Comment on dates from a resin-coated sherd from Spirit Cave, Thailand.

White, Joyce C.

Lampert et al. (2003) argue that the early Holocene dating (c. 7500 BP uncalibrated) assigned to the earliest ceramics at Spirit Cave, Thailand should be revised based on two new dates of c. 3000 BP from a single resin-coated sherd from that site. But Lampert et al.'s article does not demonstrate the stratigraphic or typological relationship between their one sherd and the sherds upon which Gorman claimed early Holocene dating for the appearance of ceramics at Spirit Cave.

Gorman excavated at Spirit Cave twice, initially in 1966 (Gorman 1969, 1972) and again in 1971, and the records are housed at the University of Pennsylvania Museum. Provenance nomenclature received in 2000 in a spreadsheet from Lampert indicates that the 20 sherds (labelled A through T) in her possession at Bradford University came from the 1971 excavation. Although the details of the 1971 excavation were never published by Gorman, the provenance information indicates that the 20 Lampert sherds came from either (1) the cave surface with no horizontal provenience (i.e., no square designation--eight sherds), or from (2) the removal of "baulk extensions." Baulk extension is the phrase used by Gorman for deposits along the cave wall left between it and some of the squares excavated from the 1966 excavations. None of the 20 Lampert sherds came from the main excavation squares of either season.

For those 12 Lampert sherds from the baulk extensions, the depth and stratigraphic information that could be gleaned from the original field records was tenuous. In the comprehensive information I sent to Lampert (White 2001) I state that the sherd that later became the one from which dates were derived (Sherd N) is "probably best considered depositionally equivalent to the sherds marked 'T19 SUR,'" (T19 is the site number Gorman gave Spirit Cave). Elsewhere I stared that sherds marked T19 SUR were "most likely to be from surface deposits with potential to have been 'kicked around.'" I noted to Lampert that only "Sherd G and Sherd F appear to come from the lowest stratigraphic positions of your set of sherds" with possible association with Hoabinhian lithics (all italics in the original White 2001). However, Sherds G and F have no resin coating.

Lampert's spreadsheet of her 20 sherds and Gorman's discussion of the 426 sherds from the 1966 excavations make clear that there is variability among the fabric, surface treatments, and decorative finishes in the ceramics recovered from Spirit Cave, so that the sherds from the overall assemblage do not have to be of similar date (the full count and character of sherds from the main 1971 excavation squares are unknown). Regarding the earliest sherds from the site, Gorman (1969:672) states, "Impressed into the surface of layer 2 were several concentrations of pottery ... The ceramic material [was] characteristically cord-marked, [and] also included a smooth, burnished ware ... The cord-marked sherds were a uniform dark reddish brown ... averaging 3 to 6 mm in thickness." The dating of these sherds to the early Holocene is noted in Lampert et al. (2003). It consists of three early Holocene dates from the same surface as the ceramics (surface of layer 2), but Lampert et al. (2003: 128) note that "it is largely on the date obtained from the sample taken from within Layer 1 [7400 [+ or -] 300 BP uncalibrated; FSU 317] that the possible debate hinges." Deposition of the charcoal providing this latter fourth date from within Layer 1 arguably superseded deposition of the sherds impressed into the surface of Layer 2.

According to Gorman (1972:96), resin coating was found on only one kind of sherd: those that had been net-impressed. In the Lampert collection, all four resin-coated sherds were also listed as net-impressed. Only two net-impressed sherds had no resin coating. Gorman provides no specific information on the stratigraphic context of the resin-coated or net-impressed sherds, but neither does he mention them as among the "early ceramics" impressed into the surface of layer 2.

The provenance information for the Lampert sherds indicates that of the six net-impressed sherds, five unambiguously are attributed to the surface and three of these have resin coating. The one remaining resin-coated net-impressed sherd is Sherd N, the one from which the late Holocene date comes. The provenience information on that sherd is ambiguous. It came from baulk extension C1-C2, and the layer Lampert indicates is written on the Ziploc bag is layer 8. However as reported to Lampert in White (2001), there is no record for the excavation of a layer 8 in baulk extension C1-C2. The deepest layer mentioned in the field notes for C1-C2 is layer 4. Moreover according to the Lampert spreadsheet, this Sherd N had been bagged with the three other resin-coated net-impressed sherds (Sherds O, P, and Q) that were marked as coming from the site's surface. I concluded that the "8" is probably a transcription error for "S" and state in White (200l) "I therefore think it is highly likely that the ... '8' for Sherd N [indicates] the deposit excavated as 'surface' or 'S' in C1-C2." I conclude that all resin-coated net-impressed sherds in Lampert's collection came from the site's surface.

In short therefore, it is probable on the basis of several lines of evidence that the resin-coated sherd dated by Lambert et al. (2003) came from the surface of Spirit Cave. Thus the sherd and its dates are not from a stratigraphic context equivalent to the sherds used by Gorman in his efforts to date the first appearance of ceramics at that Cave. Although the newly dated sherd from Spirit Cave does add interesting information to the site's chronology, depositional history, and ceramic typology and use, the claim that this date "revises" the "Spirit Cave early ceramic dates" is overstated. The AMS dates support the presence on the surface at Spirit Cave of resin-coated net-impressed pottery dating to the late Holocene c. 3000 BP uncalibrated. Even if the dating for the earliest ceramics at Spirit Cave and surrounding regions may ultimately be revised, the dated sherd reported in Lampert et al. (2003) does not provide a basis for revision.

References

GORMAN, C.F. 1969. Hoabinhian: A Pebble-tool Complex with Early Plant Associations in Southeast Asia, Science 163: 671-3.

--1972. Excavations at Spirit Cave, North Thailand, Asian Perspectives (1970) 13: 79-107.

LAMPERT, C., I.C. GLOVER, R.E.M. HEDGES, C.P. HERON, T.E.G. HIGHAM, B. STERN, R. SHOOCONDEJ & G.B. THOMPSON. 2003. Dating Resin Coating on Pottery: The Spirit Cave Early Ceramic Dates Revised, Antiquity 77:126-133.

WHITE, J.C. 2001. [Unpublished letter to Cynthia Lampert dated March 10, 2001, giving provenience information on Lampert sherds A-T from the original Spirit Cave excavation records.]

Response by C.D. Lampert (1) *, I.C. Glover (2), R.E.M. Hedges (3), C.P. Heron (1), T.F.G. Higham (3), B. Stern (1), R. Shoocongdej (4) & G.B. Thompson (1)

Joyce White notes that the resin-coated potsherd front which the dates were obtained came from "the surface of Spirit Cave". This is quite correct, as indicated in our discussion of the limitations of dating the Spirit Cave ceramics using their resinous coatings. That the newly dated sherd does not come from a sealed stratigraphic context is a valid point, and a regrettable one. It would indeed be "desirable to measure the date of resin on other samples of pottery from Spirit Cave" (Lampert et al. 2003:129), ideally from within sealed contexts. Sadly, given that of the 426 potsherds from the original excavations only 22 can now be found, this is not feasible.

In describing the Spirit Cave ceramics, Gorman (1969; 1972) reports them as a group of artefacts, with no differentiation between the typologies of those seen on the Layer 2 surface and those from Layer 1. Neither does he note whether the potsherds illustrated in his 1972 interim report were recovered from within the very shallow upper stratigraphy of the sire (c.3 to 9 cm depth) or as surface finds, which stone indisputably were.

At least two of the potsherds at Bradford are illustrated in Gorman's interim report (Plate V. d and f, 1972), thus would have come from the first excavations at Spirit Cave in 1966 and be included amongst the 'early' ceramics. However, White confirmed to the authors that some of our sherds derived from the second Spirit Cave excavation, thus we believe that the small collection of ceramics at Bradford is likely to include material recovered from both the 1966 and 1971 excavations.

White proposes that the variability in fabric, surface treatments, and decorative finishes "may vary by time depth", and implies that the recently dated resin-coated net-impressed sherd might thus be later in date than Gorman's 'early' sherds. To support her argument, White, in reviewing Gorman's description of the earliest sherds at Spirit Cave (1969:672), notes that he does not mention net-impressed ceramics in his description of the 'early' ceramics. However, the resin-coated sherd from which the AMS date was obtained (Lampert et al., 2003:127, Figure 2) appears to be typologically closely related to a sherd illustrated in the interim report on the first excavation at the site (Plate V. c, Gorman, 1972), and could come from the same vessel.

Furthermore, in his description of the 'early' ceramics, Gorman (1969:672) reports potsherds "incised with a multipronged tool". A implement of this type would almost certainly leave parallel incised lines comparable to those on two of the four resin coated sherds held at Bradford. Whilst Gorman (1972:96) appears to suggest that only net-impressed sherds were resin coated, this is not in Fact the case.

Given the ambiguity as to the extent of the sherd assemblage which is included with the 'early' ceramics described by Gorman, it is entirely plausible that typological relationships exist between the resin-coated sherds, including the net-impressed sherd from which the AMS date was obtained, and the 'early' ceramics from the site.

Clearly, a typological relationship does not offer the certainty of association that would be provided by a new date on a resin-coated sherd from a sealed context within the stratigraphy. Unfortunately, unless the remainder of the ceramics assemblage from Spirit Cave comes to light, this is unlikely to be achievable.

Higham's (1989:60) caveat regarding the early ceramic dates at this site, and similar concerns expressed by Bellwood (1999:97) do point to the need for caution in accepting the 'early' dates claimed for the ceramics from Spirit Cave. The recent dates add weight to the argument that the original early Holocene dates proposed for the ceramics assemblage at Spirit Cave may be questionable. The reliability of the AMS dates on the resin-coated sherd from Spirit Cave are supported by AMS radiocarbon dates from resinous deposits on pottery from other sites in Southeast Asia.

In conclusion, whilst Joyce White may be justified in her view that dates from a single potsherd cannot provide the basis for a complete revision, they certainly provide a starting point for reopening the debate regarding the site's chronology and the inferred appearance of ceramics during the early Holocene.

(1) Department of Archaeological Sciences, University of Bradford, UK

(2) Institute of Archaeology, University College London, UK

(3) Oxford Radiocarbon Accelerator Unit, Oxford University, UK

(4) Department of Archaeology, Silpakorn University, Bangkok, Thailand

* (Email: c.lampert@bradford.ac.uk)

References

BELLWOOD, P. 1999. Southeast Asia before History. In The Cambridge History of Southeast Asia: From Early Times to c. 1500, Vol. 1., Pt. 1. (ed. N. Tarling): 55-136. Cambridge University Press, Cambridge.

GORMAN, C.F. 1969. Hoabinhian: A Pebble-tool Complex with Early Plant Associations in Southeast Asia, Science 163: 671-3.

--1972. Excavations at Spirit Cave, North Thailand, Asian Perspectives (1970) 13: 79-107.

HIGHAM, C. 1989. The Archaeology of Mainland Southeast Asia: From 10,000 B. C. to the Fall of Angkor, Cambridge: Cambridge University Press.

LAMPERT, C., I.C. GLOVER, R.E.M. HEDGES, C.P. HERON, T.E.G. HIGHAM, B. STERN, R. SHOOCONGDEJ & G.B. THOMPSON. 2003. Dating Resin Coating on Pottery: The Spirit Cave Early Ceramic Dates Revised. Antiquity 77:126-133.

Joyce C. White (1)

(1) University of Pennsylvania Museum, 3260 South Street, Philadelphia PA 19104-6324, USA (Email: banchang@sas.upenn.edu)