6.1Two kinds of release
The two cases release different things. The Hammurabi excerpt releases a debtor from one year's payment of grain, and only for a natural failure of the crop; it releases a debtor's household from service after three years; and it releases a slave mother on repayment of her price. It does not, in the laws at hand, release any debt as a class. The Solonian accounts release existing debts as a class, persons held or sold for debt, and the land from its markers of debt. They also forbid, for the future, the kind of loan that had led to bondage.
Set against the clean-slate thesis, the first case is not a witness for it, and the second is a witness of an unexpected kind. The clean slate of the literature is a recurring act of the ruler; the Babylonian law collection at hand offers instead a recurring rule that needs no act; the Athenian account offers a single act that ends in a permanent rule. Graeber's and Hudson's notes do not say what either author makes of law 48 or of the Solonian prohibition, and the article does not supply their view. What it can say is that a reader who goes to these two cases looking for periodic royal cancellations finds, in the words that survive in these translations, something else: in Babylon, rules for bad years and limits on bondage; in Athens, one cancellation joined to a prohibition meant to make a second unnecessary.
This bears also on the side of the debate that attends to kinds of text. Westbrook's note names royal edicts that remitted debts as one of the subjects of its reference work. No such edict is among the texts at hand. The absence of a clean slate from the Hammurabi excerpt is therefore an absence from a law collection, which is not where an edict would be found. The case shows the value of the distinction rather than the answer to the question. And the modern record of the Babylonian case, which warns that enforcement would need contracts and court records, takes the same view.
6.2Who paid
On the question of the bearer, the two cases differ in a way the parent study's ledger cannot see. Law 48 does not name the creditor as bearer, but its form makes the creditor the bearer: the debtor "need not give his creditor any grain". Laws 115 and 116 put on the creditor the risk of the debtor's death in his custody. Law 117 puts on the buyer of a debtor's household the loss of their labour after three years. The Babylonian text names its bearers by the logic of its provisions, not by any statement of loss.
The Solonian accounts name their bearers explicitly and by group: the rich who "were vexed because he took away their securities for debt"; the poor who did not get the land they expected; the wealthy who lent to the friends; Solon himself with his five talents or fifteen. Androtion's version even names a group who did not lose: on his account those who accepted payment in the new coin "were no losers". The disagreement among the ancient sources is a disagreement about who bore the cost.
Neither modern record names any bearer. Across the parent study's eighty-four cases, a record that names who bore the cost of an act is rare; the Appendix gives the count. The two oldest cases in the ledger are not exceptions on the record's side. They are exceptions on the side of the texts behind the record. The article does not say the records ought to have named the bearers; they are summaries for a different purpose, and their care over the boundaries of the measures is real. It says that a reader who relies on the summary alone will not learn who paid, and that the texts tell.
6.3For how long
The question of term divides the cases along the line between rule and act. Law 48 names its term: "that year". Law 117 names its term: three years, free in the fourth. Both are terms that run from an event, the bad harvest or the sale, and both recur whenever the event recurs. They are written in advance and do not depend on anyone deciding when the term ends.
The Solonian measures have no term in the texts read here, and for a reason that the texts make plain. The cancellation of existing debts is complete when it is made; it has nothing to end. The prohibition on loans against the person is, in Aristotle's words, a liberation "once and for all"; it is meant not to end. Bordo and Kydland's escape clause, a suspension that the public expects to be reversed, does not fit the Athenian case: nothing in either account suggests that anyone expected debts to be reinstated or loans on the person to return. It fits the Babylonian law 48 better, in a limited way. The year's excuse is suspended payment, not cancelled debt, and the contract resumes with the next harvest; law 52 shows that the contract itself is not weakened. Their argument concerns convertibility and parity, and its use here for grain debts is a transfer of concept, not an application of their findings.
6.4What the texts can and cannot carry
The texts can show what the surviving words of two standard witnesses say about release, its condition, its term and its bearers. They can show that the ancient Athenian sources themselves disagreed about whether Solon's disburdenment was a cancellation. They can show that the modern records keep the boundaries of the measures and lose the bearers.
They cannot show enforcement. No contract, court record or ledger of debts cancelled is among the texts at hand for either case, and both modern records say that such documents would be needed. They cannot show frequency: one excerpt of a law collection and one episode in Athens do not count anything. They cannot show what Hammurabi's Akkadian says, only what King's English, as transcribed online and not checked against the printing, says; nor what Solon's Greek says, only what Perrin's and Kenyon's English says. And the Athenian accounts are later histories: Aristotle's, on the record's own answer, of the fourth century BC; Plutarch's, on its note, written centuries after Solon and drawing on earlier writers. When those histories report Solon's verses, they report verses the article cannot check against any earlier copy.
The clean-slate thesis therefore survives the test, in the sense that two of its witnesses do not disprove it; and it is qualified by the test, in the sense that one of the witnesses does not testify to it and the other testifies to a single act that its own ancient readers disputed. Whether specialists who dispute Graeber's and Hudson's claims do so on these grounds, the notes do not say.