Archaeology Division, National Museum of the Philippines, Manila 1000, Philippines

Transcription

1 1 The first settlement of Remote Oceania: the Philippines to the Marianas: supplementary information on radiocarbon dating of the Nagsabaran site Hsiao-chun Hung 1*, Mike T. Carson 2, Peter Bellwood 1, Fredeliza Z. Campos 3, Philip J. Piper 4, Eusebio Dizon 5, Mary Jane Louise A. Bolunia 5, Marc Oxenham 1 & Zhang Chi 6 1 School of Archaeology and Anthropology, The Australian National University, Canberra ACT 0200, Australia 2 Micronesian Area Research Center, University of Guam, Mangilao, GU 96923, USA 3 School of Humanities, University of Hong Kong, Pokfulam Road, Hong Kong S.A.R., P. R. China 4 Archaeological Studies Program, University of the Philippines, Diliman, Quezon City 1101, Philippines 5 Archaeology Division, National Museum of the Philippines, Manila 1000, Philippines 6 School of Archaeology and Museology, Peking University, 5 Yiheyuan Road, Beijing , P. R. China * Author for correspondence ( hsiao-chun.hung@gmail.com) This paper is published in full in Antiquity 85 no. 329 September 2011 ( Here we publish supplementary material. The authors compare pottery assemblages in the Marianas and the Philippines to claim endorsement for a first human expansion into the open Pacific around 1500 BC. The Marianas are separated from the Philippines by 2300km of open sea, so they are proposing an epic pioneering voyage of men and women, with presumably some cultivated plants but apparently no animals. How did they manage this unprecedented journey? Keywords: Oceania, Marianas, Neolithic Philippines, Austronesian Table 1 lists the radiocarbon dates from the 2009 excavations at Nagsabaran, and from the main pit, Pit 1, excavated in 2000 by the Philippine-Taiwan team (Tsang 2007: 94). Three additional samples dated from the 2000 excavation are not listed since they fall within the range represented by the Pit 1 dates and add nothing to the overall chronology. Basically, the Nagsabaran stratigraphy consists of two series of layers. The upper Hung, H-C., M.T. Carson, P. Bellwood, F.Z. Campos, P.J. Piper, E. Dizon, M.J.L.A. Bolunia, M. Oxenham & Z. Chi The first settlement of Remote Oceania: the Philippines to the Marianas: supplementary information on radiocarbon dating of the Nagsabaran site. Antiquity:

2 2 series is a midden formed of shells of the riverine bivalve Batissa childreni, often very loose and heavily disturbed by postholes and graves. This shell midden extends down to a maximum depth of 2.5m below the surface in Pit 1. While the dates from the midden are not in precise stratigraphic order, presumably because of the ease of disturbance of the loose shell, they nevertheless fall within a range from about 850 BC at the base of Pit 14 to about AD 500 in the upper midden layers. Two samples, ANU and 13018, are clearly much older, and presumably are on shells scuffed into the midden from earlier unknown locations. One of these, ANU was clearly Preceramic in origin. The other, ANU-13017, could be from the older Neolithic occupation discussed below. The material culture in the midden includes a black-surfaced late Neolithic and Iron Age style of pottery with shapes and decoration very different from that in the earlier Neolithic layers below the midden. Iron and glass beads occur in burials dug into the midden from above, presumably after AD 200. In typological terms of vessel shape and decoration, there appears to be a break between the pottery assemblages of the lower silts and the upper midden, but this is not entirely evident from the 14 C date distribution. A number of large postholes were dug from the top of the silt series immediately below the shell midden, and these penetrate deeply into the silts below. They represent a settlement of houses raised on stilts that was in existence when the shell midden began formation. These early postholes do not contain shell midden. A few later ones, dug from within the midden itself, do contain shell. The second series of layers consists of the alluvial deposits that lie immediately below the base of the shell midden. Apart from the upper surface with the postholes, these silty clays reveal no signs of actual human occupation no midden, stamped floors, postholes or burials. It is these lower deposits that contain the red-slipped pottery discussed in this paper, and it should be noted that this pottery style does not occur in the midden above. These lower layers are clean silty clay, of alluvial origin, with charcoal only surviving where protected beneath sherds or in postholes. Our assumption is that these silty clay layers were deposited by fluvial action on the site of the future shell midden, bringing in cultural material from areas of Preceramic and Neolithic occupation which we have not yet located through excavation. The Nagsabaran shell mound is 4.2ha in size, and less than 1 per cent of its surface has been excavated. For that reason, the dates from these lower silts are a little confusing, especially for Pit 9 where there is a great deal of stratigraphic inversion represented in the five dates listed Hung, H-C., M.T. Carson, P. Bellwood, F.Z. Campos, P.J. Piper, E. Dizon, M.J.L.A. Bolunia, M. Oxenham & Z. Chi The first settlement of Remote Oceania: the Philippines to the Marianas: supplementary information on radiocarbon dating of the Nagsabaran site. Antiquity:

3 3 in Table 1. Pit 7, however, has three dates in good order by depth. We make the following observations on the dates from the lower silty clays: 1. GX and 26711, NTU-3798, WK and 17756, ANU and all date within the time-span of the shell midden, after 850 BC, and thus probably come from materials introduced by the digging of postholes or other disturbances from immediately beneath the base of the shell midden. This could explain why they are all on charcoal. They do not come from disturbed shell midden since these lower layers have no shells, except where they have been carried down by a small number of identifiable tree root holes, these also being a major source of disturbance. 2. GX and WK are presumably from Preceramic activity, and not relevant for dating the Neolithic occupation. 3. The other five samples should relate directly to the Neolithic occupation and the redslipped and stamped pottery discussed in this paper. They begin with the domesticated (and thereby introduced) pig tooth date WK (Piper et al. 2009), and thus extend from slightly before 2000 BC to about 1300 BC. Perhaps shell sample ANU (discussed above) can also be added to this group. This is the dating range, based on a total of six 14 C samples, that we apply to the Nagsabaran red-slipped pottery in this paper. Dates on shells of Batissa childreni Dr Stewart Fallon of the AMS Radiocarbon Laboratory in the Research School of Earth Sciences at ANU kindly agreed to date five samples of modern freshwater shell for us. The results are also listed in Table 1. All have terrestrial (freshwater) δ 13 C values that are moderately negative. The four samples collected from Zabaran Creek near the Nagsabaran site have percentage modern carbon values that, according to Dr Fallon, are predictable for terrestrial/riverine samples under post-bomb conditions. The sample from the main Cagayan River (ANU-13021) is from a source that could have marginal tidal influence, so its slightly greater age (98.47 per cent modern) could be due to a very slight marine reservoir effect, although this does not come through in the δ 13 C value. Some of the tributaries of the Cagayan River, but not apparently Zabaran Creek, also run through limestone, so this would be another source of slight contamination by old carbon in the main river channel. Nevertheless, Dr Hung, H-C., M.T. Carson, P. Bellwood, F.Z. Campos, P.J. Piper, E. Dizon, M.J.L.A. Bolunia, M. Oxenham & Z. Chi The first settlement of Remote Oceania: the Philippines to the Marianas: supplementary information on radiocarbon dating of the Nagsabaran site. Antiquity:

4 4 Fallon s advice is that the dated modern samples do not show strong effects from limestone or marine sources of old carbon, and with this in mind we have calibrated the results using the same IntCal 09 data base as for the charcoal samples. Higham and Higham (2009) have recently used freshwater shells for successful dating of the occupation and mortuary sequence at Ban Non Wat in north-eastern Thailand, similar in time range to the occupation of Nagsabaran, even though the two sites are unrelated culturally. A modern shell sample dated by Higham and Higham produced a result of per cent modern, very close to the results for modern Nagsabaran shells, indicating the absence of any significant limestone or marine reservoir effect. Hung, H-C., M.T. Carson, P. Bellwood, F.Z. Campos, P.J. Piper, E. Dizon, M.J.L.A. Bolunia, M. Oxenham & Z. Chi The first settlement of Remote Oceania: the Philippines to the Marianas: supplementary information on radiocarbon dating of the Nagsabaran site. Antiquity:

5 5 Table 1. Summary of radiocarbon dating of earliest site deposits at Ritidian and Unai Bapot, Mariana Islands. Conventional Site and Lab Sample Measured 14 C δ 13 C Marine reservoir Calibrated Provenience C age reference sample material age (years BP) ( ) correction (ΔR) * 2σ probability ** (years BP) Ritidian, Guam (Carson 2010) Beta Fenceline Pit 35; m; later cultural layer Charcoal 2820± ±40 n/a BC (0.4%); (1%); (98.6%) Beta- Fenceline Pit 35, 2.50 Anadara 3030± ±40-44± BC m, earliest cultural layer, intertidal zone antiquata shell Beta- Fenceline Pit 35, 2.55 Halimeda 2980± ±40-44± BC m, earliest cultural layer, intertidal zone sp. algal bioclast Beta m, pre-dates cultural layer Heliopora sp. coral 3610± ±50-44± BC (99.7%); BC (0.3%) limestone Unai Bapot, Beta- Layer III-A, combustion Charcoal 2850± ±40 n/a BC (100%) Saipan (Carson 2008) feature, post-dates earliest cultural layer Beta- Layer IV-A, localised Anadara 3210± ±40-44± BC discard pile, earliest cultural layer antiquata shell Beta Layer IV-A, localised discard pile, earliest cultural layer Anadara antiquata shell 3320± ±50-44± BC * Marine reservoir correction of -44±41 was calculated for Anadara antiquata shells at the Ritidian site in northern Guam (Carson 2010).

6 6 ** Calibrations are by CALIB software version 6 (Stuiver &Reimer 1993), using INTCAL09 dataset for charcoal specimens and MARINE09 dataset for marine specimens (Reimer et al. 2009). Table C dates from Nagsabaran, Cagayan Valley, northern Philippines. The upper shell midden is represented by dates from Pit 1 (P1) excavated in 2000, and Pit (P 14), excavated in All dates from all pits that relate to the alluvial silt layers below the shell midden are listed in this table. The Gakashuin and National Taiwan University dates listed in this table are from Tsang 2007: 94 and we do not have measured δ 13 C values. Sample # Dated material Pit number and depth Conventional age Calibration δ 13 C below ground surface (years BP) (IntCal 09) GX Charcoal P1, 0.8m, shell midden 1470±50 AD GX Charcoal P1, 1.1m, shell midden 1670±60 AD GX Charcoal P1, 1.2m, shell midden 2120± BC AD 335 GX Charcoal P1, 1.4m, shell midden 1830±70 AD GX Charcoal P4, 1.5m, shell midden 2150± BC AD 175 GX Charcoal P1, 1.8m, shell midden 1920± BC AD 320 GX Charcoal P1, 1.8m, shell midden 1760±110 AD GX Charcoal P1, 1.5m, shell midden 1960± BC AD 245 GX Charcoal P1, 2.3m, shell midden 2260± BC AD 336 GX Charcoal P1, 2.4m, shell midden 2240± BC AD 346 GX AMS Charcoal P1, 2.5m, shell midden 1820±40 AD ANU Batissa childreni P14, 0.8m, shell midden ± BC ANU Batissa childreni P14, 1.2m, shell midden ± BC

7 7 ANU Batissa childreni P14, 1.4m, shell middem ± BC ANU Batissa childreni P14, 1.8m, shell midden ± BC ANU Batissa childreni P14, 2.1m, shell midden ± BC NTU-3799 Batissa childreni * P1, 3.1m, lower silts 3450± BC GX AMS Charcoal P2, 1.4m, lower silts 2620± BC GX Charcoal P2, 1.5m, lower silts 6610± BC GX AMS Charcoal P4, 2.1m, lower silts 2520± BC NTU-3798 Charcoal P7, 1.6m, lower silts 2670± BC GX Charcoal P7, 1.6m, lower silts 3050± BC GX Charcoal P7, 1.9m, lower silts 3390± BC WK Pig premolar ** P9, 1.4m, lower silts 3940± BC WK Charcoal P9, 1.5m, lower silts ± BC WK Animal bone P9, 1.5m, lower silts ± BC WK Charcoal P9, 1.6m, lower silts ±30 21 BC AD 127 WK Charcoal P9, 1.8m, lower silts ± BC ANU Charcoal P11, 1.7m, lower silts ± BC ANU Charcoal P14, 2.4m lower silts ± BC ANU Charcoal P14, 2.4m, lower silts ± BC ANU Batissa childreni Modern shell, Cagayan River ANU Batissa childreni Modern shell, Zabaran Creek % modern % modern

8 8 ANU Batissa childreni Modern shell, Zabaran Creek ANU Batissa childreni Modern shell, Zabaran Creek ANU Batissa childreni Modern shell, Zabaran Creek % modern % modern % modern * Sample originally published as charcoal. ** Piper et al References CARSON, M.T Radiocarbon chronology with marine reservoir correction for the Ritidian archaeological site, northern Guam. Radiocarbon 52: HIGHAM, C. & T. HIGHAM A new chronological framework for prehistoric Southeast Asia, based on a Bayesian model from Ban Non Wat. Antiquity 83: PIPER, P.J., H-C. HUNG, F.Z. CAMPOS, P. BELLWOOD & R. SANTIAGO A 4000 year-old introduction of domestic pigs into the Philippine Archipelago: implications for understanding routes of human migration through Island Southeast Asia and Wallacea. Antiquity 83: REIMER, P.J., M.G.L. BAILLIE, E. BARD, A. BAYLISS, J.W. BECK,C.J.H. BERTRAND, P.G. BLACKWELL, C.E. BUCK, G.S. BURR, K.B. CUTLER, P.E. DAMON, R.L. EDWARDS, R.G. FAIRBANKS, M. FRIEDRICH, T.P. GUILDERNSON, A.G. HOGG, K.A. HUGHEN, B. KROMER, G. MCCORMAC, S. MANNING, C. BRONK RAMSEY, R.W. REIMER, S. REMMELLE, J.R. SOUTHON, M. STUIVER, S. TALAMO, F.W. TAYLOR, J. VAN DER PLICHT, C.E. WEYHENMEYER IntCal04 terrestrial radiocarbon age calibration, 0-26 cal kyr BP. Radiocarbon 46: STUIVER, M. & P.J. REIMER Extended 14 C data base and revised CALIB C age calibration program. Radiocarbon 35: TSANG, C-H Recent archaeological discoveries in Taiwan and northern Luzon: implications for Austronesian expansion, in S. Chiu & C.

9 Sand (ed.) From Southeast Asia to the Pacific: archaeological perspectives on the Austronesian expansion and the Lapita cultural complex: Taipei: Academia Sinica. 9