Gold Artifacts at Prehistoric-Protohistoric Sites in Hồ Chí Minh City

Full text

Text extracted from the article PDF for reading and search. Tables, figures and equations are in the PDF, which remains the authoritative version.

1. Introduction

Gold occupies a distinctive position in the archaeology of late prehistoric and protohistoric Southeast Asia. Unlike stone, pottery, or iron, it is rarely represented by large quantities of utilitarian objects. Its archaeological visibility is instead concentrated in ornaments, small sheets, wires, and objects associated with the body, burial display, or ritual action. This pattern makes gold especially useful for studying technological choice, long-distance access to raw materials, changing practices of personal display and the social differentiation expressed in mortuary contexts. At the same time, the interpretive value of gold can easily be overstated. The presence of a precious material in a grave does not automatically identify a hereditary elite, a chief, or a centralized political hierarchy; nor does chemical similarity between objects from two regions demonstrate direct trade between those communities. Gold therefore requires a contextual approach in which typology and composition are examined together with burial associations, chronology, manufacturing traces, and the distribution of other exchanged materials (Demandt, 2015; Bennett, 2009) [8, 4]. The geographical scope of this study also requires clarification. The original archaeological publications described Giồng Cá Vồ and Giồng Phệt as Hồ Chí Minh City sites, Giồng Lớn as a Bà Rịa–Vũng Tàu site, and Phú Chánh as a Bình Dương site. Resolution No. 202/2025/QH15 reorganized the provincial-level administrative map by combining the former Hồ Chí Minh City, Bà Rịa–Vũng Tàu, and Bình Dương into the present-day Hồ Chí Minh City (Quốc hội, 2025) [21]. Archaeological interpretation, however, should not allow a modern administrative change to erase historical geography. For that reason, this article uses the present-day city boundary for the title and overall inventory, but retains the former provincial and district names when describing excavation history, cultural geography, and earlier scholarship. In practical terms, this means that Giồng Lớn and Phú Chánh must now be considered when discussing gold artifacts found within present-day Hồ Chí Minh City, even though their archaeological contexts were documented under the former Bà Rịa–Vũng Tàu and Bình Dương administrations.

Chronology is another issue that requires caution. Published dates for the Cần Giờ and Long Sơn assemblages do not always form a single, internally consistent sequence. The often-cited determination of 2,480 ± 50 BP for Giồng Cá Vồ and the determinations for Giồng Phệt coexist with other published dates and phase estimates, including proposals that place important parts of the Giồng Cá Vồ sequence in the third–second centuries BC and later activity into the first centuries AD (Đặng et al., 1998; Nguyễn et al., 2023; Nguyen, 2017) [7, 17, 16]. Such differences may reflect distinct samples, contexts, laboratory histories, calibration practices, or the multi-phase nature of the sites. Uncalibrated radiocarbon years BP should not be converted mechanically into calendar years by simple subtraction from 1950. Accordingly, this article uses broad chronological ranges where necessary and treats typological comparison as complementary to, rather than a substitute for, stratigraphy and absolute dating.

Methodologically, the discussion follows a chaîne opératoire perspective: raw material acquisition, preparation, hammering or rolling, joining, finishing, use, deposition, and post-depositional history are treated as linked stages rather than as isolated traits. This is important because a visually simple gold bead can embody substantial knowledge of thinning, annealing, rolling, joining, and control of a highly malleable material. Conversely, an object with elaborate morphology need not be later than a simpler form. Technological complexity may reflect different workshop traditions, craft specialists, social demands, or imported objects operating at the same time. Studies of Khao Sam Kaeo and other Southeast Asian craft centers demonstrate the analytical value of reconstructing technical traditions and production organization instead of using formal complexity as a direct chronological index (Pryce et al., 2006; Bellina & Silapanth, 2006) [20, 3].

2. Sites

2.1. Giồng Cá Vồ Site

Located on a red soil ridge about 1.5m higher than the surrounding areas (Nguyễn, 2012, p. 40) [19], in Hòa Hiệp hamlet, Long Hòa commune, Cần Giờ district, the site is situated between two rivers, on the right bank of the Dinh Bà River and the left bank of the Đồng Hòa River. The site was surveyed and excavated in 1993 and 1994, and by 2000 it was recognized as a National archaeological site. After a long time, the Giồng Cá Vồ site was re-surveyed and explored in 2018 by the Hồ Chí Minh City Museum of History. Three years later, the site was licensed for archaeological exploration and excavation over a total area of 225m2, of which 25 m2 was for exploration and 200 m2 for excavation. This excavation was coordinated by the Hồ Chí Minh City Department of Culture and Sports and the Institute of Archaeology, chaired by Mr. Nguyễn Gia Đối of the Institute of Archaeology (Bộ Văn hóa, Thể thao và Du lịch, 2021) [27]. As of 1994, a quantity of gold metal artifacts accounting for about 38% of all metal artifacts was found at the Giồng Cá Vồ site. The majority of gold metal artifacts are beads, with 66 items (Đặng et al., 1998, p. 151) [7] accounting for 82.5%, followed by flattened gold foil pieces with 13 items accounting for 16.25%, and the remaining artifact is 1 gold ring. The 2021 excavation found only one gold artifact; this item has 3 small circles spread out from a gold wire and connected together, whose function has not yet been identified (Nguyễn et al., 2023, p. 87) [17]. The dating of the Giồng Cá Vồ site, according to C14 radiocarbon isotope analysis, falls around 2,480 ± 50 years before present (BP) (Đặng et al., 1998, p. 151) [7].

The Giồng Cá Vồ assemblage is significant not only because of the number of gold objects but because gold occurs within a broader package of ornaments and exchanged materials. The site has produced stone and glass beads, agate and nephrite ornaments, and characteristic ear ornaments whose distributions extend well beyond Cần Giờ. Such associations are crucial. A gold bead by itself can indicate access to a rare material, but a recurring combination of gold, agate, glass, and nephrite is better understood as evidence that inhabitants participated in multiple, partly overlapping circulation systems. The extensive distribution of Taiwan nephrite ear ornaments around the South China Sea between approximately 500 BC and AD 500 provides an independent example of the scale on which prestige materials and finished ornaments could move through maritime networks (Hung et al., 2007) [12]. Giồng Cá Vồ should therefore be studied as a node where several material streams converged rather than as a site connected to a single external partner. The chronology of the site also deserves a more differentiated formulation than a single date. The value of 2,480 ± 50 BP cited in the primary synthesis provides important evidence for early occupation, but the 2021 excavation report discusses a broader sequence and proposes an early phase around the third–second centuries BC with later activity extending from the first century BC into the first centuries AD (Nguyễn et al., 2023) [17]. Nguyen Kim Dung (2017) [16], drawing on a different published radiocarbon determination, also presents a date of 2,350 ± 60 BP for Giồng Cá Vồ. Rather than selecting one value and treating it as the date of every artifact, these determinations should be understood as evidence that the site history is stratified and potentially prolonged. Establishing the exact position of each gold object within that sequence requires direct association with well-recorded layers or burials and, ideally, a new Bayesian model based on calibrated AMS dates from short-lived samples.

2.2. Giồng Phệt Site

From the Giồng Cá Vồ site, heading northwest about 3.2 km as the crow flies, we reach the Giồng Phệt site. This is also a type of red soil ridge similar to Giồng Cá Vồ. The Giồng Phệt site is located on the left bank of the Bà Vú River in Long Hòa commune, Cần Giờ district, Hồ Chí Minh City. The site was surveyed and subsequently excavated in 1993, over a total area of 50.5m2. The living traces of the residents here were found through hard soil floors, primitive kitchen types, and burials. Gold metal artifacts found in this excavation are bead types, 01 bamboo-node (đốt trúc) shaped bead, and 03 long tubular beads.

In 2008, a field survey of the site was conducted, and on the surface of the ridge, some pottery sherds and human bones were recorded. Nguyễn Thị Hậu reckoned that the site is favorable for a small excavation or digging some exploratory pits (Nguyễn, 2012, p. 43) [19].

The Giồng Phệt site has 2 C14 radiocarbon dating samples and one thermoluminescence date, providing dates of 2,420 ± 484 BP for the thermoluminescence result and 2,100 ± 50 BP for the C14 date (Đặng et al., 1998, p. 213) [7]. Although the quantity of gold from Giồng Phệt is small, the assemblage is analytically important because its forms overlap with those from Giồng Cá Vồ. The bamboo-node bead and the long tubular beads show that Cần Giờ communities shared at least part of a common repertoire of gold ornament. Their presence in jar-burial contexts also links metal ornaments to funerary display and bodily presentation. Because Giồng Phệt contains traces of occupation as well as burials, future analysis should distinguish objects deposited during everyday activities from those selected for funerary use. Such a distinction would help determine whether gold circulated primarily as worn ornament, as curated wealth, or as material deliberately reserved for burial.

2.3. Giồng Lớn Site

Giồng Lớn is situated on the northwestern part of Long Sơn Island, historically in Vũng Tàu city, former Bà Rịa–Vũng Tàu Province. The site was identified during archaeological survey in 2002 and excavated in 2003 and 2005. The first excavation uncovered 54 burials, including 49 earthen graves and 5 pot burials, together with pottery, stone, glass, copper, iron, and gold objects. The second excavation added 23 earthen graves and 3 pot burials. In the published synthesis, approximately 80 burials are recognized, with earthen inhumations forming the dominant rite and pot burials representing a smaller component (Trương, 2016b) [26]. Unlike Giồng Cá Vồ, where occupation traces and burials occur in the same broader site history, Giồng Lớn has generally been interpreted as a cemetery without a clearly identified habitation layer. This difference is important because nearly all artifacts from Giồng Lớn enter the archaeological record through deliberate funerary selection. Gold at Giồng Lớn is both more numerous and more morphologically diverse than at the Cần Giờ sites. The published inventories include large numbers of beads in diamond, spherical, hexagonal, spindle, wheel, and other forms; earrings; perforated gold foils; three gold masks or facial coverings; and an unusual gold-wire object interpreted by the excavators as a linga (Trương, 2015, 2016b) [24, 26]. Counts vary slightly among publications—one synthesis records 179 beads, whereas Nguyen Kim Dung (2017) [16] lists 178 beads plus 15 other gold objects—probably reflecting cataloguing conventions rather than a substantive contradiction. For quantitative comparison, it is therefore preferable to state the inventory source and counting rules instead of merging published totals without verification. The three gold masks are among the most distinctive finds. Their location in graves demonstrates deliberate placement on or near the body and suggests that gold participated in the construction of mortuary personhood, perhaps marking particular identities, statuses, ages, or ritual roles. Yet the presence of gold cannot, by itself, establish the existence of hereditary elites. To test social ranking, the distribution of gold should be compared systematically with grave architecture, age and sex estimates, other exotic materials, weaponry, vessel sets, labor investment, and spatial patterning across the entire cemetery. A concentration in a small subset of graves would support differential access to valuables; demonstrating institutionalized hierarchy would require a stronger and repeated association between wealth and social categories.

The object described as a gold linga deserves similarly careful treatment. The excavators' formal comparison is important and may indicate engagement with South Asian religious imagery, but the identification should remain provisional until the object's manufacturing traces, use wear, deposition, and regional parallels are examined in detail. A single ambiguous object is insufficient to demonstrate established Hindu ritual practice. It is more secure to treat it as evidence for the circulation or local reinterpretation of forms that may have carried South Asian meanings within a period of intensifying cross-regional contact.

2.4. Phú Chánh Site

Phú Chánh, historically in Tân Uyên district of former Bình Dương Province, also falls within the boundaries of present-day Hồ Chí Minh City after the 2025 administrative reorganization. The site is later than the earliest Cần Giờ contexts, generally dated to around 1,900–1,800 BP, and its archaeological character differs from the coastal cemeteries. Two gold objects have been reported, including a circular ring from tomb M2 with a diamond-shaped face bearing a paired flame-like motif (Bùi et al., 2010) [5]. Although the quantity is small, the find broadens the geographical distribution of gold within the current city boundary from coastal–estuarine Cần Giờ and Long Sơn toward the inland riverine zone of the former Bình Dương region.

3. Artifacts

Under the pre-2025 administrative boundary, Giồng Cá Vồ and Giồng Phệt were the principal prehistoric–protohistoric sites of Hồ Chí Minh City known to contain gold. Under the present-day boundary, Giồng Lớn and Phú Chánh must also be included. These sites should nevertheless not be treated as a homogeneous assemblage: they differ in date, site function, burial rite, artifact quantity, and degree of coastal accessibility. The following typological description retains Giồng Cá Vồ and Giồng Phệt as the core Cần Giờ assemblage, while Giồng Lớn and Phú Chánh are integrated in the comparative and interpretive sections.

3.1. Beads

Beads are the most frequently found artifact type; they are mainly made by the flattening method and then shaped by rolling. If considering only the gold part, the bead thickness ranges from 0.1 to 0.2 mm (Đặng et al., 1998, p. 152) [7]. The beads consist of forms such as: • Beads with terracotta cores: the terracotta core is drilled with a hole for stringing and covered with gold on the outside. At Giồng Cá Vồ, 13 beads were found, and Giồng Phệt also recorded one gold-covered spherical bead in jar burial 3a (Đặng et al., 1998, p. 209) [7]. • Rolled circular beads: made by the method of using a flattened gold piece and then rolling it into a circle; 16 beads were found at Giồng Cá Vồ. • Rolled cylindrical beads: rolled from a rectangular piece of gold; after rolling, the gold piece becomes cylindrical. Besides leaving them plain, craftsmen sometimes carved threaded lines or rice ear patterns on the gold beads (Đặng et al., 1998, p. 152) [7]. A number of beads were also found at Giồng Cá Vồ. Giồng Phệt found 03 beads in jar burial No. 4. • Bamboo-node shaped beads (Hạt chuỗi hình đốt trúc): long cylinders from 2 to 3cm, divided into nodes with small barbs protruding outward, accounting for the highest number among the types found at Giồng Cá Vồ. At Giồng Phệt, 01 bead with 3 nodes was found; at the middle two nodes, there are 5 pointed teeth protruding. • Rolled triangular block beads: 2 beads were found at Giồng Cá Vồ • Spherical beads with indented grooves: 3 beads were found at Giồng Cá Vồ. After rolling them into a flattened sphere, the craftsmen would press into them to create grooves.

3.2. Gold Foil

At Giồng Cá Vồ, 13 gold foils were found, mostly rolled flat into rectangular shapes. These might be gold pieces to be crafted into bead types, as the bead crafting techniques presented above are all made from flattened gold pieces.

3.3. Gold Ring

A gold ring was found as a single specimen at Giồng Cá Vồ, created by gouging and chiseling from a gold foil.

3.4. Artifacts of Unidentified Function

During the 2021 excavation at the Giồng Cá Vồ site, in layer number 2, archaeologists found a gold artifact of unidentified function. The artifact was crafted from three gold wires; the maker hammered a section of the wire flat, bent it into a circle, and then joined the three wire sections together.

3.5. Manufacturing Technology and Technological

Interpretation The Cần Giờ gold assemblage is dominated by operations that exploit the metal's exceptional malleability: hammering into thin sheet, cutting, rolling, bending, and in some cases joining. The reported thickness of approximately 0.1–0.2 mm for gold sheet used in beads indicates controlled thinning rather than casual deformation. Rolled tubular and circular beads required the maker to produce relatively even sheet, cut blanks, bring edges into a stable cylindrical form, and finish the surface sufficiently for wear. Gold-covered terracotta cores represent a different technological solution: they create the visual effect of a larger gold ornament while minimizing the quantity of precious metal. Such composite objects are technologically and socially informative because they separate visual display from raw-metal mass. The bamboo-node beads are especially relevant for technological comparison. Their segmented form and projecting points require more complex shaping than a plain tube, but complexity should not be translated directly into relative age. The original argument that the more elaborate gold of Giồng Lớn must be later than the simpler forms of Giồng Cá Vồ and Giồng Phệt is plausible as a hypothesis but is not a secure chronological method. Workshops with different skill sets can operate simultaneously; imported and locally made ornaments can coexist; and simple forms can persist for long periods because they are efficient, culturally preferred, or suited to particular modes of stringing. The technological sequence must therefore be built from dated contexts and manufacturing evidence, not from an evolutionary assumption that simple necessarily precedes complex.

This point is reinforced by archaeometallurgical research at Khao Sam Kaeo. Pryce et al. (2006) [20] emphasized the reconstruction of technological behavior and technical traditions across iron, copper-alloy, and precious-metal working rather than treating artifact appearance as a direct proxy for origin. The broader Khao Sam Kaeo research has shown that foreign specialists, local craftspeople, imported raw materials, and hybrid production could coexist within a single trading center (Bellina & Silapanth, 2006) [3]. A similar analytical framework is useful for southern Vietnam: a gold object may have been imported as a finished ornament, made locally from imported raw gold, made by a mobile craftsperson, or locally copied from an exogenous style. Typology alone cannot distinguish these possibilities. Future technological work should combine low-magnification and metallographic microscopy, high-resolution photography, micro-XRF or SEM-EDS screening, and carefully targeted LA-ICP-MS where sampling is permitted. Particular attention should be paid to seams, folded edges, tool marks, solder or mechanical joins, thickness variation, and annealing traces. Comparing these features among Giồng Cá Vồ, Giồng Phệt, Giồng Lớn, Sa Huỳnh sites, and Thai-Malay Peninsula assemblages would allow researchers to identify shared recipes of manufacture rather than merely similar shapes. This would provide a stronger basis for identifying craft learning, movement of specialists, and local imitation.

3.6. Gold in Mortuary Contexts: Value, Identity, And

Differential Access The strong association between gold and burials is one of the most important characteristics of the assemblage. Gold beads and other ornaments were not simply lost in settlement debris; many were deliberately placed with the dead, often together with glass, stone, pottery, or metal objects. Such deposition removed valuable material from circulation and transformed it into a durable statement made by the living about the deceased. Demandt (2015) [8] interprets early Southeast Asian gold ornaments as prestige goods closely connected with strategies of dress and changing elite identities. That interpretation is useful, but it should be applied at the scale supported by the evidence. At the site level, the safest inference is that gold had high symbolic and exchange value and that access to it was unequal or socially differentiated.

A stronger claim such as hereditary ranking, chiefly authority, or centralized control over trade requires additional tests. Burial wealth can vary for many reasons: age, gender, lineage, ritual role, circumstances of death, household strategies, competitive display, or the temporal accumulation of objects. In a heterarchical social configuration, several forms of status may coexist without being ordered into a single permanent hierarchy. Consequently, a cemetery-wide database should record the number and weight of gold objects per grave, artifact diversity, exotic-material counts, grave dimensions, body treatment, demographic attributes, and spatial clustering. Measures of concentration can then be compared with qualitative evidence for repeated status distinctions. Such analysis would move the discussion from an impression of 'rich graves' toward a testable model of differential access.

The suggestion that gold beads functioned as currency also requires qualification. Small ornaments can certainly store value, circulate as gifts or exchange goods, and be converted between social and economic contexts. However, calling them currency implies stronger evidence for standardized denominations, repeated weight units, widespread transactional use, or explicit monetary contexts. The present assemblages do not yet provide such evidence. It is therefore more accurate to describe gold beads as portable valuables or prestige ornaments with potential exchange value, while leaving open the possibility that some may have operated in payment-like transactions.

4. Regional Comparative Context

4.1. Southeastern Vietnam

In some synchronic sites in the Southeast region, gold artifacts have also been found, such as: • Đồng Nai Province: At the Hàng Gòn 9 site in Xuân Lộc district, scientists discovered 1 gold bead lying together with other beads and earrings in a jar burial. The gold bead is pyramidal, 8mm high, and has a base diameter of 8mm. Crafted using a flattening technique, drilling a flat pyramidal gold ingot, rolled long, with two curved sections used as wearing hooks, 4.5cm high (Lê et al., 1991, p. 160) [15]. The Hàng Gòn 9 site has a C14 date on a plant material sample of 2100 ± 150 BP. At the Phu Hoa site, not far from Hàng Gòn 9, scientists also found a gold bead, diamond-shaped, crafted with a pierced hole, measuring 7mm long and 8mm wide. Phú Hòa has C14 dates of 2,400 ± 140 BP and 2,590 ± 240 BP (Lê et al., 1991, p. 166) [15]. • Former Bình Dương Province (now within Hồ Chí Minh City): At Phú Chánh, two gold objects were recorded. The most distinctive comes from tomb M2: a circular gold ring with a diamond-shaped face containing a paired flame-like motif. The site is generally dated to approximately 1,900 - 1,800 BP (Bùi et al., 2010, p. 258) [5]. In the current administrative framework, Phú Chánh is no longer an external comparison but part of the broader citywide archaeological inventory. • Former Bà Rịa - Vũng Tàu Province (now within Hồ Chí Minh City): Giồng Lớn on Long Sơn Island is a predominantly funerary site in which earthen graves greatly outnumber pot burials. It contains the largest and most diverse gold assemblage presently documented among the late prehistoric-protohistoric sites considered here, including approximately 178-179 beads depending on cataloguing, earrings, perforated sheets, three gold masks, and the unusual gold wire object interpreted as “a linga” (Trương, 2015, 2016b; Nguyen, 2017) [24, 26, 16]. The main occupation or cemetery use is commonly placed around the later centuries BC to early centuries AD, although individual published radiocarbon determinations and typological assessments require renewed evaluation.

The regional distribution therefore shows a pronounced coastal and estuarine concentration, but not an exclusively coastal phenomenon. Giồng Cá Vồ, Giồng Phệt, and Giồng Lớn occupy settings directly connected to tidal rivers, estuaries, and the sea, whereas Hàng Gòn 9, Phú Hòa, and Phú Chánh demonstrate movement inland through riverine corridors. The original inventory calculated beads at roughly 86.25% of 291 recorded gold objects in prehistoric southeastern Vietnam, with sheet/wire and other forms comprising the remainder. Because published totals vary slightly between catalogues, these percentages should be treated as provisional until the assemblages are reconciled object by object. The robust pattern is not the exact percentage but the overwhelming dominance of personal ornaments, especially beads.

The spatial pattern supports a model in which the coast and major rivers operated as complementary rather than competing routes. Cần Giờ and Long Sơn were well positioned to receive materials moving along the South China Sea, but they were also connected upstream to Đồng Nai and other inland zones. Raw gold, finished ornaments, semi-precious stones, glass, ceramics, and metal goods could have passed through different hands and routes before reaching a burial community. Accordingly, it is premature to state that maritime exchange 'predominated' over riverine exchange solely because the coastal sites have more gold. Coastal cemeteries may simply preserve different consumption practices, or their communities may have acted as redistribution nodes linking marine and fluvial transport.

4.2. With Central Vietnam

Gold artifacts found in Central Vietnam are usually located within the group of sites belonging to the Sa Huynh culture such as Hòa Diêm, Gò Quê, Gò Mã Vôi, Gò Mùn, Lai Nghi, Bình Yên, Cồn Ràng. Gold artifacts found are often concentrated in wealthy jar burials at the locations (Nguyễn et al., 2016, p. 84) [18]. Some basic records of gold artifacts are as follows:

The dominant gold artifact is still the bead type. At Gò Mã Vôi, during the 1999 excavation, 2 gold beads shaped like a “cylindrical drum” trống cơm were found. At the Hậu Xá II location, during the 1993 excavation, 1 flat circular bead was found, and 1 flat hexagonal bead was found at A Bang in 1995. From external collections, researchers also found 14 beads with terracotta cores in the shape of two inverted cones pierced at both ends and covered with a thin layer of gold, with the gold composition reaching up to 94%; these artifacts are determined to originate from the Go Mùn site in Đại Lộc District, Quảng Nam Province (Lê, 2000, pp. 64-65) [14]. At the Hòa Diêm site, many gold beads were found with floral shapes, very small, thin sizes, and pierced holes. At the Lai Nghi site - Quảng Nam, 4 twisted grooved wire earrings were found, similar to the type at Giồng Lớn - Bà Rịa Vũng Tàu.

Gold artifacts at Lai Nghi, Gò Mùn, and Gò Mã Vôi were collected and analyzed on 22 samples, including 18 beads and 4 earrings. The analysis results showed that the gold composition from the three locations in Central Vietnam has two distinct origin groups for the bead samples. Two drum-shaped beads found at Gò Mã Vôi have a different origin from the remaining beads; except for one bead at Lai Nghi where the Platinum composition could not be determined, all the rest share a main group with a gold ratio ranging from 74 - 98% (Reinecke et al., 2014, pp. 52-63) [22]. Regarding the 4 earrings with prominent teeth/ridges, 3 of them belong to the same group with an assumed external origin, while 1 piece looks rougher and is a locally made product. This type of earring has a very wide coverage; they are found in the Northern part of Central Vietnam, the Central Highlands, Southeast Asia, South Asia, and as far as Europe, dating from around the 5th to the 2nd century BC.

The gold artifacts in the Sa Huỳnh culture are similar to the gold artifacts found in the Southeast region (Lâm, 2018, p. 153) [13]. The gold crafting technique of flattening gold foil and rolling it is similar to the gold crafting technique of Giồng Cá Vồ - Giồng Phệt. Specifically, the bead type with a terracotta core covered on the outside by a thin layer of gold is very similar to the typology of the circular bead with a terracotta core from Giồng Cá Vồ.

The Sa Huỳnh comparison is most productive when it moves beyond the question of cultural influence. Nguyen Kim Dung (2017) [16] emphasizes that Sa Huỳnh communities were active participants in regional exchange networks: they imported raw materials and ornaments but also circulated locally meaningful forms and technical practices. This perspective is important for Cần Giờ. Similar beads or ear ornaments need not indicate that one cultural group passively received a package from another. They may instead reflect a shared interaction sphere in which forms were copied, reworked, and assigned new meanings by multiple coastal communities.

This approach also fits the wider evidence for nephrite. Electron-probe analysis demonstrated that specific nephrite ear ornaments made from eastern Taiwan raw material circulated through the Philippines, East Malaysia, southern Vietnam, and peninsular Thailand within a network spanning roughly 3,000 km around the South China Sea (Hung et al., 2007) [12]. The significance for gold research is methodological rather than a claim that the two materials followed identical routes. Nephrite shows that late prehistoric communities possessed the maritime knowledge and social relationships required to move high-value ornaments across very long distances. Gold should therefore be analyzed within a pre-existing regional infrastructure of mobility, not as the single commodity that created exchange.

4.3. Wider Southeast Asia And Trans-Asian Networks

At the Khao Sam Kaeo site in Chumphon Province, Thailand, many artifacts closely resembling those from Giồng Cá Vồ were found, such as Nephrite stone two-animal-headed earrings (khuyên tai hai đầu thú), three-pointed (khuyên tai ba mấu) agate earrings, and various types of agate beads. Here, archaeologists found glass beads subsequently covered with a layer of foiled gold (Nguyễn et al., 2016, p. 83) [18], closely analogous to the gold-covered terracotta type found at Giồng Cá Vồ.

From 2009 onward, the Ancient Gold in Southeast Asia research program generated new analytical data for gold from multiple sites, including samples from Vietnam. Eighteen sampled objects from Giồng Cá Vồ were reported as being made from placer gold. Their trace-element characteristics differed from analyzed groups in central Vietnam and Cambodia while showing similarities to some samples from Bali and Thai sites including Khao Sam Kaeo and Noen U-Loke (Trương, 2016b, p. 82) [26]. These results are significant, but they should not be read as a direct map of trade routes. Provenance studies of archaeological gold are intrinsically difficult because natural deposits can overlap chemically and because melting, mixing, alloying, and recycling can alter or blur source signatures (Guerra et al., 1999; Guerra & Calligaro, 2004) [11, 10]. The most defensible conclusion is that the Giồng Cá Vồ samples belong to a broader compositional field that is compatible with connections beyond southern Vietnam and deserves comparison with expanded geological and archaeological reference datasets.

Khao Sam Kaeo is particularly important for interpreting these connections. Located on the Upper Thai-Malay Peninsula, it developed from about the fourth century BC as a major coastal and trans-peninsular node linking the South China Sea with routes toward the Bay of Bengal. Research has identified specialized production in metals, stone ornaments, and glass, together with evidence for resident or mobile craftspeople familiar with technical traditions from different parts of Asia (Bellina & Silapanth, 2006; Pryce et al., 2006) [3, 20]. In this context, similarities between Cần Giờ and Khao Sam Kaeo should not be reduced to a single traded object. They are more plausibly part of a larger system in which technologies, artisans, raw materials, finished ornaments, and aesthetic preferences circulated together but not necessarily along identical paths.

Island Southeast Asia supplies further evidence for the movement and local consumption of gold. At Pangkung Paruk in Bali, elaborate gold ornaments occur with Roman gold-glass beads and Han Chinese bronze mirrors. Calo et al. (2020) [6] interpret the assemblage within early to mid-first-millennium AD trans-Asian exchange connecting Island Southeast Asia with mainland centers, South Asia, China, and ultimately the Roman world. The Bali evidence is later than the earliest Cần Giờ contexts, but it demonstrates the long-term expansion and recombination of exchange systems into which southern Vietnam was increasingly integrated. Compositional similarities between some Giồng Cá Vồ and Bali gold should therefore be evaluated against this diachronic background rather than taken to imply a simple bilateral Cần Giờ-Bali route.

Taken together, the regional comparisons suggest a network with multiple levels. At one level were local communities moving along rivers and short coastal routes; at another were intermediary ports and craft centers such as Khao Sam Kaeo; at still another were trans-Asian flows of materials, technologies, and ideas. Giồng Cá Vồ, Giồng Phệt, and Giồng Lớn need not have been major ports equivalent to Khao Sam Kaeo to participate in this system. Their archaeological importance may lie precisely in showing how coastal burial communities consumed, recontextualized, and perhaps redistributed objects and materials that had passed through wider exchange networks.

5. Discussion

5.1. Chronology and The Problem of A Single

Developmental Sequence The gold assemblages considered here span several centuries and should not be compressed into a single evolutionary sequence. Giồng Cá Vồ and Giồng Phệt include contexts that reach back into the later first millennium BC, Giồng Lớn overlaps the late prehistoric-protohistoric transition, and Phú Chánh belongs largely to the early first millennium AD. Bennett (2009) [4] places the broader archaeological appearance of gold in Southeast Asia around the mid-to-late first millennium BC, in association with major changes that also include iron technology, glass, and semi-precious stone working. This regional timing makes the early Cần Giờ evidence important, but it does not resolve the internal chronology of individual object types.

A revised chronological program should begin with a grave-by-grave concordance of all existing radiocarbon and thermoluminescence determinations, including laboratory codes, sample materials, stratigraphic contexts, conventional radiocarbon ages, calibration curves, and calibrated probability ranges. Dates from long-lived charcoal should be distinguished from short-lived botanical samples or directly associated human and animal remains. Bayesian modeling could then test whether particular burial groups or ornament styles form discrete phases. Until such work is completed, statements such as 'Giồng Cá Vồ gold is earlier than Giồng Lớn because it is technologically simpler' should be replaced by the more limited observation that the two assemblages differ in morphology and manufacturing repertoire.

5.2. Raw Material, Production, And Provenance

No major gold source has been documented in the immediate Cần Giờ coastal zone, so the raw material was almost certainly introduced from elsewhere. However, 'elsewhere' can mean several things archaeologically. Gold could arrive as placer nuggets or dust, as ingots or scrap, or as finished ornaments. Finished objects could subsequently be repaired, melted, recycled, or recombined. Each pathway leaves different evidence. The presence of unfinished sheet, wire, droplets, crucible residues, working debris, or standardized preforms would provide stronger evidence for local manufacture than finished ornaments alone. At present, the thin foils from Giồng Cá Vồ are reasonably interpreted as possible working stock or preforms, but this hypothesis should be tested through edge morphology, cutting traces, and contextual association with tools or production debris. The published LA-ICP-MS results represent an important foundation because trace elements can characterize groups of ancient gold at concentrations not accessible through simple visual inspection. Yet analytical similarity should be expressed probabilistically. Guerra et al. (1999) [11] stressed that precious-metal provenance is difficult and usually requires multiple trace elements and, where possible, complementary isotopic or mineralogical evidence. Guerra and Calligaro (2004) [10] likewise emphasized both the possibilities and limitations of elemental characterization. For the Hồ Chí Minh City assemblages, the next step should therefore be to analyze larger, contextually selected sample sets from Giồng Cá Vồ, Giồng Phệt, Giồng Lớn, and Phú Chánh using identical protocols, while simultaneously expanding comparative data from natural gold sources in Vietnam, Thailand, Cambodia, Laos, Malaysia, and Indonesia.

A crucial distinction should also be made between provenance of metal and provenance of objects. Two ornaments made from chemically similar gold could have been manufactured in different workshops from raw material obtained through the same supply zone. Conversely, two objects made in the same workshop could contain gold from different sources if the craftsperson recycled metal. Technological fingerprints sheet thickness, joining methods, tool marks, alloy recipes, and construction sequence may therefore identify workshop traditions independently of geological source. Integrating these datasets offers a more realistic model of ancient circulation.

5.3. Gold, Funerary Differentiation, And Social

Organization Gold is often used as a shortcut for identifying elites, but the Hồ Chí Minh City assemblages illustrate why a more explicit social model is needed. Giồng Lớn contains graves with spectacular objects, including facial gold, while Cần Giờ has numerous gold beads and composite ornaments. These patterns demonstrate that some individuals or funerary groups could mobilize scarce materials and labor. They are compatible with emerging inequality, but they do not specify the political form of that inequality. Wealth could be concentrated in households, kin groups, ritual specialists, successful traders, age grades, or leaders whose authority was temporary rather than institutionalized. A useful analytical sequence is therefore wealth → differential access → repeated association with social categories → possible control. The first two steps can be investigated directly through grave assemblages. The latter steps require evidence that access was systematically restricted and reproduced across generations. If gold and other exotic materials cluster consistently with particular demographic categories, spatial sectors, or elaborate grave constructions, a model of durable ranking becomes more plausible. If wealth is distributed through several cross-cutting categories, a heterarchical interpretation may fit better. The archaeological record should decide between these alternatives rather than assuming that the arrival of long-distance trade automatically produced centralized elites. The mortuary evidence also cautions against treating all gold as equivalent wealth. A thin foil bead and a solid earring may differ dramatically in mass, labor, visibility, and social significance. Composite beads with terracotta cores maximize visual surface while minimizing gold consumption, which may indicate a strategy of display under conditions of limited raw material. Gold masks at Giồng Lớn, by contrast, transform the face and the identity of the deceased in a specifically funerary setting. Quantitative analyses should therefore measure both objects count and estimated metal mass, while qualitative analyses consider placement on the body, visibility during funerary rites, and relationships with other grave goods.

5.4. Coastal Communities as Active Nodes In Exchange

The combined evidence supports a model in which the communities of Cần Giờ and Long Sơn were active nodes rather than passive recipients at the edge of distant civilizations. Long before the first centuries AD, South China Sea communities were already linked through the movement of nephrite and other valued materials (Hung et al., 2007) [12]. By the fourth century BC, Khao Sam Kaeo demonstrates the presence of organized craft production and trans-peninsular connectivity between the Indian Ocean and South China Sea worlds (Bellina & Silapanth, 2006) [3]. Sa Huỳnh communities similarly participated in the circulation of ornaments and technologies along the Vietnamese coast (Nguyen, 2017) [16]. The gold from southeastern Vietnam is best understood within this dense regional background. Riverine connectivity should remain central to this model. The Đồng Nai-Sài Gòn river system connected inland communities to estuaries, and the Cần Giờ waterways connected those estuaries to the coast. Rather than choosing between 'maritime' and 'riverine' exchange, the archaeology points toward an integrated waterscape in which goods could shift between sea-going and riverine transport. The concentration of gold near the coast may reflect access to overseas traffic, but inland finds show that valuables continued to move through fluvial networks. A GIS-based least-cost and hydrological study, combined with settlement chronology, could test which sites were positioned at transfer points between ecological zones.

5.5. Research Priorities

Five lines of research would substantially strengthen the interpretation of gold in present-day Hồ Chí Minh City. First, the inventories should be reconciled and digitized at object level, recording provenience, dimensions, weight, material, manufacturing features, and associated finds. Second, existing absolute dates should be recalibrated and linked to specific archaeological contexts, followed by targeted AMS dating of key graves. Third, technological study should document chaîne opératoire features through microscopy and elemental screening before destructive sampling. Fourth, gold provenance research should use a larger comparative database and explicitly model mixing and recycling instead of equating compositional similarity with a single source. Fifth, mortuary analysis should compare gold distribution with demographic, spatial, and labor-investment variables to test competing models of hierarchy, heterarchy, and differential access.

6. Conclusion

The expanded evidence changes the interpretation of gold artifacts from a narrowly local inventory into a problem of chronology, technology, mortuary practice, and regional connectivity. Under the present-day administrative boundary, the principal discussion must include not only Giồng Cá Vồ and Giồng Phệt in Cần Giờ but also Giồng Lớn in the former Bà Rịa-Vũng Tàu area and Phú Chánh in the former Bình Dương area. These sites do not represent a single cultural or chronological unit; rather, they provide a sequence of different contexts through which gold entered social life from the later first millennium BC into the early centuries AD. At Giồng Cá Vồ and Giồng Phệt, thin sheet working, rolling, composite cores, and tubular or bamboo-node beads show that gold was manipulated through a developed repertoire of techniques. Giồng Lớn contains a larger and more varied funerary assemblage, including beads, earrings, facial gold, perforated sheets, and an unusual object interpreted as a linga. Phú Chánh demonstrates the continued presence of gold farther inland and at a later date. Differences among these assemblages should not be arranged into a simple progression from technologically 'primitive' to 'advanced.' Secure chronology must derive from stratigraphy and calibrated absolute dates, while technological relationships should be tested through detailed reconstruction of manufacturing sequences.

The comparative evidence places southern Vietnam within a broad interaction sphere extending through central Vietnam, the Thai-Malay Peninsula, Cambodia, and Island Southeast Asia. Sa Huỳnh ornament networks, the specialized craft and exchange center at Khao Sam Kaeo, the long-distance distribution of nephrite, and later trans-Asian assemblages in Bali all show that maritime connectivity developed through multiple overlapping networks. Published chemical analyses of Giồng Cá Vồ gold are consistent with this wider perspective, but they do not prove a direct trade route to Thailand or Bali. Gold provenance remains a difficult analytical problem in which natural compositional overlap, recycling, and mixing must be considered. Finally, the funerary concentration of gold demonstrates value, selective access, and the importance of bodily display, but it should not be used alone to reconstruct political hierarchy. The next stage of research should quantify the distribution of wealth by grave, examine the relationship between gold and demographic or spatial categories, and compare object counts with metal mass and labor investment. Such work can test whether emerging inequality was organized hierarchically, heterarchically, or through changing combinations of both. From this perspective, gold artifacts are not merely evidence of wealth or foreign contact. They are material records of how coastal communities transformed scarce substances, participated in exchange, constructed identities, and negotiated social difference during a formative period in the prehistory and protohistory of southern Vietnam.

The most productive future interpretation will therefore combine archaeology and natural science: contextual excavation, calibrated dating, technological microscopy, archaeometallurgical characterization, GIS analysis of waterways, and cemetery-wide social statistics. Only by integrating these independent lines of evidence can the gold assemblages of present-day Hồ Chí Minh City be used to distinguish direct exchange from network participation, imported objects from locally made imitations, and prestige display from durable structures of social power.