A Coin Hoard and the Limits of Economic Inference
A student is writing on the economy of a province using a hoard of 412 silver coins buried around 260 CE and never recovered. The coins come from six mints; 71% are from one mint, and the latest coins are noticeably debased. The student's draft argues: 'The hoard proves a monetary crisis, a breakdown of trade, and the insecurity of the frontier.' The teacher's feedback says the hoard shows one burial event, that hoards are found because they were lost, and that debasement may reflect fiscal policy rather than crisis. The student has one week to rewrite the analysis using only what the hoard can bear.
1. Identify the three claims in the draft that the hoard cannot support, and give a defensible alternative claim for each.
Read the worked answer
'Proves a monetary crisis.' Debasement is a fact about the coins, but its cause could be fiscal pressure, supply problems, or deliberate policy. Defensible: 'The latest coins in the hoard are debased, indicating a change in minting standards within the hoard's period of accumulation.'
'Proves a breakdown of trade.' Six mints show coins travelled; that is the opposite of breakdown. Defensible: 'Coins from six mints reached this locality, showing circulation across a wide area.'
'Proves frontier insecurity.' Burial is often read as crisis behaviour, but people hoard in peacetime too, and this hoard was never recovered - a single event. Defensible: 'One unrecovered hoard shows that someone concealed wealth and did not return; it cannot be generalised to a region.'
In each case the alternative states what the object shows and stops. The pattern to teach: replace 'proves' with 'indicates' and name the mechanism.
2. The teacher says hoards are found because they were lost. Explain the sampling problem this creates and how the student should address it in the essay.
Read the worked answer
The sampling problem: a hoard enters the archaeological record only when its owner fails to retrieve it. Recovered hoards - where the owner came back - leave no hoard. So the surviving sample is skewed toward episodes of loss, which are plausibly episodes of disruption. Concluding 'the period was insecure' from hoards therefore partly restates the condition of survival.
The student should address this explicitly: 'Hoards over-represent unrecovered deposits, so they are weak evidence for the general level of insecurity; they are strong evidence for individual episodes of concealment.' They can also note the 71% single-mint concentration as a compositional fact that may reflect regional circulation rather than crisis, and flag that a single hoard (n=1) cannot support regional generalisations at all. Naming the bias and narrowing the claim is the rewarded move.