The Copper Plates in the Corroded Ring
An assistant at a small district museum in Assam opens a donation box: five copper plates, each about 21 cm by 12 cm, strung on a corroded ring, with roughly a third of the text lost to corrosion. The readable portion records a grant of a village, a field and two slaves to a temple by a ruler whose name survives only as '...varman'. With the plates comes a handwritten note: the donor family says it has held them for 'forty generations' and that they prove the family owned land from the seventh century. Two items are already in the museum's files: a 1987 local-history pamphlet saying the village was founded in the twelfth century, and the 1905 district gazetteer describing the same village as 'ancient, with traces of a brick shrine' but giving no dates. There is no epigraphist on the staff, the acquisition budget is Rs 4,000, and the note must be filed in three weeks. The curator asks the school's history group what the plates can honestly be used to prove.
1. Classify the four items (the plates, the 1905 gazetteer, the 1987 pamphlet, the family tradition) by type of evidence, then rank them for the narrow question: did this family hold land in this village in the seventh century? Justify the ranking.
Read the worked answer
Classification
- Copper plates - primary, documentary, contemporary with the event they claim to record if genuine; but fragmentary, and their provenance is uncontrolled: no find-spot, no record of when or where they were dug up.
- 1905 gazetteer - secondary, compiled from revenue and settlement records; independent of the family's claim.
- 1987 pamphlet - tertiary: a later retelling without citations; it is evidence of what local people believed in 1987, not of the seventh century.
- Family tradition - oral; a memory claim about possession and identity, not a date claim.
Ranking for this question: plates > gazetteer > family tradition > pamphlet.
Reasoning - the criteria are (i) nearness in time to the event, (ii) independence from the person making the claim, (iii) checkability. The plates win on (i) but lose badly on (ii) and (iii): they have been in the family's hands and nobody outside has read them.
The catch - even a perfectly genuine plate proves a grant to a temple, issued by a ruler. It says nothing about the donor family's own ownership. If the family descends from the temple's managers or from the donee, that is a separate argument needing a genealogy, a later sanad, or land records. The honest answer to the curator is therefore: 'these may prove a grant, not a family estate'.
2. The family offers 'forty generations' as proof of a seventh-century date. Do the arithmetic and give two reasons why the figure cannot settle the question.
Read the worked answer
The arithmetic Take an average generation length of 25 years: 40 × 25 = 1,000 years, which from today reaches back only to about 1025 CE. Vary the assumption: at 20 years per generation the count reaches about 1225 CE; at 30 years, about 825 CE. The whole plausible band is roughly 800–1200 CE. To reach the seventh century the average would have to be about 35 years in every generation - that is, unusually late fatherhood repeated forty times without exception.
Why the count cannot be decisive
- Generation length is not a constant. It depends on marriage age, remarriage, infant mortality and how many children survived; in a family with many children the effective length is shorter than the age gap between first and last child.
- Oral genealogies are curated, not recorded. They use round, conventional numbers - 'forty' often means 'very many' - they skip or compress generations, and they are edited to attach a family to a famous ancestor or to a contested landed right. The count is a hypothesis generator, not a date.
Use it for what it is: evidence that the family has a long-standing association with the plates and the village. Then date the plates by script, not by genealogy.
3. The museum has Rs 4,000 and no epigraphist. Propose a three-step plan and write the exact label text you would accept.
Read the worked answer
Step 1 - Document, do not interpret. Photograph every plate on both faces with a scale, take ink rubbings or estampages, record measurements, weight, corrosion pattern and the ring, and write down the donor's account verbatim with date and signature. Assign accession numbers. Cost: near zero; this protects the object and the claim.
Step 2 - Desk comparison. In a college or district library, compare the script, the royal titles and the grant formulae with published corpora (Epigraphia Indica, Annual Reports on Indian Epigraphy, standard introductions to Indian epigraphy). This cannot date the plates, but it can say whether the script is plausible for a given period and whether the wording is a known type.
Step 3 - Ask an expert. Write to a university department of ancient history or epigraphy, or to the state archaeology office, enclosing the photographs. Meanwhile display a facsimile, not the original.
Label text I would accept 'Set of five copper plates, ring corroded; text incomplete. Script provisionally assigned to the second half of the first millennium CE. Authenticity, date and the identity of the granting ruler are not yet established. The family tradition of long possession is recorded but unverified.'
Why this wording - it states what is physically observable, marks the date as provisional, and refuses the words 'seventh century' and 'proves', which the evidence cannot carry.