Buddhist Boundary Markers of Thailand & Laos


A review of The Buddhist Boundary Markers of Northeast Thailand and Central Laos, 7th – 12th Centuries CE: Towards an Understanding of the Archaeological, Religious and Artistic Landscapes of The Khorat Plateau, by Stephen A. Murphy.

Stephen Murphy’s dissertation presents an in-depth analysis of the archeological and historic evidence for the earliest form of Buddhist Boundary markers (sema) in Southeast Asia. Following a preliminary discussion concerning existing views concerning sema in Southeast Asia and Northeastern Thailand in particular (Introduction – Chapter 1), he examines in depth the archaeological and historical evidence for the culture, artistic, and political milieu for the kingdom of Dvaravati (Chapter 2). He then defines what sema are and how they function (Chapter 3), which is followed by a comprehensive survey of existing sema taking into consideration their physical and material type, location, distribution by group, by cluster, and artwork (Chapter 4). The sema that have either religious imagery or narrative compositions are examined in light of the sources for the narrative within either Buddhist or Brahmanical texts and their possible meanings and associations are explored (Chapter 5). The sema are divided into types based on the above research and a tentative chronology and sequence of evolution is suggested (Chapter 6). Murphy concludes with a summary of his research and the conclusions that can be derived from it (Chapter 7) and in the appendixes presents databases in table format for the sema studied in the dissertation, a typological database of the sema, and chronological tables. There is also a glossary of terms and bibliography of works cited.

Stephen Murphy departs from the approach in studying the sema of the Khorat Plateau of Northeast Thailand by considering them in a regionally distinct political, religious, and artistic context as opposed to being a regional variation of the Dvaravati culture of Central Thailand. By considering the sema and culture of the Khorat Plateau as the product of a parallel evolution that while sharing many cultural, religious, and artistic elements in common with Central Thailand, while remaining fundamentally separate and distinct, the inherent pitfalls and bias of approaching the study of the region as derivative are avoided and a clearer and unbiased interpretation of the historical and archaeological data can be conducted. The resulting conclusions support the thesis that the culture of the Khorat Plateau was a distinct region independent of the Dvaravati culture of Central Thailand.

In the Introduction, Murphy introduces sema as the markers that mediate sacred space and the earliest archaeological evidence of these boundary markers in Southeast Asia comes from the Khorat Plateau, which today consists of modern Northeastern Thailand and parts of Central and Southern Laos. Dating from the 7th-12th centuries, they fall into the period generally referred by Thai art historians as the Dvaravati Period after the major early civilization in Central Thailand. The author then lists seven points to be covered in his dissertation research:

(1) As no comprehensive database of sema exists, the first task is to create a database containing basic information such as distribution with detailed maps, types, numbers, styles and artwork, if any. The database will then allow the creation of a typology allowing the identification, comparison, and relative dating of the sema.

(2) To remove the academic bias in archaeological and art historic literature with regards to the sema. At present the majority of literature only discusses sema with artwork or sema from a few well known sites and most of the studies conceive of and understand sema only in relation to central Thailand and Dvaravati culture in general. Murphy sets out to shift the focus to the sema of the Khorat Plateau and then to ask how much influence and appropriation is evident from the neighboring Khmer and Dvaravati civilizations. His thesis sets forth the argument that the Khorat Plateau is a distinct region in itself and that its culture, history, and archaeology should be analyzed from this perspective.

(3) The question of whether there are obvious patterns in sema distribution must be addressed, such as: Is it possible to locate any center or centers of the sema culture? What types of sites are they found at? What is the relationship between the distribution of the sema and the art depicted on them? What is the relationship of sema to the landscape, both physical and cognitive?

(4) As there is presently no comprehensive database of sema, the study of the iconography and the artwork has been done on an individual basis with few comparisons between different sites and regions. This thesis analyses the art in conjunction with the distribution analysis of sema throughout the Khorat Plateau and draws its conclusions from this perspective, raising further questions to be addressed in terms of development and relationships between regions.

(5) Several scholars have suggested that sema culture of the Khorat Plateau evolved out of a preexisting megalithic culture. The conclusions drawn from the database created in the preceding section will allow this theory to be placed in proper context and challenged.

(6) The sema represent the most comprehensive surviving evidence of early Buddhism in the region and as they are large and not easily transported any distance, remain for the greater part in their original locations. Because of this the quantitative methods and analysis of the distribution patterns of sema and their relation to settlements, the spread of Buddhism and differences in practice can be studied across the Khorat Plateau.

(7) This thesis seeks to bring together both Thai language site-reports, journal articles and conference proceedings with Western academic sources to provide a bridge towards the integration of Thai and Western scholarship on the sema culture and the Khorat Plateau.

Following the structural outline of his dissertation, he discusses the theoretical approaches underlying his research and the sources that he used to formulate his approaches. The archaeological approach broadly falls under the classification of landscape archaeology, but goes beyond the physical landscape to consider the sema in terms of the religious and cognitive landscapes in which they existed. He then discusses the two types of art historical approaches that he used, one concerned with style and dating, the other concentrating on the motifs and iconography found, with the latter being subdivided into the iconography of narratives and motifs. He concludes his introduction with a description of his fieldwork methodology and the data collected to document each sema stone. He assigned each stone a specific number, recording its current location and original site, if different. Each sema was then measured, sketched, and described, followed by a high resolution digital image. The site was then described as to its type (moated, earthen mound, etc.), along with any associated material, and finally the precise location was recorded using a GPS device. The site could then be further analyzed using Google Earth. The data recorded for each sema was then entered into an ArcGIS software application which allowed for the creation of accurate digital maps. The datasets produced are included as an appendix, the photographs of the sema and site locations are included on a DVD.

Chapter 2 explores the definitions and evidence for the cultural, artistic, political, and archaeological evidence for the culture referred to as Dvaravati in order to provide a backdrop against which to analyze the sema from the Khorat Plateau. As noted earlier, most scholars have analyzed the sema tradition as a peripheral reflection of Dvaravati art and culture. However, it can be concluded that Dvaravati was politically restricted to sites in Central Thailand. As there was no direct political, economic, or religious control exerted by Dvaravati over the Khorat Plateau, while its culture and art spread via trade, economic activity, and the spread of Buddhism, the local population freely appropriated specific elements and reshaped them to meet their own needs. What is important is that the influences that helped shaped and define the sema tradition were not imposed from a central power, but selected and adapted from within the Khorat Plateau region. The major population centers in Central Thailand controlled the immediate area around them and further away by vassal/tributary arrangements that became weaker with distance. The same basic political structure is evident among centers at the Khorat Plateau with large sites such as Muang Fa Daed and Muang Sema having considerable political reach and influence, which also extended into the art of the sema tradition. By defining the distinctive material culture of Dvaravati, the presence of Buddhism, the Dvaravati art style, and oblong sites plans the reference points needed to understand the uniqueness of the sema tradition of the Khorat Plateau can be established. Based on this, the author argues that a shift in perspective is required to fully appreciate these objects as a distinct art separate from that of the Central Thai Dvaravati civilization.

In chapter 3 Stephen Murphy draws upon a range of different types of evidence to reach an understanding of what sema are and how they functioned. Buddhist textural evidence cites a need to separate sacred space within which specific rituals could be carried out. This space was demarcated by objects, nimitta, that defined the boundary. The texts list a variety of natural features and manmade objects that could function as nimitta, but it is never precisely defined or limited to a single type of object. Turning to archaeological evidence from the Khorat Plateau during the 7th through 12th centuries, it is clear that in this region, large carved stones of sandstone, and more rarely laterite, were used to demarcate sacred space.  These stones became known later among the Thai as sema. It is unclear as to what other Buddhist centers outside of the Khorat Plateau used for nimitta, but they probably used natural features or perishable materials. One unique site in Burma has fossilized wood used for sema, at Thaton; another unique structure is surrounded by eight boundary stones.

As no texts concerning sema survive from this early period, scholars have looked to project backwards modern Buddhist practice. In recent times, the Thai have used either eight or sixteen sema to consecrate and demarcate a square or rectangle around an ordination hall and thus create a khandasema or sema defined by the placement of stones. However archaeological evidence from Dvaravati sites show differing usages, configurations and placements suggesting that at this time the function of sema was perhaps not firmly fixed and a degree of flexibility existed. Sema may have functioned to fix sacred space where no buildings were present or built of perishable materials. As there was no monumental architecture, perhaps sema played a primary role in defining the religious landscape. Inscriptional evidence is limited, but it appears that sema consecration ceremonies were sponsored by local dignitaries or rulers and the stone was sometimes donated by royalty. Brahmins may have been employed in the ritual and merit was acquired by those who made the dedications. From this it may be inferred that sema served not just a religious function, but were also served a social and political purpose. Sema have been widely reused in different contexts during the immediate past and are still considered to be important religious objects. As there is a fluidity of meaning associated with the reuse of sema today, it may be suggested that it would be wise not to restrict our understanding of sema to a narrow definition of an object used to create and demarcate sacred space when considering the Dvaravati period. Lastly, by reviewing how modern scholarship has understood and approached sema, it is clear how certain explanations, classifications and theories arose and have remained in use through the present day. With this knowledge their basic assumptions can be tested, especially the viewpoint of seeing sema as the product of Central Thailand as opposed to the Khorat Plateau and the opposing theory that these sema arose from a megalithic cult.

Chapter 4 looks at the sema in terms of the physical and cognitive landscape of the Khorat Plateau by examining the patterns revealed by their relationship to the geography of the region, to settlement patterns, and the distribution of their artwork. By dividing the sema into three distinct groups and eight clusters, a clearer picture emerges of the Khorat Plateau sema culture.

Geographically, the Chi River system has the highest proportion of both sites and sema, and also has the finest examples of carved sema, not only in terms of artistic quality and iconography, but also in terms of aesthetics. Within this region, clusters one and two represent major centers of the sema tradition. Cluster six comprising sites around Vientiane, Loei, Udon Thani, and Nongkhai is an important grouping that demonstrates the role played by the Middle Mekhong River in spreading the tradition beyond the confines of the Central Khorat Plateau. Cluster six further demonstrates that the sema and the Buddhism that they represented were by the seventh-eighth centuries an integral part of the region. This tradition continued in the Middle Mekhong region well into the eleventh and twelfth centuries, gradually displaying influence from Khmer culture. In contrast, the sema tradition of the Mun River system never developed and flourished as strongly as it did in the Chi and Middle Mekhong region. The early presence of Khmer culture and the accompanying Khmer culture meant the sema tradition and the Buddhist religion that it represented never achieved a strong foothold in the region. While Settlement Sites with moats in the Chi River system demonstrate a direct connection with sema locations, this is not the case with all site sema are found.

There are clear patterns in the distribution of motifs. Narrative art is restricted to a handful of key locations, primarily in clusters one, two, and six. In contrast, axial stupa and stupa-kumbha motifs have a much more wide distribution and might be considered to represent the “typical” motifs of the sema tradition. In conclusion, this chapter shows the pattern of distribution of Buddhism within the region. It was well established by the eighth-ninth centuries and proposes that it was centered in particular on cluster one in the Chi River system and that it spread along the major river systems and as a result developed in the lowland alluvial plains.

In chapter 5, Murphy investigates the identification and interpretation of the narrative Buddhist and Brahmanical scenes on the sema stones, following the rich range of aniconic imagery found on sema from the Khorat Plateau such as axial stupas, stupa-kumbha motifs, dharmacakra, lotus petal bands, and “cloud” motifs. Visually the most striking of all the remains of the Dvaravati Period culture of the Khorat Plateau – sema with images of the Buddha or bodhisattvas have received the most attention from both academics and members of the general public. Defining narrative art “as artwork that portrays a specific scene, event, episode or tautological element” (p. 210) he explains that for this tradition, the narrative scenes almost always portray subjects from the Jataka Stories or scenes from the Life of the Buddha. In almost all cases a single direct scene is portrayed, without continuous narration or combined subjects or interpretations.

The style of the figures in sema with narratives and figural representations is close to that of central Thai Dvaravati images, although the arched eyebrows that join at the center are not as prevalent as in central Thailand. Also, the double vitarka mudra of Dvaravati art in central Thailand is found, though it is more common to find just the single right hand raised in vitarka mudra. Another distinctive characteristic of bodhisattva images on the Khorat Plateau is the “drápe-en-poche” handling of the robe, which is unique to the Khorat Plateau and distinct from both central Thai Dvaravati and Khmer practice.

Using an approach which combines changes in style over time with distribution patterns, Murphy has arrived at three major stylistic divisions that correspond to three chronological periods. The first group is centered in the Chi river system and includes Muang Fa Daed. Spanning in time the eighth-ninth centuries, it has strong affinities with the Dvaravati art of central Thailand and there is little or no discernible Khmer influence. The second group corresponds to the tenth-eleventh centuries and is characterized by “…a fusion of Dvaravati and Khmer artistic motifs and conventions…. The drápe-en-poche of the first group gives way to a full Khmer style sampot and the facial features of the protagonists take on a more square appearance, contrasting with the oval faces of the 8th-9th centuries” (p. 213). The third group dating from the eleventh-twelfth centuries is found primarily within cluster 6 sites in the Vientiane province of Laos and Nong Khai, Loei, and Udon Thani provinces of Thailand, as well as one independent site that does not easily fit within any cluster. The style of this group have lost all stylistic characteristics of central Thai Dvaravati art and instead are executed in the style of the figures on provincial Khmer architectural lintels.

In discussing possible textual sources for the Jataka Stories depicted on sema, Murphy looks at existing scholarship regarding the identification of Jatakas in contemporaneous narrative sculpture, not only on the Khorat Plateau and central Thailand, but also sites such as Borobudur in Indonesia. He concludes:

“the idea that there was a homogeneous, uniform type of Buddhism being practiced at this time is also misleading with archaeological and art historic evidence pointing towards a plurality of religions, with Mahayana Buddhism and Brahmanism being practiced alongside Theravada doctrines. The presence and importance of the oral tradition can also not be overlooked and it is possible that many of the jataka scenes found throughout the Khorat Plateau were transmitted by word of mouth as opposed to palm leaf manuscript” (p. 216).

While using as his primary textural reference Cowell’s six volume translation of the Jataka-atthakatha for researching Jataka iconography, he makes it clear that, while there is a close correlation between the text and the Jataka narrative scenes on the sema stones, it cannot be assumed that this was indeed the text that was followed in their creation (E. B. Cowell, The Jataka or Stories of the Buddha’s Former Births. Delhi: Cosmo Publications, 1978.).

After this introduction, Murphy proceeds to discuss each narrative and figurative sema stone that represents or might represent a Jataka Story grouped by subject beginning with the Mahanipata (literally “Great Section,” the Buddha’s last ten re-births as a bodhisattva prior to his re-birth as the historical Buddha) which comprise the majority of the scenes illustrated on the sema and continuing through less common Jataka Stories (unidentified scenes that can be identified as most likely being from a Jataka). Each sema is illustrated with a clear photograph and presented with its provenance (if known); associated Jataka Story with the key identifying iconographic characteristics listed, and discussed if necessary for clarity, textual attributions or non-typical iconography or presentation; and finally a date range determined on stylistic grounds and grouping. The organization makes it possible to compare different renditions of the same subject comparatively, evaluate the attribution, and compare the regional and temporal variations for each subject.

As with the Jataka Stories, the scenes from Life of the Buddha that are depicted on sema stones are drawn from both Pali (Theravada) and Sanskrit (Mahayana) texts, and it is impossible to know precisely which text might have been the inspiration for a specific scene. Ten specific scenes from the Life of the Buddha have been identified with reasonable certainty, and are discussed in the order that they occurred during the Buddha’s life. These are followed by a group of sema stones that most likely illustrated episodes from the Life of the Buddha, although their precise identification is problematic. This portion of the dissertation is fully illustrated and organized like the preceding portion on Jataka Stories.

Many sema stones depict single images of the Buddha or a bodhisattva. These are discussed and presented in same manner as the semas illustration, Jataka Stories, and scenes from the Life of the Buddha next with good photographs, iconographic description, and stylistic analysis and dating – with the exception that sema that stylistically influenced by Khmer art are covered at the end of the section. This leads naturally into a section on sema stones that have Buddhist subjects influenced by Hindu or Brahmanical religious ideas, including images of Lakshmi, Durga, Garuda, Indra, Surya, as well scenes from the Ramayana. Again, each sema is fully illustrated, fully described, and placed in context. At the end of this section sema stones from Cambodia and Burma are also illustrated and discussed comparatively.

Chapter 5 continues with a discussion of aniconic imagery on the sema stones from the Khorat Plateau. As sema with narrative and figurative art make up just ten percent of the existing sema in the region, the discussion of sema with aniconic symbols is extremely important in achieving an overview of the sema culture of the Khorat Plateau. Murphy begins this portion of his dissertation by presenting briefly the major forms of sema stones with axial stupa designs, which are the most common iconographic motif found on sema in the Khorat Plateau. This motif is only depicted on slab type sema stones and may be present on one side only or both the front and rear. It is usually a relatively narrow ridge, perhaps ten centimeters wide bisecting the sema stone down the middle, though sometimes it is broader and more triangular. Occasionally it will rise out of a scene below.

The next most common motif is the “stupa-kumbha,” which is found at thirty-four sites throughout the Khorat Plateau. The kumbha is depicted either as a plain water pot or a more elaborate form that resembles a purnaghata or kalasha pot, which in mythology contained the elixir of life, amrita, and is symbolic of abundance and auspiciousness. On top of the kumbha, a stupa motif is depicted, sometimes with concentric rings that most likely represent the chattravali on actual stupas, or occasionally, an elaborate finial. This motif was very popular and widely distributed in at both central Thai Dvaravati sites and the Khorat Plateau in various mediums. Murphy has identified five general types of the stupa-kumbha motif. The first type is the most basic form of the motif and has no floral decoration and consists only of a plain kumbha with a plain stupa arising from it. It is found at three sites in the Chi river system and Middle Mekong. The second type has increased decoration, with the kumbha surmounted by petals and a plain stupa and is found throughout the Khorat Plateau. The third type is more elaborate and is restricted to two sites in Cluster 4. It has either an elaborate base below the kumbha or a double kumbha design. On top of the kumbha there is an elaborate foliate design which can take several forms, even rising and following the form of the stupa on top of it. The stupa is also more elaborate and has a finial that might take the form of a series of concentric rings, a trident or a four-spoked wheel that might symbolize a dharmacakra. Type four is similar to both types two and three with the exception that it has either a kumbha motif or a dharmacakra placed in the middle or upper part of the stupa, and like type three, also comes from Cluster4. Type five seems to have evolved out of types three and four and the design and depiction of the foliage has Khmer artistic traits.

Type five is restricted to a few sites in Cluster 6 of the Middle Mekhong Group. Murphy has proposed a sequence for the evolution and dating of the stupa-kumbha motif moving from the simple to the more complex, with the additional support of including the relative dating of the narrative sema from the same sites. The first is the basic axial type stupa that evolves into the basic stupa-kumbha design of types one and two, evolving to reach its apogee in types three and four. The overly ornate motifs and strong Khmer influence in type five represents the decline of the form. This is supported by comparison with sema with narrative scenes within the same clusters: “Types one, two, three, and four all come from sites with a date range of 8th-9th centuries. They are therefore contemporary with the narrative art of group one. Furthermore, types two, three, and four all come from Cluster 4. These factors emphasize a lateral relationship between the motifs. It is therefore difficult to maintain that one type of motif may have post-dated or pre-dated another” (p. 323).

The only type where it is possible to make a clear chronological separation from the other types is type five, which is an amalgamation of all four previous types and can be dated through its Khmer style foliate motifs to the late 10th-11th centuries. This would place type five contemporary with the narrative art Group 2. Murphy then presents background research into the stupa-kumbha motifs, focusing on the different interpretations of the kumbha and stupa within the Buddhist textual traditions. He proposes that the stupa-kumbha motif is: “a representation of the Buddha in all of his aspects and meanings… [the axial stupa] too simultaneously represented the image of the axis-mundi as stupa, the stupa as the Buddha, and the Buddha as Dharma. Consciously the Buddhist communities saw the stupa-kumbha and axial stupa motif as aniconic forms of the Buddha…” (p. 327).

The dharmacakra has come in Buddhism to represent the Buddha’s first sermon, the first turning of the Wheel of the Law. Three-dimensional dharmacakra raised on pillars or stambhas is one of the defining artistic and religious characteristics of the Dvaravati culture of central Thailand and has been found at all major sites. Three-dimensional dharmacakra have been found at sites on the Khorat Plateau, but in far fewer numbers than in central Thailand. Three sema in poor condition have also been found with dharmacakra motifs on the Khorat Plateau, coming from one site in Cluster 5 and two sites in Cluster 3. All three depictions are different, though their spokes appear to be stylized lotus leaves. One appears to have a dharmacakra on top of an axial stupa; the others have a dharmacakra on top of a kumbha, as if it is emerging from the it. Five sema with stupa-kumbha motifs have a dharmacakra as the finial of the stupa. Surveys of the sema with dharmacakra from the sites on Phnom Kulen, Cambodia record that there were thirty-two sema of which thirty had a dharmacakra motif on one side and most had stupa motifs on the other. Murphy suggests:

It appears therefore that the stupa-kumbha and the dharmacakra motif developed alongside each other in the Khorat Plateau from circa 8th century onwards. The examples found at Phnom Kulen would appear to post-date those from the Khorat Plateau as they illustrate a combination of fully formed stupa-kumbha and dharmacakra motifs. They most probably date from the 9th century onwards” (p. 333).

Many sema, no matter what motifs they depict and with or without narrative scenes, have their bases carved with a lotus band motif. In many cases, the lotus band is the only decorative motif on the sema, a few sema have plain bands at their bases instead. There is a variety of types and styles, with some sema from the 11th-12th centuries having a more angular, Khmer style lotus band. The lotus band is also found on many Dvaravati religious objects from central Thailand, especially dharmacakras and dharmacakra stambhas. Other motifs encountered on sema from the Khorat Plateau, but not common are ornate floral bands around the base, trident motifs, and cloud motifs.

In the conclusion to Chapter 5, Murphy suggests that by associating the sema with dates and locations, it is possible to propose the existence of workshops or schools. The first and most prolific of these was at Muang Fa Daed. Archeological evidence indicates that Muang Fa Daed was one of the most important centers in the region, its economic and agricultural resources would have supported the skilled craftsmen and supplied them with the resources they needed to produce numerous, high quality sema stones. Most of the surviving sema with narrative scenes from the Life of the Buddha and the Jataka Stories were produced there over a period of roughly four hundred years, from the 8th-11th centuries. A second major workshop must have been centered on the site of Bahn Ngong Hang, also located in Cluster 1 and most likely was an offshoot of the workshop at Muang Fa Daed. The sema stones from this site are very similar to those produced at Muang Fa Daed, so close that the same artists might have made them. They both date from the same period of time and only Muang Fa Daed and Bahn Ngong Hang produced a tapered type sema. In Cluster 2, the site of Bahn Kut Ngong was possibly a workshop, but on a much smaller scale and only during the eighth through ninth centuries. A similar time span applies to the site of Bahn Korn Sawan, which was also possibly a workshop. The few sema that were produced here are in lower relief, suggesting a small scale workshop with artisans trained at one of the larger workshops. Cluster 4 had a workshop centered around Bahn Tat Tong and Bahn Kum Ngoen which produced a number of sema stones with the stupa-kumbha motif from the eighth through ninth centuries, but no narrative art or Buddha images. Murphy suggests that, based on the variety and number of well executed stupa-kumbha motifs produced at this site, this was the workshop from which the design originated and spread to other sites in the cluster and beyond. During the tenth or eleventh centuries a workshop appears to have emerged at Bahn Nong Khuem and Bahn Pailom where both the Dvaravati and Khmer artistic traditions were merged into a distinctive and original aesthetic comprising narrative scenes from the Jatakas based on the main sema traditions, but executed with the more stylized and refined forms of the Khmer aesthetic. The site of Wang Sapung also seems to have flourished briefly producing sema with the stupa-kumbha motif of the Chi River system fused with Khmer floral conventions. Murphy concludes that, apart from these six sites, there is little evidence for workshops or schools at other locations, and while sema may have been produced at many other locations, these were most likely centered around local Buddhist monasteries and the artisans may have been the Buddhist monks themselves.

Murphy believes that these six sites, and Muang Fa Daed in particular, may have functioned similarly to the ‘restricted-centers-diffusion rule’ hypothesized by Robert Brown in which a few centers developed and transmitted the art forms throughout the region (Robert Brown, “‘Rules’ for change in the transfer of Indian art to Southeast Asia.” In Ancient Indonesian sculpture, edited by Marijke J. Klokke and Pauline Lunsingh Scheurleer. Leiden: KITLV Press, 1994, pp. 10-32, 12-14). This would explain the profusion and spread of the sema tradition from workshops in the Chi river system and Middle Mekhong throughout the Khorat Plateau.

As expressed by the author, Chapter 6 “provides a typology of sema arrived at from the study of their form, style, material, distribution, artwork and epigraphy” (p. 344). The sema are logically organized in the typology by the four major types of sema: slab, pillar, octagonal, and un-fashioned. Each main type is clearly described and defined, with their average range of measurements. The discussion of the pillar and octagonal types of sema are followed by maps showing their distribution. No maps are supplied in this chapter for the distribution of the slab type sema as they are fully mapped in Chapter 4. After each major type is introduced and defined, its dependent subtypes are presented and discussed, with nine subtypes under slab-type sema and four each under pillar- and octagonal-type sema. Each subtype is clearly defined with their geographic distribution and the range of decorative motifs present on them, if any representative photographs of the subtype are supplied. If measurements are significant to the subtype, average measurements, as well as the ratio of the measurements, are listed. The reasons for assigning dates are also clearly described and discussed, as are inter-regional and external influences, and the presence or lack of inscriptions and stylistic associations.

Two categories usually not considered by scholars are also discussed: un-fashioned and unfinished-type sema. Un-fashioned sema are very crude in their shaping, but usually conform to the rough shapes of the slab and pillar types. Some sema in this category might originally have been fashioned but then eroded over the centuries, but there do appear to be a group of sema that were minimally shaped and that span the entire region. No definite conclusions can be drawn concerning them. The unfinished type of sema seem to have been intended primarily to be fully fashioned and decorated slab sema, though a few pillar type sema also fall into this category. They are classified unfinished as the surface design was sketched on them but never fully carved.

The discussion of types concludes with a proposed dating sequence and evolution of each of the different types of sema. This is described for each one of the three major types and is illustrated with a photographic flowchart arranged in a proposed chronological order. A short discussion of the major forms of post-Dvaravati sema follows, briefly looking at sema from Angkor, and the sema of the Thai cities of Ayutthaya, Sukhothai, and Phetchburi.

Murphy then begins a discussion of the theory that a megalithic cult might have given rise to the sema culture of the Khorat Plateau. The idea that the sema of the Khorat Plateau developed from a pre-existing megalithic culture was first introduced in 1969 by Quaritch Wales (H. G Wales Quaritch, Dvaravati: The Earliest Kingdom of Siam [6th to 11th century A.D.]. London, Bernard Quaritch Ltd, 1969, 109-111). This theory was strongly criticized by Piriya Krairiksh in 1974 (Piriya Krairiksh, “Semas with Scenes from the Mahanipata-Jatakas in the National Museum at Khon Kaen” in Art and Archaeology in Thailand. Bangkok, Fine Arts Department, 1974, 35-100, 43), but revived by the artist and newspaper writer Prayuun Uluchada writing under the pen name No Na Paknam in 1981 (No Na Paknam, The Buddhist Boundary Markers of Thailand. Bangkok: Muang Boran Publishing House, 1981, 61) and expanded upon by Srisakara Vallibhotama in 1985 who went on to suggest that the sema culture that evolved on the Khorat Plateau was the result of the transformation of an indigenous megalithic burial cult with the introduction of Buddhism, and that the sema stones were the Buddhist equivalent of the megalithic stones associated with pre-Buddhist funerary practices (Srisakara Vallibhotama, “Sema Stone Boundary Markers from the Northeast: Survey and the Study on the Continuation of Megalithic Culture in the Region,” Muang Boran 11 (4), 1985: 6-33).

In presenting evidence to refute this theory, he briefly examines the colonialist “Diffusionist” theory which held that civilizations spread out from the more advanced to the more primitive, bringing with them more advanced technologies, religions, and other cultural elements. It is this theory that underlies the initial hypothesis presented by Quaritch Wales and then further elaborated upon by Thai scholars. He illustrates how this theory has come under criticism as not always being adequate or applicable to Southeast Asia by scholars such as Ian Glover (Ian C. Glover, “The Archaeological Past of Island Southeast Asia” in Messages in Stone – Statues and Sculptures from Tribal Indonesia in the Collection of the BarbierMueller Museum, edited by Jean Paul Barbier and Milan Skira. Geneva, the Barbier-Mueller Museum, 1998, 17-34, 23-25). He further explains that his fieldwork has shown that all of the “megaliths” listed in early archeological surveys of the region were actually sema whose purpose had been misidentified. The closest evidence for the presence of an authentic megalithic culture that has been properly documented is found in the highlands of Laos. These include 150 standing stones in Hua Phan province, which are accompanied with both cist and dolmen type burial that were documented by M. Colani who felt that they might predate the Plain of Jars, giving a date of 300 BCE (M. Colani, “Mégaliths du Haut-Laos,” Publications de l’Ecole Française d’Extrême-Orient [PEFEO]. Paris, Les Editions d’Art et d’Histoire XXV-XXIV, 1935.) Additional sites were later found by Kanda Keosphha where standing stone megaliths were clearly associated with burials (Kanda Keosphha, “Standing Stones in Northern Lao PDR” in Uncovering Southeast Asia’s Past-selected papers from the Tenth Biennial Conference of the European Association of Southeast Asian Archaeologists at The British Museum, London, September 2004, , edited by E. Bacus, I. Glover, and V. Pigott. Singapore, National University Press, 2006, 148-153). Murphy then concludes that the closest megalithic predating the sema culture of the Khorat Plateau is restricted to the highland culture of Laos. While Rungroj Piromanukul and Kanda Keosphha both argue that the presence of a megalithic culture in highland Laos is enough to justify the theory that the lowland sema culture of the Khorat Plateau evolved from it (Rungroj Piromanukul, Rungroj, “Les bornes rituelles du nord-est de la Thaïlande” in Dvāravatī : aux sources du bouddhisme en Thaïlande. Paris, Musée Guimet, 2009, 97-104), Murphy counters with the argument that if the megalithic culture of highland Laos influenced the development of the sema culture in the lowland areas of the Khorat Plateau, then one must ask why no comparable megaliths have been found in the Khorat plateau and that the sema culture of the Khorat plateau or a related culture did not evolve in highland Laos to succeed the megalithic culture there.

Discussing archeological evidence concerning megaliths in the immediate vicinity of the Khorat Plateau, the evolution of scholarship concerning megaliths in the region, and the theses linking a pre-existing megalithic cult to the Dvaravati sema culture of the Khorat Plateau, he concludes:

“A combination of misidentification, a rather loose and ambiguous use of the term ‘megalith’, the remnants of a colonial Diffusionist viewpoint and universalist ideas of the evolution of religions have led many scholars to accept the theory that sema evolved out of an indigenous forerunner in the shape of a widespread megalithic cult. However, the archaeological evidence on the ground paints a very different picture. No evidence whatsoever has been found to date for a megalithic culture in the Khorat Plateau with the vast majority of claims of this nature being misidentifications. Instead it appears that the sema tradition arose as a result of Buddhism entering the region and the need to demarcate sacred space. In fact, a counter-argument to the more universal ‘religious fusion/synthesis argument’ can be proposed along the lines that, due to their monumentality and visual impact in the landscape, sema did not represent a tie with the past but instead were employed to represent a clean break with former traditions and situate the new incoming religion in a dominant and permanent manner” (p. 372).

In his summary of Chapter 6, Murphy states that the origins of the sema tradition may not be solved, as evidence from Sri Lanka before the seventh century is inconclusive and there exist sema with solid dates only from after the twelfth century. The evidence for influence from a megalithic culture is inconclusive and appears unlikely. Despite compiling a database of sema stones, it has been impossible to narrow the date range beyond one or two centuries, and in some cases the date range spans two to three centuries. The only thing that is clear is that, by the seventh century, sema had become an important part of the culture of this region. By the eighth-ninth centuries it developed substantially in the Chi river system and could also be found in the Middle Mekong. Both regions were producing sophisticated designs with local characteristics. This tradition would continue to develop and sema would become an important part of Buddhism in the later Thai states of Sukhothai and Ayutthaya.

Chapter 7 is the conclusion of the dissertation which brings together the questions addressed in the study and then outlines the results achieved in the foregoing chapters. This is followed by a discussion of the insights that these results have revealed relative to both the sema tradition and the nature of Buddhism in the Khorat Plateau during the 7th-12th centuries.

To summarize, by organizing the sema into three distinct groups and eight clusters, Murphy reveals that the Chi river system was the most important area relative to the sema tradition, with Clusters 1 and 2 being the most significant centers of sema production. The concentration of sites in Cluster 6 demonstrates the strength of the tradition in this region and the role of the Middle Mekhong as a venue for its spread away from the Chi river system. The sema tradition never took hold to a significant extent in the Mun river system, most likely because of the strong influence of the predominately Hindu Khmer civilization in the region. Clear patterns have emerged regarding the distribution of motifs on sema throughout the Khorat Plateau. Narrative art is restricted to a few locations in Clusters 1, 2, and 6, while axial stupa and stupa-kumbha motifs are spread throughout the region. In identifying the iconography of the narrative scenes, seventy-one identifications are proposed. The possible symbolic significance of the axial stupa and stupa-kumbha motifs are also explored, with the suggestion made that they represent aniconic images of the Buddha and subsequently the Buddha as Dharma. Through the study of the style and evolution of motifs, combined with the distribution analysis presented in Chapter 4 and the typology presented in Chapter 6, the sema were dated. The result suggest that the sema did not evolve in a traditional linear manner, but more “laterally” when important aspects such as location, cultural influences, and settlement patterns were factored in. As a result, narrative art on sema can be divided into three groups by periods corresponding to the 8th-9th, 10th-11th, and 11th-12th centuries; and the stupa-kumbha motifs can be placed into five types dating from two periods: Types 1-4 being contemporary with narrative Group one and dating from the 8th-9th centuries, while Type 5, which demonstrates a fusion of Khmer and Dvaravati art, corresponds to narrative Group 2 dating from the 10th-11th centuries. Finally, the sema demonstrate that the Buddhist religion was present by the 8th and 9th centuries, and most likely centered on Cluster 1 in the Chi river system. It is also clear that Buddhism spread along the major river systems and thus primarily developed in the lowland, alluvial plains.

Murphy’s insights and observations from this thesis are:

Firstly: the tradition of using large stone boundary markers appears to have originated in the Khorat Plateau. This tradition spanned the 7th-12th centuries with the 8th-9th centuries being the period in which the majority of narrative art was created and the tradition began to spread throughout the region.

Secondly: The tradition flourished and reached its apogee in the Chi river system, specifically at the site of Muang Fa Daed during the 8th-9th centuries. The distribution analysis, numbers of sema present, and the study of the artwork confirms this conclusion, though it does not necessarily mean that the tradition originated at the site.

Thirdly: The sema tradition represents a unique phenomenon of the Khorat Plateau. It is clear that the aesthetic of the artwork is not derivative of Khmer or Dvaravati culture, but developed as a distinct form in the region, though it drew at times upon different cultures.

Fourthly: Sema can help illuminate how Buddhism emerged and spread through the region. It is clear that Buddhism spread along the course of the rivers and was primarily restricted to the lowlands. The rivers provided good communication networks and moated city sites supplied the patronage that Buddhism needed to thrive. Sema demonstrate that later Buddhism spread to less populated sites and even suggests the presence of a forest dwelling lineage of monks at Phu Phra Baht.

Finally: The question of which types of Buddhism were being practiced and the texts they employed can be explored, but not definitively answered. The presence of a khakkharaka (an ascetics staff) on some narrative scenes has been interpreted by some to indicate the presence of the Mulasarvastivadin sect of Sanskrit Buddhism, but the khakkharaka is not exclusive to this group alone and thus other lineages could have created one or all of these sema with this symbol. The presence of stupa symbolism on sema has led some scholars to associate it with stupa worship and thus a connection with the Apara-mahavinaseliya sect that was active at Nagarjunakonda from the 3rd century on. However, almost all lineages of Buddhism have the stupa as one of the most important symbols in their religion. While there are close connections between the episodes from the Life of the Buddha depicted on sema from the Khorat Plateau and the Pali-based text of the Nidana-katha, as well as further similarities between the Pali Jatakas and the Jataka scenes found on the sema stones, both the use of oral traditions and different texts cannot be ruled out. While we cannot be sure about the texts and precise Buddhist lineages present on the Khorat Plateau at this point in history, the presence of “Ye Dhamma…” inscriptions at a number of locations, as well as the lack of bodhisattva images carved on sema during the 8th and 9th centuries, indicate that a Pali-based Buddhism was practiced in the region. In the 10th and 11th centuries, bodhisattvas appear on sema as a result of the influence of Khmer Mahayana Buddhist practice.

One other implication emerges from the research in this dissertation, the indication that the local elites and rulers actively supported and patronized the Buddhist order on the Khorat Plateau. The fact that the sema tradition flourished at large moated sites such as Muang Fa Daed indicates that it was integrated into the society of the region. As in other areas, there was most likely a reciprocal arrangement where the rulers and elites patronized the Buddhist order and in turn their positions were legitimized by the Buddhist order.

John A. Listopad
Art Department
Sacramento State University

Dissertation Information

School of Oriental and African Studies. 2010. 479 pp. Primary Advisor: Elizabeth Moore.

Primary Sources

Thai Fine Arts Department Library – published and unpublished survey reports
Silpakorn University Library – published and unpublished reports, theses, and dissertations
Muang Boran Journal
Khon Kaen Museum – a few semas, much more of the material culture of the Khorat Plateau
Muang Fa Daed – archeological site, center of sema culture

Image: Photograph of sema by Stephen Murphy.

  1. Thanks for this thorough review. Stephen’s approach is certainly what we need more of in this field – comprehensive data gathering and reinterpretation/investigation of much of SEA’s art history bringing together recent research in archaeology other disciplines – plenty of scope for future PhD students!

Leave a Reply

Your email address will not be published. Required fields are marked *

You May Also Like