Ancient shopping lists point to widespread Bible-era literacy

Handwriting recognition algorithm suggests even lower-rank soldiers were writing.

Cathleen O'Grady – Apr 15, 2016 11:32 am | 96

The fortress in Arad.

Around 2,600 years ago, in a military fortress in Southern Judah, a man called Eliashib sent and received messages written in ink on fragments of pottery. The contents were mundane, mainly concerning food supplies, but they provide evidence of literacy that could inform the debate about when major Biblical texts were written.

Eliashib’s correspondence happened on the cusp of the fall of the Kingdom of Judah, which took place during 588-87 BCE. The date plays an important role in an ongoing debate among Biblical scholars: were the first Biblical texts produced before the fall of Jerusalem—as events were unfolding—or afterwards? One part of the debate hinges on the literacy levels at the time: if the pre-demolition population wasn’t generally literate, it wouldn’t have been likely that important historical texts were created in this era.

But Eliashib and his colleagues in the Arad military fortress provide some evidence that literacy in this era may have been more widespread than previously thought. A multidisciplinary group of researchers from Tel Aviv University have combined their expertise in applied math, Jewish history, and archaeology to assess communications from the fortress, trying to establish how many people, and of what rank, were writing messages.

The ink-inscribed pottery fragments used by Eliashib are called ostraca, and more than 100 of them have been found in the remains of the Arad fortress, which is around 50 miles south of modern Jerusalem. Of these, 16 were long enough and legible enough for computer analysis. This one is typical: “To Eliashib, and now: Issue from the wine 3 baths and Hananyahu has commanded you to Beersheba with 2 donkeys’ load and you shall wrap up the dough with them.”

After they had reconstructed the faded brush strokes, the researchers developed new machine learning methods for handwriting analysis. These methods could describe elements of the handwriting, such as the angles between different strokes, and estimate how likely it was that different messages were written by the same person. If it was less than two percent likely, the authors took the messages as having been written by different people. Based on this, they estimate that six different people were responsible for the 16 ostraca in the analysis.

Seven of the messages were addressed to Eliashib, and he wrote one of them himself. Based on the content of the messages, the researchers suggest that other obviously literate individuals must also have included the unnamed military commander who wrote one of the messages, the fortress commander addressed in another message, and a subordinate who addresses Eliashib as “my lord.”

From this, the authors suggest that literacy spread quite far down the military ranks. “Extrapolating the minimum of six authors in 16 Arad ostraca to the entire Arad corpus, to the whole military system in the southern Judahite frontier, to military posts in other sectors of the kingdom, to central administration towns such as Lachish, and to the capital, Jerusalem, a significant number of literate individuals can be assumed to have lived in Judah,” around 600 BCE, the authors write.

Of course, that’s a huge amount of extrapolation. There’s evidence to support some of it; for instance, there are ostraca from other military bases around the same time. However, it’s a leap that entails a number of assumptions: not only about the ranks held by these individual message-writers, but also that levels of literacy would be similar everywhere.

These results bring new evidence to the debate about the timing of Biblical texts, but they’re far from the final word. Even if we were to assume that the researchers’ extrapolations to the whole population of Judah are sound and that there was widespread literacy at the time, that would settle just one part of the debate—one of the conditions for pre-586 BCE Biblical writing would be met, but there’s still no evidence of actual Biblical writing from the period.

Nonetheless, the techniques developed by the researchers will undoubtedly be useful in expanding the search for ancient literacy in Judah and elsewhere. Their estimate of six different authors is conservative; there may have been more, and the same analyses applied to different places could give us a much clearer insight into literacy in the ancient world—even if all we have to go on are shopping lists.

PNAS, 2016. DOI: 10.1073/pnas.1522200113 (About DOIs).