Reading a Dry-Zone Cemetery: One Sherd, Two Dates, Six Weeks
A dry-zone mound in the north-central province is being excavated ahead of an irrigation rehabilitation scheme. Fourteen cist burials have been exposed in three trenches. Three graves contain iron implements: a knife, a spearhead and a small adze. Black-and-Red Ware occurs in all trenches; a single sherd of rouletted ware was recorded at 45 cm in square B4, beside a root channel. Two charcoal samples from grave fill gave calibrated ranges of 390-350 BCE and 210-180 BCE at 95 per cent probability. A surface collection includes a tenth-century CE potsherd and several slag fragments. The excavator's draft report describes the site as 'a megalithic iron-working settlement linked to Indo-Roman trade', and the heritage impact assessment must decide within six weeks whether to recommend preservation in situ or record-and-release. The team has no petrographic analysis and no funds for further dating.
1. What chronology can the excavator defensibly claim for the burials, and how should the two radiocarbon ranges be handled?
Read the worked answer
The two ranges do not overlap, which is the central problem: 390-350 BCE and 210-180 BCE are roughly two centuries apart. Two readings are possible, and the evidence must discriminate between them.
- A long-lived cemetery. If both samples are securely associated with sealed grave contexts, the burials were made over at least two centuries. That is a substantive finding, not a nuisance.
- A biased sample. Charcoal in grave fill can be older than the burial: old wood, or charcoal redeposited from earlier occupation on the mound. In that case a sample gives a terminus post quem for the deposit it sits in, not a date for the burial.
- What to report. Calibrated ranges at 95 per cent, the laboratory numbers, the exact stratigraphic position of each sample, and an explicit statement that the cemetery falls somewhere between the fourth and second centuries BCE and probably spans more than one generation.
- What must not be reported. A founding date, a single-phase cemetery, or a precise century. With two dates and no short-life material, the honest statement is a bracket, not a point.
If funds are truly exhausted, the report should say which material would resolve the question (charred grain, human bone, tooth enamel) so a future project can.
2. Does the rouletted ware sherd establish contact with long-distance trade? Set out what would have to be shown.
Read the worked answer
A single sherd establishes that a piece of rouletted ware reached the mound. Everything else is inference, and the inference has several links.
- Context first. The sherd lay at 45 cm beside a root channel. The first question is whether it came from a sealed deposit, a grave fill, or a disturbed horizon. If it is intrusive, it dates the disturbance, not the burials, and may postdate them by centuries.
- Then association. Even in situ, one sherd in one square is not a trade assemblage. It must be shown to be contemporary with the graves, not with later occupation on the same mound.
- Then origin. Rouletted ware was produced in several regions of South Asia, including South India and Sri Lanka itself. 'Indo-Roman' requires evidence of origin, which without petrography the team cannot supply.
- Then scale. Contact, exchange and trade are different claims. A density count per cubic metre, compared with regional assemblages, would be the minimum for any statement about volume.
The defensible sentence is narrow: 'a single sherd of rouletted ware, of uncertain context and origin, was recovered from the upper deposit.' The word 'trade' should not appear in the impact assessment.
3. The draft report also calls the site 'iron-working' on the basis of three iron objects and some surface slag. Assess this, and state what claim the evidence supports.
Read the worked answer
Three iron objects and some slag support iron use at or near the mound; they do not support iron production there.
- Objects are not production. A knife, spearhead and adze show that iron was available, curated and buried with the dead. They could have been made elsewhere and exchanged.
- Slag needs a context. The slag is from surface collection and may be modern, agricultural or redeposited. Smithing slag and smelting slag are distinguishable, and furnace lining, tuyères and hammerscale would be needed to argue production. None is reported.
- Terminology carries assumptions. 'Megalithic' describes a burial form, not a people, language or ethnic group. Using it as a culture label smuggles in an identity claim the excavation cannot test.
- What the evidence supports. 'A cemetery of cist burials, with iron grave goods and Black-and-Red Ware, used within a bracket of the fourth to second centuries BCE, with later activity indicated by a tenth-century sherd and possible ironworking debris of unknown date.'
Given the six-week deadline, the recommendation should turn on the burials, which are demonstrably significant, not on the unverified industrial claim.