaps spanning several centuries; however, that is the nature of the evidence, in particular in prehistory. Despite those limitations, there are nonetheless clear patterns within the data in some periods, and undeniable contrasts between periods as well. First, though, we can use the shifting patterns of archaeological evidence to help reconstruct how the ancient physical environment changed as a result of the interplay of natural processes and cultural agency.

Reconstructing Sha Po’s Ancient Landforms Introduction While the character and patterning of archaeological remains in different periods at Sha Po revealed many specific details of activities and social organisation through time, those same data can also reveal how landforms, land use, and, by association, vegetation cover, were also changing through time. Map 10, which uses the 1981 1:1000 map as its basis, shows a reconstruction of the changing coastal landforms of the Sha Po study area in the Late Neolithic–Bronze Age, Six Dynasties–Yuan, and late Qing periods.

The shifting shape of the backbeach landscape Based on the patterning of Neolithic remains, the first people to make use of the Yung Shue Wan area encountered a bay that looked quite different from its modern form. The shoreline was perhaps as much

Map 10:  Composite map showing changing coastal landform through time. Source: CLSO. Yung Shue Wan, Lamma Island. 1:1000 Scale Topographic Map Sheets 14-NE-10D and 14-NE-15B. Hong Kong: CLSO, 1981a and 1981b respectively. Reproduced with permission of the Director of Lands. © The Government of the Hong Kong SAR. Licence No. 59/2015.



Reconstructing Sha Po’s Landscapes and Lifeways   163

as 75 m further east than its modern position, whereas the western face of the Sha Po plateau—before it was cut back in modern times—extended somewhat further to the west. As a result the plateau, in particular in prehistory, was a far more prominent feature in the coastal landscape, thereby creating a deeper, more enclosed sandy bay to the south and a smaller beach to the north. However, our understanding of the ancient landform at the mouth of the Yung Shue Long valley is much poorer than that to the south due to a lack of archaeological work, although it seems unlikely that a backbeach ever existed there. Given the dearth of reliable data, it seemed sensible to not attempt a reconstruction of the pre-Qing coastline in that area. The plateau has no evidence for human activity before the Bronze Age and, like most of inland Hong Kong in the Neolithic, was probably covered with broad-leaved, evergreen forest. The Neolithic backbeach was a curving feature probably no more than 20–30 m wide east–west, with a shallow freshwater lagoon behind, which drained into the sea through the southern end of the sand barrier. Before Qing–modern drainage, the very low-lying area (2.4 mPD) between the Tin Hau Temple’s sand spit and the southern backbeach would have flooded during spring tides, but our understanding of the dynamic environment in that area is at present poor. As can be observed in Hong Kong today, dramatic summer monsoonal rainstorms can carve deep floodwater channels through (back)beach formations, only to be backfilled with interest by typhoon-related gales and waves. In the case of severe rainstorms and major typhoons, these transformations can occur over the space of a few hours and the effects can be very dramatic indeed. Those same cyclical processes were undoubtedly at work throughout Sha Po’s human occupation, and for periods following powerful typhoons the backbeach would have completely ‘dammed-off ’ the valley— raising the lagoon’s water level—until floodwaters during the next major rainstorm washed through the backbeach and remodelled its outline. Based on the recorded position of its sloping seaward face observed in excavations at its southern end, there was perhaps a slight westward expansion of the backbeach between the Later Neolithic and end of the Bronze Age. But the greatest change in the Bronze Age landscape involved the clearance of forest from at least the south facing flank—perhaps the entire upper surface—of the New Village plateau to create a new area for settlement and craft-working. The mapped extent of tell-tale industrial residues indicates that by the Six Dynasties–Tang period the backbeach had expanded westward to almost double its Bronze Age extent, which was also reflected in the patterning of Northern Song to Southern Song–Yuan material across more or less the same area. Interestingly, excavations on the plateau also recorded the presence of thin scatters of Tang–Song pottery, which one can interpret as an indication that the area was then open—probably as a result of industrial fuel gathering—and perhaps in use for some form of cultivation. Whether the plateau remained continuously in use from later prehistory through to the Song dynasty is, however, impossible to say. Not much is known about the Ming dynasty landscape, although the bay was no doubt still being occasionally visited by naval and merchant vessels, while a more regular presence of boat-dwelling fishing families might be anticipated. However, the fact that Lamma Island was in the Ming government’s eyes considered wasteland available for imperial grant to absentee landlords suggests that inland areas were then largely uninhabited. Returning to the geomorphology, by the end of the Ming, the backbeach seems to have expanded to the line of Yung Shue Wan Back Street while, by the later Qing, a footpath pre-dating Main Street overlooked the shore, but from a position safely beyond the reach of the highest spring tides. Given the survival of much of the Qing landscape into the modern era, its features and overall form are reserved for discussion of the later historical social landscape in the final section of this chapter.

164   Piecing Together Sha Po

Reconstructing the changing backbeach profile The Sha Po backbeach’s rather piecemeal, discontinuous pattern of archaeological investigation—mostly in areas previously subjected to modern disturbance—makes the task of reconstructing the succession of pre-modern buried landforms extremely challenging. We nevertheless felt that an attempt should be made to reconstruct—by ‘patching together’ the fragments of stratigraphy available—the backbeach profile from the Middle Neolithic to modern era, and the result is shown in Figure 32. The profile shows a north-west–south-east transverse section across the south-central backbeach (location indicated on central panel of Map 10), which stretches from the edge of the former lagoon in the south-east, across the top of the backbeach, and then down the seaward face to the approximate western edge of the Six Dynasties–Yuan backbeach to the north-west. Westward of that there has been too much modification since the late Qing to permit identification of that part of the profile. We have shown above that the shape of the Sha Po landscape—with the backbeach and plateau at its heart—has changed significantly over the millennia of its human occupancy. Next, we focus on the changing character and patterning of human activity in order to reconstruct how natural processes and cultural agency interacted through time to create a succession of contrasting social landscapes at Sha Po.

Top probably flattened to build Sha Po Old Village in c.1800 CE

mPD 5

5

Ming–Qing 4

4

Six Dynasties–Yuan

NW 3

Middle–Later Neolithic and Bronze Age (NB: top perhaps closer to 5mPD before early historical remodelling)

Coast L.Qing–E.C20th

Song–Yuan Sui–Tang Western edge of Six Dynasties–Tang backbeach: earlier historical deposits with a few disturbed prehistoric artefacts

BA

S–Y

H–T LN

Mostly Sterile (Occasional LN) West central backbeach: core of early historical industry with kilns cut into Han to Tang (H–T) middens and Later Neolithic (LN) and Bronze Age (BA) deposits, kilns backfilled with Song–Yuan (S–Y) material

3 Lagoon

5mPD

S–Y L.Qing–E.C20th

SE

4mPD

L.Qing–E.C20th 6 Dynasties–Yuan LN–BA

L.Qing–E.C20th

LN (some MN) 3mPD

M. Sterile (OLN) East central backbeach: core of prehistorical activity with domestic, craft, industrial, and funerary activity, disturbed by early historical industry and later historical to modern development and farming

Qing 6D–T Eastern edge of Tang to Qing backbeach: sloping Six Dynasties to Tang deposits with Later Neolithic beneath, Qing rubbish dumps and revetments, late Qing to WWII fill and development

Figure 32:  60  m wide east–west transect across backbeach: showing changing profile through time, with four sections showing simplified sequence of archaeological deposits



Reconstructing Sha Po’s Landscapes and Lifeways   165

Interpreting the Shifting Patterns of Human Activity Introduction In our plots of different zones constituting the changing social landscape through time we have used different colours to highlight activity areas in each period, a series of standardised symbols for human burials, bronze casting, and kiln structures, while all broader areas are circled and labelled in accordance with the activity occurring, namely food preparation and consumption, rubbish disposal (middens), stone tool/weapon workshop, stone ornament workshop, housing structures, and intense spreads of kiln debris perhaps indicative of kilns.

A kin-based Middle–Late Neolithic social landscape? The small, localised assemblage of Middle Neolithic pottery connected with cooking and eating, the use of general purpose stone tools and presence of adze rough-outs together suggest a quite generalised, probably domestic and subsistence related use of the central-eastern backbeach by one or more small, perhaps kin-based, groups of mobile fisher-hunter foragers, who probably made brief but repeated visits to the backbeach when in the area to exploit seasonally available resources, while primarily basing themselves in the next bay south at Tai Wan. The north-west Lamma community’s characteristic use of red painted pottery and other distinctive materials suggests that they were part of a wider cultural-ideological tradition evidenced by people throughout the Hong Kong–Pearl River Delta region and in adjacent areas of the mainland coast. However, it is difficult to say whether some or all of the groups sharing those material traditions did so because they also considered themselves part of some larger socio-political entity or ‘regional polity’. The general lack of obvious markers of social differentiation—for example, prestige goods, an interest in competitive display, and higher-status burials—actually suggests a generally egalitarian character to the region’s Middle Neolithic society. The geographical extent of the cultural tradition perhaps offers some clues regarding the importance of boat-based coastal mobility to the local way of life. As Map 11 shows, the known patterning of Later Neolithic activity at Sha Po occupies a 20–30 m wide curving strip down the eastern, landward side of the backbeach. This pattern—and that of Middle Neolithic and Bronze Age remains—supports the idea that the earliest build-up of storm-blown sand occurred when the prehistoric shoreline was perhaps up to 75 m further east than its current position. The western limit of the prehistoric backbeach—in other words, the ancient shoreline—was further suggested by discoveries of water-rolled Neolithic sherds at 2.6 mPD in two small trenches in 1988.1 Domestic activity relating to food preparation and storage was evidenced by coarse cooking pots and utilitarian pebble tools across much of the Later Neolithic backbeach. Generally poor organic preservation meant that no true middens were recognisable, but pig (probably wild boar), large deer, and sea turtle were clearly hunted—as was also reflected in finds of two spearheads and an arrowhead dotted across the backbeach—while fish and some shellfish were also present. A particular focus of cooking activity was suggested by concentrations of pot-stand and fire-grate fragments in the centre-east of the domestic zone [22a]. Interestingly, that potential cooking zone also overlapped with the northern end of an area within which grey schist-slate was being worked to create a variety of tools and ornaments, as was indicated by rough-outs of knives and a large spade-like adze, an arm-ring, ring-core, and waste flakes, as well as a number of whetstones used in their shaping. Some quartz earring manufacture is also suggested on the backbeach, but again as one of a myriad of low-intensity craft activities carried out while at Sha Po, rather than the more specialised, larger-scale workshop activity of the Bronze Age discussed below.

166   Piecing Together Sha Po

Map 11:  Middle to Late Neolithic social landscape. Source: CLSO. Yung Shue Wan, Lamma Island. 1:1000 Scale Topographic Map Sheets 14-NE-10D and 14-NE-15B. Hong Kong: CLSO, 1981a and 1981b respectively. Reproduced with permission of the Director of Lands. © The Government of the Hong Kong SAR. Licence No. 59/2015.

The twelve complete adzes and associated whetstones found across the Later Neolithic backbeach show the potential importance of wood and bamboo for the procurement of raw materials and manufacture of all manner of items essential to survival, for example, boats and paddles, houses and shelters, spear and arrow shafts, and a myriad of other domestic artefacts in daily use. Adzes would also be essential for harvesting carbohydrate-rich starch from palms, which research at Xincun, Taishan suggests was important to Neolithic diet at the coast. By the Later Neolithic period, people at Sha Po were also using spindle whorls to spin plant fibres such as hemp to make yarn and cordage that might have been used in woven fabrics, fishing line and nets, boat anchor ropes, and carrying handles for pots. The Later Neolithic community left us a mosaic of different subsistence-related and craft-working activities across the backbeach, which seems to reflect a genuinely undifferentiated, unspecialised social use of space. One could reasonably interpret that pattern in terms of activities occurring during many short-term visits by relatively small-scale communities of fisher-hunter-foragers, who made repeated use of the Sha Po backbeach as one of several regular stop-offs within their coastal ‘home range’. It is possible



Reconstructing Sha Po’s Landscapes and Lifeways   167

also that the pattern actually reflects several distinct phases of use, each with their own activity foci, but the millennia of natural and cultural post-depositional processes have rendered them unintelligible to archaeologists.

A Bronze Age landscape of increasing social complexity Social landscapes are places created by particular communities that can reflect many aspects of their daily life. This is particularly evident in Bronze Age Sha Po, where the community’s lifeway and associated use of the physical environment can be mapped in terms of a number of discrete zones relating to dwelling and stone ornament manufacture, food processing and consumption, rubbish disposal, stone toolweapon polishing and maintenance, bronze casting, and mortuary activity. While on the backbeach, the overall extent of human activity in this period broadly matches that of the Later Neolithic, the range and zonation of activities are rather different. There is also a suggestion of

Map 12:  Bronze Age social landscape. Source: CLSO. Yung Shue Wan, Lamma Island. 1:1000 Scale Topographic Map Sheets 14-NE-10D and 14-NE-15B. Hong Kong: CLSO, 1981a and 1981b respectively. Reproduced with permission of the Director of Lands. © The Government of the Hong Kong SAR. Licence No. 59/2015.

168   Piecing Together Sha Po

some westward expansion of the southern backbeach, presumably through natural, typhoon-related sand accumulation, into which cultural activities then expanded. The rich Bronze Age deposits at the southwestern foot of the New Village plateau are identical in composition to those found on the plateau itself and were almost certainly redeposited downslope through natural erosion and Qing dynasty agriculture. In contrast, the Bronze Age clearance and use of the New Village plateau for the construction of a ‘settlement’, which was associated with the manufacture of finely polished earrings, tells us that the local community was part of a society undergoing significant socio-political change. The three main excavated areas on the plateau present a consistent picture of a landscape cleared of trees around the middle of the second millennium BCE to permit the construction of a number of postbuilt, probably stilt-house type, shelters. A small oval structure was tentatively identified in one of two areas without intercutting or ‘double’ post-holes, which perhaps indicates a low intensity of reuse and rebuilding of individual structures, and therefore either a single phase of use, or a migration of the ‘settlement’ across the plateau through time. In contrast, the third clustering of post-holes, slightly higher up on the plateau, did include several overlapping, multiple examples that indicate refurbishment and/ or rebuilding on the same spot. Either way, it is clear from the volumes of white and clear quartz raw materials, ring rough-outs, semi-finished and completed earrings, and associated saws and whetstones that the plateau was home to a group of skilled stone ornament manufacturers, whose products were perhaps consumed far beyond their local community. There is also widespread evidence on the plateau for domestic activity relating to the preparation and consumption of food and drink. Why that location was so favoured is open to question, but proximity to suitable raw materials may have been a significant factor. Narrow veins of white quartz are a widely occurring feature in granite on Lamma, so the Yung Shue Wan area was perhaps particularly noted as a raw material source.2 On the central and southern backbeach the pottery and stone artefact assemblages suggest a striking degree of continuity in the use of that zone for the preparation and consumption of food and the finishing, repairing, and resharpening of tools and weapons such as adzes, spearheads, and arrowheads. As in the Later Neolithic, the latter craft-working activities seem to have taken place at Sha Po as part of the daily routine of subsistence-related activities, rather than within a more segregated workshop environment. This makes sense in that general stone workshops—i.e., those focused on shaping all manner of utilitarian tools from raw materials—create a great deal of mess and waste material, whereas the finer work of finishing, sharpening, and polishing tools and weapons was undoubtedly one of a myriad of ‘maintenance’ tasks carried out within a group social context. The more obvious presence of consumption wares—mostly stem cups—on the plateau and backbeach may simply reflect the harder, more robust nature of higher-fired Bronze Age pottery, but could be an indication also of the longer-term occupation of Sha Po by rather larger groups during this period. The above circumstantial evidence for food acquisition and consumption is more categorically evidenced by a localised midden spread on the southern backbeach containing shellfish and the bones of fish, dolphin, deer, crocodile, green turtle, and even porcupine. The archaeological remains tell us much about the social organisation of activities within the Sha Po sites, but the remains of food animals and the tools, weapons, and equipment used in their capture open up far bigger socio-economic landscapes. They remind us of the community’s rich and diverse environment teeming with resources from forests, coastlines, and the open sea. Arguably, the most dramatic and archaeologically important feature of the Bronze Age backbeach is Hong Kong’s single convincing case of bronze casting as evidenced by splashes of casting slag associated with sandstone bivalve moulds and possible crucible fragments. The location of industrial activity near the southern tip of the backbeach, which would have kept the noxious fumes and fire risk well away from

Figure 33:  Artist’s impression of Bronze Age landscape (Dina. B. Knight)

170   Piecing Together Sha Po

the only identifiable dwelling area on the plateau, is perhaps significant, although there is good evidence for domestic consumption and rubbish disposal nearby. The chronological resolution of our data is too coarse to identify subphases of activity within prehistoric periods—especially in the loose, sandy sediments of the backbeach—so it could be that the domestic and midden activity never occurred at precisely the same time and in the same place as the bronze casting. It is impossible to tell for sure. A similar question mark hangs over the two potential burials and one more definite example noted on the central backbeach,3 all of which are in an area also rich in craft and domestic activities. One must question whether people would bury their dead in an area where they frequently gathered to cook and consume food and carry out daily maintenance activities. So while everything discussed is notionally Bronze Age in date, we would suggest that the burials are probably either later or earlier than the other activities identified nearby. Fundamentally, though, if all three are indeed burials, then they are located in the same quite compact area of the central backbeach, which in 1972 produced the complete double-F pot that first raised the possibility of Bronze Age burials in the area. Finally, with regard to the canoes used by the mobile, seafaring-coastal population, in the end we decided to depict canoes as single dugouts without outriggers or sails, which is a format we would envisage in use for general coastal, island hopping, and estuarine-riverine travel. But double canoes, some probably with sails, may also have been used.

The earlier historical socio-economic landscape The discovery of Six Dynasties–Tang kiln debris in every excavation so far completed on the Sha Po backbeach, added to our knowledge of more extensively excavated kiln sites, suggests that the clusters of particularly intense kiln debris spreads at Sha Po probably mark the approximate positions of perhaps as many as eight further kilns to go with the seven industrial structures so far identified. The known and probable locations of kiln structures are shown respectively by circles and darker green shading in Map 13, while further kilns possibly await discovery in the untested areas in between. If that is true, then it would mark out Sha Po as one of the more important industrial complexes then operating in Hong Kong. The discovery of locally unique artefacts associated with government officials and literacy supports the idea that the industrial activity at Sha Po—and by association across Hong Kong—was managed and policed as part of the Imperial Salt Monopoly, then thought to have been administered from Nantou. With respect to non-industrial remains, the scatters of Six Dynasties–Tang domestic pottery—storage jars, bowls, and cups—are quite thin and mirror the patterns at most other contemporary industrial sites, with one notable exception.4 The Han–Six Dynasties midden remains on the southern backbeach appear to pre-date or perhaps overlap with the earliest periods of kiln use, so may at least in part reflect patterns of rubbish disposal by kiln workers. As discussed in Chapter 6, besides the unique San Tau cemetery,5 Six Dynasties–Tang burials are generally rare in Hong Kong, so the two burials at Sha Po—Graves 2 and 3—are particularly notable in that they also contained preserved human remains, in both cases of adult females. They are most interesting, though, for their almost identical orientations, which can be interpreted as evidence for shared beliefs concerning the appropriate placement of the dead in the landscape, in other words, geomancy or fung shui as we now know it. Fascinatingly, that orientation appears to be mirrored by Bronze Age Grave 1, which could indicate long-term use of the Sha Po area by maritime groups with similar ideas regarding mortuary placement. Although the scientific dating suggests that all but one of the kilns (K4) probably ceased operation by the end of the Tang dynasty, clear continuity of backbeach use is suggested by the patterning



Reconstructing Sha Po’s Landscapes and Lifeways   171

of Song–Yuan storage and consumption wares in post-abandonment spreads. So was there also some continuity of industrial production beyond the Tang dynasty—as was hinted at by the TL dates from K4—but it was so small-scale as to be almost unidentifiable archaeologically? Interestingly, thin scatters of Tang–Song material on the plateau seem to indicate some use—perhaps cultivation—of that area during the operation and post-abandonment phase of the industrial complex. The archaeological evidence creates a general impression of a busy industrial landscape with multiple— perhaps household—teams working away at their kilns, smoke and fumes filling the air, the latter being particularly the case when quicklime was being slaked to prepare or repair the large woven boiling ‘pans’. The pans themselves might have been made by specialists elsewhere or maybe were woven on site when required. The primary product’s raw material—concentrated brine—was probably at this time extracted using the leaching rather than solar evaporation method, but archaeological evidence for this crucial aspect of the process still eludes us.6 Fuel was also an essential resource and the black ashy layer found in the base of Kiln K4 hints at the use of brushwood and maybe grass from local hillsides. We can only

Map 13:  Six Dynasties–Tang social landscape. Source: CLSO. Yung Shue Wan, Lamma Island. 1:1000 Scale Topographic Map Sheets 14-NE-10D and 14-NE-15B. Hong Kong: CLSO, 1981a and 1981b respectively. Reproduced with permission of the Director of Lands. © The Government of the Hong Kong SAR. Licence No. 59/2015.

Figure 34:  Artist’s impression of Six Dynasties–Tang landscape (Dina. B. Knight)



Reconstructing Sha Po’s Landscapes and Lifeways   173

speculate whether some form of scrub or woodland management was practised; but if not, the exhaustion of fuel supplies may have been a contributory factor leading to the industry’s eventual collapse. Boats were obviously pivotal to the industry’s operation, but given the apparent persistence of non-Han Chinese culture at the coast—i.e., Yue or Yao people—and probable operation of the industry by the local population, it seems more likely that such vessels were something similar to a Philippines’ barangay, rather than an early form of Chinese sampan or junk.7 The boats would have been needed to bring in coral and shells to make lime and to carry away shipments of salt, and perhaps also lime. However, the guan brick, nature of the Sha Po kiln complex, archaeological suggestions of a potential naval military presence in the region,8 and other historical clues all suggest that more junk-like Chinese military vessels carrying imperial soldiers and administrators may also have patrolled the area. In our artist’s reconstruction drawing, we show only barangay-type vessels. Despite quite extensive excavation on some sites (e.g., Sham Wan Tsuen), no evidence for settlement associated with any Six Dynasties–Tang industrial site has ever been found, which therefore raises the possibility that the kilns were operated by a boat-based or even coastal stilt-house dwelling population.

The late Qing to mid-twentieth-century farming-fishing landscape With the arrival of rice farmers in the earlier Qing dynasty, the physical environment began to undergo significant change. In particular in the valleys, villages were built and streams were diverted and sluiced for the management of wet rice agriculture, while all accessible slopes—including the Sha Po plateau— were eventually terraced to grow vegetables. Some consideration of the landscape setting, or fung shui, was probably necessary in the placement and orientation of the first villages, around which shrines, ancient ‘spirit’ trees and fung shui groves served to enhance the auspiciousness of the locale. Then around the beginning of the nineteenth century, Sha Po Old Village was itself built on the central backbeach as a single row of houses, beside which a second parallel block was later added. The nineteenth-century expansion of Yung Shue Wan’s coastal market reflects the economic impact of Hong Kong’s rapid development, which generated a vibrant trade through Yung Shue Wan to Aberdeen in livestock—cattle, pigs, and chickens—fish, vegetables, charcoal, and fuel-grass.9 Cattle raised in south Lamma were driven to Sha Po and penned there awaiting shipment, and such pens were recorded in the Block Crown Lease Survey map of 1905 (see Map 14 below). The 1905 map also shows that the former freshwater lagoon had by then been drained and was in use as paddy fields, indeed all available land was either under crop or otherwise put to economic use. Although their respective economic interests were fundamentally land and maritime-focused, the social landscapes of farmers and Hoklo fisher-folk substantially overlapped as a result of their joint crewing of inshore fishing boats, their use of Yung Shue Wan’s market as an arena for intergroup economic exchange, and their shared religious focus on the Tin Hau temple. The temple also served an important geomantic role in the social landscapes of both communities, in that it overlooked and protected the moorings and breaming beaches used by the fishing community, and provided a spiritual defence against the negative fung shui of Sha Po Old Village’s rather exposed position. Perhaps the most significant consequence of the traditional farming way of life was its underlying sustainability, which was grounded in the intergenerational stewardship of the socio-economic landscape. That involved the practical matters of ongoing maintenance of houses and other buildings, rice-drying grounds, field systems, and water-management features. But more importantly, it was mediated through the tight-knit social structures of rice-farming communities, which were bound together by a deep respect for natural and supernatural forces, and a reverence for the beliefs, wisdom, and achievements of

174   Piecing Together Sha Po

Map 14:  Detail of 1905 map centred on Sha Po study area with cattle pens circled. Source: CLSO. Lamma Island, D.D.3 Sheet 2. 1:1980 Scale Map. Hong Kong: CLSO. Reproduced with permission of the Director of Lands. © The Government of the Hong Kong SAR. Licence No. 59/2015.

village elders and ancestral figures. That is why, despite Hong Kong’s complete transformation in the same period, the essential features of the Qing and earlier twentieth-century social landscape recorded in the 1905 map survived largely intact until the late 1960s at Sha Po.10 Our artist’s impression of the late Qing landscape seems a very appropriate place to round off our discussion of the 6,500-year story of humanenvironment interaction at Sha Po. In the last four decades the hinterlands of Yung Shue Wan have witnessed their most dramatic period of change, during which the backbeach, former paddy fields, and hillside terraces have all succumbed to modern development. Paradoxically, these most recent impacts of human agency—in effect, the

Figure 35:  Artist’s impression of late Qing to early twentieth-century landscape (Dina. B. Knight)

176   Piecing Together Sha Po

formation processes of the modern social landscape—also caused the discovery of all its earlier manifestations discussed above. In our final chapter we reflect on our research and draw some conclusions concerning our use of a social landscape approach, the local importance of Sha Po, and its wider significance within Hong Kong and Chinese archaeology.

Notes 1. Spry, ‘Sha Po Tsuen’, 12–13. 2. This comment is based on general field observations of quartz veins visible in local granite outcrops. But the small scale (1:25,000) used for most Hong Kong geological maps means that occurrences of quartz are usually too small to be mapped. Interestingly, the authors’ recent research on a small headland site at Pak Kok Tsui—the northern tip of Lamma Island and a half-hour walk from Sha Po—produced significant quantities of white quartz, which may have been quarried in the area. 3. As discussed in Chapter 5, the one convincing burial (Grave 1) seems to be firmly Bronze Age in date, which is also the case with the possible burial containing a hard pottery jar and stem cup, whereas the softer ceramics in the third perhaps suggest an Early Bronze Age date. 4. The pottery assemblage at Sham Wan Tsuen is exceptionally rich given the excavator’s suggested brief, but presumably very intense, period of Tang dynasty activity. Meacham, Chek Lap Kok Island, 197. 5. Atha, ‘San Tau, North Lantau’. 6. Hase, ‘Salt’, 66. 7. Thanks to Stephen Davies for advice on possible forms of prehistoric and early historical boats of the South China Coast. 8. Atha, ‘San Tau, North Lantau’, 215. 9. Hase, ‘Yung Shue Wan’, 30. 10. Compare the 1905 map shown in Maps 9 and 14 with the 1968 situation recorded in Map 6. That degree of continuity was also evidenced in many other areas of the New Territories over that same period.

9 Conclusions and Reflections on Sha Po

The Regional Significance of Sha Po Our study has drawn together the results of over eight decades of archaeological work at Sha Po to reveal its unique and fascinating human story. The overall coherence of that story has, we feel, been enhanced by our use of a social landscape perspective, which encouraged the development of strong narrative threads transcending periodic boundaries. In overview, Sha Po is typical of many coastal sites on islands in Hong Kong in that it evidences the activities of Middle and Late Neolithic and Bronze Age fisher-hunter-foragers, has traces of Eastern Han activity and widespread remains of Six Dynasties–Tang coastal industry with Northern Song to Southern Song–Yuan material in post-abandonment deposits. There is also a dearth of Ming dynasty remains and contrastingly huge quantities of Qing to early modern material. But the devil, as they say, is in the detail, and some material details mark out Sha Po as an area of great archaeological significance, both within and beyond Hong Kong. In the Neolithic, the backbeach is one of many sites on islands evidencing what appears to be an intermittent use by small groups targeting particular resources. In contrast, Sha Po is the only site in Hong Kong with adjacent backbeach and plateau areas apparently in contemporary use during the Bronze Age. The backbeach has the territory’s only compelling evidence for in situ bronze melting and casting, while the plateau has unique settlement remains associated with the operation of a specialised quartz earring workshop. The use of headland sites began in the Late Neolithic and may be an indication of a growing concern with coastal surveillance and defence relating to increasing competition between rival groups. In the context of Bronze Age Hong Kong, the use of the plateau may have similar connotations, while the presence of craft specialisation at Sha Po—focused on metallurgy and stone ornament manufacture— must be viewed in the context of a regional society enjoying greater access to trade and exchange networks extending far inland and perhaps, in the case of exotic bronze weapons and jades found nearby at Tai Wan, right across Lingnan. When viewed together, these changes signpost the emergence in Bronze Age Hong Kong of an increasingly complex, hierarchical, and competitive society, which had a significant focus on Lamma Island. They also indicate that some form of economic intensification and probable redistributive system had developed through which community members not involved in day-to-day subsistence (i.e., craft specialists and political leaders) could be supported. Sha Po’s two Six Dynasties burials are also regionally important as very few have been found and fewer still with preserved skeletal material. Their use of a shared orientation, which appears to mirror that of a nearby Bronze Age burial, perhaps provides significant insights into early use of geomancy and beliefs surrounding human placement of the dead, and by association the settlements of the living, in the landscape.

178   Piecing Together Sha Po

The discovery of four Six Dynasties–Tang industrial structures during our 2008–10 excavations and watching briefs on the Sha Po backbeach increased the overall number identified to seven. Based on our further assessment of intense kiln debris spreads during this research, we would presently suggest that perhaps at least another eight kilns await detection in future fieldwork. That prospect makes Sha Po potentially one of the largest and most important kiln complexes in Hong Kong. Furthermore, as we argued in Chapter 6, our combined reanalysis of the kiln structures, kiln furniture, and types of lime residues suggests to us that the industry was most likely focused on the production of salt using limecoated basins. We attempted to address the crucial issue of determining how many kilns might have been in contemporary use—and therefore the scale of the industry—by thermoluminescence dating kiln remains associated with the seven structures and other particularly intense deposits of kiln-related materials. The results provided a clear indication that the kiln industry began operating in the Six Dynasties–Sui period and continued into the mid-late Tang (three kilns), latest Tang–Nanhan (two kilns), or in one case perhaps somewhat later. Multiple kilns overlap in date and the use of the term ‘complex’ therefore seems appropriate. The scale of operations and inferred connection with salt production together suggest a connection with the Imperial Salt Monopoly. Significant extra weight is added to that argument by the guan (‘official’) brick, lattice pattern brick, and the Tang green glazed ink palette, which were all discovered in association with the northernmost kiln. Together, the objects suggest imperial authority, higher social status, and, of course, literacy, and seem to be a further indication of the importance of the Sha Po complex at that time. Overall, then, our study of early historical Sha Po provides Hong Kong’s clearest evidence yet that the widespread kiln industry was to some degree planned and probably imperially controlled. Although modest when compared with the recent major discoveries at To Kwa Wan and Tai Kong Po, the remains left by Sha Po’s Northern Song–Yuan inhabitants nevertheless serve to emphasise the contrasting character and trajectories of communities on the outlying islands, with those near the important military-administrative centre in Kowloon Bay, and in rice-farming villages in the northern and western New Territories. Fundamentally, landscape archaeology is inclusive and embraces material evidence from all periods of human endeavour. We therefore felt it was extremely important to ensure that the later historical archaeology, which at Sha Po really means that from the Qing dynasty to World War II, was integrated with the historic landscape and documentary evidence to create a more nuanced and humanistic exploration of what is locally a much-neglected period (in archaeological writing at least). Indeed, the archaeology of the Qing–World War II period, in particular that of farming and fishing communities, remains something that for many archaeologists in Hong Kong and South China seems to be too ‘everyday’, ‘recent’, or ‘commonplace’ to warrant serious study. Many archaeological reports therefore offer scant discussion of any material from the period and quickly move on to discuss earlier ‘more interesting’ periods. However, the Qing–World War II period will one day be ancient history and we would argue that it is about time archaeologists started taking it a bit more seriously. So while we accept that Qing–World War II Sha Po was a fairly unremarkable place, the period actually provided by far the richest and most diverse data resource, which consequently yielded some fascinating insights into daily life and was probably, in truth, the most enjoyable part to write.

The Bigger Picture For seafaring prehistoric people, the sea was the centre of a world dotted with islands and fringed by coastal sites and estuaries. Indeed, for five-and-a-half of the six millennia covered by our study, the



Conclusions and Reflections on Sha Po   179

patterning of human activity is clearly focused around the coastline and islands, and boats must have been of singular economic and socio-political importance to local communities. The coastal channels, bays, estuaries, and major rivers provided access to much of their food and served to interconnect communities throughout the region. By the Later Neolithic and, in particular Bronze Age, the coastal waters and major rivers were also conduits for trade and exchange as evidenced in the latter period by the presence across the archipelago of occasional exotic jades and bronzes and the ubiquitous hard pottery, which at present appears to have been manufactured some 200 km away by water up the East River in inland Guangdong. Based on our assessment of Sha Po’s kiln complex and its likely role as one part of a region-wide imperially controlled salt-lime industry, which seemingly made use of every available backbeach in the territory, we can imagine a quite spectacular coastal vista with smoke rising from innumerable kilns. There would be lots of activity between the kilns and shoreline, where boats were beached or tied up, while others were plying to and fro with cargoes of coral, salt, or lime. War junk patrols and any larger merchant vessels en route to or from Guangzhou would no doubt complete the scene. Even at the height of Hong Kong’s post-war manufacturing boom, there were never such widespread or prominent signs of industrialisation around the territory’s coastline, and certainly not all focused on the same products. The Qing–World War II historic landscape of Sha Po and its environs reflects a pattern repeated across rice-farming areas throughout Hong Kong, but is also unique in the details of its arrangement, which reflect a combination of local socio-political factors, topographic constraints, and broader geomantic concerns governing placement of human elements in the landscape. This local-regional characteristic of Hong Kong’s historic agricultural landscapes makes them particularly interesting in that the detailed study of one place generates a situated understanding of its particular social landscape, while at the same time allowing broader insights into drivers of change, whose effects might be observable throughout the region.

Looking Forward The importance of ‘thinking landscape’ This book would have looked very different if our study had been designed and implemented as a multiperiod and interdisciplinary landscape study, but none of the investigations at Sha Po were carried out within such a research framework. However, our attempts to understand and explain the dizzying variety of human activity evidenced during the six millennia of Sha Po’s inhabitation were nonetheless bolstered by the use of a holistic approach and consistent discursive framework—the social landscape—within which all periods and remains were afforded equal attention. Of course, the levels of interpretive detail and richness of resulting narratives varied depending upon the quality and resolution of the data available for each period. But our synthetic reanalysis has nonetheless provided a window on Sha Po’s many different communities and their contrasting lifeways and interactions with the local landscape over a period of around 6,500 years. That in itself is, arguably, an important step forward, which will hopefully encourage others to reconsider Hong Kong’s past from more holistic perspectives. If that occurs then it could only be positive for the future management of archaeological heritage in Hong Kong.

Predicting future discoveries One of the undoubted highlights of Sha Po’s archaeological story is the evidence for its Six Dynasties– Tang industry. Looking back, it seems surprising to note that the many separate excavations carried out

180   Piecing Together Sha Po

between 1972 and 2000 only identified one kiln (K1), whereas two projects focusing on footpaths have more recently found five kilns and a unique ‘working floor’. What this confirms is that when trying to identify discrete structures such as kilns—rather than archaeological deposits of much wider extent— widely spaced small test pits are relatively ineffective, whereas long thin trenches (transects) are a costeffective alternative to large box-grid or open-area excavations. Although housing or utilities are likely to determine the positions of future investigations at Sha Po, the large almost entirely untested area on the northern backbeach would benefit from an initial geophysical survey, should the opportunity for investigation there arise. Although archaeological geophysics remains in its infancy in Hong Kong, the authors’ recent research at San Tau—in collaboration with Dr. Wallace Lai and his team at the Hong Kong Polytechnic University—has clearly demonstrated the effectiveness of ground penetrating radar (GPR) in backbeach conditions, most interestingly including the identification of buried kiln debris. In Chapter 8, we predicted the locations of as many as eight further kilns, but the untested northern area may contain yet more. Based on the one small trench so far excavated in that northern area—which has remained undeveloped to this day—there appeared to be stratified Bronze Age and Neolithic deposits, while deposits of those periods will almost certainly exist down the entire eastern side of the backbeach. Turning to the plateau, the results of the three larger investigations so far conducted there suggest that remains of the Bronze Age stilt-house settlement and associated ornament workshop will also be identified in areas between and around the periphery of those tested so far. Further fascinating insights into late Qing domestic life are to be anticipated in rubbish dumps located to the rear of Yung Shue Wan Back Street and, in particular, east of Sha Po Old Village. Staying with rubbish, it might also be prudent at Sha Po—and similar backbeach sites with former lagoons—to check for potential waterlogged deposits at the interface between the backbeach and lagoon, which may have been an area used for rubbish dumping in prehistory and earlier historical times. Generally speaking, in Hong Kong archaeological work is recommended for developments affecting rich cultural deposits on backbeach sites, but adjoining areas at the interface with low-lying former wetlands tend to be dismissed as archaeologically unpromising, which may be a mistake. But we will not know for sure unless somebody actually checks.

Addressing more general gaps in knowledge Our attempts to clarify the character and duration of prehistoric occupation at Sha Po, as at many other Hong Kong prehistoric sites, are inevitably influenced by the underlying idea that communities pursued diversified subsistence strategies, which presumably involved a fair degree of—some would argue seasonal—mobility. But little has yet been done to test the latter notion, for example using seasonality indicator species such as shellfish. If further shell-rich midden deposits are encountered in future, for example on Sha Po’s southern backbeach, we would recommend that analysis of shellfish growth rings be carried out to see whether a consistent season of collection is indicated. Where food remains survived at Sha Po, they comprised shellfish and bones of fish, land and marine mammals, and marine reptiles, but no trace of plant foods was even sought, never mind found. The research at Xincun in Taishan, Guangdong provides a powerful lesson regarding the value of prehistoric stone tools not just as potential indicators of food processing tasks, but also as ‘containers’ for plant starch and phytolith remains. The extension of such research into Hong Kong archaeology is strongly advocated and could revolutionise our understanding of prehistoric subsistence in the territory. It may



Conclusions and Reflections on Sha Po   181

even be worth expanding the scope of testing to include adzes to see if it is possible to determine which plants or trees were being harvested and worked using the tools. The analysis of food residues—typically fatty acids or starch grains—inside storage and cooking pots is another area of research that again has been neglected in Hong Kong, but has yielded fascinating insights into ancient cuisine and diet in other parts of the world. The artefacts can certainly tell us much about Sha Po’s ancient way of life, but would it not be fabulous, for example, to learn what foods they were cooking and eating, but also whether the same foodstuffs were being left as offerings in mortuary jars? By the Bronze Age, we see evidence of expanding trade and exchange networks involving exotic bronze weapons and mass-produced hard pottery, but there has so far been no comprehensive scientific research into raw material provenance—sources or quarries—or patterns of workshop production, distribution, and consumption of the stone tools and ornaments that typify local prehistory. Based on the results of such studies elsewhere, if we can establish how different categories of raw material and stone artefacts were being moved around the landscape, then that should produce insights into trade and exchange networks. That in turn might allow a deeper understanding of developments in social complexity, political relations, and interconnectedness within prehistoric communities inhabiting the Hong Kong–Pearl River Delta region. Returning to our landscape approach and the holism that implies, backbeach sites like Sha Po are by far the commonest and most intensively investigated category of prehistoric site in the Hong Kong region, but woefully few have seen anything like the same intensity of sampling applied to their adjoining clay-rich lower hill slope areas. There is therefore a need for a synthetic review of the evidence for prehistoric patterns of dwelling in different geomorphological coastal settings, in particular around backbeach sites with clay-rich slopes behind and backbeach sites with lagoons fringed by similar slopes at their inland edge. In several instances where lower slopes have been investigated, a rich and contrasting range of activities to those evidenced on the backbeach has been identified. This therefore begs the question: Was there really a sudden interest in settlement in Bronze Age Sha Po, or were Middle and Late Neolithic groups also living in the area, but exploiting different lower hillside locations around the inland edge of the lagoon? Ultimately, of course, our aim should be to interrogate and understand entire multi-period social landscapes spanning a range of topographic zones, including uplands, a goal which will be significantly more achievable if we fully exploit the amazing results of the 2010 government-commissioned territory-wide LIDAR survey of Hong Kong. The ground-based and aerial remote sensing technologies and data are available, the multi-period social landscape awaits, and we just need the imagination and vision (and funding) to exploit their potential.

Final Words There are clearly many factors to be considered when attempting to judge how representative our archaeological discoveries are of the full range of human activities occurring at any given site in the various periods contributing to its human story. Many gaps in our Sha Po story certainly remain to be filled, while future discoveries may require significant revision of some chapters. However, we hope our landscape perspective has allowed you—the reader—to recognise that people at all times in the past interacted with, experienced, shaped, and were in turn shaped by their environment, which they perceived not as a static backdrop to their daily lives, but rather as a dynamic landscape full of social meaning. For each of the several hundred generations that have inhabited Sha Po, their social landscape connected them with

182   Piecing Together Sha Po

their ancestors, framed and gave meaning to their community’s present, and signposted their way forward into the future. Ultimately, we hope that this book has provided enough archaeological detail to satisfy students and scholars of the subject, while also offering non-specialists an interesting and entertaining journey of discovery as we have painstakingly pieced together the six-millennium-long human story of ancient Sha Po.

Appendices

Appendix 1: Catalogue of Selected Finds

Introduction Most of the objects presented below have never before appeared in publications, and the collection includes a number of genuine rarities—some indeed unique in Hong Kong—which warranted more detailed discussion than could be offered in the main text of the book. The artefacts are presented in groups by material type—bone, non-vessel fired clay, metal, pottery, and stone. The catalogue entries are fairly standardised in format, although some items needed only a quite brief entry whereas others required more in-depth discussion. In each case, details of comparanda and full references are provided in the footnotes. For each item, we provide the site code/archive numbers as allocated during the original fieldwork, as well as our own site investigation numbers in square brackets as used throughout the text (e.g., for Item 8, which was found in 2008 in trench AA5C-MH in context 510, the entry reads thus: ‘2008, AA5C-MH, Co. 510 [34:510]’). Please refer to Map 5 for details of our numbering sequence of site investigation areas and see also Chapter 2 for fuller details of the various excavation and watching brief areas within which the catalogue items were discovered.

The Catalogue Bone artefacts Personal ornaments (1) Fish vertebrae beads

Non-vessel fired clay artefacts (2) Lattice ‘tomb’ brick (3) Guan ‘tomb’ brick

Metal artefacts Iron tools (4) Socketed iron axe-head (5) Two iron hoes (cha)

186   Appendix 1

Personal ornaments (6) Burial assemblage, Grave 2: (a) silver hairpin, and (b) finger-rings Copper alloy coins (7) Northern Song coins

Pottery artefacts (8) Coarse corded cooking pot (9) Pot-stand (10) Possible burial assemblage: (a) shallow bowl and (b) globular jar (11) Possible burial assemblage: (a) hard pottery storage jar and (b) stem cup (12) Hard pottery lid (13) Hard pottery jar (14) Four-lugged storage jar (15) Green glazed bowl from Grave 3 (16) Green glazed lid (17) Green glazed lotus bowl (18) Green glazed bowl (19) Green glazed cup (20) Green glazed stem cup with flower decorations (21) Green glazed ink palette (22) Green glazed bowl (23) Wine jar sherd with stamp (24) Blue-and-white spouted wine/water jar

Stone artefacts (25) Rotary drill or ring-polishing stone (26) Bivalve fish-hook mould (half ) (27) Adze (28) Burial assemblage from Grave 1: four knife rough-outs (29) Chipped pebble tool (30) Whetstone (31) Polishing stone or mortar (32) Polishing stone (33) Ring ornaments and core (34) Adze rough-out (35) Stepped adze (36) Sceptre (zhang) (37) Soapstone net-weights (38) Ink-stone



Catalogue of Selected Finds   187

Bone Artefacts (1) Fish vertebrae beads At least fifteen carved and pierced fish vertebrae were discovered in earlier historical middens on the Sha Po backbeach, comprising four of sea-bass and eleven small shark.1 Carving is visible around the periphery, in particular on the four sea-bass beads, and all had central perforations. As will be noted below, although they had a Han–Six Dynasties context of discovery, such objects are commonly considered to be prehistoric in date. It is not unusual to find carved and perforated fish vertebrae in prehistoric coastal sites in Hong Kong, such as Shek Pik and Sham Wan, or in the Pearl River Delta region at sites such as Xiangnancun, or even in Taiwan (e.g., Nanke, South Taiwan).2 In terms of function, they are usually interpreted as ornaments: based on the observation of two similar perforated fish vertebrate discovered lying near the wrist of a Bronze Age burial in Shek Pik, Meacham suggested that these ornaments could possibly be worn as a series in bracelets or necklaces, or even used individually as pendants.3 Similar artefacts found in prehistoric sites in Taiwan were identified as ‘bone beads’.4 Two similar perforated fish vertebrae found in Late Neolithic (or Early Bronze Age) burial C7 at Tung Wan Tsai North in 1997 were interpreted as possible personal ornaments or spindle whorls belonging to the deceased.5 Further examples excavated at Xiangnancun in Shenzhen were also interpreted as spindle whorls,6 but this use seems unlikely as they have insufficient mass for the purpose and fired clay or stone are better suited to the task. Interestingly, over one hundred worked fish vertebrae were also excavated in late Ming cultural layers in the northern New Territories, but it was unclear whether they were of that date or possibly associated with disturbed burials.7 Dimensions: diameter c. 2.0–3.0 cm, thickness 1.5–2.2 cm AA5E, Co. 535 and 537 [34:535, 537] Han–Six Dynasties (possibly disturbed prehistoric)

1. HKAS 1994 archive; AAL, ‘Sha Po Tsuen Rescue Excavation’, 2003; AAL, ‘Rescue Excavations at Sha Po Tsuen’, 2011; ERM, ‘Rescue Excavation at Small House Lot No. 1575s.B’, 2005. 2. Schofield, Shek Pik, 58; Meacham, Sham Wan, 229; Ye, Yang et al., ‘Excavation of Xiangnancun Site in Nanshan, Shenzhen’, Kaogu 6 (1997): 85, T8:189; Tsang Cheng-Hwa, Li Kuang-Ti, and Zhu Zhengyi, Archaeological Discoveries in Nanke (Tainan: Tainan Cultural Council, 2006), 96. 3. Meacham, Sham Wan, 229. 4. Tsang Cheng-Hwa, Li Kuang-Ti, and Zhu Zhengyi, Nanke, 96. 5. Chau, ‘North Tungwantsai’, 6, 13, and figure 9. 6. Ye Yang et al., ‘Xiangnancun’, 85. 7. Mo Zhi, ‘North District Sewerage, Stage 2 and Stage 3: Archaeological Investigation’, unpublished report, 2001.

188   Appendix 1

Plate 44:  Carved and perforated fish vertebrae beads: (a) 7 small shark, (b) 4 sea bass

Non-vessel Fired Clay Artefacts (2) Lattice ‘tomb’ brick This low-fired creamy orange coloured brick was discovered in post-abandonment deposits associated with a Six Dynasties–Tang kiln (our reference K1) at the northern end of the backbeach. The brick is rectangular in plan with one broken edge and has impressed lattice patterns on one face. As shown in the rubbing, the surface was stepped into three different levels, the lower two being impressed with lattice patterns. Although this was a rare find in Hong Kong, a number of similar bricks with impressed lattice motifs found in Guangdong were dated to between the Han and Southern Dynasties. For example, the Southern Dynasties lattice tomb bricks from Meihuacun and Dongshan in Guangdong,8 the M68 reddish-yellow tomb bricks from Zhongxing Primary School site in Guangzhou,9 the early Southern Dynasties reddish

8. Chen Chunli, ‘Preliminary Report on the Excavation of a Southern Dynasties Tomb at Meihuacun, Dongshan, Guangdong’, in Guangzhou Wenwu Kaogu Ji, ed. Institute of Cultural Relics and Archaeology of Guangzhou (Beijing: Wenwu Publishing), 189, 192. 9. Yi Xibing and Ma Jingguo, ‘Preliminary Report on the Excavation of Southern Dynasties Tombs at Zhongxing Primary School, Taojin Road’, in Archaeological Discoveries and Research in Guangzhou Vol. I, ed. Institute of Cultural Relics and Archaeology of Guangzhou (Beijing: Wenwu Publishing), 135.



Catalogue of Selected Finds   189

tomb bricks from Baoan, Shenzhen,10 and the mid-late Eastern Han tomb brick from Honghuayuan Han tomb, Shenzhen.11 These types of lattice-patterned tomb bricks were also excavated in nearby provinces: for example, the Western Jin (296 CE) tomb bricks from Fujian with impressed lattice on both sides,12 and the Eastern Han tomb brick from Hunan.13 Despite the differences in date and origins, the above mentioned bricks were all identified as tomb bricks and with dimensions 34–41 cm (length), 13–19 cm (width), and 4–7 cm (thickness). The Sha Po brick was found in a Tang–Song post-abandonment spread, but based on the above comparanda should itself date to the Southern Dynasties at the latest. Here, the Baoan brick is interesting as it was found with a ceramic ink palette very similar to the one backfilled in Kiln K1 (see No. 26 below). Based on the evidence of Sha Po thermoluminescence and pottery dating, we would associate the brick with the earliest phase of kiln-based industrial activity on the backbeach, which occurred during the Six Dynasties period. Dimensions: length (surviving) 13 cm, width 15 cm, thickness 4 cm 2000, T1L3 [16:T1L3] Six Dynasties

Figure 36:  Rubbing of ‘tomb’ brick with impressed lattice decoration

10. Wen Benheng and Rong Daxian, ‘Preliminary Report on the Excavation of a Southern Dynasties Tomb in Baoan, Shenzhen’, Wenwu 11 (1990): 11. 11. Yang Hao and Yang Yaolin, ‘Preliminary Report on the Excavation of Honghuayuan Han Tomb in Nantou, Shenzhen’, Wenwu 11 (1990): 32. 12. Lin Zhonggan and Chen Ziwen, ‘Preliminary Report on the Excavation of a Jin Tomb in Luchuwu, Pucheng, Fujian’, Kaogu 10 (1988): 922, 924. 13. Wang Shizhen and Wang Shancai, ‘Report on the Excavation of an Eastern Han Tomb in Dongchengqu, Suizhou, Hubei’, Wenwu 7 (1993): 56.

190   Appendix 1

Plate 45:  Lattice ‘tomb’ brick

(3) Guan ‘tomb’ brick This dark grey brick was discovered in 2000 among Six Dynasties–Tang materials in pre-kiln deposits near Kiln K1. The brick is rectangular in plan and curves slightly along its length. A raised motif in mirror image of the character ‘官’ (guan, meaning literally ‘government official’) is impressed or moulded on one long edge. There are other examples of bricks with mirror image inscriptions, most from tombs, hence the commonly used label ‘tomb brick’. Jin dynasty examples include tomb bricks from He M1 in Qujiang (items 7 and 8) and Shixing (item 11). Southern Dynasties examples include He M3 and Nanbuasi M16 in Qujiang (items 30 and 31 respectively) and Long M10, Lianxian and Sanjiang in Liannan (items 32 and 53 respectively).14 An obvious question, then, is: What does the character guan mean when used on bricks in this way? First, it should be noted that guan motifs are also found on a range of artefacts including tiles, bricks, 14. Lam, Archaeological Finds from Jin, 65, 67.



Catalogue of Selected Finds   191

pottery, lacquerware, and metal products dating from as early as the Three Kingdoms period.15 The meaning of guan or xinguan characters inscribed on Late Tang to early Northern Song white glazed stoneware has been widely debated but no consensus has yet been reached. However, in the case of Tang– Song tiles and bricks with guan motifs, previous research revealed that guan motifs mostly referred to the name of the official craftsperson (e.g., ‘官匠X X’ from Luoyang Ducheng), the indication of official tiles [on buildings] (e.g., ‘官瓦’ and ‘官作’ from Nanzhao and Dali), or production marks from governmentassociated workshops (e.g., ‘官窰’ from Zhenjiang).16 Stylistically speaking, the guan character impressed on the Sha Po brick is very different from those used on Han dynasty artefacts, such as the tiles found in the Southern Yue Palace site in Guangzhou.17 On the other hand, its combination of a top stroke sloping down to the right, elongated vertical stroke to the left, and round cornered rectangular mouth radicles is rather similar to guan characters used on tiles from Nanzhao sites (738–937 CE) in Yunnan Province, such as ‘官買’, ‘官尚方’ from Dali Jinsuo Island, ‘官’ from Dali City West, and ‘官’, ‘保官’ from Yaoan Zhugesi.18 It is believed that such Nanzhao tiles were produced by official workshops or were used in state-owned buildings.19 Further cases of similar guan inscriptions on tiles were found in the Tang Dynasty Huaqing Palace site in Lintong, Xian, Shaanxi Province, such as ‘六官泉南’, ‘北六官泉’, and ‘天六官瓦’.20 Later examples include a Southern Han (907–960 CE) ‘官’ inscribed celadon plate from Guangzhou21 and tiles inscribed ‘北務楊達官’ and ‘內 西賈□官’ from a Northern Song (960–1127 CE) palace site in Luoyang.22 However, it should be noted that different contemporary craftspeople may have produced different styles of the same character, as shown by the range of guan characters found on Sui–Tang tiles within the same site in Luoyang.23 So how should we interpret this locally unique artefact? As we have noted, guan bricks are sometimes found in tombs but the character seems most likely to refer to an official workshop, artisan, or building. The most interesting and important feature of the brick, therefore, is its connection with ‘official’ (i.e., imperial business). In the context of early historical Sha Po that is arguably most likely to relate to the kiln industry, which we have suggested was probably part of the historically attested Imperial Salt Monopoly then operating in Hong Kong. If that is true, then it is quite remarkable that, beyond the stamped bricks used in the construction of the Lei Cheng Uk Han tomb, no other bricks of either guan or lattice type (No. 3 above), which were both found in close proximity at Sha Po, have been noted elsewhere in Hong Kong. Finally, in terms of date, the guan brick was found in deposits containing mid-Tang pottery dating perhaps as late as the eighth century CE, but it could, of course, be somewhat earlier and even overlap in date with the lattice brick. Dimensions: length 15.4 cm, width 10 cm, thickness 4 cm 2000, T1L5 [16:T1L5] Tang or somewhat earlier 15. Xie Mingliang, ‘A Few Queries on the Meanings of Guan and Xinguan Inscriptions on White Glazed Porcelain’, The National Palace Museum Research 5 (2) (1987): 3–4. 16. Xie Mingliang, ‘Guan and Xianguan Inscriptions’, 5. 17. Cultural Council of Guangzhou City, ed., Three Major Archaeological Findings of the Qin and Han Periods in Guangzhou (Guangzhou: Guangzhou Publishing, 1999), 169. 18. Zhang Zengqi, ‘Tile-ends with Characters of Nanzhao and Dali State’, Wenwu 7 (1986): 39, 40, 42 19. Zhang, ‘Nanzhao and Dali State’, 43. 20. Luo Xizhe, ‘Excavation of the Living Quarter Royal Musicians and Bathroom in the Huaqing Palace of the Tang Dynasty’, Wenwu 3 (1999): 36, 38. 21. Institute of Cultural Relics and Archaeology of Guangzhou, ed., Zhu Ji Cun Lei: A Selection of Excavated Finds from Quangzhou in the Last Decade (Beijing: Wenwu Publishing, 2005b), 193. 22. Chen Lianglei, ‘Preliminary Report on the Excavation of a Tang Song Site at Tanggong Road (North) in Luoyang’, Kaogu 12 (1999): 50–51. 23. Chen Jiuheng, ‘Investigation and Excavation on the Sui–Tang Dongdou City Site’, Kaogu 6 (1978): 363, 365.

Figure 37:  Rubbing of guan ‘tomb’ brick

Plate 46:  Guan ‘tomb’ brick



Catalogue of Selected Finds   193

Metal Artefacts (4) Socketed iron axe-head This socketed iron axe-head was discovered in a Han–Six Dynasties layer in 1994.24 According to Zhu, this is a typical iron axe form dating from the Warring States to Qin/Han period. A similar and better preserved example of a Warring States iron axe (T10:2) was found at the Dieshishan site in Shenzhen, Guangdong.25 Dimensions: length 9.25 cm, width 8.25 cm, thickness 3 cm 1994, SF30, C.16.122, Square Kg, Layer 2 [13:L2] Warring States to Han

Plate 47:  Socketed iron axe-head. © Antiquities and Monuments Office, reproduced with permission.

(5) Two iron hoes (cha) According to archaeological findings, it was during the Qin–Han dynasties that iron tools in a variety of forms became more commonly available in Guangdong.26 Two such iron tools were found on the Sha Po backbeach in 2000 (a) and 2005 (b) and both were identified as cha (hoes), which are usually connected with cultivation or related activities such as the construction of irrigation channels.27 The socketed form was designed to be mounted on long wooden handles, some of which are known to have had 24. Zhu Hairen, ‘Han Archaeology in Hong Kong’, 48. 25. Rong Daxian, ‘Preliminary Report on the Excavation of Dieshishan Site in Shenzhen’, in The Archaeological Discoveries and Research of Shenzhen, ed. Shenzhen Museum (Beijing: Wenwu Publishing, 1994), 44; Shenzhen Cultural Heritage Management Office, Shenzhen Museum, and Shenzhen Municipal Institute of Cultural Relics and Archaeology, eds., 7000 Years of Shenzhen: Finds Catalogue (Beijing: Wenwu Publishing, 2005), 84. 26. Yang Shiting, ‘A Few Questions on the Early Iron Tools in Guangdong’, Kaogu 2 (1977): 104. 27. Yang Kuan, Development History of China’s Ancient Smelting Technology (Shanghai: Shanghai People’s Publishing, 1982), 297; Dong Shouxian, ‘Research of Iron Tools of the Han Dynasty’ (MA diss., Zhengzhou University, China, 2010); Yang Cong, ‘Iron Agricultural Tools Excavated from a Han Site in Chengcun, Suian, Fujian’, Agricultural Archaeology 1 (1990): 264.

194   Appendix 1

paddle-shaped heads.28 Two types of iron cha were common: one was a ‘straight-lined cha’ and the other a ‘U-shaped cha’.29 Both of the Sha Po examples are U-shaped cha, although their simple rounded form is quite different from the common Han type, which has ‘wings’ projecting wider than the main blade, as seen in examples from Huaqiao, Guangzhou, and Shahuqiao, Changsha.30 The Sha Po cha are, however, somewhat more similar in shape to the Eastern Han cha (DP87C:03) discovered in Dongpingcun, Xichang, Sichuan,31 and the Type I (T314) Han cha from Suiancun, Fujian.32 Overall, though, there are no close comparanda in the archaeological data and the Sha Po iron cha seem quite unusual members of the Han family of iron tools. Although iron cha are commonly known as agricultural tools, the actual function and date of the Sha Po examples remains unclear. As observed by Li,33 when compared with the standardised Han forms, the size and styles of agricultural tools became more diversified during the Six Dynasties period. Regional traditions developed, not only in the Central Plain but also in frontier areas, to suit different local usage and land types. It is interesting to note that the two Sha Po cha were discovered near a kiln structure (K1)34 and near a kiln debris concentration.35 It is therefore possible that the tools at Sha Po were associated with the kiln industry. Whatever the case, it is very unusual to discover relatively complete, delicate iron implements such as socketed cha in Hong Kong, where our corrosive ground conditions often destroy iron objects. (a) Iron hoe (cha) A relatively complete—fragmented but refitted—cast iron socketed hoe found in a Sui–Tang layer earlier than or contemporary with the construction and use of Kiln K1. This cha has a rounded D-shape in plan with a square ‘cut-out’, lentoid cross-sectioned socket and blade on all sides but the top edge. Dimensions: length 13.5 cm, width 14 cm, thickness (over socket) 2.5 cm AMO 2000, T1L5:9 [16:T1L5] Han to Tang

(b) Iron hoe (cha) A broken fragment of a cast iron socketed hoe with serious corrosion, found in a disturbed layer along with Six Dynasties–Tang kiln debris, Song pottery, and Qing materials.36 A large enough stub of the socket wall survives to indicate that it is curving towards a central square ‘cut-out’, which in turn suggests that the fragment was probably part of a U-shaped cha very similar in form to that found in 2000 (see suggested dashed outline in Figure 39). Dimensions (surviving): length 6.5 cm, width 4.25 cm, thickness 0.35 cm ERM 2005, T1L1, SF46 [26:L1] Han to Tang 28. Yang, Smelting Technology, 133. 29. Bai Yunxiang, Research on the Archaeological Findings of Pre-Qin to Han Dynasty Iron Works (Beijing: Science Publishing, 2005), 190; Dong, ‘Iron Tools’. 30. Yang, ‘Iron Tools in Guangdong’, 100, No. 6, Fig. 5, No. 5, Fig. 3. 31. Liu Shixu and Zhang Zhengning, ‘Investigation of a Han Dynasty Bronze Smelting Site in Dongpingcun, Xichang, Sichuan’, Kaogu 12 (1990):1073–74; Dong, ‘Iron Tools’. 32. Yang, ‘Iron Agricultural Tools’, 265. 33. Li Jinghua, ‘Ancient Iron Agricultural Tools in Henan’, Agricultural Archaeology 1 (1985): 61. 34. Au, ‘Sha Po Tsuen’, 2001 (kiln). 35. ERM, ‘Rescue Excavation at Lot No. 1575s.B’, 15–16. 36. ERM, ‘Rescue Excavation at Lot No. 1575s.B’, 10.

Figure 38:  Iron hoe (cha) ‘A’

Figure 39:  Iron hoe (cha) ‘B’

Plate 48:  Iron hoes (cha): (a) from 2000 to 2001 [16], (b) from 2005 [27] (front and back views)



Catalogue of Selected Finds   197

(6) Burial assemblage from Grave 2: (a) silver hairpin and (b) finger-rings In 1989, a long U-shaped hairpin and pair of finger-rings were found with a well-preserved Jin dynasty female skeleton in what appeared to be a very shallow grave in the southern part of the backbeach.37 All three items were identified as silver alloy through preliminary scientific analysis,38 but were dull grey in colour when found due to oxidation upon re-exposure to the air. Unpublished site photographs showed that one of the near-identical pair of finger-rings was excavated near the left hand, while the other may have been disturbed but was found nearby. The plain U-shaped hairpin was excavated above the skull, where the hair would have originally been, and was therefore probably in situ. A similar combination of Eastern Jin burial objects comprising golden hairpins and wire-thin fingerrings was discovered in Tomb 9, Xiangshan, Nanjing.39 Another similar silver hairpin dated to Eastern Jin was found in Pingshigang, Zhaoqing, Guangdong,40 although the Sha Po piece does not have a hook at the end. A study of ancient Chinese hairpins indicated that the gap between the legs of U-shaped hairpins widened during the Jin, Southern, and Northern Dynasties, while the head became less rounded and increasingly square in shape.41 (a) Silver alloy hairpin

Dimensions: length 17.40 cm, width 2.4 cm, thickness 0.25 cm 1989, C.16.12, Square CCg, c. 70 cm from surface [11: Grave 2] Jin

(b) Two silver alloy finger-rings

Dimensions: diameter 1.65 and 1.68 cm, thickness 0.5–0.7 mm 1989, C.16.13, Square CCgx, c. 72 cm from surface [11: Grave 2] Jin

Figure 40:  Silver hairpin from Jin dynasty Grave 2 37. Meacham, ‘Sha Po Tsuen’, 40, 46. 38. HKAS archives (unpublished archives from the 1989 excavation, now stored at HKAS). 39. Jiang and Zhang, ‘Excavation of Tombs 8, 9 and 10 at Xiangshan ’, 10–12. 40. Qiu, ‘An Eastern Jin Tomb at Pingshigang’, 259–60, M1:15. 41. Zhou Xun and Gao Chunming, Chinese Women’s Ornaments through the Ages (Hong Kong: Publishing Ltd., and Shanghai: Shanghai Xuelin Publishing, 1988), 55.

198   Appendix 1

Plate 49:  Jin silver hairpin and finger-rings from Grave 2

(7) Northern Song coins (a) Xianping Yuanbao coin 咸平元寶 A Northern Song Xianping Yuanbao coin was found in a disturbed context at the southern edge of the backbeach.42 Such kai-scripted (楷) Xianping Yuanbao coins were minted during the Xianping reign (998–1003 CE) of Emperor Zhenzong (真宗 997–1022 CE) and should be read clockwise from the top.43 Records showing that Xianping Yuanbao were also found in Sheung Shui (1), Shek Pik (7), Mai Po (5),44 and Kellett Island (20).45 Dimensions: diameter: 2.5 cm, perforation 0.55 cm2 1988, Box F, Co. 23, F23 [7:23] Minting date: 1010–1063 CE (Northern Song)

(b) Mingdao Yuanbao coin 明道元寶 A kai (楷) calligraphy script Northern Song Mingdao Yuanbao coin (two joiners) was found in the backfill of Tang kiln K1 at the northern end of the backbeach. The coin was minted during the Mingdao reign (1032–1033 CE) of Emperor Renzong (仁宗 1022–1063 CE) and is a type of dui qian (對錢) or dui zi qian (對子錢) (literally, ‘matching money’). Northern Song dui qian refers to the short-lived practice of issuing coins bearing the same reign title inscription in pairs with two or more different calligraphy styles, such as kai (楷) and zhuan (篆) in the case of Mingdao Yuanbao. Apart from the styles of scripts, these coins have identical size, thickness, perforation details, and metal contents. The use of dui qian can be traced to as early as the Nantang Kindom (937–975 CE) of the Five Dynasties and Ten Kingdoms period, became fully developed in the Northern Song, and eventually ceased to be minted in 1190 ce in the Southern Song.46 The number of Mingdao Yuanbao coins discovered in Hong Kong is small when compared with other Northern Song coins. For example, only seven out of over two thousand Song coins found in a pot hoard on Kellett Island in 1991 were Mingdao Yuanbao, of which three were kai scripted and four were zhuan. 42. Spry, ‘Sha Po Tsuen’, 24. 43. Zhang Yan, ‘A Preliminary Study on Song Dynasty Currency and Its Culture’ (MA diss., Minzu University of China, 2006). 44. James R. Crawford, ‘Southern Sung Coin Hoards from Hong Kong’, JHKAS Vol. XI, 1984–85 (1986): 91. 45. AMO, ‘Hong Kong Archaeological Archive System: Kellett Island’, accessed 20 August 2012. 46. Zhang, ‘Song Dynasty Currency’.



Catalogue of Selected Finds   199

Another kai-scripted Mingdao Yuanbao coin was found in So Kwun Wat in 2000.47 Five other Mingdao Yuanbao coins discovered in Sheung Shui (2), Shek Pik (2), and Mai Po (1) were also recorded.48 Dimensions: diameter: 2.1 cm, perforation 0.8 cm × 0.7 cm 2000, Kiln Y1 backfill [16:K1] Minting date: 1032–1033 CE (Northern Song)

Figure 41:  Rubbing of Northern Song Mingdao Yuanbao coin

Plate 50:  Northern Song coins: (a) Xianpin Yuanbao, (b) Mingdao Yuanbao

47. AMO, ‘Hong Kong Archaeological Archives System: So Kwun Wat’, accessed 20 August 2012. 48. Crawford, ‘Sung Coin’, 91.

200   Appendix 1

Pottery Artefacts (8) Coarse corded cooking pot A complete side of a round-bottomed coarse corded pottery fu was unearthed in large articulated pieces in a Later Neolithic–Early Bronze Age stratum just south of the Old Village. The pot has a quite unusual flat-topped triangular rim with a shallow groove that may have served as a lid seating. Cord decoration was applied over the entire exterior below the rim and the widest diameter was at the middle (belly) of the pot. Much of the lower body of the pot is blackened, which may reflect its use for cooking over an open fire. The overall form is quite similar to a Shang dynasty round-bottomed coarse corded pot found in the Cuntou Site, Guangdong,49 although the latter is larger in size and does not have a lid seating. Dimensions: rim diameter 24 cm, belly diameter 25.5 cm, height 20.5 cm 2008, AA5C-MH, Co. 510 [34:510] Later Neolithic

Figure 42:  Later Neolithic coarse corded cooking pot

Plate 51:  Later Neolithic coarse corded cooking pot 49. Cultural Heritage Management Office of Shenzhen, Shenzhen Museum, and Shenzhen Municipal Institute of Cultural Relics and Archaeology, 7000 Years of Shenzhen, 63, No. 89.



Catalogue of Selected Finds   201

(9) Pot-stand A number of coarse pottery ‘pot-stand’ fragments were discovered in Sha Po, both on the backbeach and plateau. Most were plain but occasionally, like the example shown here, some were decorated with feint cord imprints. The example illustrated below is a good example of the round lipped and flaring rim form typical of Later Neolithic of pot-stands. Similar types of pot-stands include one found in Baojingwan, which was dated to the ‘later phase of Late Neolithic’.50 Dimensions: rim diameter 11.2 cm, height 13.5 cm 2008, AA2B, Co. 204 [32:204] Later Neolithic

Figure 43:  Later Neolithic pot-stand

Plate 52:  Later Neolithic pot-stand

50. Guangdong Provincial Institute of Cultural Relics and Archaeology and Zhuhai Museum, eds., Excavation Report on a Prehistoric Island Site in Baojingwan, Zhuhai (Beijing: Wenwu Publishing, 2004), 80–81, T11A:56.

202   Appendix 1

(10) Possible burial assemblage: (a) shallow bowl and (b) globular jar (a) Shallow bowl A plain shallow bowl made of fine clay was discovered in the south-eastern part of the backbeach in a disturbed Early Bronze Age context just above a small globular burial jar (Item 18) made of identical material. Although no evidence of a burial feature survived, the two objects type and close association strongly suggests that they were originally grave goods within an Early Bronze Age burial. The bowl’s pinched, slightly incurving rim, and rounded base are similar to other Early Bronze Age bowls found in Hong Kong, for example, a reddish clay bowl (SF75) and coarse bowl (SF94) from Tung Wan Tsai North,51 and a Bronze Age bowl from Hai Dei Wan.52 Dimensions: rim diameter 12 cm, height 3.8 cm 2008, AA2B, Co. 204 [32:204] Early Bronze Age

(b) Globular jar This small plain globular jar in yellowish-brown fine clay was found immediately below bowl 17. The jar is only c. 6.5–7 cm in height by c. 9 cm in width and has an S-shaped form with small pointed lip and a rounded body. The jar bears a strong resemblance to the small globular burial jar recorded in Shek Pik, Lantau Island by Schofield in 1937. According to Schofield, the small jar could be related to a Bronze Age burial with preserved skeleton.53 Dimensions: rim diameter 7 cm, height (surviving) 6.3 cm 2008, AA2B-MH, Co. 205 [32:205] Early Bronze Age

Figure 44:  Possible Early Bronze Age grave goods: (a) globular jar and (b) bowl

51. Chau Hing-wah, Wu Yaoli, and Li Langlin, ‘Tungwantsai North’, 12–13; AMO, ‘Archaeological Archives System: Tung Wan Tsai’, accessed 2 September 2013. 52. AMO, ‘Archaeological Archives System: Hai Dei Wan’, accessed 2 September 2013. 53. Schofield, Shek Pik, 40, 71 and S.P.75, Plate CXX Nos. 3 and 4.

Plate 53:  Possible Early Bronze Age grave goods: soft fine clay (a) bowl and (b) jar

204   Appendix 1

(11) Possible burial assemblage: (a) hard pottery storage jar and (b) stem cup These were found in 2002 on the central-eastern backbeach and tentatively identified as grave goods in a Bronze Age burial.54 The grave goods assemblage comprised the two pots and a large unworked rock oyster shell, associated with several fragile bone fragments. No cut feature could be defined, but the evidence collectively suggests the presence of a Bronze Age grave. Shells are often found as grave goods in prehistoric cemeteries along the South China coast and in island and mainland South East Asia.55 Shells were found locally in several graves at Tung Wan Tsai North56 and accompanying a child’s burial at the Yinzhou shell midden site in the Pearl River Delta.57 (a) Hard pottery jar Small hard pottery round-bottomed jar with sharply everted, flaring rim with rounded lip. Between the shoulder and belly, the vessel has combed dot decoration between a series of incised parallel cordons, while the rounded bottom has an impressed diamond lattice decoration. Several comparable small jars were found in graves at Henglingshan cemetery in Boluo County, Guangdong.58 Dimensions: rim diameter 14.7 cm, belly diameter 15.6 cm, height 12.5 cm 2002, Unit 6.1, Co. 02 [22a:02] Bronze Age

(b) Hard pottery stem cup Hard pottery stem cup with angular carinated form, flaring rim, and pointed lip. The vessel is mostly plain but there is a single row of combed dot decorations around the exterior lower body, while inside the hollow pedestal there is a ‘potter’s mark’ comprising two sets of five incised curving parallel lines. A smaller set of three incised lines may be a further intentional mark. A very similar stem cup was identified at Henglingshan.59 Dimensions: rim diameter 11.8 cm, base diameter 5.5 cm, height 6.5 cm 2002, Unit 6.1, Co. 02 [22a:02] Bronze Age

54. AAL, ‘Sha Po Tsuen’, 36. 55. Charles Higham, personal communication. 56. Chau, Wu, and Li, ‘Tungwantsai North’, 5–7. 57. Li Ziwen, ‘Preliminary Report on the Excavation of the Shell Midden Site at Yinzhou, Sanshui City, Guangdong’, Kaogu 6 (2000): 27. 58. Guangdong Provincial Institute of Cultural Relics and Archaeology, ed., Henglingshan in Boluo County: Excavation Report of Shang and Zhou Period Cemetery in 2000 (Beijing: Science Publishing, 2005), M068.1, M110.4, and M111.4 etc. 59. Guangdong Provincial Institute of Cultural Relics and Archaeology, Henglingshan, M272.2:3.

Figure 45:  Possible Bronze Age grave goods: (a) hard pottery jar and (b) stem cup

Plate 54:  Possible Bronze Age grave goods: hard pottery (a) jar and (b) stem cup. © Antiquities and Monuments Office, reproduced with permission.

206   Appendix 1

(12) Hard pottery lid Also found in Bronze Age deposits washed-down from the plateau, this hard pottery lid fragment again has the typical grey fabric with black inclusions. The form has vertical exterior walls that taper on the inside to create a triangular section and small pointed lip. The walls are plain but the slightly convex upper surface is decorated with a series of concentric inscribed lines over which a series of combed dots (in lines of six) were applied. Complete examples of similar lids excavated in Hong Kong and Guangdong show that they usually had a handle in the top-centre. For example, one from Hai Dei Wan has a cylindrical handle,60 while a glazed hard pottery lid from the Yingang site in Boluo, Guangdong has a small loop handle.61 Both were dated to the Bronze Age, with the latter more specifically being placed in the Warring States (Late Bronze Age) period (Yingang Phase II). Dimensions: diameter 16 cm, height 3.5 cm 2000, T2L5-21 [16:T2L5] Bronze Age

Figure 46:  Bronze Age hard pottery lid

Plate 55:  Bronze Age hard pottery lid 60. AMO, ‘Archaeological Archives System: Hai Dei Wan, No. 1968.001.00006’, accessed 4 April 2014. 61. Gu Yunquan, Li Ziwen, and Deng Hongwen, ‘Excavation of the Yingang Site at Boluo in Guangdong’, Wenwu 7 (1998): 26, II T0108:27.



Catalogue of Selected Finds   207

(13) Hard pottery jar Several large refitting sherds of a brownish-orange high-fired stoneware jar were unearthed at the southern end of the backbeach in 2009.62 The jar has an everted, dished rim and a rounded body, and throwing marks are visible on the inside and outside of the body. A similar jar was discovered in an early Eastern Han tomb in Lechang City, Duimianshan, Guangdong, while comparable forms were also noted in Han jars from Tung Wan Tsai South.63 Dimensions: rim diameter 28 cm, belly diameter 25 cm, height (surviving) 16 cm 2009, AA5F, Co. 536 [34:536] Eastern Han

Figure 47:  Han hard pottery jar

Plate 56:  Han hard pottery jar 62. AAL, ‘Rescue Excavations at Sha Po Tsuen’, 44. 63. Qiu Licheng and Gu Yunquan, ‘Eastern Zhou, Qin, and Han Tombs in Lechang City, Duimianshan, Guangdong’, Kaogu 6 (2000): 48, 60, M152:12; Rogers et al., ‘Tung Wan Tsai South’, Fig. 5.15.

208   Appendix 1

(14) Four-lugged storage jar A purplish-red slipped four-lugged storage jar with poorly defined shoulder, upright double-ridged rim, and a flat base. The four evenly spaced horizontal lugs were positioned using a line incised around the shoulder of the vessel. A similar purplish-brown slipped four-lugged storage jar was unearthed in a Jin dynasty tomb in Shixing, Guangdong.64 Another similar four-lugged bucket-shaped jar with general Eastern Han Guangdong pottery characteristic was discovered in a Three Kingdoms hoard site in Guangzhou.65 In terms of form and fabric, this Sha Po example resembled much to the purplish-red slipped fourlugged storage jar (M10:1) found in a Southern Dynasties tomb in Middle Kangle Road in Zhaoqing, Guangdong, although the Zhaoqing example is slightly more bucket-shaped along the shoulder line.66 Dimensions: rim diameter 16 cm, base diameter 18.5 cm, height 18.5 cm 2000, T1L5 [16:T1L5] Three Kingdoms to Southern Dynasties

Figure 48:  Six Dynasties four-lugged storage jar

Plate 57:  Three Kingdoms–Southern Dynasties four-lugged storage jar

64. Liao Jinxiong, ‘Preliminary Report on the Excavation of Jin, Southern Dynasties and Tang Tombs in Shixing, Guangdong’, Kaogu 2 (1990): 117, 119, Type I pottery jar (M6:4). 65. Feng and Zhang, ‘Three Kingdoms Hoarding Site’, 117, Type Ca pot J9:13, 119. 66. Shang Jie, ‘Middle Kangle Road Cemetery in Zhaoqing’, in Ancient Cemeteries in Zhaoqing District, ed. Guangdong Provincial Institute of Cultural Relics and Archaeology (Beijing: Science Publishing, 2008), 27–133, M10.



Catalogue of Selected Finds   209

(15) Green glazed bowl from Grave 3 This complete and well-preserved bowl was part of the grave goods assemblage67 of an early historical burial and had been placed by the skull of the deceased woman. The bowl has a low disc-shaped foot, upright walls tapering to a rounded lip—defined on the exterior by a single incised groove—and is decorated with a green crackle glaze everywhere except under the foot. The inside of the base has an incised circle within which there are three kiln stacking spurs. Comparanda were found in Guangdong Southern Dynasties tombs, for example, at Taojin Road in Guangzhou68 and others at Zeqiaoshan, Ruyuan, Guangdong.69 Dimensions: rim diameter 10.9 cm, height 4.5 cm 1997, TP2, Co. 07 [15:07] Southern Dynasties

Figure 49:  Southern Dynasties green glazed bowl from Grave 3

Plate 58:  Southern Dynasties green glazed bowl from Grave 3

67. Also present was a glass or crystal necklace, which was reported by the excavator to be in very poor condition, but this item was not available for study. 68. Yi and Ma, ‘Southern Dynasties Tombs’, 140, M69:17. 69. Guangdong Provincial Institute of Cultural Relics and Archaeology, ed., Burials of the Six-Dynasties and the Sui and Tang Dynasties at Zeqiaoshan, Ruyuan (Beijing: Wenwu Publishing, 2006), M51:11, M56:9, M57:13.

210   Appendix 1

(16) Green glazed lid This green crackle glazed lid from a wine or water jar was discovered in 2009 at the southern end of the backbeach.70 The lid is heavily potted and has a central lug handle within an inscribed circle. Although the lid mimics Han forms, the glaze is more characteristic of Southern Dynasties to Sui date. Dimensions: diameter 16 cm, height 4.5 cm 2008, AA5G, Co. 541 [34:541] Southern Dynasties to Sui

Figure 50:  Sui dynasty green glazed lid

Plate 59:  Sui dynasty green glazed lid

70. AAL, ‘Rescue Excavations at Sha Po Tsuen’, 49.



Catalogue of Selected Finds   211

(17) Green glazed lotus bowl This deep, round bodied bowl has an incurving rim and incised lotus petal decoration on the exterior. All surfaces carry a green crackle glaze. Bowls of similar form were found in Six Dynasties tombs in Guangdong,71 for example, the No. 3 meng from a Six Dynasties grave in Shaoguan.72 Lotus leaf decoration is, in particular, a feature of Southern Dynasties bowls and jars—as seen at Meihuacun, Guangzhou73—and green glazed pottery from Hongzhou Kiln in Jiangxi was renowned for the use of such decoration. The lotus is a well-known symbol of Buddhism, and its widespread use on Southern Dynasties green glazed pottery can be related to the growth and popularity of the religion at that time.74 Dimensions: diameter 16 cm, surviving height 10.5 cm 2008, AA5CMH, Co. 505 [34:505] Southern Dynasties

Figure 51:  Southern Dynasties green glazed lotus bowl

Plate 60:  Southern Dynasties green glazed lotus bowl

71. Huang Xiaowen, ‘Study on the Tombs of Guangdong Area during Six Dynasties’ (MA diss., Nanjing University, 2011). 72. Liang Mingshen, ‘Preliminary Report on the Six Dynasties, Sui, and Tang Tombs in Shaoguan, Guangdong’, Kaogu 5 (1965): 232. 73. Chen, ‘Southern Dynasties Tomb at Meihuacun’, 194, M2:8. 74. Li Meitian, ‘Study on the Phasing of Six Dynasties, Sui, and Tang Green Glazed Ware in Central Yangtze River Region’, Huaxia Kaogu 4 (2000): 95.

212   Appendix 1

(18) Green glazed bowl This green glazed bowl with rounded bead rim has a pair of parallel incised lines on the exterior. The glaze was applied on all surfaces except the foot and there are several dark spots of glaze on the surface. Similar examples dated to the Six Dynasties period were found in both Guangdong and other provinces, for example, a Jin/Southern Dynasties bowl from Huangjin Plaza, Guangzhou,75 and a green glazed bowl M109:37 from an Eastern Jin tomb in Hanjiang, Jiangsu.76 Dimensions: rim diameter 15.5 cm, base diameter 8.7 cm, height 5.5 cm 1989, C.16.9, Squares CC a and d, Layers 3 [11:L3] Eastern Jin

Figure 52:  Eastern Jin green glazed bowl

Plate 61:  Eastern Jin green glazed bowl 75. Yi Xibing, ‘A Han to Six Dynasties and Tang Song Site in Huangjin Plaza, Zhongshan, Liulu, Guangzhou’, in Archaeological Discoveries and Research in Guangzhou Vol. I, ed. Institute of Cultural Relics and Archaeology of Guangzhou (Beijing: Wenwu Publishing, 2005), 208, T1b:36. 76. Li Cebin and Wu Wei, ‘An Eastern Jin Tomb in Liuli, Gangquan, Hanjiang, Jiangsu’, Southeast Culture 2 (1986): 24.



Catalogue of Selected Finds   213

(19) Green glazed cup Four complete or near-complete green glazed cups with deep bodies and upright rims were found on the Sha Po backbeach: one in 1989,77 two from 2000,78 and one in 2005.79 All of them show similar features: a pair of parallel incised lines below the rim on the exterior, a high solid disc-foot with an incised circle underneath, and green crackle glaze inside and out except on the lower part of the exterior. Such cups are commonly found in burials dated to Sui–early Tang in Guangdong, such as those found at Zhixin Secondary School in Guangzhou80 and at Xihe, Shaoguan.81 A further example was found at Qinzhou in Guangxi.82 Dimensions: rim diameter 9 cm, surviving height 6.3–6.5 cm 2000, T2L5:26 [16:T2L5] Sui to early Tang

Figure 53:  Sui to early Tang cup

Plate 62:  Sui to early Tang cup 77. Meacham, ‘Sha Po Tsuen’, 50, Fig. 17. 78. Au (AMO), 2001 Sha Po Tsuen archives. 79. ERM, ‘Rescue Excavation at Lot No. 1575s.B’, Fig.4.7d: SF81. 80. Kuang Guirong, ‘Preliminary Report on the Excavation of Sui–Tang Tombs at Zhixin Secondary School’, in Archaeological Discoveries and Research in Guangzhou Vol. I, ed. Institute of Cultural Relics and Archaeology of Guangzhou (Beijing: Wenwu Publishing, 2005), 151, no. 4 (M26:1). 81. Lam, Finds from Pre-Qin sites, 182–83, Tomb No. 11. 82. Wei Renyi and Tong Xianren, ‘Sui–Tang Tombs in Qinzhou, Guangxi Zhuang Autonomous Region’, Kaogu 3 (1984): 257, M1:1.

214   Appendix 1

(20) Green glazed stem cup with flower decorations This green glazed stem cup has a hollow pedestal foot and was decorated on the inside with impressed flowers and concentric diamonds. Green glaze was applied all over except for the lower part of the pedestal. Another local example (in this case, dated to the Tang dynasty) was discovered in Sha Tsui Tau, Lantau Island.83 Although the impressed motifs were not identical, the arrangement of one in the centre surrounded by a group of 4 to 5 others was very similar. Such arrangements are a common feature of Sui to early Tang green glazed dishes: for example, from Sui grave 98M2 in Zeqiaoshan, Ruyuan, Guangdong,84 Sui Grave 33 in Shaoguan, Guangdong,85 and a Sui–early Tang grave in Qinzhou, Guangxi.86 Dimensions: rim diameter 11 cm, base diameter 6.8 cm, surviving height 4 cm 2005, T1L4, SF62 [26:L4] Sui

Figure 54:  Sui dynasty green glazed stem cup with flower motifs

Plate 63:  Sui dynasty green glazed stem cup with flower decorations

83. AMO, ‘Archaeological Archives System: Sha Tsui Tau, C074.95’, accessed 20 August 2012. 84. Guangdong Provincial Institute of Cultural Relics and Archaeology, ed. Zeqiaoshan, Ruyuan, 249. 85. Liang, ‘Shaoguan’, 233. 86. Wei and Tong, ‘Sui–Tang Tombs in Qinzhou’, 257, M1:14.



Catalogue of Selected Finds   215

(21) Green glazed ink palette This large yellowish-green crackled glazed ink palette has stepped pedestal base, an upright peripheral rim, and unglazed and smooth domed upper surface, which was thus designed for the grinding and mixing of ink. The underside is rough, concave, and covered in glaze. When laid flat, the unglazed convex surface is higher than the rim. These elements are all typical of Six Dynasties–Tang Pi Yong–style ink palettes, except for the absence of small supporting legs, of which there might be anything between three and thirty.87 However, scars on the lower periphery of the base suggest that such legs were present originally. Similar examples of ink palettes have been found: with five small legs in a Southern Dynasties tomb in Hezhuishan, Luoding, Guangdong,88 with ten legs from Tomb No. 4 at Hanguang, Yingde,89 and with scars of broken legs from an early Tang minting site in Xihu Road, Guangzhou.90 According to recent typological research on excavated Tang–Song ink-stones, Pi Yong–style ceramic ink palettes were dated from the late sixth to early eighth century CE, but thereafter Bo Ji–style inkstones dominated the findings.91 Dimensions: diameter 26 cm, height 4.8 cm 2000, Kiln Y1 backfill [16:K1] Early Tang

Figure 55:  Tang dynasty green glazed ink palette (missing legs sketched in)

Plate 64:  Tang dynasty green glazed ink palette 87. Zhang Yue, ‘A Preliminary Research on the Inkstone during the Tang and Song Dynasties’ (MA diss., Jilin University, 2013). 88. Chen Dayuan, ‘Southern Dynasties Tomb in Hezuishan, Luoding, Guangdong’, Kaogu 3 (1994): 218. 89. Lam, Finds from Pre-Qin sites, 198–99. 90. Feng Yongju, Zhang Jinguo, and Huang Zhaoqiang, ‘Three Kingdoms Hoarding Site and Tang Dynasty Minting Site in Xihu Road, Guangzhou City’, in Archaeological Discoveries and Research in Guangzhou Vol. I, ed. Institute of Cultural Relics and Archaeology of Guangzhou (Beijing: Wenwu Publishing, 2005), 128, T4‰d:7. 91. Zhang, ‘Inkstone’.

216   Appendix 1

(22) Green glazed bowl Green glazed bowl with solid disc foot, rounded body, and upright rim with plain rounded lip. The flaky glaze was applied by dipping and does not extend onto the lower exterior wall or foot. The flat underside of the foot has characteristic spiral cutting-off marks, which suggest that the finished pot was removed while the potter’s wheel was still turning. Three clay wads—kiln stacking spurs—were fused onto the bowl’s interior during firing. The bowl’s form is very similar to an example from the Tang dynasty cemetery at San Tau, Lantau Island.92 Dimensions: rim diameter 15 cm, base diameter 5.5 cm, height 8.7 cm 2000, T1L5:27 [16:T1L5] Mid-late Tang

Figure 56:  Tang dynasty green glazed bowl

Plate 65:  Tang dynasty green glazed bowl 92. Atha, ‘Further Archaeological Investigations at San Tau’, SF190.



Catalogue of Selected Finds   217

(23) Wine jar sherd with stamp A yellowish-brown glazed wine jar body fragment with a surviving horizontal lug and stamped inscription. The wine jar was possibly produced at the Qishi kiln near Shiwan, Foshan in Guangdong and carried the inscription ‘梁宅酒’ (literally means ‘Leung family wine’) set inside a double-lined rectangular border. A review of Song–Yuan pottery found in Hong Kong shows that quantities of Shiwan pottery with various finishes have been found in excavations. Moreover, incised or stamped decoration and inscriptions are commonly found on Shiwan pottery, especially around the shoulder region of large jars.93 Based on the above classification of Shiwan green glazed wares, this piece falls somewhere within Categories 1 or 3. Some variation is apparent in the character ‘宅’ (‘family’), which perhaps relates to different potters, and the Sha Po example is stylistically different from two Category 3 examples found in the 1976 investigation at Shiwan. Also, according to the 1976 investigation report, the Qishi village site contained multiple kilns, which suggests it was a major centre of pottery production within the Shiwan region during Northern Song.94 Dimensions: width 12 cm, length 13 cm 2000, T2L3, 2000.20.17 [16:L3] Northern Song

Figure 57:  Northern Song wine jar sherd with stamp

Plate 66:  Northern Song wine jar sherd with stamp 93. Wong, ‘Ceramics of the Song–Yuan’, 54. 94. Chen Zhiliang, ‘Investigation on Ancient Shiwan Kiln Sites in Guangdong’, Kaogu 3 (1978): 197, 199.

218   Appendix 1

(24) Blue-and-white spouted wine/water jar This is a blue-and-white porcelain wine or water jar with elegant rounded shoulder carrying four small vertical looped lugs and a spout. The jar has a light bluish-grey crackle glaze over blue cloud decoration, which occurs below the neck and extends from the shoulder and down the body. This wine/water jar is similar to an example excavated at the Wun Yiu kiln site.95 Its form also resembles several ‘bulbous jugs’ unearthed from the Qing dynasty Tung Lung Fort site.96 The jar is possibly of Wun Yiu or Wuhua manufacture. Dimensions: rim diameter 6.8 cm, surviving height 9.5 cm 2008 WB, Area E, Co. 103 [42:103] Qing

Figure 58:  Qing blue-and-white spouted wine jar

Plate 67:  Qing blue-and-white spouted wine jar 95. Au et al., Wun Yiu, Tai Po, 69, WYA2SS. 96. Rogers, ‘Tung Lung Fort’, 37 and Figs. 26 and 27.



Catalogue of Selected Finds   219

Stone Artefacts (25) Rotary drill or ring-polishing stone This sandstone rotary drill or polishing stone has three spiral grooves around its tip. One end is missing and the object may originally have had a polishing ‘bit’ at both ends. Such objects are conventionally associated with the manufacture of ornamental rings, but could be used to produce a circular hole in a variety of objects. A double-ended example—labelled a drill—was found at Yung Long near Tuen Mun and dated to the earlier stage Late Neolithic.97 Other ring-polishing stones were found in Baojingwan in Zhuhai.98 They were dated to the Later Neolithic and Early Bronze Age/Shang respectively. The ring-polishing stones identified in Baojingwan were complete and all had concentric-ringed points on either or both ends. A range of different raw materials—quartz, granite, and sandstone—were used in their manufacture.99 Dimensions: length 4.9 cm, width 4.1 cm, surviving height 6.3 cm 1997, TP2, Co. 09 [15:09] Later Neolithic

Figure 59:  Later Neolithic rotary drill or polisher

Plate 68:  Later Neolithic stone rotary drill or polisher 97. Chau, Ancient Yue, 207. 98. Guangdong Provincial Institute of Cultural Relics and Archaeology and Zhuhai Museum, Baojingwan, T29A:17 and T12:60. 99. Ibid., 101.

220   Appendix 1

(26) Bivalve fish-hook mould (half ) This is a fragment of a fine sandstone bivalve mould for casting a pair of bronze fish-hooks. Although found within a Southern Dynasties burial, the mould was almost certainly disturbed from an underlying Bronze Age layer, but could have been reused as a component of the grave goods assemblage. Although a good number of bronze fish-hooks were found in the region, discoveries of moulds used to cast them are rare indeed.100 Evidence for bronze casting at the southern end of the Sha Po backbeach comprised four stone moulds for bronze axes—two a matching pair—together with casting slag and possible crucible fragments.101 The finished hooks would have a form of barb similar to examples discovered in Sham Wan, Lamma Island,102 although the shank of the Sha Po ones would be relatively longer. Dimensions (surviving): length 3.5 cm, width 3.2 cm, thickness 2.4 cm 1997, TP2, Co. 07 [15:07] Bronze Age

Figure 60:  Bronze Age bivalve mould for fish-hooks

Plate 69:  Bronze Age bivalve mould for fish-hooks. © Antiquities and Monuments Office, reproduced with permission. 100. For example, two similar bronze fish-hook moulds were excavated in Sha Kong Pui, Lantau Island by Chen Kung-che in 1938. Three bronze fish-hooks were also recovered from the same dig evidencing bronze smelting activities: Chen, ‘Archaeological Excavations in Hong Kong’, 16, Plate 5, No. 4 (Items 119 and 120), 14, Plate 6, No. 14 (Items 129–31). 101. Meacham, ‘Sha Po Tsuen’, 51. 102. Meacham, Sham Wan, 227.



Catalogue of Selected Finds   221

(27) Adze This is a trapezoid-shaped stone adze made of fine-grained grey volcanic rock, with bevelled sides and round-cornered straight cutting edge and rounded, bevelled butt-end. All surfaces were smoothed and polished. Another adze of very similar form and size—but in this case made from felsite—was found nearby in 2008.103 Many other close comparanda have been found in Neolithic deposits throughout Hong Kong, for example, those found in Kwo Lo Wan and Sha Ha.104 Dimensions: length 11.5 cm, width 6.1 cm, thickness 2.6 cm 2002, Unit 6.6, Co. 01 [22a:01] Later Neolithic

Figure 61:  Later Neolithic adze

Plate 70:  Later Neolithic adze 103. AAL, ‘Rescue Excavations at Sha Po Tsuen’, 136. 104. AMO, ‘Archaeological Archive System: Kwo Lo Wan (1990.005.00073) and Sha Ha (2002.031.01917)’, accessed 10 April 2014.

222   Appendix 1

(28) Burial assemblage from Grave 1: Four knife rough-outs Another probable Bronze Age burial—also discovered in 2002 on the central-eastern backbeach—was tentatively identified by the presence of four schist or slate knife rough-outs within an area of dark-stained soil.105 No other grave goods were identified in association with the knives or feature. Similar stone knife rough-outs were found in Later Neolithic–Bronze Age deposits at Sham Wan on Lamma Island106 and in Late Neolithic deposits at Fu Tei on Chek Lap Kok.107 (a) Dimensions: length 15 cm, width 4.9 cm, thickness 3.2 cm (b) Dimensions: length 12.5 cm, width 5.1 cm, thickness 1.8 cm (c) Dimensions: length 11.7 cm, width 4 cm, thickness 2.1 cm (d) Dimensions: length 7.6 cm, width 3.1 cm, thickness 1.1 cm 2002, Unit 6.9, Co. 02 [22a:02] Bronze Age

Figure 62:  Bronze Age grave goods from Grave 1: four stone knife rough-outs 105. AAL, ‘Sha Po Tsuen Rescue Excavation’, 36. 106. Meacham, Sham Wan, 193, Fig. IX-10. 107. Meacham, Chek Lap Kok, 83; AMO, ‘Archaeological Archive System: Fu Tei, Chek Lap Kok (1990.004.00043)’, accessed 10 April 2014.

Plate 71:  Bronze Age grave goods from Grave 1: stone knife rough-outs

224   Appendix 1

(29) Chipped pebble tool These were the general purpose tool of prehistoric Hong Kong and occurred in virtually all Late Neolithic and Bronze Age contexts at Sha Po. Pebble tools mostly take the form of choppers and picks and the triangular example shown below is of the latter type. This rhyolite or dacite pebble was chipped down two sides to create a point as well as two useful chopping or scraping edges. Along with other types of stone artefacts, they were possibly used for ‘hard hammer’ shaping of stone tool rough-outs and would have also been useful for any number of food gathering and processing applications. Dimensions: length 10.5 cm, width 9.9 cm, thickness 4.5 cm 2002, Unit 6.12, Co. 04 [22a:04] Later Neolithic

Figure 63:  Later Neolithic pebble tool

Plate 72:  Later Neolithic pebble tool



Catalogue of Selected Finds   225

(30) Whetstone Significant numbers of whetstones were excavated from the backbeach and plateau at Sha Po. They were made of fine sandstone or siltstone and used for shaping, polishing, and sharpening stone tools, weapons, and ornaments. The particularly well-preserved example shown below is rectangular in shape with wear facets on all sides and signs of heavy use. The broad dished facets suggest that this example may have, among other things, been used for shaping, polishing, and sharpening adzes. Similar rectangular whetstones of Neolithic date were found at Fu Tei Wan on Chek Lap Kok and Sha Ha at Sai Kung.108 Dimensions: length 12.8 cm, width 4.7 cm, thickness 1.6 cm 2008, AA5CMH, Co. 507 [34:507] Later Neolithic

Plate 73:  Later Neolithic whetstone

108. AMO, ‘Archaeological Archive System: Fu Tei Wan, Chek Lap Kok (1990.003.00024) and Sha Ha (2002.031.02168)’, accessed 11 April 2014.

226   Appendix 1

(31) Polishing stone or mortar This fragment of rectangular red sandstone polishing stone or mortar has a wear facet with a curving edge, which might relate to the shaping of rings or could indicate its use as a mortar with a stone pestle for rotary grinding. Such wear could also reflect the processing of plant foods of some sort. Examples of polishing stones made of reddish sandstones were found at Yung Long North and Yung Long South.109 Dimensions: length 4.2 cm, width 4 cm, thickness c. 1.5 cm 2002, Unit 7.8, Co. 02 [23:02] Bronze Age

Figure 64:  Bronze Age polishing stone or mortar

Plate 74:  Bronze Age polishing stone or mortar

109. AMO, ‘Archaeological Archive System: Yung Long North (1992.001.01227) and Yung Long South (1992.002.00050)’, accessed 10 April 2014.



Catalogue of Selected Finds   227

(32) Polishing stone Red sandstone polishing stone with linear groove, which may relate to rough-shaping the edges of artefacts such as schist-slate knives, but is more commonly associated with the smoothing of arrow shafts or other wood or bamboo rods. Many grooved polishing stones found in Hong Kong have long, straight grooves, which are perhaps more convincing in that regard. Late Neolithic examples were recovered from Lung Kwu Chau and Tai Long Wan, Chi Ma Wan.110 Bronze Age examples include those from Man Kok Tsui and So Kwun Wat.111 These polishing stones were made of various materials but sandstone is common. Dimensions: length 5 cm, width 3.5 cm, thickness 2.6 cm 2002, Unit 7.8, Co. 02 [23:02] Bronze Age

Figure 65:  Bronze Age polishing stone

Plate 75:  Bronze Age polishing stone 110. AMO, ‘Archaeological Archive System: Lung Kwu Chau (1994.009.00026) and Tai Long Wan, Chi Ma Wan (1970.002.00001)’, accessed 11 April 2014. 111. AMO, ‘Archaeological Archive System: Man Kok Tsui (1958.001.00152) and So Kwun Wat (2008.016.00137)’, accessed 11 April 2014.

228   Appendix 1

(33) Ring ornaments and core (a) Arm-ring or bangle Several stone rings made from the same grey schist or slate as projectile points were excavated on the plateau and the backbeach. The arm-ring or bangle shown here is quite typical in that it is polished all over and with a completely smooth inner face and grooved periphery. Local examples were dated to both the Late Neolithic112 and Bronze Age.113 Dimensions: outside diameter 8.6 cm, inside diameter 6.2 cm, x thickness 0.9 cm 2008, AA2C, Co. 204 [32:204] Later Neolithic to Early Bronze Age

(b) Ring core This schist or slate disc-shaped ring core has a tapering cross-section—smaller at the top and wider at its base—and two peripheral grooves, which are typical features of such objects. The grooving may relate to the core’s removal using three distinct episodes of drilling using a hollow-pointed drill, perhaps made of bamboo. The cores—like the rings they produced—occur in a range of diameters and thicknesses relating to the production of larger bangles or much smaller rings. Wong observed that large ring cores might be reworked one or more times until they were too small to produce another ring and finally became debris.114 Ring cores are a common find in Hong Kong and the Pearl River Delta region, especially in prehistoric stone workshop sites such as Pak Mong where 189 were found and Baojingwan.115 Among the many possible comparanda are Bronze Age examples from Sha Ha at Sai Kung and Sheung Pak Nai on Deep Bay, while a Late Neolithic example came from Kwo Lo Wan on Chek Lap Kok.116 Dimensions: diameter 5 cm, thickness 1.2 cm 2008, AA5A, Co. 505 [34:505] Bronze Age

(c) White quartz ring A fragment found in 2002 with a schist-slate stone ring core and ring fragment. Similar stone ring fragments dated to the Bronze Age were recorded at Man Kok Tsui.117 Dimensions: reconstructed outside diameter 6 cm, inside diameter 2.4 cm, thickness 0.5 cm 2002, Unit 6.10, Co. 02 [22a:02] Bronze Age

112. AMO, ‘Archaeological Archive System: Yung Long North (1992.001.00019)’, accessed 11 April 2014. 113. AMO, ‘Archaeological Archive System: Man Kok Tsui (1958.001.00065)’, accessed 11 April 2014. 114. Wong, ‘Pak Mong’. 115. Guangdong Provincial Institute of Cultural Relics and Archaeology and Zhuhai Museum, Baojingwan, 142–44. 116. AMO, ‘Archaeological Archive System: Sha Ha (2002.031.00838), Sheung Pak Nai (1999.007.00286) and Kwo Lo Wan (1990.005.00041)’, accessed 20 August 2012. 117. AMO, ‘Archaeological Archive System: Man Kok Tsui (1958.001.00083 and 1958.001.00084)’, accessed 10 April 2014.

Figure 66:  Later Neolithic–Bronze Age ring ornaments and core

Plate 76:  Later Neolithic–Bronze Age stone rings and core

230   Appendix 1

(34) Adze rough-out This schist rough-out is interesting for its exceptionally large size and unusual shape combining almost straight sides, a rounded rectangular cutting edge, and triangular butt end. Its cutting edge profile is very much ‘adze-like’ in that it has one flat face and the other is roughly bevelled; however, when finished to shape and polished, its size, material, and weight would presumably make it an impractical tool for wood-working. It may therefore be a rough-out for spade or similar tool, or could be an object of more ceremonial rather than practical use. Given its unusual size and shape, no close comparanda could be found, although several ‘short-edged knives’ from Sham Wan on Lamma Island show some similarities.118 Dimensions: length 22.7 cm, width 9.5 cm, thickness 3.1 cm 2008, AA5C-MH, Co. 510 [34:510] Later Neolithic

Figure 67:  Later Neolithic adze rough-out

Plate 77:  Later Neolithic adze rough-out 118. Meacham, Sham Wan, 190–91.



Catalogue of Selected Finds   231

(35) Stepped adze This small trapezoid-shaped stepped adze was the smaller of two such artefacts found on the Sha Po backbeach. The adze has poorly defined sloping shoulders, a gently rounded butt end, and cutting edge, which is worn but nevertheless still sharp. It is made from a fine-grained hard volcanic rock. The other stepped and shouldered adze found at Sha Po in 1989 has asymmetrical shoulders and a proportionally longer stepped blade.119 Stone adzes of similar form are commonly found in the wider Hong Kong region, for example, of Bronze Age date at So Kwun Wat and Sha Ha,120 and Later Neolithic date at Fuchuanling, Guangdong.121 It is generally accepted that stepped and shouldered adzes appeared around 3000–1500 BCE in Hong Kong and most were probably locally produced.122 Dimensions: length 5.5 cm, width 3.1 cm, thickness 1.5 cm 1997, TP1, Co. 04, SF1 [14:04] Later Neolithic to Early Bronze Age

Figure 68:  Later Neolithic–Early Bronze Age stepped adze

Plate 78:  Later Neolithic–Early Bronze Age polished stepped adze 119. Meacham, ‘Sha Po Tsuen’, 52; Tang Chung and Au Ka-fat, eds., Archaeological Finds from the Pearl River Delta in Guangdong, China (Hong Kong: Chinese University Press, 1991), 45. 120. AMO, ‘Archaeological Archive System: So Kwun Wat (2000.019.00070) and Sha Ha (2002.031.01863)’, accessed 11 April 2014. 121. Li Ziwen, ‘Excavation of Fuchuanling Site in Renhua, Guangdong’, Wenwu 7 (1998): 34, T201:45. 122. Fu Xianguo, ‘Neolithic Stepped Stone Adzes and Shouldered Stone Implements of Hong Kong’, a Lord Wilson Heritage Trust Funded Project, 2006; Fu Xianguo, ‘The Relationship between Prehistoric Hong Kong, South China, South East Asia and South Pacific Regions Based on Stepped Stone Adzes and Shouldered Stone Tools’, a Lord Wilson Heritage Trust Funded Project, 2006.

232   Appendix 1

(36) Sceptre (zhang) Part of perforated stone implement was found in a disturbed context in 2005.123 Judging from the surviving form, it seems to be the butt-end of a stone zhang or yazhang, but a butt fragment from a ge-like blade is another, albeit less convincing, possibility. The zhang was a ritual object of the Shang culture of the Central Plains, where they were usually made of jade, and had an elongated form with a square butt with a circular hole and an often tapering or curved blade. In the case of yazhang, there were also usually a series of teeth-like protruding decorations between the butt and the blade.124 A jade yazhang was unearthed in the next bay south from Sha Po at Tai Wan.125 Similar stone implements were also found in Guangdong, for example, a polished stone zhang fragment found at Zengjiang, Zengcheng126 and a ge-shaped stone implement collected from Longxuecun, Zhongshan.127 The excavated stone yazhang from Guangdong were dated to Early to Middle Bronze Age.128 Dimensions: length (surviving) 5.5 cm, width 6.5 cm 2005, SF84, T2L1 [27:L1] Bronze Age

Plate 79:  Bronze Age stone sceptre (zhang). © Antiquities and Monuments Office, reproduced with permission. 123. ERM, ‘Yung Shue Wan’, 13. 124. Tang, Archaeological Past, 88. 125. Au Ka-fat et al., ‘Preliminary Report on the Excavation of Tai Wan site, Lamma Island’, in Cultures of South China and Neighboring Regions: Essays in Honor of Professor Cheng Te-k’un on the Occasion of the Sixtieth Anniversary of His Academic Career, ed., Tang Chung (Hong Kong: Chinese University Press, 1994), 195–209. 126. Institute of Cultural Relics and Archaeology of Guangzhou, Zhu Ji Cun Lei, 266. 127. Yang Shiting, ‘Preliminary Discussion on Yazhang from Guangdong and Hong Kong: Sings of South-Bound Transmitting of Xia Shang Zhou Culture’, in Cultures of South China and Neighboring Regions: Essays in Honor of Professor Cheng Te-k’un on the Occasion of the Sixtieth Anniversary of His Academic Career, ed., Tang Chung (Hong Kong: Chinese University Press, 1994), 301, 305. 128. Xu Hengbin, ‘Problems and Suggestions Related to Archaeological Research of Sandbar Sites, and Discussion on the Dating of the Stone Yazhang Unearthed from Tai Wan, Lamma Island’, in Archaeology in Southeast Asia, ed. Yeung Chuen-tong and Li Wai-ling (Hong Kong: The University Museum and Art Gallery, the University of Hong Kong, 1995), 244.



Catalogue of Selected Finds   233

(37) Soapstone ‘net-weights’ Three unusual carved stone objects with grooved ends were discovered on the backbeach in a disturbed Six Dynasties–Tang layer, together with a number of Bronze Age finds. Although apparently some form of laminated stone, the three objects were surprisingly light. One (c) appeared to be intact, but the other two had split in half longitudinally along natural laminations in the rock. Based on recent results of nondestructive scientific analyses and the opinion of a prominent geological specialist, it was established that the objects were carved from soapstone.129 Similar items from sites in mainland China are usually identified as ‘ceramic net-weights’, but many may also be soapstone. It seems that these net-weights were used for a prolonged period, with the earliest being from the Dayishan site in Guanyan, Jiangsu dated to the Neolithic (4500 bce).130 Bronze Age examples include a ‘grey clay net-weight’ from a Shang dynasty city site in Shanxi,131 various styles and sizes of ‘ceramic net-weights’ from the Shang dynasty Wucheng site in Jiangxi,132 a ‘black clay net-weight’ (late Western Zhou, T3ƒ:7) and a ‘dark brown clay net-weight’ (late Western Zhou to early Spring and Autumn) from Fengxingzuishan in Yueyang City, Hunan,133 and ‘ceramic net-weight’ of Late Western Zhou to early Spring and Autumn date.134 Also, large quantities of ‘ceramic net-weights’ were excavated in a Han site in Juting, Gansu.135 It is notable that most of the above examples have a ‘waisted’ profile caused by deep grooves on both sides of their midsection. (a) Pinkish colour, polished, one flat surface

Dimensions: length 2.2 cm, width 0.9 cm, thickness 0.6 cm, weight 1.5 g SF10, Accession No. 1997.13.51

(b) Dark grey, polished

Dimensions: length 2.3 cm, width 1 cm, thickness 0.35 cm, weight 2 g SF8, Accession No. 1997.13.48

(c) Dark grey; polished; worn surfaces on both sides, perhaps originally convex like the other two; refitted from three fragments Dimensions: length 2.4 cm, width 1.2 cm, thickness 0.7 cm, weight 3 g (SF8, 1997.13.48) All from 1997, TP2, Co. 03 [15:03] Pre-Tang, possibly Bronze Age

129. LCSD Central Conservation Section, ‘Analysis Report for Historic Relics’, unpublished report, 2012. 130. Wu Rongqing, ‘Excavation of the Dayishan Site at Guanyan, Jiangsu in 1986’, Wenwu 7 (1991): 22–23, GD: 46. 131. Wang Rui and Tong Weihua, ‘Excavation of the Shang Dynasty City Site at Nanguan, Gucheng, Yuanqu, Shanxi in 1988–89’, Wenwu 10 (1997): 22, H405:9. 132. Jiangxi Provincial Institute of Cultural Relics and Archaeological and Zhangshu City Museum, eds., Wucheng: 1973–2002 Excavation Report (Beijing: Science Publishing, 2005), 338–44. 133. Fu Xuan and Zhang Biwu, ‘Excavation of the No. 1 Tomb at Hill Fengxingzuishan in Yueyang City, Hunan’, Wenwu 1 (1993): 26, T3B:8. 134. Fu and Zhang, ‘Fengxingzuishan’, 26, T3B:9. 135. Chu Shibin and Ren Buyun, ‘The Discovery of a Han Site in Juting, Gansu and the Newly Excavated Scripts’, Wenwu 1 (1978): 18.

Figure 69:  Soapstone ‘net-weights’ (pre-Tang, possibly even Bronze Age)

Plate 80:  Soapstone ‘net-weights’ (pre-Tang, possibly even Bronze Age)



Catalogue of Selected Finds   235

(38) Ink-stone This heavily used ink-stone—used in fact to the point where it had worn through—was decorated with a well-executed melon and leaf motif. On the reverse, there is an inscription of the character ‘有’ (meaning ‘have’), which could possibly relate to the ownership of the object. The material is a fine-grained shale or slate. This is an unusual find for a rural site. Dimensions: length 6.5 cm, width 6.5 cm, thickness 0.5 cm 2008, AA6A, Co. 605 [36:605] Qing

Figure 70:  Qing ink-stone with melon motif and inscription on reverse

Plate 81:  Qing ink-stone with melon motifs and inscription on reverse

Appendix 2: Glossary

absolute dating 絕對年代: the determination of age in calendar years based on scientific testing (e.g., Radiocarbon C14) (cf. relative dating). adze 錛: an asymmetrical stone cutting tool of the Neolithic and Bronze Age, usually chipped, ground, and polished. It is designed to be hafted (mounted) on a handle with the blade horizontal (cf. axe). aerial photography 航拍照片: photographs taken from aircraft, either vertically down or from an oblique angle used in archaeology for the identification and mapping of archaeological sites and landscape change. archaeological record 考古資料: a static, ‘dead’ material record reflecting past dynamic human behaviour and post-depositional processes (both cultural and natural). archaeology 考古: the study of past human societies through the scientific analysis of material remains they left behind. artefact 遺物: any portable object that has been made (e.g., a pot) or modified (a stone pebble tool) by humans (cf. ecofact). assemblage 器物組合: a collection of related materials representative of human activity. auger 鑽探: manual or mechanical hollow-headed drill with T-bar handle, used to remove soil cores thereby revealing the subsurface sequence of deposits. axe 斧: a symmetrical cutting tool of Bronze Age (bronze) or later date (iron) designed to be hafted (mounted) on a handle with the blade vertical (cf. adze). backbeach 後灘: raised coastal features above and behind the active shoreline comprising layers of sand and sometimes coarser material deposited by seasonal storms (a.k.a. storm beach). Many of the most important prehistoric and earlier historical sites in Hong Kong are on such topographic features. Bronze Age 青銅時代: prehistoric period dated in Hong Kong to 1500–221  bce (between Late Neolithic and Qin dynasty), but in reality the period is seldom dated on Hong Kong sites beyond 500 bce (hence our use of the date range 1500–500 BCE). context 考古單元: the basic units of archaeological analysis—cuts, fills, layers, and structural elements— representing individual events of varying duration that occurred on site. corded 繩紋: a characteristic form of impressed decoration that was used on coarse pottery throughout Hong Kong prehistory from the Middle Neolithic to Bronze Age. cultural landscape 文化景觀: the product of the interaction between human agency and environmental processes (cf. landscape). cultural layer 文化堆積層: a soil stratum with evidence for human activity (cf. sterile layer).

Glossary   237

cultural remains 文化遺存: any material residues left behind by people. cut 打破面: a term used to define past events involving digging: e.g., cutting a post-hole, pit, wall foundation trench, grave, or well shaft. dating 年代: analysis of finds and other excavated materials to provide dates for the finds themselves and associated contexts and features (can be absolute or relative). desk-based assessment 卓面評估: a review of all published and unpublished map, photographic, and documentary evidence about a development location (an early stage of impact assessment process). disturbed layer 擾亂地層: a stratum with evidence for mixing of material of different dates. ditch 溝: a linear feature, for example, cut to create a boundary or for drainage purposes. double-F 夔紋: characteristic motif of high Bronze Age (c. 1200–500 bce) hard geometric pottery. ecofact 動植物遺存: ecofacts are the organic and environmental remains found on an archaeological site, mostly food residues; they include bone, seeds, shells, and plant phytoliths (cf. artefact). excavation 發掘: the main data-gathering method in archaeology; the systematic process of identification, definition, and removal of cultural deposits in the reverse order of their deposition. fabric 陶質: the material from which pottery is made (e.g., throughout Hong Kong prehistory two main types of fabric were in use: fine clay and coarse gritted). feature 遺跡: a group of contexts reflecting a sequence of interrelated events that led to the creation of something (e.g., a pit, post-hole, or ditch) comprising a cut and one or more fills. fill 填土: the contents of a cut. finds 遺物: artefacts (things made by people) and ecofacts (mainly remains of things eaten by people). flotation 浮選: a technique of wet sieving devised to remove carbonised material such as seeds, from samples of soil. form 器型: the shape of pottery vessels which, together with fabric and decoration, is usually diagnostic for date and, in historical periods, even to a specific kiln or production region. geometric 幾何紋飾: pottery decoration comprising repeating pattern characteristic of later prehistory of the Hong Kong–Pearl River Delta, in particular in the Later Neolithic and Bronze Age. geophysical survey 地球物理: the systematic identification of buried archaeology through recordable contrasts in levels of magnetism, material density, and resistance to electrical current flow (moisture content) between the natural background and archaeological features. grid 格網: a regular pattern of squares with known coordinates within which archaeological fieldwork takes place. in situ 原位: a Latin term meaning ‘in its original position’. When objects can be shown to have been undisturbed since their date of deposition, they are referred to as being in situ. interpretation 詮釋: the ‘art’ of creating scientifically robust stories about a site’s dynamic human past based on the static material remains and data recovered. kiln 窰: a fired clay and/or stone structure designed to provide a controlled high-temperature environment within which certain chemical and physical changes in materials can occur (e.g., converting clay into pottery; shells and coral into lime; tin, copper, and lead into bronze; and brine into salt). landscape 景觀: an area as perceived and experienced by people (cf. cultural landscape).

238   Appendix 2

landscape archaeology 景觀考古: a cross-disciplinary approach that focuses on multi-period investigation, analysis, and interpretation of archaeological remains within topographically defined areas and using variety of complementary techniques applied across a range of spatial scales. Late Neolithic 新石器時代晚期: prehistoric period in Hong Kong dated to 2700–1500  bce and falling between Middle Neolithic and Bronze Age, divided into LN1 (2700–2400 bce) and LN2 (2400–1500 bce). layers 堆積: extensive deposits that can accumulate gradually over longer timescales, such as middens (old rubbish dumps), occupation deposits, or agricultural soils; or can be created more quickly, such as a demolition or site-levelling deposits. LIDAR (Light Detection and Ranging) 激光雷達: remote sensing technique using beams of light fired at the ground from aircraft. Light bounces back from the tree canopy, underlying natural topography, and any artificial features on the ground surface. The time taken for reflection allows calculation of the distance involved and the creation accurate 3D models of surface features. lime 石灰: a white caustic alkaline substance consisting of calcium oxide and produced and/or used during the operation of earlier historical (Sui–Tang dynasties) kilns in Hong Kong. midden 貝丘: a build-up of ancient rubbish typically comprising food remains (shells, fish, and mammal bones), as well as broken pots and other artefacts. Middle Neolithic (MN) 新石器時代中期: prehistoric period in Hong Kong dated to 4500–2700 bce and falling between the controversial Pre-pottery Neolithic (c. 5000–4500 bce) and Late Neolithic; divided into MN1 (4500–3500 bce) and MN2 (3500–2700 bce). Neolithic 新石器時代: New Stone Age; period of development of pottery use, ground and polished stone tool technologies, emergence of agriculture, and domestication of animals, but in Hong Kong fishing, hunting, and gathering appear to have been dominant (cf. Palaeolithic). palaeobotany 古植物學: study of ancient plant remains. Palaeolithic 舊石器時代: Old Stone Age; period dominated by use of chipped stone technology and no pottery, hunting, and gathering lifeway (present in mainland China, claimed at Wong Tei Tung in Hong Kong but as yet not widely accepted; cf. Neolithic). pit 灰坑: a hole dug in the ground (e.g., to quarry clay or stone, or to bury something). pollen and phytolith analysis 孢粉和植物硅酸體分析: study of microscopic plant remains as evidence for past human exploitation of wild and domesticated species. post-excavation processing and analysis 發掘後的資料整理和分析: the work of converting mudcovered finds and field records into a submission-ready site archive and a site report. post-hole 柱洞: the cylindrical hole that is left when a wooden or bamboo post has decayed to leave a (usually) subcircular or suboval patch of darker soil which can be identified during excavation (often contains stones originally placed to keep the post in place = packing stones). preservation by record 透過記錄作保存: refers to excavating archaeological remains and ‘preserving’ them in the records comprising site archive and report. preservation in situ 原位保留: in other words, leaving remains in the ground. radiocarbon dating 碳-14年代測定法: the scientific method of dating used for organic materials. The technique measures how much of the gradually decaying radiocarbon isotope C14 remains in a sample, which then denotes how long has elapsed since the organism died (and by association the date of the archaeological deposit/feature in which it was found).

Glossary   239

recording 考古記錄: the process of preserving the excavated site using written records, drawings, and photographs. relative dating 相對年代: the ordering of layers and features on site based on their relative positions in the stratigraphic sequence (the law of superposition) and the diagnostic type fossils they contain (cf. absolute dating). remote sensing 考古遙感: the investigation of archaeological sites and landscapes using non-invasive methods (no digging) such as satellite imagery, aerial photography, LIDAR, and ground-based geophysical survey. rescue excavation 搶救發掘: the excavation of sites that would otherwise be destroyed by construction (see also preservation by record). resources 考古資源: archaeological remains. section 剖面: the vertical face at the trench edge showing details of any layers and features present. sherd 陶片: a piece of broken pottery. site 遺址: a spatially discrete cluster of archaeological remains reflecting a focus of human activity. sterile layer 貧瘠層: a soil stratum (layer) with no evidence of human activity (cf. cultural layer). stilt-house 杆欄式建築: houses with raised floors supported on a pattern of posts set into the ground (still common throughout South East Asia). strata 地層: layers (either cultural or natural) making up the sequence of deposits on a site (singular stratum). stratigraphy 地層學: a method of placing excavation units (contexts and features) in the correct chronological order based on their spatial location and physical relationships with each other (lower strata being earlier than upper strata). structure 建築遺存: anything built by people, such as a wall, paved floor, or agricultural terrace. subsistence strategy 經濟形態: people’s approach to procuring food and materials necessary for life. surface collection 地表採集: a survey technique involving the collection of finds from the surface of an archaeological site (ideally systematic and within a grid). survey 調查: a method of quite rapidly but systematically sampling large areas in order to assess their archaeological potential—in Hong Kong usually surface collection, augering, and test pit rather than geophysical survey or larger-scale excavation. terraced field 梯田: classic Hong Kong–South China agricultural landscape feature comprising flat terraces cut in steps into sloping ground and usually faced with stones. test pit 探方: a small, often square, trench excavated to reveal the date and nature of any archaeological deposits present on site. topography 地形: the arrangement of natural and artificial (man-made) features of an area. type fossil 斷代物: an artefact that is characteristic of a particular period or date, such as ‘double-F’ hard geometric pottery, which is diagnostic of the Bronze Age. watching brief 考古監察: a method of archaeological fieldwork that occurs in the construction phase of projects when archaeologists monitor the contractors’ works and check to see whether any archaeological remains are encountered. Any archaeological remains are recorded by the archaeologist and, if significant, mitigation measures may then be designed and implemented (e.g., preservation in situ or by record).

Bibliography

AECOM Asia Co. Ltd. ‘Consultancy Agreement No. C8016, Environment Team Consultancy for Express Rail Link, Final Archaeological Rescue Excavation Report’. Unpublished excavation report, 2014. Akazawa, T., and C. M. Aikens, eds. Prehistoric Hunter-Gatherers in Japan: New Research Methods. Tokyo: University of Tokyo, 1986. Antiquities and Monuments Office. ‘Sha Po Old Village, Lamma Island, 1997’. Unpublished draft excavation report, 1997. ———. ‘Sha Po Old Village’. Unpublished excavation report, 2001. ———, ed. The Ancient Culture of Hong Kong: Archaeological Discoveries in Sha Ha, Sai Kung. Hong Kong: LCSD, 2005. ———. ‘Preliminary Report on the 2004 Fat Tong Chau Archaeological Investigation’. Huaxia Kaogu 4 (2007): 35–41, 55. [古物古蹟辦事處:〈香港佛頭洲2004考古調查報告〉,《華夏考古》,2007年第4期,頁35–41、55。] Archaeological Assessments Ltd. ‘Sha Po Tsuen Rescue Excavation’. Unpublished excavation report, 2003. ———. ‘Archaeological Rescue Excavations at Sha Po Tsuen’. Unpublished excavation report, 2011a. ———. ‘Archaeological Watching Brief at Sha Po Tsuen’. Unpublished watching brief report, 2011b. Ashmore, Wendy. ‘Social Archaeologies of Landscape’. In A Companion to Social Archaeology, edited by Lynn Meskell and Robert W. Preucel, 255–71. Oxford: Blackwell, 2004. Atha, Mick. ‘Late Iron Age Regionality and Early Roman Trajectories (100 BC–AD 200): A Landscape Perspective from Eastern Yorkshire’. PhD thesis, University of York, 2008. ———. ‘A Neglected Heritage: Towards a Fuller Appreciation of the Landscapes and Lifeways of Hong Kong’s Rice Farming Past’. Asian Anthropology 11 (1) (2012): 129–56. ———. ‘Archaeological Investigation at San Tau, Lantau Island’. Unpublished report of investigation conducted on behalf of the Hong Kong Archaeological Society, 2012. ———. ‘Further Archaeological Investigation at San Tau, Lantau Island’. Unpublished report of investigation conducted on behalf of the Hong Kong Archaeological Society, 2013. ———. ‘A Military and Civilian Cemetery of the Mid to Late Tang Maritime Trade? Ground Penetrating Radar (GPR) Surveys and Excavations at San Tau, North Lantau, Hong Kong’. In Collected Essays of the International Conference on Historical Imprints of Lingnan: Major Archaeological Discoveries of Guangdong, Hong Kong and Macao, edited by Susanna L. K. Siu, Rosa S. K. Yau, Joseph C. L. Chow, Priscilla S. H. Wong, and Kwok Hoiching, 176–217. Hong Kong: Hong Kong Museum of History, 2014. ———. ‘Survey-cum-Excavation at Pak Kok Tsui, Lamma Island’. JHKAS XVII (2009–2014) (forthcoming). ———. ‘Survey-cum-Excavation on Yim Tin Tsai, Sai Kung’. JHKAS XVII (2009–2014) (forthcoming). Atha, Mick, and Steve Roskams. ‘Pre-medieval Transitions at Wharram Percy’. In Wharram: A Study of Settlement on the Yorkshire Wolds, XIII, edited by Stuart Wrathmell, 63–82. York: University of York Press, 2012. Atha, Mick, and Kennis Yip. ‘A Tang Dynasty Cemetery at San Tau, North Lantau, Hong Kong: Summary of the 2011–2012 Archaeological Survey-cum-Excavation’. JHKAS XVII (2009–2014) (forthcoming in Chinese). [范旼澔(Mick Atha)、葉可詩:〈香港大嶼山䃟頭唐代墓葬2011至2012年度考古調查發掘簡報〉,載 《香港考古學會會刊》,第17期(2009–2014)(待出版)。] Au, Ka-fat. ‘Report on the Archaeological Survey on the Project of Small House Development at Yung Shue Wan Archaeological Site, Lamma Island’. Unpublished report, 2001a.

Bibliography   241 [區家發:〈南丫島榕樹灣小型鄉村屋宇建築工程考古調查工作報告〉(未刊報告),2001a。]

Au, Ka-fat. ‘Rescue Excavation at Lot 1687, Back Street, Sha Po Tsuen’. Unpublished excavation report (LI22), 2001b. [區家發:〈南丫島榕樹灣村屋建築工程搶救發掘工作報告〉(未刊報告),2001b。] Au, Ka-fat, and Tang Chung. ‘A Preliminary Excavation Report on Tung Wan Beach Site in Hong Kong’. In Treatises in the Commemoration of the 30th Anniversary of the Discovery of the Maba Human Cranium, edited by the Guangdong Provincial Museum and Museum of the Qujiang County, 208–18. Beijing: Cultural Relics Publishing House, 1988. [區家發、鄧聰:〈香港大嶼山東灣新石器時代沙丘遺址的發掘〉,載廣東省博物館、曲江縣博物館 編:《紀念馬壩人化石發現三十周年文集》。北京:文物出版社,1988,頁208–18。]

Au, Ka-fat, Feng Yongju, Li Guo, Tang Chung, and Shang Zhitan. ‘Preliminary Report on the Excavation of Tai Wan Site, Lamma Island’. In Cultures of South China and Neighboring Regions: Essays in Honor of Professor Cheng Te-k’un on the Occasion of the Sixtieth Anniversary of His Academic Career, edited by Tang Chung, 195–209. Hong Kong: Chinese University Press, 1994. [區家發、馮永駒、李果、鄧聰、商志(香覃):〈香港南丫島大灣遺址發掘簡報〉,載鄧聰編:《南中 國及鄰近地區古文化研究:慶祝鄭德坤教授從事學術活動六十週年論文集》。香港:中文大學出版 社,1994,頁195–209。]

Au, Ka-fat, Zeng Guangyi, Chow Sai-wing, Yu Ka-tung, Chan Chung-fat, Hui Siu-mui, and Yam Tak-kau. Report of the Investigation and Research on Wun Yiu, Tai Po Blue-and-White Porcelain Site. Hong Kong: Regional Council, 1997. [區家發、曾廣億、周世榮、余家棟、陳宗發、許小梅、任德球:《香港大埔碗窰青花瓷窰址調查及研 究》。香港:香港區域市政局,1997。]

Au, Ka-fat, and Mo Zhi. ‘Working Paper on the Ng Ka Yuen, Ha Pak Nai Sandbar Site Excavation’. JHKAS XIV (1993–1997) (1998): 5–40.

[區家發、莫稚(香港考古學會):〈元朗下白泥吳家園沙丘遺址調查試掘工作報告〉,《香港考古學會 會刊》,第14期(1993–1997),1998,頁5–40。]

Au, Ka-fat, and Tung Po-ming. ‘Ancient Salt Industry in Hong Kong and Shenzhen Region’. JHKAS XIV (1993– 1997) (1998): 81–87. [區家發、佟寶銘:〈香港深圳地區之古代煮鹽業〉,《香港考古學會會刊》,第 14 期( 1993–1997 ), 1998,頁81–87。] Au, Ka-fat, and Raymond Lee. ‘Archaeological Investigation at Yung Shue Wan, Lamma Island, 2000–2001’. Unpublished report, 2001. [區家發、李偉文:〈南丫島榕樹灣考古調查2000–2001〉(未刊報告),2001。] Au, Ka-fat, and Mo Zhi. ‘Archaeological Investigation Report on the Main Drainage Channels for Ngau Tam Mei Project, Phase II’. Unpublished report, 1999. [區家發、莫稚:〈元朗牛潭尾主渠道第二期工程考古調查工作報告〉(未刊報告),1999。] Bai, Yunxiang. Research on the Archaeological Findings of Pre-Qin to Han Dynasty Iron Works. Beijing: Science Publishing, 2005. [白雲翔:《先秦兩漢鐵器的考古學研究》。北京:科學出版社,2005。] Baker, Hugh D. R. ‘The Five Great Clans of the New Territories’. JRASHKB 6 (1966): 26. Bard, Solomon M. ‘Chung Hom Wan’. JHKAS VI (1975): 9–25. ———. ‘Archaeology in Hong Kong: A Review of Achievement’. In Conference on Archaeology in Southeast Asia, edited by Chun-tong Yeung and Wai-ling Li, 383–96. Hong Kong: University Museum and Art Gallery, the University of Hong Kong, 1995. ———. ‘Report on Archaeological Survey at Tai Fu Tai, Phase II’. Unpublished report, 2001. Barrett, C. 1973. ‘Tai Wan Reconsidered’. JHKAS, Volume IV: 53–59. Barrett, John C. ‘Fields of Discourse: Reconstituting a Social Archaeology’. Critique of Anthropology 7 (3) (1988): 5–16. Bu, Gong. ‘Archaeological Observations on the Development of Lingnan Culture’. Journal of Historical Anthropology 3 (2) (2005): 1–23. [卜工:〈嶺南文明進程的考古學觀察〉,《歷史人類學學刊》,2005年第3卷第2期,頁1–23。] Cameron, Hugh. ‘Lo So Shing’. JHKAS VII (1979): 125–29. ———. ‘Yi Long’. JHKAS VII (1979): 118–24.

242   Bibliography ———. ‘Discussion of Tang Lime Kilns at Sham Wan Tsuen’. In Archaeological Investigations on Chek Lap Kok Island, Hong Kong Archaeological Society Journal Monograph IV, edited by William Meacham, 223–30. Hong Kong: HKAS, 1994. Cameron, Hugh, and B. V. Williams. ‘Sham Wan Tsuen: Chek Lap Kok’. JHKAS X (1984): 10–54. Chan, Emily H. Y. ‘More Than Meets the Eyes: Slotted Rings and Social Complexity in Bronze Age Hong Kong’. Unpublished final-year project, Department of Anthropology, the Chinese University of Hong Kong, 2012. Chan, William. ‘Identification and Analysis of Excavated Fish Remains’. In Sham Wan, Lamma Island: An Archaeological Site Study, Journal Monograph III, edited by William Meacham, 248–57. Hong Kong: HKAS, 1978. Chang, Kwang-chih. The Archaeology of Ancient China. New Haven and London: Yale University Press, 1986. Chau, Hing-wah, ed. Collected Essays on the Culture of the Ancient Yue People in South China. Hong Kong: Urban Council, 1993. [鄒興華編:《嶺南古越族文化論文集》。香港:香港市政局,1993。] ———. ‘Periodization of Prehistoric Culture of Pearl River Delta area’. In Collected Essays on the Culture of the Ancient Yue People in South China, edited by Chau Hing-wah, 40–55. Hong Kong: Urban Council, 1993. [鄒興華:《珠江三角洲史前文化分期〉,載鄒興華編:《嶺南古越族文化論文集》。香港:香港市政 局,1993,頁40–55。]

Chau, Hing-wa, Wu Yaoli, and Li Langlin. ‘A Preliminary Report on the Excavation of a Prehistoric Site in Tung Wan Tsai North, Ma Wan, Hong Kong’. Kaogu 6 (1999): 1–17. [鄒興華、吳耀利、李浪林(古物古蹟辦事處):〈香港馬灣島東灣仔北史前遺址發掘簡報〉,《考古》, 1999年第6期,頁1–17。]

Chen, Chun. ‘Salt Production’. In Ao Bo Tu, Volume 2 (reproduced from the collection in possession of the National Palace Museum). Taipei: Taiwan Commercial Press, 1984. [(元)陳椿:〈上滷煎圖〉,《熬波圖》。臺北:臺灣商務印書館,1984,下卷第四十二(國立故宮博 物館藏本)。]

Chen, Chunli. ‘Preliminary Report on the Excavation of a Southern Dynasties Tomb at Meihuacun, Dongshan, Guangdong’. In Guangzhou Wenwu Kaogu Ji, edited by the Institute of Cultural Relics and Archaeology of Guangzhou, 188–96. Beijing: Wenwu Publishing, 1998. [陳春麗:〈廣州東山梅花村南朝墓發掘簡報〉,載廣州市文物考古研究所編:《廣州文物考古集》。北 京:文物出版社,1998,頁188–96。]

Chen, Dayuan. ‘A Southern Dynasties Tomb in Hezuishan, Luoding, Guangdong’. Kaogu 3 (1994): 216–20. 陳大遠:〈廣東羅定縣鶴咀山南朝墓〉,《考古》,1994年第3期,頁216–20。 Chen, Jiuheng. ‘Investigation and Excavation on the Sui–Tang Dongdou City Site’. Kaogu 6 (1978): 361–79.

[陳久恆(中國社會科學院考古研究所洛陽工作隊):〈隋唐東都城址的勘查和發掘續記〉,《文物》, 1978年第6期,頁361–79。]

Chen, Kung-che. ‘Archaeological Excavations in Hong Kong’. Acta Archaeologica Sinica 4 (1957): 1–16. [陳公哲:〈香港考古發掘〉,《考古學報》,1957年第4期,頁1–16。] Chen, Liangwei. ‘Preliminary Report on the Excavation of a Tang–Song Site at Tanggong Road (North) in Luoyang’. Kaogu 12 (1999): 45–53. [陳良偉(中國社會科學院考古研究所洛陽唐城工作隊):〈河南洛陽唐宮路北唐宋遺址發掘簡報〉, 《考古》,1999年第12期,頁45–53。]

Chen, Moxi, Mu Yinghang, and Chen Hengan. ‘Report on the Excavation of a Han Tomb in Pingba, Qingzhen, Guizhou’. Acta Archaeologica Sinica 1 (1959): 85–103. [陳默溪、牟應杭、陳恆安:〈貴州清鎮平壩漢墓發掘報告〉,《考古學報》, 1959 年第 1 期,頁 85– 103。] Chen, Wen. ‘Bronze Age of Hong Kong and Its Periodisation’. Huaxia Kaogu 3 (2002): 73–77. [陳文:〈香港青銅時代及其分段〉,《華夏考古》,2002年第3期,頁73–77。] Chen, Zhiliang. ‘Investigation on Ancient Shiwan Kiln Sites in Guangdong’. Kaogu 3 (1978): 195–99. [陳智亮(佛山博物館):〈廣東石灣古窰址調查〉,《考古》,1978年第3期,頁195–99。] Cheung, Sidney C. H. ‘The Politics of Wetlandscape: Fishery Heritage and Natural Conservation in Hong Kong’. International Journal of Heritage Studies 17 (1) (2011): 36–45.

Bibliography   243 Chinese Academy of Social Science and the Institute of Cultural Relics and Archaeology of Guangzhou, eds. International Conference on Southern Yue Kingdom, Western Han Archaeology and Han Culture. Beijing: Science Publishing, 2010. [中國社會科學院考古研究所、廣州市文物考古研究所編:《西漢南越國考古與漢文化國際學術研討 會》。北京:科學出版社,2010。]

Chinese Ceramics Society, The, ed. Zhongguo tao ci shi. Beijing: Cultural Relics Publishing House, 1982. [中國硅酸鹽學會編:《中國陶瓷史》。北京:文物出版社,1982。] Chu, Shibin, and Ren Buyun. ‘The Discovery of a Han Site in Juting, Gansu and the Newly Excavated Scripts’. Wenwu 1 (1978): 1–25. [初士賓、任步雲(甘肅居廷考古隊):〈居廷漢代遺址的發掘和新出土的簡冊文物〉,《文物》,1978 年第1期,頁1–25。] Cooney, Gabriel. ‘Social Landscapes in Irish Prehistory’. In The Archaeology and Anthropology of Landscape, edited by P. J. Ucko and R. Layton, 46–64. London: Routledge, 1999. Council of Europe. European Landscape Convention. Florence: Council of Europe, 2000. CPAM of Guangzhou, the Municipal Museum of Guangzhou, and the Institute of Archaeology, Chinese Academy of Social Science, eds. Excavations of the Han Tombs at Guangzhou. Beijing: Cultural Relics Publishing House, 1981. [廣州市文物管理委員會、廣州市博物館、中國社會科學院考古研究所編:《廣州漢墓》。北京:文物 出版社,1981。]

Crawford, James R. ‘Po Yue Wan’. JHKAS XI (1986): 64–79. ———. ‘Southern Sung Coin Hoards from Hong Kong’. JHKAS XI (1986): 80–94. Crown Lands and Survey Office. Lamma Island, D.D.3 Sheet 2. 1:1980 Scale Map. Hong Kong: Crown Lands and Survey Office, 1905. ———. Yung Shue Wan: Lamma Island. 1:1200 Scale Topographic Map Sheet C-229-SW-C / C-246-NW-A. Hong Kong: Crown Lands and Survey Office, 1963. ———. Yung Shue Wan: Lamma Island. 1:1000 Scale Topographic Map Sheet 14-NE-10D. Hong Kong: Crown Lands and Survey Office, 1981a. ———. Yung Shue Wan: Lamma Island. 1:1000 Scale Topographic Map Sheet 14-NE-15B. Hong Kong: Crown Lands and Survey Office, 1981b. Cultural Council of Guangzhou City, ed. Three Major Archaeological Findings of the Qin and Han Periods in Guangzhou. Guangzhou: Guangzhou Publishing, 1999. [廣州市文化局編:《廣州秦漢考古三大發現》。廣州:廣州出版社,1999。] Davis, S. ‘Archaeology on Lamma Island’. JHKAS I (1968): 33–34. Davis, S. G., and Mary Tregear. ‘Man Kok Tsui Archaeological Site 30, Lantau Island, Hong Kong’. Asian Perspectives 4 (1960): 183–212. Dong, Shouxian. ‘Research of Iron Tools of the Han Dynasty’. MA dissertation, Zhengzhou University, 2010. [董守賢:〈漢代鐵質農具研究〉。鄭州大學歷史學碩士論文,2010。] Drewett, Peter L. ‘Neolithic Sha Lo Wan’. Antiquities and Monuments Office Occasional Paper No. 2. London: Archetype, 1995. Empson, Hal. Mapping Hong Kong: A Historical Atlas. Hong Kong: Government Printer, 1992. ERM. ‘Rescue Excavation at Small House Lot No. 1575s.B. in D.D. 3 at Yung Shue Wan, Lamma Island’. Unpublished excavation report, 2005. Feng, Yongju, Zhang Jinguo, and Huang Zhaoqiang. ‘Three Kingdoms Hoarding Site and Tang Dynasty Minting Site in Xihu Road, Guangzhou City’. In Archaeological Discoveries and Research in Guangzhou Vol. I, edited by the Institute of Cultural Relics and Archaeology of Guangzhou, 107–32. Beijing: Wenwu Publishing, 2005. [馮永駒、張金國、黃兆強:〈廣州市西湖路三國錢幣窖藏和唐代鑄幣遺址〉,載廣州市文物考古研究 所編:《羊城考古發現與研究(一)》。北京:文物出版社,2005,頁107–32。]

Field Archaeological Co. ‘Footpath Excavation at Sha Po Tsuen, Lamma Island’. Unpublished excavation report, 2001. [田野考古事務所:〈南丫島沙埔舊村行人小徑考古調查報告〉(未刊報告),2001。] Finn, Daniel J. ‘Archaeological Finds on Lamma Island near Hong Kong, Part I’. The Hong Kong Naturalist 3 (3–4) (1932): 226–46.

244   Bibliography ———. ‘Archaeological Finds on Lamma Island near Hong Kong, Part II’. The Hong Kong Naturalist 4 (1) (1933a): 55–70. ———. ‘Archaeological Finds on Lamma Island near Hong Kong, Part III’. The Hong Kong Naturalist 4 (2) (1933b): 132–55. ———. ‘Archaeological Finds on Lamma Island near Hong Kong, Part IV’. The Hong Kong Naturalist 5 (1) (1934a): 46–53. ———. ‘Archaeological Finds on Lamma Island near Hong Kong, Part V’. The Hong Kong Naturalist 5 (2) (1934b): 123–45. ———. ‘Archaeological Finds on Lamma Island near Hong Kong, Part VI’. The Hong Kong Naturalist 5 (3) (1934c): 197–209. ———. ‘Archaeological Finds on Lamma Island near Hong Kong, Part VII’. The Hong Kong Naturalist 5 (4) (1934d): 282–303. ———. ‘Archaeological Finds on Lamma Island near Hong Kong, Part VIII’. The Hong Kong Naturalist 6 (1) (1935a): 40–61. ———. ‘Archaeological Finds on Lamma Island near Hong Kong, Part IX’. The Hong Kong Naturalist 6 (2) (1935b): 117–31. ———. ‘Archaeological finds on Lamma Island near Hong Kong, Part X’. The Hong Kong Naturalist 6 (3–4) (1935c): 240–71. ———. ‘Archaeological Finds on Lamma Island near Hong Kong, Part XI’. The Hong Kong Naturalist 7 (1) (1936a): 37–60. ———. ‘Archaeological Finds on Lamma Island near Hong Kong, Part XII’. The Hong Kong Naturalist 7 (2) (1936b): 163–83. ———. ‘Archaeological Finds on Lamma Island near Hong Kong, Part XIII’. The Hong Kong Naturalist 7 (3–4) (1936c): 257–68. Frost, R. J. ‘Sha Chau’. JHKAS VI (1975): 37–50. Fu, Xuan, and Zhang Biwu. ‘Excavation of the No. 1 Tomb at Hill Fengxingzuishan in Yueyang City, Hunan’. Wenwu 1 (1993): 17–38. [符炫、張必武(岳陽市文物工作隊):〈湖南省岳陽市郊毛家堰:閻家山周代遺址發掘簡報〉,《文 物》,1993年第1期,頁17–38。]

Fuller, Dorian Q., Nicole Boivin, Tom Hoogervorst, and Robin Allaby. ‘Across the Indian Ocean: The Prehistoric Movement of Plants and Animals’. Antiquity 85 (2011): 544–58. Fyfe, J. A., R. Shaw, S. D. G. Campbell, K. W. Lai, and P. A. Kirk. The Quaternary Geology of Hong Kong. Hong Kong: Geotechnical Engineering Department, 2000. Geotechnical Engineering Office. HGM20 Series Solid and Superficial Geology Sheet 14 Cheung Chau (1st Ed). Hong Kong: Geotechnical Control Office, 1995. Greenwood, Roberta S. ‘Old Approaches and New Directions: Implications for Future Research’. In Hidden Heritage: Historical Archaeology of the Overseas Chinese, edited by Priscilla Wegars, 375–403. New York: Baywood Publishing Company, 1993. Gu, Yunquan, Li Ziwen, and Deng Hongwen. ‘Excavation of the Yingang Site at Boluo in Guangdong’. Wenwu 7 (1998): 17–30. [古運泉、李子文、鄧宏文(廣東省文物考古研究所):〈廣東博羅銀崗遺址發掘簡報〉,《文物》, 1998年第7期,頁17–30。]

Guangdong Provincial Institute of Cultural Relics and Archaeology. ‘Archaeological Investigation of Lamma Island, Cheung Chau, Tai A Chau, Siu A Chau, and Shek Kwu Chau’. Unpublished report, 1997. [廣東省文物考古研究所:〈香港南丫島、長洲島、大鴉洲島、小鴉洲島及石鼓洲島考古調查報告〉 (未刊報告),1997。]

Guangdong Provincial Institute of Cultural Relics and Archaeology, and Zhuhai Museum, eds. Excavation Report on a Prehistoric Island Site in Baojingwan, Zhuhai. Beijing: Science Publishing, 2004. [廣東省文物考古研究所、珠海市博物館編:《珠海寶鏡灣:海島型史前文化遺址發掘報告》。北京: 科學出版社,2004。]

Guangdong Provincial Institute of Cultural Relics and Archaeology, ed. Henglingshan in Boluo County: Excavation Report of Shang and Zhou Period Cemetery in 2000. Beijing: Science Publishing, 2005.

Bibliography   245 [廣東省文物考古研究所編:《博羅橫嶺山:商周時期墓地 2000 年發掘報告》。北京:科學出版 社,2005。]

Guangdong Provincial Institute of Cultural Relics and Archaeology, ed. Burials of the Six Dynasties and the Sui and Tang Dynasties at Zeqiaoshan, Ruyuan. Beijing: Wenwu Publishing, 2006. [廣東省文物考古研究所編:《乳源澤橋山六朝隋唐墓》。北京:文物出版社,2006。] Guangdong Provincial Museum, and Foshan City Museum, eds. Hedang Site in Foshan: Excavation Report from 1977 Winter to 1978 Summer. Guangzhou: Guangdong People’s Publishing House, 2006. [廣東省博物館、佛山市博物館編:《佛山河宕遺址:1977年冬至1978年夏發掘報告》。廣州:廣東人 民出版社,2006。] Guangdong Provincial Museum, and Museum of the Qujiang County, eds. Treatises in the Commemoration of the 30th Anniversary of the Discovery of the Maba Human Cranium. Beijing: Cultural Relics Publishing House, 1988. [廣東省博物館、曲江縣博物館編:《紀念馬壩人化石發現三十周年文集》。北京:文物出版社, 1988。]

Gui, Wencan, ed. ‘Vol. 13 Xinan County’, Guangdong Tu Shuo (Reprint of the late Qing [1862–1874] Tongzhi Cui Wen Tang edition). Yangzhou: Jiangsu Guangling guji keyinshe, 1993. http://goo.gl/7hs6FN Guo, Fei. ‘Coastal Map of Kwang Tung’. In Yue Da Ji. Place and publisher unknown, c. 1598. 郭棐:〈廣東沿海圖〉,《粵大記》。出版地及出版社不詳,約1598。 Hase, Patrick H. ‘Some Notes on the History of Lamma, Especially Yung Shue Wan’. Unpublished paper, 2002. ———. ‘Salt in Hong Kong’. Unpublished paper, 2003. ———. Custom, Land and Livelihood in Rural South China: The Traditional Land Law of Hong Kong’s New Territories, 1750–1950. Hong Kong, Hong Kong University Press, 2013. Hase, Patrick, and Man-yip Lee. ‘Sheung Wo Hang Village, Hong Kong: A Village Shaped by Fengshui’. In Chinese Landscapes: The Village as Place, edited by R. G. Knapp, 79–94. Honolulu: University of Hawai‘i Press, 1992. Hayes, James. ‘The Hong Kong Region: Its Place in Traditional Chinese Historiography and Principal Events Since the Establishment of Hsin-An County in 1573’. JRASHKB 14 (1974): 108–35. ———. The Great Difference: Hong Kong’s New Territories and Its People 1898–2004. Hong Kong: Hong Kong University Press. 2006. Heanley, C. M. ‘Fields of Hong Kong’. Hong Kong Naturalist 6 (2–3) (1935): 233–38. Heanley, C. M., and J. L. Shellshear. 1932. ‘A Contribution to the Prehistory of Hong Kong and the New Territories’. In Praehistorica Asiae Orientalis, 63–76. Hanoi: s.n. Herklots, Geoffrey, A. C. ‘Obituary: The Rev. Father Daniel J. Finn, S.J’. The Hong Kong Naturalist 7(3–4) (1936): 269–70. Higham, Charles, and Tracey Lie-Dan Lu. ‘The Origins and Dispersal of Rice Cultivation’. Antiquity 72 (1998): 867–77. Hill, J. D. Ritual and Rubbish in the Iron Age of Wessex. BAR British Series Vol. 242. Oxford: Tempus Reparatum, 1995. HKU Architecture Department. ‘The Lime Kiln Industry of Sai Kung’. JHKAS VII (1979): 139–40. Ho, Ching-hin. ‘Sha Po Tsuen’. JHKAS IV (1973): 19–20. ———. ‘Preface’. In Collected Essays on the Culture of the Ancient Yue People of South China, edited by Chau Hingwah, 14–15. Hong Kong: Urban Council, 1993. Hong Kong, Colonial Secretariat. A Gazetteer of Place Names in Hong Kong, Kowloon and the New Territories. Second Edition. Hong Kong: Government Printer, 1969. Hong Kong Institute of Archaeology. ‘Working Paper on the Rescue Excavation at Yuen Leng Tsai’. Unpublished excavation report, 2003. [中港考古研究室:〈圓嶺仔遺址搶救發掘工作報告〉(未刊報告),2003。] Huang, Xiaowen. ‘Study on the Tombs of Guangdong Area during Six Dynasties’. MA dissertation, Nanjing University, 2011. [黃筱雯:〈廣東地區六朝墓葬研究〉。南京大學考古學及博物館學碩士論文,2011。] Hung, Hsiao-chun, Mike T. Carson, Peter Bellwood, Fredeliza Z. Campos, Philip J. Piper, Dizon Eusebio, Mary Jane Louise A. Bolunia, Marc Oxenham, and Chi Zhang. ‘The First Settlement of Remote Oceania: The Philippines to the Marianas’. Antiquity 85 (2011): 909–26.

246   Bibliography Institute of Cultural Relics and Archaeology of Guangzhou, ed. Guangzhou Wenwu Kaogu Ji. Beijing: Wenwu Publishing, 1998. [廣州市文物考古研究所編:《廣州文物考古集》。北京:文物出版社,1998。] ———, ed. Archaeological Discoveries and Research in Guangzhou. Beijing: Wenwu Publishing, 2005a. [廣州市文物考古研究所編:《羊城考古發現與研究》。北京:文物出版社,2005a。] ———, ed. Zhu Ji Cun Lei: A Selection of Excavated Finds from Guangzhou in the Last Decade. Beijing: Wenwu Publishing, 2005b. [廣州市文物考古研究所編:《銖積寸累:廣州考古十年出土文物選萃》。北京:文物出版社, 2005b。]

Jen, Yu-Wen. ‘The Travelling Palace of Southern Sung in Kowloon’. JRASHKB 7 (1967): 21–38. Ji, Chengming. ‘Analysis on Several Research Materials Available on Tang Dynasty Salt Production Techniques’. Salt Industry History Research 1 (1993): 44–46. [吉成名:〈有關唐代海鹽生產技術的幾條材料分析〉,《鹽業史研究》1993年第1期,頁44–46。] Jiangxi Provincial Institute of Cultural Relics and Archaeology, and Zhangshu City Museum, eds. Wucheng: 1973– 2002 Excavation Report. Beijing: Science Publishing, 2005. [江西省文物考古研究所、樟樹市博物館編:《吳城:1973–2002年考古發掘報告》。北京:科學出版 社,2005。] Jiang, Linhai, and Zhang Jiuwen. ‘Excavation of Tombs 8, 9, and 10 at Xiangshan near Nanjing’. Wenwu 7 (2000): 4–20. [姜林海、張九文(南京市博物館):〈南京象山8號、9號、10號墓發掘簡報〉,《文物》,2000年第7 期,頁4–20。] Knapp, R. G., ed. Chinese Landscapes: The Village as Place. Honolulu: University of Hawai‘i Press, 1992. Ko, Tim-keung. Hong Kong Now and Then. Hong Kong: Joint Publishing, 2005. [高添強:《香港今昔》(新版)。香港:三聯出版社,2005。] Koike, H. ‘Prehistoric Hunting Pressure and Palaeobiomass: An Environmental Reconstruction and Archaeozoological Analysis of a Jomon Shellmound Area’. In Prehistoric Hunter-Gatherers in Japan: New Research Methods, edited by T. Akazawa and C. M. Aikens, 27–53. Tokyo: University of Tokyo, 1986. Krahl, Regina, John Guy, Julian Raby, and Keith J. Wilson, eds. Shipwrecked: Tang Treasures and Monsoon Winds. Washington D.C.: Smithsonian Institution, 2011. Kuang, Guirong. ‘Preliminary Report on the Excavation of Sui–Tang Tombs at Zhixin Secondary School’. In Archaeological Discoveries and Research in Guangzhou Vol. I, edited by the Institute of Cultural Relics and Archaeology of Guangzhou, 147–60. Beijing: Wenwu Publishing, 2005. [鄺桂榮(廣州市文物考古研究所):〈執信中學隋唐墓發掘簡報〉,載廣州市文物考古研究所編:《羊 城考古發現與研究(一)》。北京:文物出版社,2005,頁147–60。]

Lai, Alice C. S., Anthony C. T. So, Samuel K. C. Ng, and David Jonas. ‘The Territory-Wide Airborne Light Detection and Ranging Survey for the Hong Kong Special Administrative Region’. Paper presented at the 33rd Asian Conference on Remote Sensing, November 2012. Lam, Peter Y. K., ed. Archaeological Finds from Pre-Qin Sites in Guangdong. Hong Kong: CUHK Art Gallery, 1984. ———, ed. Archaeological Finds from the Jin to the Tang Periods in Guangdong. Hong Kong: CUHK Art Gallery, 1985. ———, ed. Archaeological Finds from Five Dynasties to the Qing Periods in Guangdong. Hong Kong: CUHK Art Gallery, 1989. Larsen, Greger, Ranran Liu, Xingbo Zhao, Jing Yuan, Dorian Fuller, Loukas Barton, Keith Dobney, Qipeng Fan, Zhiliang Gu, Xiao-Hui Liu, Yunbing Luo, Peng Lv, Leif Andersson, and Ning Li. ‘Patterns of East Asian Pig Domestication, Migration, and Turnover Revealed by Modern and Ancient DNA’. PNAS 107 (17) (2010): 7686–91. LCSD Central Conservation Section. Analysis Report for Historic Relics. Dated 25 October 2012. Lee, Peter. Opium Culture: The Art and Ritual of Chinese Tradition. Rochester: Park Street Press, 2006. Li, Cebin, and Wu Wei. ‘An Eastern Jin Tomb in Liuli, Ganquan, Hanjiang, Jiangsu’. Southeast Culture 2 (1986): 22–26. [李側斌、吳煒(揚州博物館):〈江蘇邗江甘泉六里東晉墓〉,《東南考古》,1986年第 2期,頁 22– 26。]

Bibliography   247 Li, Hairong. ‘Xiantouling Site in Shenzhen’. In Major Archaeological Discoveries in China 2006, edited by SACH, 34–37. Beijing: Wenwu Publishing, 2007. Li, Hairong, and Liu Junxiong. ‘A Late Neolithic Site in Xiantouling, Shenzhen, Guangdong’. Kaogu 7 (2007): 9–15. [李海榮、劉均雄(深圳市文物考古鑒定所、深圳市博物館):〈廣東深圳市咸頭嶺新石器時代遺址〉, 《考古》,2007年第7期,頁9–15。]

Li, Jinghua. ‘Ancient Iron Agricultural Tools in Henan (Cont’d)’. Agricultural Archaeology 1 (1985): 55–65. [李京華:〈河南古代鐵農具(續)〉,《農業考古》,1985年第1期,頁55–65。] Li, Long-lam. ‘Studies on 5–10 Centuries’ Brine-kilns in Hong Kong’. JHKAS XVI (2003–2008) (2009): 94–103. [李浪林:〈香港南朝至五代煮鹽遺存的研究〉,《香港考古學會會刊》,第16期(2003–2008),2009, 頁94–103。] Li, Meitian. ‘Study on the Phasing of Six Dynasties, Sui, and Tang Green Glazed Ware in Central Yangtze River Region’. Huaxia Kaogu 4 (2000): 83–99. [李梅田:〈長江中游地區六朝隋唐青瓷分期研究〉,《華夏考古》,2000年第4期,頁83–99。] Li, Shui Cheng. ‘Ten Years of Salt Archaeology in China’. Kaogu Xue Yanjiu 9 (2012): 362–80. [李水城:〈中國鹽業考古十年〉,《考古學研究》,2012年第9期,頁362–80。] Li, Yan. ‘The Second Season of Excavation of the Cuntou Site at Dongguan, Guangdong’. Wenwu 9 (2000): 25–34. [李岩(廣東省文物考古研究所、東莞市博物館):〈東莞村頭遺址第二次發掘簡報〉,《文物》,2000 年第9期,頁25–34。] Li, Ziwen. ‘Excavation of Fuchuanling Site in Renhua, Guangdong’. Wenwu 7 (1998): 31–47. [李子文(廣東省文物考古研究所):〈廣東仁化覆船嶺遺址發掘〉,《文物》,1998年第 7期,頁 31– 47。] ———. ‘Preliminary Report on the Excavation of the Shell Midden Site at Yinzhou, Sanshui City, Guangdong’. Kaogu 6 (2000): 24–36. [李子文(廣東省文物考古研究所、北京大學考古學系、三水市博物館):〈廣東三水市銀洲貝丘遺址 發掘簡報〉,《考古》,2000年第6期,頁24–36。]

Liang, Mingshen. ‘Preliminary Report on the Six Dynasties, Sui, and Tang Tombs in Shaoguan, Guangdong’. Kaogu 5 (1965): 230–35. [梁明燊(廣東省文物管理委員會):〈廣東韶關六朝隋唐墓葬清理簡報〉,《考古》,1965年第5期,頁 230–35。] Liao, Jinxiong. ‘Preliminary Report on the Excavation of Jin, Southern Dynasties, and Tang Tombs in Shixing, Guangdong’. Kaogu 2 (1990): 116–22. [廖晉雄(始興縣博物館):〈廣東始興縣晉、南朝、唐墓清理簡報〉,《考古》, 1990 年第 2 期,頁 116–22。] Lin, Zhonggan, and Chen Ziwen et al. ‘Preliminary Report on the Excavation of a Jin Tomb in Luchuwu, Pucheng, Fujian’. Kaogu 10 (1988): 921–28, 960. [林忠幹、陳子文等(福建省博物館、浦城縣文化館):〈福建浦城呂處塢晉墓清理簡報〉,《考古》, 1988年第10期,頁921–28、960。]

Lisowski, F. P. ‘Human Remains and Burials at Sham Wan’. In Sham Wan, Lamma Island: An Archaeological Site Study, edited by William Meacham, 232–46. Hong Kong: Hong Kong Archaeological Society, 1978. Liu, Chengji. ‘Report on the Rescue Excavation for a Small House Development at Sha Po New Village, Yung Shue Wan, Lamma Island’. Unpublished excavation report, 2004. 劉成基(古物古蹟辦事處):〈2004南丫島沙埔新村小型屋宇地考古發掘報告〉(未刊報告),2004。] ———. ‘Excavation on the Ancient Site at Sha Po Tsuen, Lamma Island, Hong Kong’. Kaogu 6 (2007): 10–29. [劉成基(古物古蹟辦事處):〈香港南丫島沙埔新村遺址發掘簡報〉,《考古》, 2007 年第 6 期,頁 10–29。] Liu, Chengji, and Yang Shaoxiang. ‘Excavation on the Kiln Site at Meihuadun, Yuanzhou, Boluo, Guangdong’. Kaogu 7 (1998): 28–43. [劉成基、楊少祥(廣東省文物考古研究所、博羅縣博物館):〈廣東博羅縣圓洲梅花墩窰址的發掘〉, 《考古》,1998年第7期,頁28–43。]

Liu, Mao. ‘The Study of the Siu Lam Kiln Site in Hong Kong’. Unpublished report, 2002. [劉茂:〈香港小欖窰址研究〉(未刊報告),2002。]

248   Bibliography Liu, Shixu, and Zhang Zhengning. ‘Investigation of a Han Dynasty Bronze Smelting Site in Dongpingcun, Xichang, Sichuan’. Kaogu 12 (1990): 1069–75. [劉世旭、張正寧:〈四川西昌市東坪村漢代煉銅遺址的調查〉,《考古》,1990 年第 12 期,頁 1069– 75。] Liu, Xun. Ling Biao Lu Yi (with footnotes and commentary added by Lu Xun between 1911 and 1912). Guangzhou: Guangdong People’s Publishing, 1983. [The supplementary is on pp. 37–40 but the relevant page of salt is p. 37.] [(唐)劉恂(著),魯迅(校勘):〈嶺表錄異補遺〉,《嶺表錄異》。廣州:廣東人民出版社,1983, 頁37–40。] Liu, Zhihua, Zheng Tingyi, Chen Youding, and Xiang Anqiang. ‘Research on the Excavated Ancient Cultivated Rice Species Discovered in Guangdong’. Guangdong Agricultural Sciences 9 (2010): 52–64. [劉志華、鄭庭義、陳友訂、向安強:〈考古發現的廣東栽培稻品種研究〉,《廣東農業科學》,2010年 第9期,頁52–64。] Lo, Hsiang-lin. Hong Kong and Its External Communications Before 1842. Hong Kong: Institute of Chinese Culture, 1963. Lu, Tracey L-D. ‘The Management of Cultural Heritage in Hong Kong’. Hong Kong Institute of Asia-Pacific Studies, Occasional Paper No. 137 (2003): 11–12. ———. ‘Natural Resources and Subsistence Strategies in Prehistoric Hong Kong’. Kaogu 6 (2007a): 36–45. [呂烈丹:〈香港史前的自然資源和經濟形態〉,《考古》,2007a年第6期,頁36–45。] ———. ‘Report on the Date of the Wong Tei Tung Archaeological Assemblage’. Unpublished AMO Archaeological Research Report, 2007b. Lu, Tracey L-D, Zhijun Zhao, and Zhuo Zheng. ‘The Prehistoric and Historic Environments, Vegetations and Subsistence Strategies at Sha Ha, Sai Kung’. In The Ancient Culture of Hong Kong: Archaeological Discoveries in Sha Ha, Sai Kung, edited by AMO, 57–64. Hong Kong: AMO, 2005. Luo, Hongjie, and Li Jiazhi. ‘Preliminary Discussion on the Definition of Proto-porcelain’. Kaogu 7 (1998): 69–72. [羅宏杰、李家治:〈試論原始瓷器的定義〉,《考古》,1998年第7期,頁69–72。] Luo, Xizhe. ‘Excavation of the Living Quarter of Royal Musicians and Bathroom in the Huaqing Palace of the Tang Dynasty’. Wenwu 3 (1999): 25–42. [駱希哲(唐華清宮考古隊):〈唐華清宮梨園、小湯遺址發掘簡報〉,《文物》, 1999 年第 3 期,頁 25–42。] Maglioni, Rafael. Archaeological Discovery in Eastern Kwantung, Journal Monograph II. Hong Kong: HKAS, 1975. Mai, Yinghao. ‘Preliminary Report on the Excavation of a Qin Tomb in Luogang, East of Guangzhou’. Kaogu 8 (1962): 404–6. [麥英豪(廣州市文物管理委員會):〈廣州東郊羅崗秦墓發掘簡報〉,《考古》, 1962 年第 8 期,頁 404–6。] Mai, Yinghao, Li Jin, and Chen Weihan. ‘Summaries on the Two Seasons of Excavations on the Qin Dynasty Shipyard Site’. In Three Major Archaeological Findings of the Qin and Han Periods in Guangzhou, edited by Cultural Council of Guangzhou City, 5–42. Guangzhou: Guangzhou Publishing, 1999. [麥英豪、黎金、陳偉漢:〈秦代造船工場遺址兩次試掘綜述〉,載廣州市文化局編:《廣州秦漢考古三 大發現》。廣州:廣州出版社,1999,頁5–42。]

Meacham, William. ‘Tung Kwu: Phase 4’. JHKAS VI (1975): 55–66. ———, ed. Sham Wan, Lamma Island: An Archaeological Site Study. Hong Kong Archaeological Society Journal Monograph III. Hong Kong: Hong Kong Archaeological Society, 1978. ———. ‘Lo So Shing’. JHKAS VII (1979): 16–26. ———. ‘Pui O’. JHKAS X (1984): 60–71. ———. ‘Hac Sa Wan’. JHKAS XI (1986): 106–8. ———. ‘Sha Po Tsuen’. JHKAS XIII (1993a): 33–54. ———. ‘Report on Salvage Excavations at Lung Kwu Sheung Tan 1990’. JHKAS XIII (1993b): 13–32. ———, ed. Archaeological Investigations on Chek Lap Kok Island. Hong Kong Archaeological Society Journal Monograph IV. Hong Kong: Hong Kong Archaeological Society, 1994. ———. ‘Report on an Archaeological Survey and Excavation of Northern Kau Sai Chau’ (on behalf of the Hong Kong Archaeological Society, December 1993–February 1994). Unpublished report, 1995a.

Bibliography   249 ———. ‘Middle and Late Neolithic at “Yung Long South”’. In Conference on Archaeology in Southeast Asia, edited by C. T. Yeung and W. L. Li, 467–78. Hong Kong: University Museum and Art Gallery, the University of Hong Kong, 1995b. ———. The Archaeology of Hong Kong. Hong Kong: Hong Kong University Press, 2009a. ———. Rock Carvings in Hong Kong. Hong Kong: William Meacham, 2009b. Medley, Margaret. The Chinese Potter. Oxford: Phaidon, 1989. Meskell, Lynn, and Robert W. Preucel, eds. A Companion to Social Archaeology. Oxford: Blackwell, 2004. Mo, Zhi. ‘New Findings on Guangdong Archaeological Investigations and Excavations’. Kaogu 12 (1961): 666–68. [莫稚:〈廣東考古調查發掘的新收穫〉,《考古》,1961年第12期,頁666–68。] ———. ‘North District Sewerage, Stage 2 and Stage 3: Archaeological Investigation’. Unpublished report, 2001. [莫稚:〈新界北區污水渠第二、三工程考古調查工作報告〉(未刊報告),2001,參考編號ND20。] Mo, Zhi, and Chen Zhiliang. ‘A Neolithic Shell Midden Site in Dongxing, Guangdong’. Kaogu 12 (1961): 644–49, 688. [莫稚、陳智亮:〈廣東東興新石器時代遺址貝丘遺址〉,《考古》,1961年第12期,頁644–49、688。] Mo, Zhi, and Li Ziwen. ‘Archaeological Excavations on Hok Chau, Mong Tseng Wai’. JHKAS XV (1999–2002) (2002): 1–17. [莫稚、李子文:〈輞井鶴洲嶺遺址發掘報告〉,《香港考古學會會刊》,第15期(1999–2002),2002, 頁1–17。] Munsell Color. Munsell Soil-Color Charts. Grand Rapids, MI: Munsell, 2009. Ng, Steven W. H. ‘The Spatial Pattern of Prehistoric Sites around Estuary of Pearl River’. JHKAS XIV (1993–1997) (1998): 41–60. [吳偉鴻:〈珠江口史前遺址分佈規律〉,《香港考古學會會刊》,第 14 期( 1993–1997 ), 1998 ,頁 41–60。] Ng, Steven W. H., Ngan Kong-zhou, Chou Peter K. W., Wong Fu, Chan Kwan-shan, and Cheng Kai-ming. ‘Archaeology Investigation at Lam Tsuen Valley, Hong Kong’. JHKAS XV (1999–2002) (2002): 33–47. [吳偉雄、顏孔超、周國偉、黃虎、陳君山、鄭啟明:〈林村谷考古調查報告〉,《香港考古學會會 刊》,第15期(1999–2002),2002,頁33–47。]

Palmer, Michael J. E. ‘The Surface-Subsoil Form of Divided Ownership in Late Imperial China: Some Examples from the New Territories of Hong Kong’. Modern Asian Studies 21 (1) (1987): 1–119. Pang, Quanmin, Huang Wenming, Huang Xiaohong, and Feng Yongju. ‘Preliminary Excavation Report on the Xiantouling Sand Dune Site in Dapeng, Shenzhen City’. In Archaeological Discovery and Research of Shenzhen, edited by Shenzhen Museum, 28–41. Beijing: Wenwu Publishing, 1994. [彭全民、黃文明、黃小宏、馮永駒:〈深圳市大鵬咸頭嶺沙丘遺址發掘簡報〉,載深圳博物館編:《深 圳考古發現與研究》。北京:文物出版社,1994,頁28–41。]

Peacock, Brian A. V., and Taryn J. P. Nixon. The Hong Kong Archaelogical Survey: Subsurface Investigation Reports. Hong Kong: AMO, 1986. Philip J. Piper, Hsiao-chun Hung, Fredeliza Z. Campos, Peter Bellwood, and Rey Santiago. ‘A 4000-Year-Old Introduction of Domestic Pigs into the Philippine Archipelago: Implications for Understanding Routes of Human Migration through Island Southeast Asia and Wallacea’. Antiquity 83 (2009): 687–95. Potter, Jack M. Capitalism and the Chinese Peasant: Social and Economic Change in a Hong Kong Village. Berkeley and Los Angeles: University of California Press, 1968. Qiao, Xiaoqin. ‘The Maritime Expansion of Yue Culture and the Related Archaeological Evidences’. In Collected Essays on the Culture of the Ancient Yue People in South China, edited by Chau Hing-wah, 34–39. Hong Kong: Urban Council, 1993. [喬曉勤:〈古越族向海洋的拓展及考古證據〉,載鄒興華編:《嶺南古越族文化論文集》。香港:香港 市政局,1993,頁34–39。]

Qian, Xianhe, and Fang Jianneng, eds. Prehistoric Jade Carving Techniques. Taipei: National Taiwan Museum, 2003. [錢憲和、方建能編:《史前琢玉工藝技術》。臺北:國立臺灣博物館,2003。] Qiu, Licheng. ‘An Eastern Jin Tomb at Pingshigang of Zhaoqing City, Guangdong’. In Huanan Archaeology, edited by the Guangdong Provincial Institute of Cultural Relics and Archaeology and Guangzhou City Institute of Cultural Relics and Archaeology, 248–64. Beijing: Wenwu Publishing, 2004.

250   Bibliography [邱立誠(廣東省文物考古研究所、肇慶市博物館):〈廣東肇慶市坪石崗東晉墓〉,載廣東省文物考 古研究所、廣州市文物考古研究所、深圳博物館編:《華南考古》。北京:文物出版社, 2004,頁 248–64。]

Qiu, Licheng, and Gu Yunquan. ‘Eastern Zhou, Qin and Han tombs in Lechang City, Duimianshan, Guangdong’. Kaogu 6 (2000): 37–61. [邱立誠、古運泉(廣東省文物考古研究所、樂昌市博物館、韶關市博物館):〈廣東樂昌市對面山東 周秦漢墓〉,《考古》,2000年6期,頁37–61。]

Qiu, Licheng, and Liu Chengji. ‘Preliminary Findings of Cuntou Site in Dongguan’. Journal of Guangdong Provincial Museum 2 (1991): 70–73. [邱立誠、劉成基:〈東莞村頭遺址發掘的初步收獲〉,《廣東省博物館館刊》, 1991年第2期,頁70– 73。] Renfrew, Colin, and Paul Bahn. Archaeology: Theories, Methods, and Practice. London: Thames and Hudson, 2012. Robinson, Tony, and Mick Aston. Archaeology Is Rubbish. London: Channel 4 Books, 2002. Rogers, Pamela R. ‘Tung Lung Fort: A Qing Dynasty Fortification on Tung Lung Island, Hong Kong’. Unpublished excavation report, 1985. Rogers, Pamela R., Nan W. Leininger, S. Mirchandani, Julie van den Bergh, and Ellen A. Widdowson. ‘Tung Wan Tsai: A Bronze Age and Han Period Coastal Site’. AMO Occasional Paper No. 3. Hong Kong: AMO, 1995. Rong, Daxian. ‘Preliminary Report on the Excavation of Dieshishan Site in Shenzhen’. In The Archaeological Discoveries and Research of Shenzhen, edited by Shenzhen Museum, 42–50. Beijing: Wenwu Publishing, 1994. [容達賢:〈深圳市疊石山遺址發掘簡報〉,載深圳博物館編:《深圳考古發現與研究》。北京:文物出 版社,1994,頁42–50。]

Rössler, Mechtild. ‘World Heritage Cultural Landscapes: A UNESCO Flagship Programme 1992–2006’. Landscape Research 31 (4) (2006): 333–53. Sando, R. A., and D. L. Felton. ‘Inventory Records of Ceramics and Opium from a Nineteenth Century Chinese Store in California’. In Hidden Heritage: Historical Archaeology of the Overseas Chinese, edited by Priscilla Wegars, 151–76. Amityville, NY: Baywood Publishing, 1993. Sas, Robert J., Jr., Jason P. Yu, Carol Y. Y. Pau, and Kevin A. Styles. ‘Detection of Old Agricultural Terraces in Steep, Vegetated Terrain Suing Airborne LIDAR: Case Studies from Hong Kong’. Paper presented at the 33rd Asian Conference on Remote Sensing, November 2012. Sauer, Carl O. ‘The Morphology of Landscape’. In Land and Life: A Selection of the Writings of Carl Ortwin Sauer, edited by J. Leighly, 315–50. Berkeley: University of California Press, 1965 [1925]. Schofield, Walter. An Archaeological Site at Shek Pik: Excavation Report and Related Papers by Walter Schofield (1888–1968). Journal Monograph I. Hong Kong: Hong Kong Archaeological Society, 1975. Shang, Jie. ‘The Discovery of the Mount Shixiong Ancient City Site in Wuhua, Guangdong Province and Its Initial Research’. South-East Culture 1 (2013): 23–30.

[尚杰(廣東省文物考古研究所):〈廣東五華獅雄山秦漢城址的發現與初步研究〉,《東南文化》, 2013年第1期,頁23–30。]

Shang Zhitan. ‘Preface’. In Collected Essays on the Culture of the Ancient Yue People of South China, edited by Chau Hing-wah, 10–11. Hong Kong: Urban Council, 1993. ———. Collected Essays on Hong Kong Archaeology. Beijing: Wenwu Publishing, 2000. [商志(香覃):《香港考古論集》。北京:文物出版社,2000。] Shang, Zhitan, and Chen Shilong. ‘Characteristics and Problems of the Prehistoic Dune Sites near the Mouth of the Zhujiang River’. Wenwu 11 (1990): 48–52. [商志(香覃)、諶世龍:〈環珠江口史前沙丘遺址的特點及有關問題〉,《文物》,1990年第11期,頁 48–52。] Shenzhen Cultural Heritage Management Office, Shenzhen Museum, and Shenzhen Municipal Institute of Cultural Relics and Archaeology, eds. 7000 Years of Shenzhen: Finds Catalogue. Beijing: Wenwu Publishing, 2006. [深圳市文物管理委員會辦公室、深圳博物館、深圳市文物考古鑒定所編:《深圳7000年:深圳出土文 物圖錄》。北京:文物出版社,2006。] Shenzhen Municipal Institute of Archaeology, ed. Xiantouling Site in Shenzhen: Report on the Excavation in 2006. Beijing: Cultural Relics Press, 2013. [深圳市文物考古鑒定所編:《深圳咸頭嶺:2016年發掘報告》。北京:文物出版社,2013。] Shenzhen Museum. Archaeological Discovery and Research of Shenzhen. Beijing: Wenwu Publishing, 1994.

Bibliography   251 [深圳博物館編:《深圳考古發現與研究》。北京:文物出版社,1994。]

Sinn, Elizabeth. Pacific Crossing: California Gold, Chinese Migration, and the Making of Hong Kong. Hong Kong: Hong Kong Univeristy Press, 2012. Siu, Anthony K. K. ‘A Study of the Ch’ing Forts on Lantau Island’. JRASHKB 19 (1979): 195. Song, Yingxing. ‘Salt Production’. In The Heavenly Creations, Volume 1, 99–125. Shanghai: Guoxue zhenglishe: Shijie shuju yinxing, 1936. [(明)宋應星:〈卷上:作鹹第五〉,《天工開物》。上海:國學整理社:世界書局印行, 1936 ,頁 99–125。] Spry, Nigel. ‘Sha Po Tsuen’. JHKAS XII (1990): 7–28. Stokes, Edward. Hong Kong’s Wild Places. Hong Kong: Oxford University Press, 1995. Styles, K. A., and M. H. Law. ‘Some Observations about Man-Made Features on Natural Terrain in Hong Kong’. Quarterly Journal of Engineering Geology and Hydrology 45 (2012): 131–38. Sun, Kevin T. W. 2002. ‘Research on the Prehistoric Remains in the Pearl River Delta’. MA dissertation, Peking University, 2002. [孫德榮:〈珠江三角洲地區史前遺存的研究〉。北京大學考古學與博物館學碩士論文,2002。] ———. ‘Recent Archaeological Discoveries and Studies in Hong Kong’. Kaogu 6 (2007): 3–9. [孫德榮(古物古蹟辦事處):〈香港近年的考古發現與研究〉,《考古》,2007年第6期,頁3–9。] Sun, Tianjian. ‘The Invention of Proto-porcelain and Its Significance as a Milestone’. China Ceramics 39 (3) (2003): 60–62. [孫天健:〈原始瓷器的發明及其里程碑意義〉,《中國陶瓷》,2003年第39卷第3期,頁60–62。] Survey and Mapping Office. Explanatory Notes on Geodetic Datums in Hong Kong. Hong Kong: Lands Department, 1995. ———. 1:1000 Topographic Map, Sheet No.14-NE-10D. Hong Kong: Lands Department, 2015a. ———. 1:1000 Topographic Map, Sheet No.14-NE-15B. Hong Kong: Lands Department, 2015b. Tang, Chung. A Journey into Hong Kong’s Archaeological Past. Hong Kong: Regional Council, 1991. ———. ‘Preliminary Report and Submission of the Pa Tau Kwu Site Excavation’. Unpublished excavation report, 1992. ———. ‘Study of the Bronze Dagger with Bow-Shaped Guard and Human Mask Design Discovered at Shek Pik of Lantau Island in Hong Kong’. In Collected Essays on the Culture of the Ancient Yue People in South China, edited by Chau Hing-wah, 86–101. Hong Kong: Urban Council, 1993. [鄧聰:〈香港石壁出土人面弓形格銅劍試釋〉,載鄒興華編:《嶺南古越族文化論文集》。香港:香港 市政局,1993,頁86–101。]

———. ‘Investigation Report on the 1994 Pa Tau Kwu Areas I and II Excavation’. Unpublished excavation report, 1994a. 鄧聰:〈1994年扒頭鼓遺址I、II區發掘,調查工作報告〉(未刊報告),1994a。 ———, ed. Ancient Cultures of South China and Neighboring Regions: Essays in Honor of Professor Cheng Te-k’un on the Occasion of the Sixtieth Anniversary of His Academic Career. Hong Kong: Chinese University Press, 1994b. [鄧聰編:《南中國及鄰近地區古文化研究:慶祝鄭德坤教授從事學術活動六十週年論文集》。香港: 中文大學出版社,1994b。]

———. ‘New Discoveries of Hong Kong’s Ancient History’. Historical Research 3 (1997a): 32–52. [鄧聰:〈古代香港歷史的新發現〉,《歷史研究》,1997a年第3期,頁32–52。] ———. ‘Notes on Geomorphology of the Tai Wan Culture in the Circum-Pearl River Delta’. Tropical Geography 17 (2) (1997b): 179–83. ———. ‘Prehistoric Tubular Boring Technique in East Asia’. In Prehistoric Jade Carving Techniques, edited by Qian Xianhe and Fang Jianneng, 145–56. Taipei: National Taiwan Museum, 2003. [鄧聰:〈東亞史前玉器管鑽技術試釋〉,載錢憲和、方建能編:《史前琢玉工藝技術》。臺北:國立臺 灣博物館,2003,頁145–56。]

———. ‘A Study of Maritime Diffusion of Prehistorical Mongoloid’. Southeast Culture 11 (2000): 6–22. ———. ‘Archaeology and the Origins of Hong Kong’. New Asia Life Monthly 32 (6) (2005): 6–8. [鄧聰:〈考古與香港尋根〉,《新亞生活月刊》,2005年第32卷第6期,頁6–8。] ———. Forget Me Not: The Historical Roots of Hong Kong (II): Origins of Clothes – Barkcloth. Hong Kong: Centre for Chinese Archaeology and Art, 2011.

252   Bibliography Tang, Chung, and Au Ka-fat, eds. Archaeological Finds from the Pearl River Delta in Guangdong, China. Hong Kong: Chinese University Press, 1991. [鄧聰、區家發(編):《環珠江口史前文物圖錄》。香港:中文大學出版社,1991。] Tang, Chung, and Cheng Wai-ming. Relatório das Escavações Arqueológicas de Hac Sá, Coloane, Macau. Hong Kong: Chinese University Press and Macao Foundation, 1996. [鄧聰、鄭煒明:《田野考古報告專刊(一)澳門黑沙》。香港:中文大學出版社及澳門基金會, 1996。]

Tang, Chung, Shang Zhitan, and Wong Wan-cheung. ‘Preliminary Report on the Excavation of Pak Mong Site on Lantau Island’. Kaogu 6 (1997): 54–64. [鄧聰、商志(香覃)、黃韻璋:〈香港大嶼山白芒遺址發掘簡報〉,《考古》, 1997年第6期,頁54– 64。] Taylor, Kenneth. ‘Cultural Landscapes and Asia: Reconciling International and Southeast Asian Regional Values’. Landscape Research 34 (1) (2009): 7–31. Tsang, Cheng-hwa, Li Kuang-Ti, and Zhu Zhengyi. 2006. Archaeological Discoveries in Nanke. Tainan: Tainan Cultural Council. [臧振華、李匡悌、朱正宜:《先民履跡:南科考古發現專輯》。臺南:臺南縣政府文化局,2006。] Tsang, Steve. A Modern History of Hong Kong. Hong Kong: Hong Kong University Press, 2004. Ucko, P. J., and R. Layton, eds. The Archaeology and Anthropology of Landscape. London: Routledge, 1999. Wang, Changsui, Li Wenjing, and Chen Yue. ‘The Re-exploration of the Concept of “Proto-porcelain” and the Origin of the Celadon’. Kaogu 9 (2014): 86–92. [王昌燧、李文靜、陳岳:〈「原始瓷器」概念與青瓷起源再討論〉,《考古》,2014年第 9期,頁 86– 92。] Wang, Di. ‘Analysis of Spindle-Whorl: From Neolithic Age to Bronze Age in Shandong Province’. MA dissertation, Shandong University, 2009. [王迪:〈新石器時代至青銅時代山東地區紡輪淺析〉,山東大學考古及博物館學碩士論文,2009。] Wang, Rui, and Tong Weihua. ‘Excavation of the Shang Dynasty City Site at Nanguan Gucheng, Yuanqu, Shanxi in 1988–89’. Wenwu 10 (1997): 12–29. [王睿、佟偉華(中國歷史博物館考古部):〈 1988–89 山西垣曲古城南關商代城址發掘簡報〉,《文 物》,1997年第10期,頁12–29。] Wang, Shizhen, and Wang Shancai. ‘Report on the Excavation of an Eastern Han Tomb in Dongchengqu, Suizhou, Hubei’. Wenwu 7 (1993): 52–65. [王世振、王善才:〈湖北隨州東城區東漢墓發掘報告〉,《文物》,1993年第7期,頁52–65。] Watson, James L. Emigration and the Chinese Lineage: The Mans in Hong Kong and London. Berkeley: University of California Press, 1975. Watson, Rubie S. ‘Fengshui, Landscape, and History in Rural Hong Kong’. Symbols, Spring (2007): 3–7. Watt, James. ‘Excavations at Yung Shu Wan, November 1967’. JHKAS I (1968): 36. ———. ‘A Brief Report on Sung-Type Pottery Finds in Hong Kong’. JRASHKB 11 (1971): 142–50. Wegars, Priscilla, ed. Hidden Heritage: Historical Archaeology of the Overseas Chinese. Amityville: Baywood Publishing, 1993. ———, ed. ‘Asian American Comparative Collection: Artifact Illustations’. http://webpages.uidaho.edu/aacc/illus. htm. Accessed 14 August 2010. Wei, Renyi, and Tong Xianren. ‘Sui–Tang Tombs in Qinzhou, Guangxi Zhuang Autonomous Region’. Kaogu 3 (1984): 249–63. [韋仁義、佟顯仁(廣西壯族自治區文物工作隊):〈廣西壯族自治區欽州隋唐墓〉,《考古》,1984年 第3期,頁249–63。] Wen, Benheng, and Rong Daxian. ‘Preliminary Report on the Excavation of a Southern Dynasties Tomb in Baoan, Shenzhen’. Wenwu 11 (1990): 39–43. [文本亨、容達賢(深圳博物館):〈廣東深圳寶安市南朝墓發掘簡報〉,《文物》,1990年第11期,頁 39–43。] Williams, Bernard. ‘Hai Dei Wan’. JHKAS VIII (1980): 27–51. Wong, J. C. F., and R. Shaw. High-Level Coastal Deposits in Hong Kong, GEO Report No. 243. Hong Kong: Geotechnical Engineering Office, 2009.

Bibliography   253 Wong, Sharon W. Y. ‘A Preliminary Study of the Distribution and Consumption of Ceramics in Hong Kong during the Song–Yuan Period’. Bulletin of the IPPA 26 (2006): 140–46. ———. ‘A Preliminary Study of the Ceramics of the Song–Yuan Period Unearthed from Hong Kong’. Kaogu 6 (2007): 46–58. [黃慧怡:〈香港出土宋元瓷器的初步研究〉,《考古》,2007年第6期,頁46–58。] Wong, Wan-cheung. ‘Discussion of the Slit Ring Manufacturing Technology in the Pearl River Delta: A Case Study of Pak Mong Site, Hong Kong’. MA dissertation, Xiamen University, 2009. [黃韻璋:〈環珠江口玦飾製作工藝探討:以香港白芒遺址為例〉,廈門大學考古學與博物館學碩士論 文,2009。]

Wu, Rongqing. ‘Excavation of the Dayishan Site at Guanyun, Jiangsu in 1986’. Wenwu 7 (1991): 10–27. [吳榮清(南京博物館、連雲港市博物館、灌雲縣博物館):〈江蘇灌雲大伊山遺址 1986年的發掘〉, 《文物》,1991年第7期,頁10–27。] Xie, Mingliang. ‘A Few Queries on the Meanings of Guan and Xinguan Inscriptions on White Glazed Porcelain’. The National Palace Museum Research 5 (2) (1987): 1–14. [謝明良:〈有關官和新官款白瓷官字涵義的幾個問題〉,《故宮學術季刊》, 1987 年卷 5 第 2 期,頁 1–14。] Xu, Hengbin. ‘Problems and Suggestions Related to Archaeological Research of Sandbar Sites, and Discussion on the Dating of the Stone Yazhang Unearthed from Tai Wan, Lamma Island’. In Archaeology in Southeast Asia, edited by Yeung Chun-tong and Li Wai-ling, 241–48. Hong Kong: The University Museum and Art Gallery, the University of Hong Kong, 1995. [徐恆彬:〈沙丘遺址發掘研究的問題和建議兼談南丫島大灣遺址的「牙璋」時代問題〉,載楊春棠、 李惠玲編:《東南亞考古論文集》。香港:香港大學美術博物館,1995,頁241–48。]

Yang, Cong. ‘Iron Agricultural Tools Excavated from a Han Site in Chengcun, Suian, Fujian’. Agricultural Archaeology 1 (1990): 262–65. 楊琮:〈福建祟安城村漢代城址出土的鐵農具〉,《農業考古》,1990年第1期,頁262–65。 Yang, Hao, and Yang Yaolin. ‘Preliminary Report on the Excavation of Honghuayuan Han Tomb in Nantou, Shenzhen’. Wenwu 11 (1990): 29–38. [楊豪、楊耀林(廣東省博物館、深圳博物館):〈深圳市南頭紅花園漢墓發掘簡報〉,《文物》,1990 年第11期,頁29–38。] Yang, J. K., and V. R. Hellmann. ‘What’s in the Pot? An Emic Study of Chinese Brown Glazed Stoneware’. Proceedings for the Society for California Archaeology 11 (1998): 59–66. Yang, Jianjun. ‘Issues on Burial Customs of Shang and Zhou Dynasties in the Lingnan Area’. Jianghan Kaogu 2 (2010): 71–81. [楊建軍:〈嶺南商周時期埋葬習俗的若干問題〉,《江漢考古》,2010年第2期,頁71–81。] Yang, Kuan. Development History of China’s Ancient Smelting Technology. Shanghai: Shanghai People’s Publishing, 1982. [楊寛:《中國古代冶鐵技術發展史》。上海:上海人民出版社,1982。] Yang, Shiting. ‘A Few Questions on the Early Iron Tools in Guangdong’. Kaogu 2 (1977): 97–106. [楊式梃:〈關於廣東早期鐵器的若干問題〉,《考古》,1977年第2期,頁97–106。] ———. ‘Preliminary Discussion on Yazhang from Guangdong and Hong Kong: Sings of South-Bound Transmitting of Xia Shang Zhou Culture’. In Cultures of South China and Neighboring Regions: Essays in Honor of Professor Cheng Te-k’un on the Occasion of the Sixtieth Anniversary of His Academic Career, edited by Tang Chung, 173–82. Hong Kong: Chinese University Press, 1994. [楊式梃:〈淺說粵港牙璋及相關器物:夏商周文化南傳跡象採微〉,載鄧聰編:《南中國及鄰近地區 古文化研究:慶祝鄭德坤教授從事學術活動六十週年論文集》。香港:中文大學出版社,1994,頁 173–82。]

Yang, Xiaoyan, Huw J. Barton, Zhiwei Wan, Quan Li, Zhikun Ma, Mingqi Li, Dan Zhang, and Jun Wei. ‘Sago-Type Palms Were an Important Plant Food Prior to Rice in Southern Subtropical China’. PLoS One 8 (5) (2013): e63148. http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0063148 Ye, Hongming, and Cao Heming. ‘Views on the Origin of Chinese Porcelain’. Wenwu 10 (1978): 84–87. [葉宏明、曹鶴鳴:〈關於我國瓷器起源的看法〉,《文物》,1978年第10期,頁84–87。] Ye, Yang, Shi Hongwei, Liu Junxiong, Huang Xiaohong, and Gao Aiping. ‘Excavation of Xiangnancun Site in Nanshan, Shenzhen’. Kaogu 6 (1997): 77–86.

254   Bibliography [葉楊、史紅衛、劉均雄、黃小宏、高愛萍(深圳市文管會辦公室、深圳市博物館、南山區文管會辦公 室):〈深圳市南山向南村遺址的發掘〉,《考古》,1997年第6期,頁77–86。]

Yeung, Chun-tong, and Li Wai-ling, eds. Archaeology in Southeast Asia. Hong Kong: The University Museum and Art Gallery, the University of Hong Kong, 1995. [楊春棠、李惠玲編:《東南亞考古論文集》。香港:香港大學美術博物館,1995。] Yi, Xibing. ‘A Han to Six Dynasties and Tang Song Site in Huangjin Plaza, Zhongshan Liulu, Guangzhou’. In Archaeological Discoveries and Research in Guangzhou Vol. I, edited by the Institute of Cultural Relics and Archaeology of Guangzhou, 202–42. Beijing: Wenwu Publishing, 2005. [易西兵(廣州市文物考古研究所):〈廣州市中山六路黃金廣場漢六朝唐宋遺址〉,載廣州市文物考古 研究所編:《羊城考古發現與研究(一)》。北京:文物出版社,2005,頁202–42。]

Yi, Xibing, and Ma Jianguo. ‘Preliminary Report on the Excavation of Southern Dynasties Tombs at Zhongxing Primary School, Taojin Road’. In Archaeological Discoveries and Research in Guangzhou Vol. I, edited by the Institute of Cultural Relics and Archaeology of Guangzhou, 133–46. Beijing: Wenwu Publishing, 2005. [易西兵、馬建國(廣州市文物考古研究所):〈廣州市淘金東路中星小學南朝墓發掘簡報〉,載廣州市 文物考古研究所編:《羊城考古發現與研究(一)》。北京:文物出版社,2005,頁133–46。]

You, Xiaolei. ‘The Preliminary Study on Henglingshan Cemetery and the Relative Problems’. MA dissertation, Shandong University, 2011. [游曉蕾:〈橫嶺山墓地及相關問題初步研究〉,山東大學碩士論文,2011。] Young, S. A Gazetteer of Placenames in Hong Kong, Kowloon and the New Territories. Second Edition. Hong Kong: Colonial Secretariat, 1969. Zhang, Chi, and Hsiao-Chun Hung. ‘The Neolithic of Southern China: Origin, Development and Dispersal’. Asian Perspectives 47 (2) (2008): 299–329. Zhang, Senshui, and Steven W. H. Ng, eds. 2006 The Important Discovery of Hong Kong Archaeology: Wong Tei Tung Palaeolithic Site. Hong Kong: China Review Academic Publishers, 2010. [張森水、吳偉鴻編:《2006年香港考古重大發現:西貢黄地峒舊石器時代晚期遺址》。香港:中國評 論學術出版有限公司,2010。] Zhang, Yan. ‘A Preliminary Study on Song Dynasty Currency and Its Culture’. MA dissertation, Minszu University of China, 2006. 張燕:〈試論宋代貨幣與錢幣文化〉,中央民族大學碩士論文,2006。 Zhang, Yue. ‘A Preliminary Research on the Ink-Stone during the Tang and Song Dynasties’. MA dissertation, Jilin University, 2013. [張悅:〈唐宋時期出土硯台初步研究〉,吉林大學歷史學碩士論文,2013。] Zhang, Zengqi. ‘Tile-Ends with Characters of Nanzhao and Dali State’. Wenwu 7 (1986): 39–46. [張增祺:〈南詔、大理國時期的有字瓦:兼談白族歷史上有無「白文」的問題〉,《文物》,1986年第 7期,頁39–46。] Zhang, Rongfang, and Huang Miaozhang. History of Southern Yue Kingdom (Second Edition), edited by the Lingnan Wenku Editorial Committee and the Guangdong Association of Chinese Culture. Guangzhou: Guangdong People’s Publishing, 1995. [張榮芳、黄淼章(嶺南文庫編輯委員會、廣東中華民族文化促進會合編):《南越國史》。廣州:廣東 人民出版社,1995。]

Zhou, Xun, and Gao Chunming. Chinese Women’s Ornaments through the Ages. Hong Kong: Joint Publishing and Shanghai: Xuelin Publishing, 1998. [周汛、高春明:《中國歷代婦女妝飾》。香港:三聯書店(香港);上海:學林出版社,1988。] Zhu, Feisu. ‘Excavation and Phasing of the Shixia Site’. A Collection of Studies on Archaeology 10 (2013): 595–602. [朱非素:〈石峽遺址發掘與分期〉,《考古學研究》,2013年第10期,頁595–602。] Zhu, Feisu, Peng Ruce, and Liu Chengde. ‘On Geometric Pottery from Shixia, Maba’. Wenwu Jikan III (1981): 225–43. [朱非素、彭如策、劉成德:〈談談馬壩石硤遺址的幾何印紋陶〉,《文物集刊》, 1981 年第 3 期,頁 225–43。] Zhu, Hairen. ‘Research on Han Archaeology in Hong Kong’. In International Conference on Southern Yue Kingdom, Western Han Archaeology and Han Culture, edited by the Chinese Academy of Social Science and the Institute of Cultural Relics and Archaeology of Guangzhou, 43–62. Beijing: Science Publishing, 2010.

Bibliography   255 [朱海仁:〈香港漢代考古發現與研究〉,載中國社會科學院考古研究所、廣州市文物考古研究所編: 《西漢南越國考古與漢文化國際學術研討會》。北京:科學出版社,2010,頁43–62。]

Internet Sources AMO 2011. Hong Kong Archaeological Archive System: Hai Dei Wan. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail.jsp?artefactID=69&lang=1. Accessed 2 September 2013. AMO 2011. Hong Kong Archaeological Archive System: Kellett Island. http://hkaas.lcsd.gov.hk/hkaas/artefactlist.jsp. Accessed 20 August 2012. AMO 2011. Hong Kong Archaeological Archive System: Lung Kwu Chau. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail.jsp?artefactID=46153&lang=1. Accessed 11 April 2014. AMO 2011. Hong Kong Archaeological Archive System: Man Kok Tsui. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail.jsp?artefactID=6140&lang=1. Accessed 11 April 2014. AMO 2011. Hong Kong Archaeological Archive System: Man Kok Tsui, Lantau Island. http://hkaas.lcsd.gov.hk/ hkaas/artefactdetail.jsp?artefactID=6227&lang=1. Accessed 11 April 2014. AMO 2011. Hong Kong Archaeological Archive System: So Kwun Wat. http://hkaas.lcsd.gov.hk/hkaas/artefactlist.jsp. Accessed 20 August 2012. AMO 2011. Hong Kong Archaeological Archive System: So Kwun Wat. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail.jsp?artefactID=46583&lang=1. Accessed 11 April 2014. AMO 2011. Hong Kong Archaeological Archive System: Tai Long Wan, Lantau Island. http://hkaas.lcsd.gov.hk/ hkaas/artefactdetail.jsp?artefactID=46519&lang=1. Accessed 11 April 2014. AMO 2011. Hong Kong Archaeological Archive System: Tung Wan Tsai. http://hkaas.lcsd.gov.hk/ hkaas/artefactdetail.jsp?artefactID=69&lang=1 and http://hkaas.lcsd.gov.hk/hkaas/artefactdetail. jsp?artefactID=424&lang=1. Accessed 2 September 2013. AMO 2011. Hong Kong Archaeological Archive System: Yung Long North. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail.jsp?artefactID=4930&lang=1. Accessed 11 April 2014. AMO 2014. Hai Dei Wan lid viewed on AMO website. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail. jsp?artefactID=101&lang=1. Accessed 4 April 2014. AMO Hong Kong Archaeological Archive System: Fu Tei, Chek Lap Kok, Lantau Island. http://hkaas.lcsd.gov.hk/ hkaas/artefactdetail.jsp?artefactID=3874&lang=1. Accessed 10 April 2014. AMO Hong Kong Archaeological Archive System: Fu Tei Wan, Chek Lap Kok. http://hkaas.lcsd.gov.hk/hkaas/ artefactdetail.jsp?artefactID=3921&lang=1. Accessed 11 April 2014. AMO Hong Kong Archaeological Archive System: Kwo Lo Wan. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail. jsp?artefactID=4005&lang=1. Accessed 20 August 2012. AMO Hong Kong Archaeological Archive System: Kwo Lo Wan, Lantau Island. http://hkaas.lcsd.gov.hk/hkaas/ artefactdetail.jsp?artefactID=4037&lang=1. Accessed 10 April 2014. AMO Hong Kong Archaeological Archive System: Man Kok Tsui. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail. jsp?artefactID=6179&lang=1. Accessed 20 August 2012. AMO Hong Kong Archaeological Archive System: Man Kok Tsui, Lantau Island. http://hkaas.lcsd.gov.hk/hkaas/ artefactdetail.jsp?artefactID=6158&lang=1. Accessed 10 April 2014. AMO Hong Kong Archaeological Archive System: Sha Ha. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail. jsp?artefactID=5531&lang=1. Accessed 20 August 2012. AMO Hong Kong Archaeological Archive System: Sha Ha. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail. jsp?artefactID=5505&lang=1. Accessed 11 April 2014. AMO Hong Kong Archaeological Archive System: Sha Ha, Sai Kung. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail.jsp?artefactID=5521&lang=1. Accessed 10 April 2014. AMO Hong Kong Archaeological Archive System: Sha Ha, Sai Kung. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail.jsp?artefactID=46201&lang=1. Accessed 11 April 2014. AMO Hong Kong Archaeological Archive System: Sha Tsui Tau. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail. jsp?artefactID=374&lang=1. Accessed 20 August 2012. AMO Hong Kong Archaeological Archive System: Sheung Pak Nai. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail. jsp?artefactID=4477&lang=1. Accessed 20 August 2012.

256   Bibliography AMO Hong Kong Archaeological Archive System: So Kwun Wat. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail. jsp?artefactID=46627&lang=1. Accessed 20 August 2012. AMO Hong Kong Archaeological Archive System: So Kwun Wat. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail. jsp?artefactID=4589&lang=1. Accessed 11 April 2014. AMO Hong Kong Archaeological Archive System: Yung Long North. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail.jsp?artefactID=46327&lang=1. Accessed 10 April 2014. AMO Hong Kong Archaeological Archive System: Yung Long South. http://hkaas.lcsd.gov.hk/hkaas/artefactdetail.jsp?artefactID=5176&lang=1. Accessed 10 April 2014. Chinesecol.com, ‘A North American Pioneer Chinese Virtual Museum’. http://www.chinesecol.com/treasure10. html. Accessed 14 August 2010. Fu, X. G. 2006. ‘Neolithic Stepped Stone Adzes and Shouldered Stone Implements of Hong Kong’ (A Lord Wilson Heritage Trust Funded Project). http://www.lordwilson-heritagetrust.org.hk/chi/news/Prof_Fu.pdf. Accessed 25 June 2011. [傅憲國:〈香港地區史前時期有段石錛和有肩石器的類型、年代及演化〉,2006。] 傅憲國:〈香港史前文化與華南、東南亞及南太平洋地區的文化關係:以有段石錛和有肩石器為 例〉,2006。http://www.lordwilson-heritagetrust.org.hk/chi/news/Prof_Fu_1.pdf. Accessed 25 June 2011.

Index

Page numbers in italics refer to figures, maps, plates, and tables. boat, xviii, 34, 35n17, 39, 41, 44, 46, 52–53, 57n1, 57n6, 61n75, 93, 127, 138–39, 141, 153, 155, 157n60, 163, 165–66, 173 bone, 21, 24, 41, 43, 49, 50, 57, 59n29, 59n31, 59n33, 59n42, 61n81, 66, 70, 75, 77, 84n71, 94, 96–97, 100, 126, 132n68, 168, 180, 185, 187, 204 Bronze Age, 6, 10, 13, 17 map 5, 18–25, 25 plate 9, 26, 40–43, 45 map 7, 46, 48–49, 51, 54, 56–57, 58n19, 60n46, 60n51, 61n73, 61n79, 63–69, 70 plate 12, 70–72, 72 fig. 10, 73–74, 74 fig. 11, 75, 76 fig. 12, 76–77, 77 fig. 13, 78–79, 78 fig. 14, 79 plate 14, 80–81, 82n1, 82n3, 82n4, 82n18, 83n20, 83n31, 83n36, 83n51, 84n70, 84n74, 87, 90, 94, 96, 99, 108, 161, 163, 164 fig. 32, 165, 167, 167 map 12, 168, 169 fig. 33, 170, 176n3, 177, 179–81, 187, 200, 202, 202 fig. 44, 203 plate 53, 204, 205 fig. 45, plate 54, 206, 206 fig. 46, 206 plate 55, 219–20, 220 fig. 60, 220 plate 69, 222, 222 fig. 62, 223 plate 71, 224, 226, 226 fig. 64, 226 plate 74, 227, 227 fig. 65, 227 plate 75, 228, 229 fig. 66, 229 plate 76, 231, 231 fig. 68, 231 plate 78, 232, 232 plate 79, 233, 234 fig. 69, 234 plate 80 bronze casting, 13, 21, 74, 79, 80 fig. 15, 81, 165, 167–68, 170, 220 burial, 8, 12, 21–22, 24, 30, 32, 39, 42, 43, 44, 54, 59n33, 59n42, 59n43, 60n46, 61n81, 62n87, 65–67, 74, 76–79, 81, 82n3, 83n31, 85n83, 85n89, 86, 90, 94, 98–100, 102, 126, 130n6, 131n58, 132n64, 132n65, 132n66, 132n81, 132n83, 165, 170, 176n3, 177, 186–87, 197, 202, 204, 209, 213, 220, 222 Chinese diaspora, 151, 155 craft (domestic or activity or working), 6 (workshop), 40, 42, 44, 46, 49, 52, 67, 71, 75, 78, 163, 164 fig. 32, 165–66, 168, 170 craft specialisation (or specialist), 13, 41, 63, 72–74, 81, 82n10, 177 craftspeople (person), 10, 191

domesticated (plants or animals), 39, 41–42, 49, 59n31, 61n75, 69, 83n51, 132n67 farming, 8, 10, 13, 15n22, 29–32, 34, 39, 44, 58n3, 58n8, 58n10, 65, 69–70, 91, 99, 129, 134–36, 138–39, 141, 153–55, 164, 173, 178–79 figurine (animal), 18, 65, 68–70, 69 fig. 9, 70 plate 12, 82n12, 83n52 Finn, Daniel, v, xvii, 5, 12, 14, 15n10, 15n26, 16, 18 plate 5, 18–19, 21, 24, 26, 60n51, 68–69, 69 fig. 9, 72, 74, 79, 83n39, 83n40, 84n60, 84n63, 84n66 fishing, 8, 34, 41, 43, 49, 50 fig. 4, 52, 54, 57, 57n1, 60n56, 63, 65–66, 74 fig. 11, 75, 91, 97, 99, 129, 135, 138, 152–55, 157n60, 163, 166, 173, 178 food, 13, 34, 40–44, 46, 49–52, 50 fig. 4, 56, 59n31, 61n76, 65–66, 68, 74–76, 83n28 (foodstuffs), 90, 94, 96–98, 126, 132n68, 139, 144–45, 155, 165, 167–68, 170, 179–81, 224, 226 foraging, 41, 49, 50 fig. 4, 54, 60n56, 63, 65, 97 fung shui, 10, 15n22, 30–32, 99, 138, 170, 173 grave, 31, 39, 41–43, 54, 59n43, 60n51, 61n80, 61n81, 64–66, 76–78, 78 fig. 14, 82n3, 82n18, 85n81, 85n82, 85n83, 85n89, 93, 98–103, 101 fig. 18, 102 plates 20–21, 113 fig. 21, 130n1, 132n66, 132n76, 132n81, 170, 176n2, 186, 197–98, 197 fig. 40, 198 plate 49, 202 fig. 44, 203 plate 53, 202–4, 205 fig. 45, 205 plate 54, 209, 209 fig. 49, 209 plate 58, 211, 214, 220, 222, 222 fig. 62, 223 plate 71 guan, 12, 107, 107 plate 23, 128, 173, 178, 185, 190–91, 192 plate 46, 192 fig. 37 Guangdong, xviii, 6, 6 map 2, 39, 41, 58n10, 58n12, 58n13, 58n19, 60n46, 61n69, 64–65, 69–71, 82n12, 87, 90–91, 93, 101, 108, 119, 125, 129n1, 135, 138, 148–49, 157n52, 179–80, 188, 193, 197, 200, 204, 206–9, 211–15, 217, 231–32 Guangzhou, 86–87, 90, 129n1, 130n9, 143, 179, 188, 191, 194, 208–9, 211–13, 215 Guye, 42, 44

258   Index Hac Sa Wan, 40, 58n15 Hai Dei Wan, 63, 65, 83n21, 83n52, 202, 206 Han (dynasty), 13, 22, 61n79, 63–64, 69, 81, 82n4, 82n19, 85n77, 86–87, 88 fig. 16, 88 plate 15, 89 plate 16, 90, 92, 94, 96, 98–99, 108, 121, 129, 130n2, 130n5, 130n9, 153, 164 fig. 32, 170, 173, 177, 187–89, 191, 193–94, 207, 207 fig. 47, 207 plate 56, 208, 210, 233 Hase, Patrick, xviii, 15n11, 15n13, 90–91, 134–35, 138, 156n2, 157n43 Hayes, James, 134, 156n2, 157n43 Henglingshan, 64, 66, 68, 77, 82n18, 85n82, 204 Hung Shing Ye (Yeh), 3, 4 map 1, 18, 21, 79, 82n8, 83n39 hunting, 41, 49, 50 fig. 4, 54, 60n56, 63, 65–66, 74 fig. 11, 75, 97 impact assessment, 5, 16, 22, 28n25 Jin (dynasty), 21, 56, 62n87, 96, 99, 101 fig. 18, 101–3, 107, 113 fig. 21, 132n66, 132n85, 132n88, 132n90, 189–90, 197, 197 fig. 40, 198 plate 49, 208, 212, 212 fig. 52, 212 plate 61 Kau Sai Chau, 87, 90 kiln, 10, 12, 13, 15n24, 19, 21–23, 23 plate 8, 24–25, 25 plate 9, 26, 49, 60n58, 64, 68–70, 82n12, 82n18, 83n40, 86, 90, 92, 93 fig. 17, 94, 96, 97, 99, 102–3, 105 table 1, 106 table 2, 107, 108 plate 24, 111, 112 plate 26, 112, 115, 117, 118 plate 29, 119, 121 plate 30, 122 plate 31, 123, 124 plate 32, 124–29, 131n47, 131n48, 131n51, 133n112, 133n124, 140, 148–49, 149 plate 39, 155, 157n52, 164 fig. 32, 165, 170–71, 173, 178–80, 188–91, 194, 198–99, 209, 211, 215–18 Kwo Lo Wan, 42, 63, 65–66, 85n89, 221, 228 landscape, xvii, 6, 8, 10, 12–14, 15n21, 15n22, 15n25, 16, 27, 27n1, 29–32, 34–35, 35n1, 35n3, 43–44, 45 map 7, 48, 56 67, 80–81, 92, 95 map 8, 99, 124, 129, 137 map 9, 139, 154, 159, 161, 163–65, 166 map 11, 167, 167 map 12, 168, 169 fig. 33, 170–71, 171 map 13, 172 fig. 34, 173–74, 175 fig. 35, 176–79, 181 Late Neolithic, 27n14, 39–44, 48, 54, 56, 61n83, 63, 161, 165, 177, 187, 201, 219, 222, 224, 227–28 Later Neolithic, 13, 20–22, 24–26, 43, 46, 48–49, 50 fig. 4, 51, 51 fig. 5, 52 plate 11, 52–53, 53 fig. 6, 54, 55 fig. 7, 55 fig. 8, 56–57, 61n84, 63, 65–67, 75, 163, 164 fig. 32, 165–68, 179, 181, 200, 200 fig. 42, 200 plate 51, 201, 201 fig. 43, 201 plate 52, 219, 219 fig. 59, 219 plate 68, 221, 221 fig. 61,

221 plate 70, 222, 224, 224 fig. 63, 224 plate 72, 225, 225 plate 73, 228, 229 fig. 66, 229 plate 76, 230, 230 fig. 67, 230 plate 77, 231, 231 fig. 68, 231 plate 78 lime, 10, 13, 21, 23 plate 8, 62n96, 62n87, 86, 92, 94, 98, 100, 105 table 1, 111, 115 plate 27, 117, 119, 123–24, 126–29, 131n49, 131n51, 133n124, 133n125, 139, 171 (quicklime), 173, 178–79 Lingnan, xvi, 8, 35n17, 63–66, 90, 92, 129n1, 131n54, 133n125, 177 Lo So Shing, 3, 4 map 1, 5, 14n5, 48, 61n71, 103, 117, 131n48, 131n49, 131n50, 133n102, 133n103, 133n105 mammal, 21, 24, 49, 57, 59n29, 59n33, 66, 75, 84n71, 96–98, 180 Man Kok Tsui, 63, 65–66, 83n21, 84n74, 227–28 maritime, 5, 34, 39, 44, 53–54, 56, 67, 86–87, 90, 92–94, 98, 103, 129, 136, 138, 141, 170, 173 Meihuadun, 64–65, 68–69, 82n12 metallurgy, 10, 13, 66, 80–81, 82n10, 177 midden, 21–22, 24, 41, 43, 49, 56–57, 59n23, 59n29, 61n84, 62n86, 65, 75, 84n70,86, 94, 96–99, 97 plate 19, 121, 123, 132n66, 144, 148, 150–51, 164 fig. 32, 165, 168, 170, 180, 187, 204 Middle Neolithic, 5, 13, 15n20, 22, 24, 27, 27n14, 34, 39–43, 46, 27 fig. 2, 48, 53, 56–57, 57n2, 164, 165, 181 Ming (dynasty), 8, 13, 23, 91–92, 98, 126, 129, 134–35, 139–40, 140 fig. 28, 141 plate 33, 141, 150, 154, 156n1, 157n42, 163, 164 fig. 32, 177, 187 Nanhan, 91, 119, 125 fig. 27, 128, 178 Neolithic, 24, 27n14, 39–42, 44, 45 map 7, 46, 49, 51, 53, 54, 56, 58n3, 58n10, 58n13, 59n23, 59n24, 59n38, 61n75, 61n83, 65, 77, 78 fig. 14, 83n31, 161, 163, 164 fig. 32, 165–66, 166 map 11, 177, 180, 221, 225, 233 opium, 138, 146 fig. 31, 151–52, 152 plate 41, 155, 157n57 ornament: ornament (workshops), 18, 24, 41, 54, 55 fig. 7, 59n24, 61n76, 68, 72, 75–76, 80, 165, 167, 168 (manufacturers), 177, 180, 219. See also workshop (personal ornament); ornament (personal ornamentation), 13, 41, 43, 57, 61n76, 65–67, 72–73, 82n10, 84n62, 100, 102, 165, 181, 185–87, 219, 225, 228, 229 fig. 6 Pearl River Delta, xvi, 8, 13, 34, 35n16, 39–44, 54, 56, 57n1, 58n12, 63, 67, 85n81, 86, 96, 132n69, 165, 181, 187, 204, 228

Index   259 plant foods, 34, 41–42, 50, 73, 97, 180, 226 post-holes, 24, 26, 42, 52–53, 68, 71, 72 fig. 10, 81, 84n55, 84n58, 126, 168 potter’s marks, 64, 76 fig. 12, 204 prestige goods, 63–64, 66, 67, 81, 82n2, 165 Qin (dynasty), 66, 82n1, 86–87, 98, 129n1, 130n1, 130n2, 193 Qing (dynasty), 8, 12–13, 24–25, 29, 49, 100, 121, 125–26, 129, 134–36, 137 map 9, 138–40, 142, 144 fig. 29, 146 figs. 30–31, 147 plate 37, 148 plate 38, 149 plate 39, 150–51, 155, 156n1, 157n42, 157n50, 157n60, 161, 163–64, 164 fig. 32, 168, 173–74, 175 fig. 35, 177–80, 194, 218, 218 fig. 58, 218 plate 67, 235, 235 fig. 70, 235 plate 81 Qishi, 108, 217 Quanzhou, 108, 111, 125, 125 fig. 27 rescue excavation, 14n9, 24, 26, 27n1, 103, 117 rice-farming, 8, 10, 29–32, 34, 58n8, 91, 129, 135–36, 141, 153, 173, 178–79 salt, 10, 13, 61n72, 86–87, 90–92, 94, 125–26, 128–29, 131n52, 131n54, 131n125, 139, 170, 173, 178–79, 191 San Tau, 93–94, 119, 132n76, 132n83, 133n119, 170, 180, 216 Sha Ha, 10, 12, 15n23, 35n2, 42, 44, 57, 58n8, 58n14, 58n18, 59n24, 61n68, 63, 65, 83n26, 84n59, 221, 225, 228, 231 Sham Wan, 4 map 1, 5, 10, 12, 15n23, 35n14, 41–42, 48, 56, 58n14, 58n18, 59n29, 61n68, 63, 65, 84n74, 97, 187, 220, 222, 230 Sham Wan Tsuen, 90, 92, 103, 127, 132n65, 132n70, 132n81, 173, 176n4 Shek Pik, 58n14, 60n51, 63, 65–67, 82n8, 82n9, 83n31, 84n76, 85n80, 85n81, 85n83, 187, 198–99, 202 shellfish, 49–50, 56–57, 61n84, 65, 75, 96–97, 165, 168, 180 Shiwan, 149, 157n52 Six Dynasties, 6, 10, 12–13, 15n24, 19, 21–25, 29, 81, 86, 90–94, 95 map 8, 96, 98–100, 103, 108, 112, 113 fig. 21, 117, 119, 121, 123–24, 127–29, 132n64, 132n81, 132n87, 161, 163–64, 164 fig. 32, 170, 171 map 13, 172 fig. 34, 173, 177–79, 187–90, 193–94, 208 fig. 48, 211–12, 215, 233 skeletal remains, 12, 66, 77, 98, 100 slaking, 92, 105 table 1, 115, 126–27, 131n49 social landscape, 6, 8, 12, 14, 15n25, 16, 29, 30, 35n1, 67, 80, 159, 161, 163–65, 166 map 11, 167, 167 map 12, 171 map 13, 173–74, 176, 177, 179, 181

socio-economic, 6, 13, 32, 39, 56, 60n63, 63–64, 67, 70, 126, 154, 168, 170, 173 socio-historical, 30, 134, 139, 155, 156, 161 So Kwun Wat, 61n79, 63, 130n6, 199, 227, 231 Song (dynasty), 13, 21, 23–25, 58n8, 86, 87, 91–94, 95 map 8, 97, 108, 109 plate 25, 111–12, 114 fig. 22, 117, 123, 125, 125 fig. 27, 126, 128–29, 131n60, 132n65, 13n119, 135, 139–41, 163, 164 fig. 32, 171, 177–78, 186, 189, 191, 194, 198–99, 199 fig. 41, 199 plate 50, 215, 217, 217 fig. 57, 217 plate 66 Southern Dynasties, 79, 99, 102, 102 plate 21, 107, 113 fig. 21, 123, 128, 132n85, 132n90, 188–90, 208, 208 plate 57, 209, 209 fig. 49, 209 plate 58, 210–11, 211 fig. 51, 211 plate 60, 212, 215, 220 Southern Yue/Yue, 63–64, 86, 90, 130n2, 130n5, 173, 191 stilt-house, 10, 24, 42, 67, 71, 84n56, 93, 138, 153 plate 42, 153, 168, 173, 180 subsistence, 26, 29, 39, 41–42, 46, 49, 56–57, 60n56, 65, 67, 75, 81, 86, 94, 97–98, 138, 165–66, 168, 177, 180 Sui (dynasty), 24, 87, 90, 96, 101, 107, 113 fig. 21, 123, 128, 164 fig. 32, 178, 191, 194, 210, 210 fig. 50, 210 plate 59, 213, 213 fig. 53, 213 plate 62, 214, 214 fig. 54, 214 plate 63 Tai Wan, 3, 4 map 1, 5, 8, 15n17, 16, 18, 18 plate 5, 21, 42, 44, 46, 48, 51, 57, 58n12, 63, 65–66, 79–80, 82n3, 83n21, 83n39, 83n51, 136, 165, 177, 232 Tang (dynasty), 6, 10, 13, 15n24, 19, 21–25, 29, 62n87, 81, 82n19, 86–87, 90–94, 96–99, 101, 103, 107–8, 109 plate 25, 111–12, 113 fig. 21, 114 fig. 22, 115, 117, 119, 123–29, 131n58, 132n64, 132n76, 132n83, 132n87, 133n119, 133n125, 140, 145, 163, 164 fig. 32, 170–71, 171 map 13, 172 fig. 34, 173, 176n4, 177–79, 188–91, 194, 198, 213, 213 fig. 53, 213 plate 62, 214–15, 215 plate 64, 215 fig. 55, 216, 216 plate 65, 216 fig. 56, 233, 234 fig. 69, 234 plate 8 thermoluminescence (TL) dating, 10, 13, 96, 99, 103, 105 table 1, 106 table 2, 107–8, 112, 117, 119, 123, 126, 128–29, 133n94, 171, 178, 189 To Kwa Wan (Kowloon Bay–Kai Tak), 93–94, 132n65, 178 Tung Lung Fort, 147, 151, 155, 157n41, 157n50, 157n52, 218 Tung Wan Tsai, 41, 43–44, 58n18, 63, 65–66, 87, 90, 130n7, 187, 202, 204, 207 watching brief, 15n9, 17 map 5, 24–25, 25 plate 9, 26, 27n1, 178, 185 woodworking, 53

260   Index workshop (personal ornament), 6, 15n10, 18, 24, 41, 43, 54, 61n76, 61n77, 65, 68, 72–74, 73 plate 13, 81, 165, 168, 177, 180–81, 228. See also craft workshop (pottery, tiles or bricks), 64, 147, 191 (tiles or bricks) workshop (tools and weapons), 57n2, 59n24, 61n76, 75, 84n66, 165, 168, 181 Wun Yiu, 146 fig. 30, 149, 155, 157n43, 218 Xincun, 42, 57, 61n69, 65, 68, 166, 180 Yingang, 64, 68, 82n12, 206 Yuan (dynasty), 13, 25, 86, 90–94, 108, 111–12, 117, 123, 125, 125 fig. 27, 126, 129, 131n52, 133n119, 141, 161, 163–64, 164 fig. 32, 171, 177–78, 217