Labelling a riverbank assemblage: Roman contact or Satavahana palimpsest?
Trustees of a district museum in the Krishna delta have three weeks to decide how to label a collection recovered during a fish-pond excavation near a river confluence. The assemblage: 47 lead coins carrying a bow-and-arrow and a legend in early Brahmi, six silver denarii (four heavily worn, one struck under Tiberius), 31 sherds of Rouletted Ware among 213 sherds, iron slag, a fragment of a baked-brick wall, and two terracotta bead moulds. No stratigraphy was recorded; the coins were bought from a labourer who says they came from 'one pit'. A local heritage group wants the label 'Roman trading colony, 1st century CE'. A retired epigraphist has written to the newspaper that the lead coins are 3rd-century Satavahana issues and that the whole assemblage is 'a rubbish heap'. The ponds will be deepened in six weeks. The museum has no radiocarbon budget this year but can spend about 40,000 rupees.
1. Set the most defensible date range for the assemblage and state exactly what the museum may and may not print on the label.
Read the worked answer
What each class of evidence can carry
- Rouletted Ware has a broad currency in peninsular India, roughly 2nd century BCE to 2nd century CE. It brackets a period; it does not date a layer.
- The Tiberius denarius (struck 14-37 CE) gives a terminus post quem only for the loss or arrival of that coin. Worn Roman silver circulated for generations in India, so it cannot date the wall.
- Bow-and-arrow lead coins of the Satavahana series run into the 3rd century CE, so the epigraphist is probably right about their date.
Consequence With no stratigraphy and coins bought from a labourer, this is a palimpsest, not a phase. The defensible label is: 'Surface assemblage, c. 1st century BCE - 3rd century CE; evidence of exchange with the Roman world. Provisional.'
Print 'Roman trading colony' and you assert settlement, status and a single date that two denarii and 31 sherds cannot support. Say 'contact' and 'exchange', mark the provenience as disturbed, and put the confidence level on the label itself. That is not hedging; it is the honest description of a surface scatter.
2. With six weeks and a small budget, design the salvage programme. What must be done first, and what must be refused?
Read the worked answer
First, fix the context before it disappears. Clean and photograph the exposed pond section, record its height and bedding, and lay a 5 m grid over the spoil to collect ceramics and slag by square. Use a hand-held GPS and a level; a total station is better if the engineering college lends one. Two days of a draughtsman's time will cost less than the coins are worth.
Second, convert 'bought' into 'reported'. Take a signed, dated statement from the labourer at the pit itself, with photographs. That establishes a find-spot; it does not make the coins excavated finds.
Third, go through the law. Report the find to the state department of archaeology. Buried coin finds fall under the Treasure Trove Act 1878, and the Antiquities and Art Treasures Act 1972 governs their sale and registration; the museum cannot quietly accession purchased coins.
Refuse: chemical cleaning of the coins, destructive sampling, radiocarbon on a brick with no context, and any payment structured as a purchase.
Publish a short interim note, so the record survives even if the ponds are deepened. Most of the budget should go to survey labour and photography.
3. The heritage group and the epigraphist are now quarrelling in public. Draft the museum's response, showing how you separate evidence from inference.
Read the worked answer
Concede what is correct. The lead coins are very likely 3rd-century Satavahana issues. Say so first, in the epigraphist's own terms. Credibility is built by yielding on the point you cannot defend.
Then separate the two claims. A 3rd-century coin date does not make the site a rubbish heap. A riverside scatter with Roman silver, Rouletted Ware, slag and bead moulds indicates a location in a exchange network over several centuries - which is more interesting than a single-phase 'colony' and better supported.
Explain the inference in plain language. 'Roman contact' means goods and silver moved; 'Roman colony' means Romans lived there under some authority. The Periplus describes trade at named ports, and Arikamedu is the standard comparison; neither licenses the second claim from this evidence.
Invite a joint field visit with both critics, and offer to print their dissent in the catalogue. Publish the label with graded confidence: established / probable / possible.
Do not say 'the public wants a story'. The museum's job is to be the institution that does not simplify when pressured.