Southeast Asian Archaeological Site Reports

The archaeological research program at St. Andrew’s site was carried out in five stages:

1. Library and archival research
2. Pedestrian survey
3. Auger testing
4. Controlled excavations of test units in the impact zone, prioritizing the area directly affected by construction
5. Salvage excavations

The STA site research area is divided into four quarters: north, south, east and west. The existing paved roads in the Cathedral’s grounds represent the boundaries of each zone. The impact zone is the west lawn and it is also given priority because the north quarter had been excavated previously in the 1980s due to the construction of the City Hall MRT station and the train lines. The east and west quarters have similarly been affected by the same construction.

Stages 1 and 2: Library and archival research and pedestrian survey

Stages 1 was completed in May 2003 and Stage 2 of the project began with the placement of the datum point at the site. The datum point was placed at the northwest corner of the west quarter (N1º 17’ 58” E103º 51’62”). There is a structure which forms a part of an existing drainage system or series of man-holes. The team ran a north-south datum line (TNS -1) from the datum point using a compass and tripod and a 50m measuring tape. This line is extended to the south quarter. The line also acts as a survey line for auger testing and is flagged at 5m intervals. A total of 7 north-south (TNS-1 to 7) and 5 east-west (TEW- 0 to 8) survey lines were run across the entire site.

The second stage of pedestrian survey is a systematic surface survey of the grounds which took place in one day. Few surface finds except occasional objects which can be dated to the modern period. No artifacts from the 14th-16th century has been collected.

Stage 3: Auger testing and soil core samples collection

The third stage of the research involved taking soil core samples from the west lawn by augering at 10-meter intervals along seven transects 50 meters long. Auger testing was undertaken for two reasons: 1) to allow the team to predict density of subsurface and for palaeo-environmental reconstruction in order to determine where test excavations can be carried out. A total of 49 locations were sampled: 39 in the west quarter; 8 in the south quarter; and 2 in the east quarter. No auger tests were carried out in the north quarter.

A two-person team carried out auger testing. The team manually drove an auger with a bucket measuring 5cm. A drill bit forms the end of the bucket. The handle extends to 2m in length and is marked at 20 cm intervals. All materials extracted were collected. Soil samples representative of each stratigraphic layer were also collected.

Following the auger testing along the main survey line (TNS-1), testing was expanded to include the 6 north-south survey lines in the south and east quarters, and 5 east-west lines in the west quarter. In the impact zone, which was the west quarter, the team adopted a more targeted approach. Auger holes were placed at 5m intervals. This is due to the more complex stratigraphy of this quarter.

Three major soil types were observed. Generally, the first 30cm extracted is primarily humus. In some locations the first 5-10cm of humus is followed by a 2-3cm layer of yellow clay, commonly used in Singapore for planting grass and trees. Underneath this layer is 10-20cm of humus. The third soil type consists of sand of varying colors. This soil type extends to the water table.

The color of the sand between 30 and 50cm below surface varies at different locations Generally, sand in the lower layers is light in color, and coarser due to inclusions of crushed quartzite. The sand turns a dark brown about 10-20cm above the water table.

Excavations show that homogeneous soil color from the surface to the water table indicates that this resulted from fill from construction work such as the laying of water pipes or electrical lines, or past construction work.

Very limited quantities of artifacts were recovered from auger testing. No pottery was recovered from 95% of the auger samples. Interesting when test units were excavated at the same locations, dense deposits were encountered. The team concluded that auger testing has limited use and is unreliable, unless combined with another method such as test excavations. In terms of determining the presence or absence of subsurface deposits, the team observed in many of the excavation units which were dug in Stage 4, subsurface deposits were recovered in numerous areas where the auger had yield no materials. This suggests that the type of auger used for the testing might have compounded the problem for the absence of artifacts. The drill bit forming the end of the bucket likely led to the crushing of any materials and the downward spiral motion of the drill blades forced materials into the walls of the drill hole, especially in this case, where the predominant soil type was sand. The type of auger and soil likely led to the absence of items recovered in the process.

Stage 4: Excavations

The subsurface soil layer of the STA site is primarily sand. The team decided to restrict the dimensions of the test units to 2x1 meter taking into account safety concerns due to the sand as the predominant soil type. In some cases, an additional unit adjacent to the first was excavated to provide more accurate recording of the stratigraphy. These are also 2x1m units. All test units which are 2x1m bear the prefix STA by default.

A total of 60 test units were excavated by 12 December 2003. This included fifty-seven 2x1m units and three, 1x1m units. Of the 57, four units were in the south quarter and 53 in the west quarter. All four units in the south quarter are 2x1m. Five excavation units, STA 1-5, were placed in a number of different locations based on the results of auger testing. This included a unit in the south quarter. These were excavated to recover subsurface deposits which can be used to help the team better understand the pre-colonial topography of the site. The results from these excavations informed the strategy for further excavations. The STA site was excavated according to natural layers. The upper layers of a majority of excavated units are mixed and contain colonial, precolonial and modern artefacts. In cases where there is a precolonial artefacts bearing layer, the assemblage is homogeneous.

The following section provides detailed descriptions of a small number of earlier units excavated. As this report represents a preliminary report of the excavation and small sampling of the materials found at the STA site, the selections of units described as well as the artifacts selected in this report and database are made arbitrarily to provide a detailed look at the excavation work and strategy and the range of materials present at the St. Andrew’s site.

STA 1

This unit had four layers. Layer 1 extended to 84cm below surface and consisted of a mix of sand, clay, and humus. The first 10cm was sterile. Materials recovered comprised pottery, slate, tiles, brick fragments, and crushed granite. Granite is naturally occurring in Singapore, but not in the STA area, it was hence brought into this area. Slate deposits were most dense at the transition to Layer 2. The bricks found were local bricks made prior to World War II.

Layer 2 was white sand and extended to 100cm below surface. This layer was void of artifacts. Layer 3 comprised black sand which appeared to be oily; it also did not contain artifacts. Layer 3 continued to the water table, which was 166 cm below surface.

A simple test was done to ascertain whether the sand from Layer 3 was infused with oil. Some sand was mixed with water in a bottle. The mixture was shaken and allowed to settle. There was no evidence of oil present.

Layer 1 yielded 14th-century artifacts but these were mixed with modern and other materials.

STA 2

This unit consisted primarily of fill, probably from previous construction works. A pipe approximately 70cm in diameter was found at 115cm below surface. The water table was at 160cm below surface. There were some variations in soil color, but generally homogenous dark brown with streaks of grey. The pipe is made of a synthetic material and probably put in place no more than 30 years before the present. Materials recovered include brick fragments, asphalt fragments, tiles, slate, glass, stoneware, and other building materials. The deposits comprised a mix of materials from the site deposited at an earlier date and waste material from the laying of the pipe.

STA 3 and 3a

STA 3 comprised six stratigraphic layers. Layer 1 extended to approximately 23cm below surface and consisted generally of humus, clay, and sand. There was a layer of debris from approximately 10-20cm below surface comprising brick and tile fragments and slate. Layer 2 extended to 45cm below surface and contained trash from the colonial period. This included a cache of glass and ceramic bottles. Immediately below this cache was a large circular artifact of unfired clay. The artifact measures 42cm in diameter and has a height of 17cm.

Layer 3 extended to approximately 80cm below surface and was a mixture of colonial and precolonial artifacts. Layers 4 and 5 were the precolonial layers from which Chinese stoneware and porcelain were recovered. Unidentifiable iron objects and a worked piece of sandstone were also excavated. Layer 6 was void of artefacts.

The materials and stratigraphy in STA 3a were similar to STA 3. There was a posthole in Layer 5, at the transition to Layer 6. It extended from 123cm to 133cm below surface. The upper part of the posthole was 8cm in diameter and the lower portion 5cm in diameter. The water table for STA 3 and 3a was 177 cm below surface.

STA 4

The volume of cultural materials and the stratigraphy of this unit were similar to that of STA 1.

STA 5

This unit was the first to be excavated in the south quarter and it contained four natural layers. Layer 1 was divided into 1 and 1b. Both 1 and 1b comprised dark sand, with the Layer 1b differing slightly in color. There was a modern trash pit at the southern half of the unit, and this extended into Layer 1b. Artifacts from Layer 1 consisted of one tile and brick fragments.

Layer 2 consisted of a darker colored sand. This Layer contained 14th-century artifacts and approximately 1.5 kg of rocks each measuring 4-5cm in diameter. Immediately below this layer, at 58cm below surface was a mottled sand layer, which marked the transition between Layer 2 and 3. This mottled soil extended to 60cm below surface.

Layer 3 was a very light colored marine sand which extended to 95cm below surface. This layer was sterile and did not contain any artifacts. Layer 4 was a dark sand layer, which was typical of the sand encountered just above the water table. The water table in this unit was at 145cm below surface.

Phase 2 of Stage 4 Excavations

The finding that STA 3 and 3a contain dense and undisturbed layers yielding colonial and 14th-16th-century artifacts led the team to focus the next phase of systematic excavation on the northern section of the west quarter. A total of four clusters were identified as sites for excavations: three clusters in the west quarter: Clusters 1, 2 and 3; and one in the south quarter, Cluster 4. The north and east quarters were not selected for further investigation in this phase.

A single test unit, STA 8, in the southwestern part of the west quarter was also investigated. This unit was to provide a representative spatial coverage of the quarter. The upper layers of this unit comprised modern and colonial artifacts. There was an intact precolonial layer of dark grey sand. However, the total volume of materials in this unit was much less than in STA 3, 3a and 5.

Cluster 1

Initially, units were selected for systematic excavation in Cluster 1 to gain maximum spatial coverage. This strategy paid off as it revealed a more complex spatial distribution of materials than expected. The density of subsurface artifacts varied between units.

In many of the units in this cluster, the two uppermost layers comprised generally building materials from the modern and colonial periods. These included bricks, tiles, slate, insulation material, asphalt etc. Ceramics, glass, coins (British and Dutch), and military paraphernalia characterized the range of colonial artifacts. Soldiers who were brought to the hospital on the church’s grounds during World War II probably deposited the military artifacts. Military paraphernalia included buttons and other insignia bearing metal objects, and bullets. The 14th-16th-century artifacts comprised the items recovered from the precolonial layer.

The remains of a colonial period brick pavement was also uncovered about 5-10cm below the surface in excavation units STA 28, 29 and 32. Interestingly, the pavement was constructed using bricks made during the precolonial period as these were significantly thinner than modern bricks. There is no mention of a pavement in the west quarter of the church in any of the literature available.

The soil stratigraphy in the cluster was similar to STA 3 and 3a. Overall it corroborates the idea that the landscape in the early colonial and precolonial periods was one that was undulating.

Cluster 2

This was a 1m wide east-west transect along which 2x1m units were excavated at 2m intervals. The upper layers of these units consisted of a mixture of modern, colonial, and precolonial artifacts , including broken building materials. The precolonial layer was identified at 40-50cm below surface.

Two units in the westernmost section of this cluster, STA 39 and 47, contained the most dense layers of 14th-16th-century artifacts. The soil stratigraphy was consistent along the transect. The precolonial sand layer consisted of a dark grey sand similar to that found in Cluster 1.

There was a posthole in STA 31, a unit to the north of the transect. The profile suggested a post of a greater diameter than the posthole found in STA 3.

Cluster 3

This cluster was the southernmost cluster in the west quarter. It was also a 1m wide east-west transect. Subsurface artifacts was less dense compared to the other clusters. The dark grey, precolonial sand layer was also absent in the units located in the eastern half of the transect.

Cluster 4

Cluster 4 was a 1m wide north-south transect. Modern and colonial artefacts were very sparse in the units. There were however dense deposits of precolonial artifacts in all of the units excavated, and these occurred in a dark grey sand. The upper layers consisted of a mix of modern, colonial, and precolonial artifacts, which had been noted to be consistently observed in other units in the site. The stratigraphy in STA 5 was representative of subsurface topography of this cluster.

Across the STA site, ground water was reached at a depth of 2 meters. The team initially took soil samples from the south and east lawns for comparison. By the end of the project the team was allowed to excavate 5 units measuring 2x1 meters on the south lawn. Eighty-two other units in total were dug on the west lawn. The archaeological team aims to send the soil samples collected over the duration of the excavation for analysis in the future.

In the final phase of excavation beginning in February 2004, an area of 30x20 meters was excavated using the rapid digging method in order to salvage as many artifacts as possible before the site would be destroyed by construction. Besides bagging large artifacts, the team rented a mechanical excavator to transport soil from the top 30 cm of the rapid digging location to the east lawn for sifting to recover smaller and more recent artifacts.