A response to Lemaire and Delorme
In their article, “Mesha’s Stele and the House of David” (BAR, Winter 2022), André Lemaire and Jean-Philippe Delorme argue that the reading btdwd (“House of David”) can now be “confirmed once and for all” (p. 40) thanks to new imaging of both the inscription itself and the squeeze that was produced in 1869. In our response, “Set in Stone? Another Look at the Mesha Stele” (BAR, Spring 2023), we presented arguments in favor of continuing to view this reading as a hypothesis—not an unreasonable proposal, but nor one that could be judged “confirmed.”
Lemaire and Delorme replied to our article on Bible History Daily, to which we have been invited to respond. We wish to express, first and foremost, our deep respect for our interlocutors and our appreciation for their work on this important inscription. The brief remarks offered here are intended to clear up any misunderstanding and correct several misrepresentations of our views.
First, we wish to restate a point made in our first communication, but which the response of Lemaire and Delorme leads us to believe we need to state again as explicitly as possible. Our contention is not that the reading btdwd is impossible or even implausible, but rather that it cannot yet be identified as a confirmed reading, and that it remains hypothetical. A claim to have confirmed a reading is a claim to have removed all room for reasonable doubt. A reading that is identified as hypothetical, on the other hand, can continue to be debated. In order to speak of a confirmed reading, a high evidential bar has to be cleared. Our BAR article argued that bar has not yet been reached.
Lemaire and Delorme argue that “the work of an epigrapher does not stop at identifying clear and complete letters. S/he also has to interpret, within certain limits, what are sometimes incomplete and partial letters.” We could not agree more! Our point was not that the incisions should be crystal clear and, as they are not, they are impossible. Rather, we argued simply that they are far from being clear enough to allow scholars to speak of confirmed readings; we concluded that we found “no solid evidence” (p. 57) for the proposed readings. We agree that epigraphers are often forced to make judgment calls—and Lemaire especially has a long track record of very good judgment—but we would place the emphasis more heavily on the “certain limits” to which Lemaire and Delorme allude. And the burden of proof naturally rests on the scholars who defend hypothetical readings.
In fact, in their printed article in BAR, Lemaire and Delorme acknowledge differences in the degrees of probability they ascribe to the reading of different letters. Although they speak of “practical certainty” with regard to the disputed dalet (p. 38), they admit that the reading taw “remains uncertain” (p. 37) and “somewhat unclear” (p. 40).
Taw (X-shaped letter)
It is difficult to understand why Lemaire and Delorme now write, in the conclusion of their online post, “that the new images of the stela and the squeeze render the identification of taw as highly probable,” since they do not put forward any new evidence for it. Most of their online response regarding the taw is not devoted to explaining why there is actually some evidence favoring it, but rather why we should not expect to find such evidence, due to the poor state of preservation of the stone and of the squeeze. While doing so, they implicitly admit the lack of solid evidence.
We argued that there are no traces of scribal strokes belonging to a taw, only striations and small depressions (see Fig. 1). Lemaire and Delorme claim that we “do not consider that these observations may derive from the very poor condition of the stone itself.” In fact, our article does note that the Mesha stone is “heavily abraded in many areas” (p. 54, quoting Nathaniel Greene and Heather Parker). More importantly, while the damaged state of the inscription and squeeze could account for the lack of probative evidence for a taw, we must consider another equally possible explanation for this absence, namely, that a taw was never there. Absence of evidence is not evidence of absence, but neither should such absence be construed as positive evidence.
Lemaire and Delorme add that their reconstructed taw fits well “the standard measurements of this letter” in the same inscription. But since this concerns the reconstructed taw, that is, the letter as they draw it, not as it could be empirically observed, this only shows that such a reading is theoretically possible; it is not evidence for that reading.
Dalet (triangle-shaped letter)
Turning to the reading dalet, Lemaire and Delorme criticize our use of an image (bottom of p. 56 in our article) as evincing inappropriate method (i.e., use of direct rather than raking light; use of an image of the face of the squeeze that was not in contact with the stone) and an argument from silence. This criticism is not directed toward the argument we made, and as it touches on an important point of method, we wish to clarify our approach here.
The misunderstanding is due in part to the fact that space constraints made it impossible to print the fuller image we originally submitted to BAR. We present the fuller image below (in two versions, with and without annotation; Figs. Figs. 2a-b and 3a-b), but first an explanatory remark on the nature of the evidence afforded by the squeeze is in order. A squeeze like that produced by Yaqub Karavaca is executed by laying a thick piece of paper on the surface of the inscription, dampening it, and pressing it against the inscription. The result is that the surface of the paper in contact with the inscription is distorted as it is pressed into the incisions. This is unlike a cast, in which liquid plaster is applied to the surface and “fills in” the incisions, resulting in a negative of the inscription on one side but a hardened, flat surface on the other. In contrast, the distortion of the paper produced in a squeeze provides evidence of two sorts for the original incisions.
The first are the actual physical distortions of the paper following the shape of the incisions. These distortions are visible not only on the side of the paper that was pressed against the incision but, very often, on the other side as well. You can see this by comparing the two frontlit images below (with tracing [Fig. 3b] and without tracing [Fig. 2b]): the triangular head of the bet and the tail that descends and sweeps to the left can be seen even on the upper (obverse) surface of the paper (i.e., the surface not in contact with the incision).
The second type of evidence accompanies the distortion just described: As the paper is pressed into the incisions, it becomes compacted where the incisions are and is consequentially thicker and denser than in surrounding areas, decreasing its translucence. As a result, the application of backlighting allows one to perceive shadows where the paper is thicker from being pressed into the incisions. We don’t always get both sorts of evidence at the same time. Consider the bet in the image above. Its triangular head leaves almost no shadow in the backlit image (Fig. 2a) but is clearly visible in the distortions it has left on the surface of the paper (Fig. 2b), even on the side of the paper not originally in contact with the incised stone. Conversely, the waw is clearly visible in the backlit image (Fig. 2a) but has left very little distortion visible on the top surface of the paper (Fig. 2b).
The difficulty is that both sorts of evidence are secondary. They are not traces of a workman’s chisel but simply indicators that the paper has been distorted and thickened in a particular area. Examination of the originals shows that both surface distortions and shadow-yielding compacting has happened not only where letters were originally engraved, but also in places on the stone where no letter was ever present. Consider, for example, the “ghost letter” that appears just to the right of the waw in the backlit image above (labeled A in Fig. 3a). Surface distortions and shadows, in other words, can be the result not only of an originally incised letter but also of cracks or other forms of damage to the original or even of secondary stress to the squeeze itself.
This means that the epigrapher has to determine not only whether a shadow is visible in any given case but also whether it is meaningful. To justify the assumption that it is meaningful, it should belong clearly enough to a recognizable sign (as in the case of the letters bet or waw above). What we observed, however, was that the shadows present between bet and waw (where Lemaire and Delorme read taw and dalet) do not belong clearly enough to recognizable signs.
Lemaire and Delorme state that we “misled” readers with the image on p. 56, suggesting that we used it to imply that nothing was present in this space. In fact, this was not the argument we made. The image on p. 56 was not used to show what is not between the bet and the waw but to show what is visibly present on the surface of the paper. Our point is that the backlit image shows a long diagonal line and that a wrinkle of the same proportions and placement is also visible on the surface of the squeeze, boxed in red in the frontlit and backlit images above (Figs. 3a and 3b). Significantly, both the shadow and the corresponding distortion coincide exactly, which makes it highly likely that the “small gap” pointed out by Lemaire and Delorme in the long line is incidental—similar “small gaps” occur in the forms of undisputed letters on the same backlit image of the squeeze, and entire parts of such letters simply do not show at all. More importantly, in order to arrive at the reading of taw and dalet, one is required to assume that certain portions of the shadow and wrinkle “count” while others do not. In our view, this introduces a level of interpretation incompatible with a claim to have arrived at a certain or confirmed reading.
Finally, we did in fact examine this surface feature with both direct and raking light, and a raking-light image only brings the surface wrinkle into sharper focus. We also examined and photographed with both direct and raking light the side of the squeeze in contact with the stone, noting the absence of any meaningful trace on that side of the suggested dalet (and taw). i Examination of RTI images from the West Semitic Research Project (Inscriptifact database) yielded similarly negative results.
Word Divider (dot)
Finally, regarding the word divider, the situation is similar to that of taw: Lemaire and Delorme note that both the stone and the squeeze are “heavily deteriorated” here, and they claim that the unclear evidence reflects this fact, rather than the absence of a word divider in the original inscription. They also minimize, in our view, the differences between the dot here and assured word dividers elsewhere on the stone.ii Here again, whatever happened, the basic fact is the lack of solid evidence.
We wish to stress once more that our observations are not incompatible with the reading btdwd (“House of David”), but that, in our view, it is premature to suggest that there is solid evidence for that reading. We are talking here of epigraphic evidence, that is, material evidence; this is not the same as possible reconstructions (filling in the lacunae of missing letters) based on contextual arguments. Some scholars might well hypothesize the reading “House of David” based on the literary context of line 31, or on the historical context as they understand it (another area of debate). But this is not the same at all as arguing that this reading is based on material evidence. We hope that further developments in imaging technology may yet allow a greater degree of certainty regarding the appropriate reading of line 31 of this fascinating inscription. Until then, debate and discussion are likely to continue.
[i] For images and discussion, see Matthieu Richelle, “A Re-Examination of the Reading BT DWD (‘House of David’) on the Mesha Stele,” in Shmuel Ahituv, Hannah Cotton, and Matthew Morgenstern, eds., Ada Yardeni Volume, Eretz-Israel 34 (Jerusalem: Israel Exploration Society, 2021), pp. 152*–159*.
[ii] For images illustrating this point, see Figs. 2 and 4 in Richelle 2021.
Sign up to receive our email newsletter and never miss an update.