2. Excavation Methodology
The EMP excavation comprises three main components: the evaluation test pits, the watching brief and the rescue excavation.
2.1. Evaluation Test Pits
A three-day excavation was carried out in June 2014 to profile the soil stratigraphy of the site and to determine the presence (or absence) of archaeological deposits (Figure 11). The manual excavation of two test pits measuring 2 metres by 1 metre was spaced randomly apart—one on each end of the western and eastern sections of the lawn—to estimate the horizontal extent of archaeological deposits, if any. At about 170 to 200 centimetres below the surface, a fine black sandy stratum—typical of Temasek deposits and containing a variety of 14th-century ceramics (Figure 12)—was encountered in both test pits. Sieving of the soil heap also uncovered small finds. More importantly, the test pits provided archaeologists with information on how far these deposits were from the surface.
2.2. Watching Brief
Fullerton Road bounds the EMP excavation site to the north, while the Empress Place (road), the Dalhousie Obelisk and the Victoria Theatre and Concert Hall mark the southern, eastern and western limits respectively (Figure 13). To accommodate the eight rain trees that were originally planted along Fullerton Road, the developer assigned arbitrary Zones 1 to 9 (west of the obelisk) for transplanting. This area will later become the main excavation area for archaeologists. The other areas affected by earthworks consist of Zone 10 (east of the obelisk) and Zones 11 to 13 (north-end of the site).
The archaeological team took on a watching brief at the start of construction, during the removal of topsoil for the insertion of sheet piles into the ground (Figure 14). The 3-metre depth needed for transplanting meant that sheet piles are a safety requirement for all of the zones. Other than scant traces of the more recent colonial period, nothing remarkable was seen during the initial stages of the watching brief. Underlying much of the landscaping topsoil is the historical roadway and construction fill that is contained in a matrix of clayey sand. After much of this modern fill material was removed, archaeologists became concerned when mechanical excavators dug up fine black Temasek sand deposits (Figure 15). The machines were prevented from digging deeper and at least 20 cm of the overlying soil and debris was left in situ to protect the archaeological deposits.
Following the establishment of the Singapore Land Authority (SLA) precise level benchmark (80153)—located at the Cavenagh Bridge—as the archaeological site datum, the start depth of pre-modern deposits was determined at 130–150 CMBD (centimetres below datum). Other than the main excavation area, archaeologists were not granted access to work in the other zones. This is regrettable as archaeological deposits were later found in these areas as well as in a series of trenches that were dug for drainage and utility diversion. Not particularly helpful were the random finds collected by construction workers.
2.3. Rescue excavation
As previously mentioned, the main excavation area, measuring 70 metres by 13 metres, is made up of zones 1 to 9 (Figure 16). Archaeologists excavated only zones 3, 4, 7, 8 and 9, which were divided into 23 excavation units measuring either 5 metres by 5 metres or 5 metres by 4 metres. With regard to zones 1, 2, 5 and 6, contractors destroyed these areas before archaeologists could excavate them (Figure 17). Fortunately (or unfortunately), salvage screening recovered about 600 kilograms of artefacts—with no primary provenience—from the spoil heaps of zones 5 and 6. This artefact mass constitutes 20% of the entire EMP artefact yield.
After gridding the main excavation area, the controlled digging of 20-cm spits using handheld tools and trowels started at a depth of 130 CMBD. When time allowed for the more detailed documentation of interesting features, diggers excavated in spits of 10 cm. Screening of the spoils using 6 mm and 3 mm wire mesh began once the upper protective layer was removed or when the pre-modern sandy stratum was encountered. Excavation forms, scaled drawings, and photography completed the documentation, and this work cycle proceeded until no more artefacts were found. Being first and foremost a construction site, disruption to archaeology was frequent. For example, work in the excavation units was halted whenever a tree was being transplanted. On one occasion breached water mains inundated several excavation units, causing damage and work to be halted for half a day. It is clear that for future projects, archaeology must precede construction. Out of the total area affected by construction work (2,273 m²), only 19.57% (445 m²)—inside the main excavation area—was systematically excavated and documented. Of the areas not excavated by archaeologists, only 18.48% (420 m²) was supervised by limited amounts of watching briefs. This meant that in addition to the area already intruded by underground services, more than half of the total area affected by earthworks was not documented by archaeologists.