Hostname: page-component-745bb68f8f-kw2vx Total loading time: 0 Render date: 2025-02-06T08:02:29.779Z Has data issue: false hasContentIssue false

Travel and landscape: the Zuo River Valley rock art of Guangxi Zhuang Autonomous Region, China

Published online by Cambridge University Press:  05 June 2015

Paola Demattè*
Affiliation:
Rhode Island School of Design, Two College Street, Providence, RI 02903–2784, USA (Email: pdematte@risd.edu)
Rights & Permissions [Opens in a new window]

Abstract

Rock art, especially in China, has often been associated with the non-literate, non-Chinese periphery. It is often thought of very broadly as a universal phenomenon rooted in religion and sharing a widely recognised iconography. This paper challenges both of these assumptions. Its focus is the rock art of the Zuo River in the Guangxi Province of China and in particular its landscape location and visibility. Broadening the parameters of what we categorise as rock art, such art is shown to have multi-layered meanings that spoke to different groups in different ways. The content, location and visibility of images along this arterial waterway reveal how rock art played a significant role in life and death in Late Bronze Age and Iron Age Guangxi.

Type
Research
Copyright
Copyright © Antiquity Publications Ltd, 2015 

Introduction

Over centuries of investigation, various totalising theories (animism, hunting and fertility magic, structuralism, shamanism) have been offered to explain the ‘meaning’ of rock art (Bahn Reference Bahn1998). At their core, these explanations all propose that rock art is a unitary and unchanging phenomenon that originates in religious belief and shares a basic iconography. These attempts are founded on at least two problematic (and circular) assumptions that are rarely acknowledged. The first is that rock art is a coherent, global phenomenon with a shared ideology and iconography; the second is that by virtue of this supposed uniformity, meaning can be discerned from rock art imagery that further supports the idea of its universality.

This is not to say that what we call rock art is not united by some characteristics, or that the iconography of individual sites cannot be understood. The problem is one of logic: manifestations whose commonality lies in the medium in which they are produced (i.e. carving or painting on rocks in the open landscape), rather than in iconography or culture, cannot be analysed en masse with the expectation that their material contiguity equates to a shared iconography or culture that can be extrapolated from one site to the next.

If the rock landscape is what globally unifies rock art, surely it is rock landscapes that require analysis. The objective of this research is therefore to eschew questions relating to the identification of meaning in iconography, and instead to engage with defining elements of the rock art phenomenon, such as landscape choices, spatial organisation and visibility.

Analysis of space and place can yield tangible information about the relationship of rock art to aspects of a territory and its communication routes (Tuan Reference Tuan1977). Examining the use of landscape in rock art production can expand our notions and definitions of the phenomenon, moving it beyond the implicit ‘primitive-civilised’ dichotomy upon which the concept of rock art has been based. Here, I examine the rock art of the Zuo River valley in Guangxi, China, as a case study to assess the use of landscape in rock art production (Figure 1).

Figure 1. The Zuo River sites (adapted from Google Earth).

Northern and southern traditions of Chinese rock art

Territory plays a role at various levels in the identification and assessment of rock art. Within the boundaries of the People's Republic of China, rock art, as conventionally defined, is concentrated at the northern, western and southern edges of the fertile lands dissected by the country's two major waterways, the Yellow and the Yangtze Rivers.

The Central Plain and the coast, areas considered the so-called ‘cradle of Chinese civilisation’ and literary tradition, have been traditionally understood as devoid of rock art, even though this is not strictly true. Deep circular holes, known as copules, and geometric patterns carved on boulders have been discovered on the Juci Mountain and elsewhere in Henan Province (Tang Reference Tang2012), while faces are known from cliffs at Lianyungang in coastal Jiangsu (Gai Reference Gai1995). There are also Buddhist, Daoist or Confucian carvings in landscapes similar to those where rock art is found outside of the ‘cradle of civilisation’. These suggest that some rock art may have been obscured by the accretion of historic events. Evidence of such accretion is shown at some Chinese sites, such as Heishan (Gansu) and Helankou (Ningxia), where petroglyphs intermingle with Buddhist symbols and inscriptions in a palimpsest of imagery (Demattè Reference Demattè, Moser and Feldman2014). The perceived scarcity of imagery in this region is therefore related to the assumption that rock art is a phenomenon of the non-literate, non-Chinese periphery.

Within this context, rock art in China is generally considered to belong to one of two main branches—north-western and southern—which can be further divided into regional clusters. Although this is not intended to be a rigid geographic distinction, the two traditions do differ in style, subject matter and cultural affiliations (Song Reference Song1998).

Southern Chinese rock art, such as that of the Zuo River, is associated with the cultures of south-eastern and south-western agriculturalists who have historically inhabited the provinces of Guangxi, Yunnan, Guizhou and to a lesser degree Sichuan. This tradition is dominated by painted (pictographs) rather than carved images, but exceptions do exist. Style and content vary, but pictographs are usually painted in red or reddish brown. The dates of these productions range from as early as 3000 BC to as late as the sixth or seventh century AD. In Yunnan, more than 10 open-air sites have been discovered in Cangyuan, an area in the Awa Mountains on the Sino-Burmese border. At site number 7, dark red paintings depict hunting, herding, planting and ritual performances with acrobats and jugglers, an activity thought to be associated with the ancestors of the local Wa and Thai minorities (Wang Reference Wang1985a & b). Another significant concentration of rock paintings featuring large animals is found in the Jinsha River valley, and it is believed to be very early, possibly dating to the local Neolithic (Taçon et al. Reference Taçon, Li, Yang, May, Liu, Ji, Aubert, Curnoe and Herries2010, Reference Taçon, Aubert, Li, Yang, Liu, May, Fallon, Ji, Curnoe and Herries2012). In Guizhou, one of the most remarkable sites is a painted cliff at Wushan (Longli County), which is thought to date to the Bronze Age (Li Reference Li2006). Other Guizhou sites, including Guanling Niujiaojing and Kaiyang Huamayan, unusually feature petroglyphs (Song Reference Song1998). In Sichuan, at Gongxian Matangba, images painted on vertical cliffs are associated with hanging coffins, a burial practice common in pre-modern southern China (Sichuan Provincial Museum & Gongxian Culture Office 1978; Zeng Reference Zeng1990).

In contrast, the northern and western tradition of Chinese rock art, which is distributed in the steppe, mountains and desert landscape that stretch from Liaoning in the east through Inner Mongolia, Ningxia, Gansu, Xinjiang, Qinghai and Tibet to the west, is characterised by carved images (petroglyphs). Their content for the most part relates to the pastoralist and hunting lifestyle of Asian nomads, and it is linked with the activities of northern and western nationalities of Turco-Mongolian and Tibetan heritage, mainly from the first millennium BC to the fifth century AD. Sites are concentrated on boulders, outcrops and canyons of Yinshan (Inner Mongolia) and Helanshan (Ningxia) and in the Hutubi (Xinjiang) (Gai Reference Gai1995; Demattè Reference Demattè2004).

The rock art of the Zuo River valley of Guangxi Province

One of the largest concentrations of painted rock art in China is found in the Zuo River valley of Guangxi, a south-eastern autonomous province that borders Vietnam and is home to the Zhuang people and other minority nationalities. The region, which is lush and mountainous, is defined by a capillary river system that becomes the Pearl River (Zhujiang) on entering the South China Sea near Guangzhou (Canton). The Zuo River (also known as Lijiang), which historically was the main communication route in the area, is one of the waterways that constitute the Xijiang (West River), the western tributary of the Pearl (Figure 2).

Figure 2. Guangxi Zhuang Autonomous Region and Pearl River system.

Landscape, with its travel routes and natural landmarks, plays a dominant role in the decision to create and situate open-air art. In the fluvial landscape of the Zuo Valley, the river with its cliffs and riverine villages appears to provide the principal organisational reason for the phenomenon. Pictographs are distributed over 80 sites in the territories of 4 counties—Ningming, Longzhou, Chongzuo and Fusui—that are crossed by the Zuo River system. Most are on the upper and middle Zuo River course and on the lower course of the Ming River in Ningming County, a large tributary of the Zuo. A few sites in these counties and in nearby Pingxiang are in hills and mountains more distant from the main river system (Qin et al. Reference Qin, Qin, Hu and Yu1987).

The painted sites are concentrated on the vertical cliffs of limestone (calcite and dolomite) outcrops that line the course of both rivers, particularly at river bends. These mountains stand 150–400m above the river and are covered top and bottom by dense vegetation. Their cliffs, white with black stains, are steep and forbidding and interspersed with karstic fissures and caves, which were often used for burials (Guangxi Zhuang Autonomous Area Cultural Relics Team 1993). Pictographs are conspicuously painted in the lighter areas halfway up the cliffs and frequently appear at the sides of, or above, cave openings.

For this research, I investigated sites on the Ming river in Ningming County (Zhushan, Longxia, Daning, Gaoshan, Huashan and Hongshan), on the upper course of the Zuo River in Longzhou County (Chenxiangjiao, Baojianshan, Da’anshan, Loutiyan, Yanhuaishan, Xiabaiyue, Chaochuanshan, Duchuanshan, Dazhoutou and Longzhou Huashan, also known as Mianjiang), and in Pingxiang City away from the river system. Here, I describe only the largest (for more examples see Qin et al. Reference Qin, Qin, Hu and Yu1987).

Zhushan (in Ningming County) is a mountain oriented east-west on the left bank of the Ming River, facing the fields of Zhulian village. There are several caves, at varying heights, on the north-facing cliffs, as well as 22 painted panels with 169 images (Figure 3). One panel features several large figures shown frontally, with arms raised and legs apart, surrounded by a greater number of smaller figures represented in profile. Some of the large figures wear swords across their hips and one appears to stand atop a quadruped, possibly a dog or ox.

Figure 3. Zhushan cliffs from the fields of Zhushan village, Ningming, Guangxi (arrows indicating painted panels).

Also in Ningming County, on the right bank at a bend in the Ming River, the L-shaped Gaoshan Mountain, with two peaks (north and south), has 18 painted panels. One panel, under a rock overhang on the north peak, is next to a square opening—probably a modified natural cave used for burial. Both the cave and the paintings are accessible by a ledge. The paintings feature the theme of a large, splayed figure standing on a quadruped and surrounded by smaller figures, some possibly in a boat. On the eastern peak, various panels are concentrated on the left side of the cliff 20m above the river. One features large and small figures and drums, and an image of a stand with bronze bells of a type excavated from local Late Bronze Age tombs (Chinese National Museum & Guangxi Zhuang Museum 2006: 56–57) (Figure 4).

Figure 4. Painted panel and cave at Gaoshan, Huashan, Ningming, Guangxi.

The largest rock art site of the Zuo Valley is at Huashan (Hua Mountain), Ningming County, where a triangular-shaped peak rises from the right bank of the Ming River, reaching a height of c. 230m (345m above sea level). The pictographs are painted on the western face of Huashan and are visible from both the river and Bayue, a Zhuang village on the opposite bank. Over 1800 pictographs from 111 painted panels distributed in 9 locations cover an area that measures c. 210m in length and 40m in height. Two locations, side by side on the same cliff, account for most of Huashan's paintings and show hundreds of human figures mixed with animals and ritual implements. People are depicted both in frontal or side view, and they often have their arms raised and their knees bent. Some wear or brandish weapons; others appear to stand on or next to quadrupeds. Often, among the human figures, there are circles decorated with star or solar patterns that probably represent bronze drums. In one case, a group of people appears to stand on a boat with their arms raised; in another, figures appear to engage in sexual activities. In 2010, a portion of the site was covered by scaffolding set up for conservation work. This gave a sense of the height of the structures that would have been needed to paint the images and also allowed for a close examination of the paintings. Although some are small (0.2m), the majority are large (0.9–1.0m), and some are over 2m high. Iconographic and stylistic analysis of the paintings at this and other Zuo River sites in comparison with archaeologically excavated material suggest that, as a general rule, the earliest paintings are at a lower level and the most recent are at a higher level (Figures 5 & 6).

Figure 5. Huashan site in 2010, Ningming, Guangxi (arrows indicating painted panels).

Figure 6. Painted panel next to scaffolding, Huashan, Ningming, Guangxi.

Chenxianjiao (Longzhou County), a mountain on the left bank of the Zuo opposite Heting village, has eight painted panels. One, featuring the standard splayed dancing figures, is painted inside a natural arched niche above a group of caves, 20m higher than the river. Another panel, covered by calcite encrustations, shows splayed figures and dogs in vertical arrangement (Figure 7).

Figure 7. Chengxiangjiao, Longzhou, Guangxi.

In Longzhou County, Yanhuaishan, a cliff face that runs for about 700m east–west on the right bank of the Zuo River, has eight painted panels divided between two locations. Sanzhoutou, which is upstream, has six, and Sanzhouwei, downstream, has two. At Sanzhoutou, a panel follows the form of the V-shaped ledge onto which it was painted: a frontally depicted figure stands splayed at the point of the ‘V’ while smaller figures walk down towards it and then up away from it. Another panel is next to a cave showing splayed figures and two people in profile playing drums (Figure 8).

Figure 8. Yanhuaishan, Longzhou, Guangxi.

Chaochuanshan, on the right bank of the Zuo, and Duchuanshan, on the left bank, are both in Longzhou County. In addition to the panels with classic splayed figures, there are also images that show people on boats. At Duchuanshan, one is painted above a cave. Interestingly, the names of both sites reference boats. Owing to its shape, Duchuanshan is ‘Ferry Mountain’, and Chaochuanshan is ‘Facing Boat Mountain’. The theme may be connected with the mountains’ names or perhaps with an ancient ferry linking the river's two banks (Figures 9 & 10).

Figure 9. Chaochuanshan, Longzhou, Guangxi (arrow indicating painted panel).

Figure 10. Paintings and cave at Duchuanshan, Longzhou, Guangxi.

Longzhou Huashan, in Longzhou County, is a 200m-high mountain on the right bank of the Zuo River, opposite Mianjiang village, from which the paintings are easily visible.On the cliffs, which are 400m long and at a significant bend in the river, there are 18 panels with 298 figures. The largest is a composition of over 100 figures that covers an area of 12 × 20m above a large cave at the southern end of the cliff. It features human figures, with raised hands brandishing daggers, surrounded by animals and drums. Another painted panel is next to a cave at the north end of the cliff. Caves at Longzhou were used for burial (Guangxi Zhuang Autonomous Area Cultural Relics Team 1993) (Figure 11).

Figure 11. Cave and paintings at Longzhou Huashan, Longzhou, Guangxi.

Two sites near Pingxiang City differ stylistically from Zuo River rock art, and although dating to later historic times, they are connected with that tradition (Qin et al. Reference Qin, Qin, Hu and Yu1987: 215–16). Neither is on the Ming-Zuo River system, but both are associated with grottos. The sites face each other and are a few hundred metres from a stream and the railway tracks, suggesting that akin to the Zuo Valley paintings, they were close to travel routes. The first, Haimashan, is below a rock overhang and features red interlocking patterns that may represent a scaly animal or a net. At the second, Lushan, a red standing figure is painted next to a gourd-shaped grotto (Figure 12).

Figure 12: Gourd-shaped grotto at Lushan, Pingxiang, Guangxi.

Dating, histories and identities

Guangxi has a long history of human occupation, ranging from the Middle Palaeolithic sites of the Bose basin (Hou et al. Reference Hou, Potts, Yuan, Guo, Deino, Wang, Clark, Xie and Huang2000), to the Upper Palaeolithic cave sites of Tongtianyan and Gaitou (Shen et al. Reference Shen, Wang, Wang, Zhao, Collerson, Zhou and Tobias2002, Reference Shen, Wang, Cheng and Edwards2007), to sites documenting the transition to sedentism, cultivation and pottery making. At Miaoyan, Dingsishan and Liyuzui, pottery has been dated to a range from 20 000 to 11 000 cal BP; at Zengpiyan cave there is equally early evidence of experimentation with rice planting, and fully developed rice agriculture appears c. 4000 BC (see Pearson Reference Pearson2005; Prendergast et al. Reference Prendergast, Yuan and Bar-Yosef2009; Zhang & Hung Reference Zhang and Hung2010).

The Zuo River valley itself was occupied in the Palaeolithic and Neolithic. Late Palaeolithic choppers were discovered in the 1960s at Ai grotto in the Luching mountains of Chongzuo (Qin et al. Reference Qin, Qin, Hu and Yu1987). Neolithic sites include shell mounds and caves, as well as evidence of ceramic production and agricultural activities, in the form of specialised agricultural tools such as stone shovels (Zheng & Tan Reference Zheng and Tan2006: 13–40; various authors in Guangxi Institute of Archaeology and Cultural Heritage 2007). The Zuo River rock art now visible does not appear to date to the Palaeolithic or the Neolithic. The quantity of painted sites, their growth into palimpsests and the synergy among paintings, caves and settlements over a wide area, point, however, to a historically deep relationship between inhabitants and the landscape in the Zuo Valley.

Based on stylistic comparisons with archaeological evidence, the paintings have been linked to the Guangxi Late Bronze Age and Iron Age, and they have been assigned to threephases falling between the fifth century BC and the fourth or fifth century AD (Wang et al. Reference Wang, Qiu and Chen1988). The making of rock paintings appears to have stopped in the fifth century, although a few sites distant from the Zuo Valley date to the late imperial period (fifteenth–seventeenth centuries AD). The reasons for the disappearance of this tradition have not been determined, but it is likely that the increasing influence of Chinese culture and of organised religions, such as Buddhism and Daoism, in the area from the early centuries of the first millennium AD may have caused a shift in religious beliefs and cultural practices that led to its abandonment.

This interpretation is partially supported by four radiocarbon dates obtained from calcium carbonate at the centre and edges of calcite encrustations that covered two paintings at Ningming Huashan. The first, which was at a height of 15m from the foot of the mountain, returned a date of 1570 ± 60 BP at the centre (BK 83039-a) and 690 ± 70 BP at the edge (BK 83039-b); the second, which was at a height of 2.5m, gave a date of 5580 ± 80 BP at the centre (BK 83038-a) and 4250 ± 80 BP at the edge (BK 83038-b). The investigators place the paintings between 4200 and 1680 years ago, but they suggest caution because of possible contamination from carbonate bioliths in the cliff, which could make samples appear older than their true age (Chen & Zhu Reference Chen, Zhu, Wang, Qiu and Chen1988). The later date range is confirmed by the 14C dating of a wooden burial casket from a cave at Longzhou Huashan to 1512 ± 50 BP (Guangxi Zhuang Autonomous Area Cultural Relics Team 1993: 61), and it is in line with historical interpretations. The earlier date range is less clear and it is possible that some paintings are older than the fifth century BC and also that older paintings may have existed that are now lost.

According to ancient Chinese texts, during this time south-eastern China was inhabited by the Bai Yue (Hundred Yue), a generic name used by Chinese polities in the north to indicate the heterogeneous populations of the south-east that they associated with the southern Yue kingdom (sixth century–334 BC), even though before the third century BC southern Guangxi was not under Chinese control. Historians have suggested that among the Bai Yue, it was the Xi Ou and Luo Yue, groups ancestral to the Zhuang people who now populate the area, that were responsible for the material culture of the Zuo River valley (Barlow Reference Barlow, Tötösy de Zepetnek and Jay1997; Gao Reference Gao2013). These populations occupied southern Guangxi and northern Vietnam (where the Zhuang are known as Nung) and were part of a larger system related to the Dongson culture, which had ties with the Au Lac polity of Bac Bo whose capital was at Co Loa, in northern Vietnam (Higham Reference Higham1996: 108–35; Kim et al. Reference Kim, Toi and Hiep2010).

The landscape of Zuo River rock art: content, visibility, routes and access

Analysis of the sites shows regularities in landscape use and uniformity in technique, style and iconography. Images are painted in thick layers of deep red that contain iron oxide fromores such as haematite, which is readily available in the Zuo Valley. Adhesive proteins point to an animal-based binder, possibly ox-hide glue, historically used by the Zhuang, but also ox blood, which was used in nearby Yunnan (Wang Reference Wang1985a & b; Wang et al. Reference Wang, Qiu and Chen1988: 231). Most images are large, varying between 1 and 2m in height; some are smaller, and a few reach over 3m. Colour and size make these paintings highly visible from a distance, although today some are faded by weathering.

Paintings are organised in narrative panels around similar iconographic themes. Most feature groups of stylised human figures, shown frontally or in profile, interspersed with animals, weapons and circles with stars inside that probably represent drums. Some paintings show people in boats. Bronze daggers and drums with decoration similar to the painted icons have been excavated in the area and were used both by the ancestors of the Zhuang and the Zhuang themselves in performances and ceremonies (Chinese National Museum & Guangxi Zhuang Museum 2006: 44–48, 84–85, 235–53). The paintings are thought to represent ritual dances relating to mythical narratives and ancestral worship. Ethnographic evidence indicates that the Zhuang, and other local groups such as the Maonan, performed dances and rituals centred around bronze drums that resemble depictions on the cliffs (Chinese Association for The Study of Ancient Drums 1988: 129; Jiang Reference Jiang1999: 222).

Most paintings are clustered on cliffs overlooking the river, where they are visible as red patches that contrast with the light colour of the rock. Generally, the sites tend to face south or east, and while some face other directions, very few face directly north. Cliffs with high numbers of painted signs (such as Ningming Huashan) are concentrated at river bends. According to Gao (Reference Gao2013), river bends had a special significance in Zhuang lore because they were thought to harbour water monsters. The reference to water monsters may relate to the presence of dangerous currents at the curves of the river. On the other hand, it is also probable that the tall and steep cliffs that modified the course of the river, creating major bends, were highly visible and probably the best places for painting images that were required to be visible.

Given their colour, size and positions, it is clear that the Zuo River painted panels were made to be visible from nearby settlements and particularly from the river. A systematic archaeological survey of the ancient settlement pattern is not available for the area; however, assuming that the position of ancient settlements is not remarkably different from the contemporary layout, it is very probable that most painted sites had close ties to nearby village communities. In several instances, such as at Zhushan, Huashan and Longzhou Huashan (Mianjiang), the painted cliffs face neighbouring hamlets and are positioned for optimal viewing from their agricultural fields, probably to signal the presence of sacred places, such as the ancestral burials in caves.

On the other hand, the extensive use of the river system as a container and framework for displaying rock art implicitly suggests that movement and travel on the river are also fundamental elements for analysis of the phenomenon. The river, which even today remains the fulcrum of local life, was in antiquity the principal communication route that linked the villages of the Zuo Valley to each other, to major Chinese centres to the north and east and to southern territories that are now part of Vietnam. Boats, which are represented occasionally in the paintings, were the life of the valley and the ideal way to view rock art.

This analysis has highlighted some points that aid our understanding of the way the landscape was used to exhibit rock art. The creation of painted panels facing the river and the villages indicates that the images, far from being secret, were meant to be seen, by people on land and by people on water (fishermen, traders and other navigators). The inclusiveness of the images, which were visible to both locals and visitors, suggests that this river-wide tradition was broadly shared and that the paintings were part of an extended public landscape that broadcast multi-layered messages to different audiences.

The messages in the images probably varied in relation to the recognised importance of a site. Smaller sites in remote or secluded areas, such as at Pingxiang, might have been more private, whereas larger and more visible ones, including Ningming Huashan and Longzhou Huashan, were meant to attract the attention of all those who travelled the Zuo River. In the latter case, the paintings on the cliffs may have signalled that the place was sacred or proclaimed the power of local villages; equally, they could have been warnings to travellers about dangerous areas of the river that were particularly treacherous after heavy rains and during flooding. The regional uniformity suggests that this was a communication code that was probably shared by all the people of the Zuo region, but it might not have been recognised by outsiders, giving the locals an advantage over those on the river with hostile intent.

If the Zuo River was instrumental in fostering peaceful communications and trade within the region and beyond, it is also clear that this shared path of communication could generate local clashes and introduce threats from the outside. Conflicts may therefore also have played a role in the decisions to inscribe the cliffs. The river system included a large number of villages dominated by different clans that recognised the cultural unity of the Zuo Valley but that were occasionally in competition with each other. Painted sites seemed to have worked on two fronts: on one hand, to mark the individuality of each village or group of related villages; and on the other, to indicate to outsiders the existence of a river community that, in case of external threats, would come together under the protection of supernatural forces (Qin et al. Reference Qin, Qin, Hu and Yu1987).

Although clearly visible, the images were not necessarily meant to be easily approachable. A glance at the steep cliffs shows how treacherous access was to the site. Archaeologists have suggested that painters reached the cliffs by climbing, using scaffolding, abseiling down from the top of the mountain or by boat at times of flooding. Careful examination of the cliffs reveals that painted locations, as for example at Longzhou Huashan, are often near rock overhangs, ledges, crevices or natural grottos that may have helped the painters reach their target.

Most of the caves near the paintings were probably used for burial purposes and associated rituals. Some had their openings reshaped and were closed with simple stone constructions; others appear to have been left in their natural state. Wooden coffins were discovered inside a few, but many others remain unexplored (Guangxi Zhuang Autonomous Cultural Relics Team 1993). The presence of burials in some does not exclude the possibility that others, such as the gourd-shaped grotto of Pingxiang City, may have had different uses.

Burials in some of the caves associated with the paintings might have been a reason to maintain a distance from the cliffs, which might have been seen as liminal places between the living and the dead. The general impression is that even though access was arduous, perhaps even dangerous, the cliffs were crossed by pathways that might have been trodden by several people on special occasions, such as funerals or ancestral rituals.

Conclusions

The content, location, visibility and access to the painted panels of the Zuo River suggests that they had multi-layered meanings and addressed the many audiences that depended on this fluvial route. These elements also highlight the tensions between public and private, sacred and secular, that characterise many rock art sites. Whether in northern, western or southern China, the majority of rock art sites tend to be in high places that are difficult to reach but at the same time enjoy a certain visibility. Many are also near water sources, shrines, burials and ancient travel routes. A few are more secluded.

These patterns of use of the landscape and space are not remarkably different from those of established religions or political institutions whose artistic production in the landscape is generally not understood as ‘rock art’. In China, historic Daoist or Buddhist rock carvings and Confucian cliff inscriptions are also associated with travel routes (land, sea or river) or remarkable landscape features (mountains, cliffs, shores) that have acquired religious or political significance (Harrist Reference Harrist2008; Demattè Reference Demattè, Moser and Feldman2014). An example of a similar pattern of distribution for Buddhist and Daoist temples and historic inscriptions on a river course is found in the Three Gorges area on the Yangtze River (Demattè Reference Demattè2012). If we also consider these to be historic sites, ‘rock art’ in China will not be solely concentrated in areas peripheral to the Central Plain that are associated with China's minority populations, but it will be more evenly distributed. This shared approach to landscape use indicates that the ‘primitive-civilised’ dichotomy often implied in the term ‘rock art’ is a fallacy; it also highlights the danger of approaching rock art as a universal phenomenon without considering its cultural background.

Acknowledgements

I would like to thank Professor Chen Xingcan of the Institute of Archaeology, Chinese Academy of Social Sciences and Dr Lin Qiang, Director of the Guangxi Institute of Archaeology and Cultural Heritage, for facilitating this research. I am extremely grateful to Mr Zhu of the Ninming County Cultural Bureau for guiding me to the various sites of the Zuo River valley. Without their help this research would not have been possible. This research was supported by the Humanities Fund of the Rhode Island School of Design.

References

Bahn, P. 1998. Cambridge illustrated history of prehistoric art. Cambridge: Cambridge University Press.Google Scholar
Barlow, J.G. 1997. Culture, identity and early weapons systems on the Sino-Vietnamese frontier, in Tötösy de Zepetnek, S. & Jay, J.W. (ed.) East Asian cultural and historical perspectives: 115. Edmonton: University of Alberta.Google Scholar
Chen, T.M. & Zhu, F.B.. 1988. Ningming Huashan yanhua tanshisi niandai ceting [14C dating of Huashan rock art in Ningming], in Wang, K.R., Qiu, Z.L. & Chen, Y.Z. (ed.) Guangxi Zuojiang yanhua [Guangxi's Zuo River rock art]: 244–47. Beijing: Wenwu.Google Scholar
Chinese Association for the Study of Ancient Drums. 1988. Zhongguo gudai tonggu [Ancient Chinese Bronze drums]. Beijing: Wenwu.Google Scholar
Chinese National Museum & Guangxi Zhuang Museum. 2006. Ou Luo yicui. Beijing: Zhongguo Shehui Kexue.Google Scholar
Demattè, P. 2004. Beyond shamanism: landscape and self-expression in the petroglyphs of Inner Mongolia and Ningxia (China). Cambridge Archaeological Journal 14: 120.CrossRefGoogle Scholar
Demattè, P. 2012. After the flood. Cultural heritage and cultural politics in the Chongqing municipality and the Three Gorges areas, China. Future Anterior 9 (1): 4964.Google Scholar
Demattè, P. 2014. Itinerant creeds: the Chinese northern frontier, in Moser, C. & Feldman, C. (ed.) Locality of sacrality: theoretical approaches to the emplacement of religion: 5773. Oxford: Oxbow.Google Scholar
Gai, S.L. 1995. Zhongguo yanhua xue [Study of Chinese rock art]. Beijing: Shumu Wenxian.Google Scholar
Gao, Q. 2013. The Huashan rock art site (China): the sacred meeting place for sky, water and earth. Rock Art Research 30: 2232.Google Scholar
Guangxi Institute of Archaeology and Cultural Heritage. 2007. Guangxi kaogu wenji [Collected essays on Guangxi archaeology: volume 3]. Beijing: Wenwu.Google Scholar
Guangxi Zhuang Autonomous Area Cultural Relics Team. 1993. Guangxi yandong zang diaocha baogao [Report on the investigation of cliff burials in Guangxi]. Wenwu 1993 (1): 5161.Google Scholar
Harrist, R.E. 2008. Landscape of words. Seattle: University of Washington.Google Scholar
Higham, C. 1996. The Bronze age of Southeast Asia. Cambridge & New York: Cambridge University Press.Google Scholar
Hou, Y.M., Potts, R., Yuan, B.Y., Guo, Z.T., Deino, A., Wang, W., Clark, J., Xie, G.M. & Huang, W.W.. 2000. Mid-Pleistocene Acheulean-like stone technology of the Bose basin, south China. Science 287: 1622–26. http://dx.doi.org/10.1126/science.287.5458.1622 Google Scholar
Jiang, T.Y. 1999. Gudai tonggu tonglun [General survey of ancient bronze drums]. Beijing: Forbidden City Publishing House.Google Scholar
Kim, N.C., Toi, L.V. & Hiep, T.H.. 2010. Co Loa: an investigation of Vietnam's ancient capital. Antiquity 84: 1011–27. http://dx.doi.org/10.1017/S0003598X00067041 Google Scholar
Li, F. 2006. Shilun Guizhou Longli Wushan yanhua [On the rock art at Wushan, Longli, Guizhou]. Sichuan Wenwu 3: 5458.Google Scholar
Pearson, R. 2005. The social context of early pottery in the Lingnan region of south China. Antiquity 79: 819–28. http://dx.doi.org/10.1017/S0003598X00114954 Google Scholar
Prendergast, M.E., Yuan, J. & Bar-Yosef, O.. 2009. Resource intensification in the Late Upper Palaeolithic: a view from southern China. Journal of Archaeological Science 36: 1027–37. http://dx.doi.org/10.1016/j.jas.2008.12.002 Google Scholar
Qin, S.M., Qin, C.L., Hu, M.F. & Yu, R.Y.. 1987. Guangxi Zuojiang liuyu yanbihua kaocha yu yanjiu [Investigation and study of the rock art of the Zuojiang River Valley, Guangxi]. Nanning: Guangxi Minzu.Google Scholar
Shen, G.J., Wang, W., Wang, Q., Zhao, J.X., Collerson, K., Zhou, C.L. & Tobias, P.V.. 2002. U-series dating of Liujiang hominid site in Guangxi, southern China. Journal of Human Evolution 43: 817–29. http://dx.doi.org/10.1006/jhev.2002.0601 Google Scholar
Shen, G.J., Wang, W., Cheng, H. & Edwards, R.L.. 2007. Mass spectrometric U-series dating of Laibin hominid site in Guangxi, southern China. Journal of Archaeological Science 34: 2109–14. http://dx.doi.org/10.1016/j.jas.2007.02.008 CrossRefGoogle Scholar
Sichuan Provincial Museum & Gongxian Culture Office. 1978. Sichuan Gongxian “Boren” xuanguan ji yanhua diaocha ji [Investigation of the Bo hanging coffins and rock art]. Wenwu ziliao congkan 2: 187–95.Google Scholar
Song, Y.L. 1998. Zhongguo yanhua kaocha [Investigation of Chinese rock art]. Taibei: Lianjing.Google Scholar
Taçon, P.S.C., Li, G., Yang, D.C., May, S.K., Liu, H., Ji, X.P., Aubert, M., Curnoe, D. & Herries, A.I.R.. 2010. Naturalism, nature and questions of style in Jinsha River rock art, north-west Yunnan, China. Cambridge Archaeological Journal 20: 6786. http://dx.doi.org/10.1017/S0959774310000053 CrossRefGoogle Scholar
Taçon, P.S.C., Aubert, M., Li, G., Yang, D.C., Liu, H., May, S.K., Fallon, S., Ji, X.P., Curnoe, D. & Herries, A.I.R.. 2012. Uranium-series age estimates for rock art in south-west China. Journal of Archaeological Science 39: 492–99. http://dx.doi.org/10.1016/j.jas.2011.10.004 Google Scholar
Tang, H.S. 2012. New discovery of rock art and megalithic sites in the central plain of China. Rock Art Research 29: 150–70.Google Scholar
Tuan, Y.F. 1977. Space and place. Minneapolis: University of Minnesota Press.Google Scholar
Wang, K.R., Qiu, Z.L. & Chen, Y.Z.. 1988. Guangxi Zuojiang yanhua [Guangxi's Zuo River rock art]. Beijing: Wenwu.Google Scholar
Wang, N.S. 1985a. Rock art paintings of Yunnan, China. Expedition 27 (1): 2533.Google Scholar
Wang, N.S. 1985b. Yunnan Cangyuan yanhua de faxian yu yanjiu [The rock paintings of Cangyuan, Yunnan: discovery and research]. Beijing: Wenwu.Google Scholar
Zeng, Z.M. 1990. Gongxian ‘Boren’ xuanguan yanhua yanliao de jianding [Identified pigments in rock art and hanging coffin of ‘Bo People’ nation, Gong County]. Kaogu yu wenwu 2: 107–10. http://dx.doi.org/10.1017/S0003598X00099737 Google Scholar
Zhang, C. & Hung, H.-C.. 2010. The emergence of agriculture in southern China. Antiquity 84: 1125.Google Scholar
Zheng, C.X. & Tan, F.. 2006. Zhuangzu lishi wenhua de kaoguxue yanjiu [Archaeological research of Zhuang history and culture]. Beijing: Minzu.Google Scholar
Figure 0

Figure 1. The Zuo River sites (adapted from Google Earth).

Figure 1

Figure 2. Guangxi Zhuang Autonomous Region and Pearl River system.

Figure 2

Figure 3. Zhushan cliffs from the fields of Zhushan village, Ningming, Guangxi (arrows indicating painted panels).

Figure 3

Figure 4. Painted panel and cave at Gaoshan, Huashan, Ningming, Guangxi.

Figure 4

Figure 5. Huashan site in 2010, Ningming, Guangxi (arrows indicating painted panels).

Figure 5

Figure 6. Painted panel next to scaffolding, Huashan, Ningming, Guangxi.

Figure 6

Figure 7. Chengxiangjiao, Longzhou, Guangxi.

Figure 7

Figure 8. Yanhuaishan, Longzhou, Guangxi.

Figure 8

Figure 9. Chaochuanshan, Longzhou, Guangxi (arrow indicating painted panel).

Figure 9

Figure 10. Paintings and cave at Duchuanshan, Longzhou, Guangxi.

Figure 10

Figure 11. Cave and paintings at Longzhou Huashan, Longzhou, Guangxi.

Figure 11

Figure 12: Gourd-shaped grotto at Lushan, Pingxiang, Guangxi.