Ration Tablets and the Size of a Palace Workforce
A palace archive yielded 320 administrative tablets from a single room, of which 90 are legible ration records. They list workers by category: 40 men receiving 2 sila of barley daily, 25 women receiving 1.5 sila, and 15 apprentices receiving 1 sila. A separate tablet records a monthly delivery of 4,800 sila of barley to the palace 'for the workers'. The archive covers roughly one year, but 30 percent of tablets are fragmentary, and the room may have held more tablets that decayed. A cylindrical seal impression on 12 tablets matches seals found in a storeroom 200 m away. The excavator estimates 600 workers; a colleague says 120.
1. Test the ration figures against the monthly delivery and decide which population estimate the numbers support.
Read the worked answer
First, compute the daily barley need of the listed workers. 40 men × 2 sila = 80. 25 women × 1.5 = 37.5. 15 apprentices × 1 = 15. Total = 132.5 sila/day, or about 3,975 sila per 30-day month.
The recorded delivery is 4,800 sila/month, about 20 percent more than the listed workers require. That surplus is consistent with the listed workforce plus a margin for waste, seed, or unlisted dependants. It does not support 600 workers: 600 at even 1 sila/day would need 18,000 sila/month, nearly four times the delivery.
The 120 estimate is close to the listed 80 workers (40+25+15), which is a reasonable floor. Scaling for the 30 percent fragmentary tablets and possible decay, the true workforce is plausibly 100-160.
Judgement: the delivery tablet is the harder constraint and it caps the workforce well below 600. The excavator's figure would require either a much larger delivery or far lower rations than any tablet records.
2. The seal impression links the archive to a storeroom 200 m away. Explain how this affects confidence in the ration data and what risk it introduces.
Read the worked answer
The seal link is strong corroboration. It ties the ration records to an actual storage facility, meaning the barley in the tablets plausibly corresponds to grain physically held and disbursed nearby. That raises confidence that the records are operational rather than literary or symbolic, and that the delivery figure reflects real throughput.
The risk is double counting or conflation. The storeroom may have served several offices, so its grain total is not necessarily the same barley as the archive's ration list. If the same official sealed both sets of documents, the archive and storeroom are one administrative system; if different officials used the same seal type, the link is weaker than it appears. Seal impressions show authorisation, not identity of the commodity batch.
A second risk: a well-connected archive may over-represent one bureau's records while other workers were fed from unrecorded sources. The 4,800 sila figure could cover only this bureau's staff.
Practical conclusion: the seal link strengthens the case that the ration numbers are real and internally consistent, but it does not license treating the storeroom total as the whole palace's grain supply. The 100-160 estimate should be described as the workforce of this bureau, not the entire palace.