Olive Pits and Alef-Bets: Notes on the Qeiyafa Ostracon

There seems to be a general feeling that there is something about the Qeiyafa Ostracon that counts against a strong minimalist understanding of the history of the Israelite monarchy. I personally think that a strong minimalist position is highly unlikely even without the evidence of the ostracon, so evidence of this kind would not surprise me. Nevertheless it might be interesting, now that the ostracon has been officially published, to give it a brief assessment of its nature and significance.

One of the key tenets of a minimalist stance is that there is little evidence of a strong state in Israel from the 10th century BCE, the period when the United Monarchy is usually placed. Since Khirbet Qeiyafa is dated to the early 10th century BCE (based on the evidence of pottery and carbon-14 dating of olive pits at the site), this causes any finds from the site to have some bearing on the question of the situation in Judah at the time. Even without the ostracon, the fact that there was a strongly fortified town within the borders of ancient Judah at the time evinces some degree of organized building activity, suggestive of political unity. This is a point that the excavators themselves make.

Another trait of state centralization often said to be lacking during the crucial 10th century is literacy, or at least the evidence of a writing tradition in epigraphic artifacts. Here is where the ostracon comes in. The more the ostracon can be taken as evidence of literacy at some level, or as the product of a scribal culture, the more the idea of a centralized state and its educational system becomes probable. Can it be taken in this way?

The ostracon itself is hard to read because of the faded letters, and even the letters that are clear do not yield a connected text or even hint at a possible genre, at least in the readings contained in the official publication. (The interpretation of Gershon Galil, so far available only in a press release, seems to depend on restoring crucial letters or filling in key lacunae in a way that does not carry conviction.) Although the reading of Misgav and his associates agrees in several points with the alternate reading provided by Ada Yardeni, the two readings also diverge from each other in important ways, and neither reading is conducive to meaningful consecutive translation.

So what do we have?

(1) A text written from left-to-right. Although some may choose to withhold judgment at this point, the fact that the second line swerves upward at the right side of the ostracon confirms the general impression of left-to-right direction: a writer swerves upward when he runs out of room at the end of a line.

(2) A text written in the Old Canaanite form of the alphabet, the form that the letters took before (but more about this later) the evolution of national scripts. The closest analogue among previous discoveries is the Izbet Sartah Ostracon, also written left-to-right. But several of the Qeiyafa letters are still unidentified, and for others there are different opinions about their values. Even for those where there is agreement, there are some violent changes of letter stance within the text, which is more typical of earlier forms of this script but not of later (e.g. Izbet Sartah does not have the same variations in stance). My impression is that this would count against rather than for the idea that there was a widespread scribal culture: surely a guild of national scribes would regularize the letter shapes and stances?

(3) A text whose language, although North-West Semitic, is still undetermined. The key sequence of letters is the first five letters: אלתעש. This has plausibly been interpreted as Hebrew אל תעש, "do not do!" Among the NW Semitic languages, the verbal root עשה 'to do' is diagnostic of Hebrew (and its congeners, such as Moabite). Although HALOT gives a few other languages where the root may appear, in this location at this date, only Hebrew is a viable candidate, if the interpretation is correct.

However, the sequence may be interpreted otherwise, and I will come back to it. Other sequences that have been plausibly read are שפט, line 2, עבד, line 1, בעל, line 3, נקם, line 4, מלך, line 4, and possibly חרם, line 5. The roots give us no help for language identification, since all of them are attested throughout North-West Semitic (although נקם and שפט are less common in Aramaic). The lack of a clear text is a handicap, needless to say.

One key sequence is found at the end of line 4. Misgav et al. read the last few letters as יסד מלך גת, "YSD king of Gath," although the last letter is restored. Yardeni, however, reads it differently. Leaving the yod aside, the last five letters she reads as בדמלך. The letter she reads as bet Misgav et al. take as either samekh or ḥet. But the letter in question, although it does not look like the other bets of the text, does look like the bets of some other Old Canaanite texts. If Yardeni is correct, then this sequence is crucial, because בדמלך is a Phoenician name, Bōd-Milk, and the name element "Bod-" ("in the hand of") is not found in Hebrew.

If this is a Phoenician name, then we might explore, heuristically, whether the other readable sequences also suggest names. In line 1, we can take עבדא as Abda, a name attested in several Phoenician inscriptions (Benz 148), as is Bodmilk (Benz 75). The first three letters of line 2 are שפט and this can be taken as Shaphat, an extremely common Phoenician name (Benz 182-184). In Yardeni's reading, the sequence beginning line 3 is גרבעל which could be "Gerbaal," an attested Phoenician name (Benz 103). The letters נקמי in line 4, while not attested, could be a hypocoristic (shortened) form of a name with the verbal form naqam, "to avenge."

Finally, we may reconsider the form אלתעש in this light. The verbal root עוש has been identified in a number of Semitic names (including Biblical Yeush, Gen 36:5 and elsewhere). If the root does occur in the present sequence of letters, then we might take the letters אלת as the goddess Ilat; the name would mean "Ilat helps" or "Ilat, help!" Ilat is a component of some Phoenician names.

If the names are Phoenician, this might mean that the text is Phoenician, but it doesn't prove it. Phoenician names could be mentioned in a text of another language. But the text as a whole would have to be understood in a convincing way before this could be argued. Unfortunately, we are still far from understanding what kind of text the ostracon is overall.

(4) The most significant fact about the ostracon, in my view, is the date. If the dating of the level it was found in is correct – late 11th/early 10th century BCE – then the use of this Old Canaanite script is surprising. Within a century or less of the ostracon's writing, another inscription would be made in ancient Israel of a very different sort. I refer to the Tell Zayit abecedary, which dates at the latest to the late 10th century BCE. But unlike the Qeiyafa ostracon, the Tell Zayit text is written from right-to-left and already has many of the distinctive letter shapes that would characterize Hebrew inscriptions after that time. It is on the way to becoming the Hebrew national script. If any text by itself indicates the presence of a literate culture or scribal guild, it is not Qeiyafa, but Tell Zayit. The chronological gap between the two epigraphs is not very large, but the contrast is dramatic, and may indicate a correspondingly rapid dramatic change in the cultural situation in 10th century Judah. Was this change caused by the monarchy? That's too big a conclusion from this limited data, but the possibility is tantalizing.

BIBLIOGRAPHY: Misgav, H., Garfinkel, Y. and Ganor, S., "The Ostracon." In Garfinkel, Y. and Ganor, S., Khirbet Qeiyafa Vol. 1. Excavation Report 2007-2008 (Jerusalem: Israel Exploration Society, 2009), pp. 243-257; Yardeni, A. 2009. "Further Observations on the Ostracon." In Garfinkel, Y. and Ganor, S. Khirbet Qeiyafa Vol. 1. Excavation Report 2007-2008, pp. 259-260; Franz Benz, Personal Names in the Phoenician and Punic Inscriptions (Rome, 1972).