Friday, March 05, 2004

THOUGHTS ON FREEDMAN'S ARTICLE: Might the "Jehoash Inscription" be genuine after all? Here are a few off-the-cuff responses to Professor Freedman's points. I haven't canvassed the literature on this, online or elsewhere, and, although these are all my own observations, others may have gotten there before me. If anyone has already made the same points, please e-mail me with references or links so I can give you credit. My conclusions are based only on the specific points discussed below. I'm not taking a position on the geological analysis (which I am not qualified to evaluate) or the script (which would take me a lot more time to evaluate than I'm willing to invest).

Here's one point Freedman and I agree on:
Authenticated inscriptions frequently challenge our knowledge of Biblical Hebrew, containing syntax, vocabulary and orthography (spelling) that differ from Biblical usage. Of the very few inscriptions of any kind from this period, including those of neighboring nations, every one provides something novel and sometimes disturbingly surprising about a language we may think we know but don?t always fully grasp. And the surer we are, the more surprised we are likely to be by what comes out of the ground.

The study of ancient Northwest Semitic inscriptions is made very difficult because our corpus is so small. So we should be very cautious about how much we claim to know. That said, I do think there are some things that we can know with reasonable certainty.

There are two main arguments in this article which I want to address. The first is:
Perhaps the most fought-over issue so far is the inscription's use of the word bedeq in line 10. In the Bible, bedeq means "crack" or "fissure." However, the inscription, because it combines bedeq with the verb 'asah ("to do," line 9) appears to mimic not the Hebrew of the Bible but rather the modern Hebrew phrase 'e'aseh 'et bedeq, which means "I made repairs." It therefore looks as if the inscription betrays a knowledge of modern Hebrew. However, it is unlikely that bedeq means "repair" in the inscription. The word is actually part of a construct chain that joins it together with the word habbayit ("the House," line 10), forming the expression "bedeq habbayit" or "the bedeq of the House." If we interpret bedeq as "the repair," then we would have to join it similarly to the other definite nouns ("walls," "ledge," "lattices," etc.) and read the text as "the repair of the walls," "the repair of the ledge" and so forth. But this is not possible. The appearance of the definite direct object marker 'et in line 11, though it does not appear before every noun, is an indication that all the nouns function as direct objects of the verb 'asah (i.e., "I did the walls," "I did the ledge" and "I did the lattices," etc.). It would seem that 'asah alone is the verb used in the inscription to mean "repair," not bedeq, which most likely carries its Biblical meaning "crack." Use of 'asah to mean "make new" or "remake" is unusual, but not unimaginable (see Deuteronomy 21;12; 2 Samuel 19:25).

In a nutshell, the expression used in the inscription looks identical to a modern Hebrew expression that means "to make a repair," and this usage is not attested in Biblical Hebrew. In biblical Hebrew the phrase would mean "to make a breach/crack." Freedman, however, suggests that bedeq in the inscription means "breach" or "crack" as in BH but that (asah', which normally means "to make" or "to do" is used in specialized sense meaning essentially "to repair." As evidence he gives Deut 21:12 (the captive woman "does" her nails, i.e., trims them) and 2 Sam 19:25 (Mephiboshet does not "do", i.e. dress?, his crippled feet or "do," i.e. trim, his moustache ).

I'm not convinced. The examples Freedman gives have to do with attending to bodily upkeep. I cannot find a case of (asah being used of repairing a building. The expressions I would expect would be chizzeq bedeq, "to strengthen a breach" (cf. 2 Kgs 22:5); or chiddesh, "to renew (the House)" (cf. 2 Chr 24:4); or just chizzeq "to strengthen (the House)" (cf. 2 Kgs 22:6); or badaq "to repair (the House)" (only in 2 Chr 34:10). These are the biblical expressions. Is it possible that the inscription preserves an otherwise unattested sense of 'asah as "to repair"? Sure. But there's no evidence for it.

Here's another line one could take to defend the inscription. The verb badaq, noted above, is obviously related to bedeq, "breach, fissure, crack" and means "to repair a breach, fissure, crack." So the idea of repair was associated with the root in the biblical period. That leaves open the possibility that there was a noun meaning "repair" from the same root, either as an alternate sense of bedeq or perhaps another noun with a different vocalization. If that's right, then (asah bdq may have been an acceptable usage in Iron Age Hebrew and the identical formation in modern Hebrew could be a later, independent back formation.

Conclusion: the expression (sh bdq could possibly be a legitimate ancient formation and our lack of information about ninth century Hebrew should make us hesitant to insist that it couldn't be. Nevertheless, there is no positive evidence for such an expression and the identical modern phrase should make us suspicious.

Second argument:
More problematic is the word 'mw ("his people," line 15), which is indeed unusual for this time and appears to many to be a clear anachronism. The normal indication of a singular masculine possessive suffix attached to a singular masculine noun is by a he at the end, with the letter representing the sound -hu. The use of the waw, for the suffix, doesn't turn up in the archaeological record until the oldest Qumran manuscripts dating from the third-second century B.C.E, meaning that the shift from he to waw came sometime between the sixth and third centuries B.C.E. So how can we explain the waw in the Jehoash inscription? The Siloam Tunnel Inscription, from the eighth century B.C.E., contains a noun with the suffix waw�the word r'w ("his fellow"). This unusual form is explained as a contraction of the archaic r'hw with synacope loss of the he. Although there is no example of "his people" spelled with a he anywhere in the Bible or extrabiblical sources, there is little doubt that in early periods the word would actually have been pronounced with a -hu suffix, making it possible that the word 'mw in the Jehoash inscription, like r'w of the Siloam inscription, is a contraction of an archaic form 'mhw.

I don't see the two cases as comparable, for reasons that unfortunately require me to go into excruciating comparative-Semitic detail. (In what follows I must be more rigorous in transliteration and short vowels are indicated as lower case and long vowels as capitals. If you'd like to read a brief discussion of the orthography of ancient Hebrew and Aramaic, see my article "Orthography" in Schiffman and VanderKam, Encyclopedia of the Dead Sea Scrolls [Oxford: OUP, 2000], 625-28.) Freedman's case sounds plausible initially only because he cites all the forms without vocalization and he does not give ordered sound changes to get from the Proto-Semitic forms to the forms in the inscriptions. He seems to be floating a general possibility without having a specific route in mind to get to the form in question. I don't think there is one.

As for the Siloam inscription: the vocalized form of R(HW would have been *Ri(iHU (with the normal u-vowel before the suffix assimiliated to the i of the noun, probably under the influence of the gutteral. Cf. BH RE(EHU). This appears to have collapsed into the form Ri(iW, which is a reasonable possibility. (It may also be that the word is plural, in which case the orthography would be unremarkable.)

But the geminate (MW in the "Joash Inscription" would be vocalized as (aMMO (from Proto-Semitic *(aMMuHu, with the -uHu suffix collapsing to long O by the biblical period). In the ninth century this word would have been written (MH, since final long O was always written with a he in this period. I can't figure out any way that *-uHu could have collapsed in Hebrew to anything that could have been spelled with a waw in the ninth century B.C.E. The use of waw to mark final long O (standard in the Masoretic Text) is very late, Persian period or later. Although our knowledge of ancient orthography is imperfect, we do understand the history of the orthography of official Jerusalem Hebrew pretty well, and it boggles my mind to think that a final long O could have been written with a waw in the ninth century.

Conclusion: it looks to me as though the inscription was written by someone who was not thinking in ancient Hebrew (Modern Hebrew seems most likely) and who tried to adjust the vocabulary and orthography to fit the ancient language, but who did not entirely succeed.

Bottom line: if I had to bet, based on the two cases [now three - see below] that Freedman deals with, I would bet the inscription is a forgery.

UPDATE (6 March): I forgot to discuss one more argument that Freedman offers:
The only instance of a possible medial vowel letter in the inscription is the waw in the word lwlm ("staircase," in line 12). It is possible, however, that the original form had a diphthong. (lawlim) or simply a consonant (lewulim).

The issue here is that medial or internal vowel letters are generally agreed to have been introduced into Judahite Hebrew spelling much later than the ninth century. The earliest certain case I know of is the form )RWR, "cursed be," in the Siloam Tomb inscription (c. 700 B.C.E., although the place name ZYP may be another example from sometime in the eighth century. Thus waw to indicate long U seems to be an anachronism in this inscription.

Freedman posits two explanations. The first is that the original form had the diphthong -aw, in which case the waw would have been written as in the inscription. The problem with this idea is that the biblical Hebrew form for this word is LUL , with a long U. (I know of no cognates in other Semitic languages.) If the original form was a diphthong, the form should be LOL, with a long O. Hebrew long U comes from Proto-Semitic long U, not from a diphthong. The first explanation does not work.

His second explanation is that the waw was a consonant, and he suggests the form *lewulim. I'm not sure what to make of this form; it just doesn't look possible to me. There was no long or short e vowel in Proto-Semitic and I don't see how to get to one in this sort of environment in the Hebrew of the ninth century. In any case this looks like a triphthong, which is a very unstable form and which had collapsed to long E or long U in proto-Hebrew. Perhaps he has in mind the waw being doubled: *lawwalim or the like. This is possible. The problem is that you can't get from a doubled waw in this form to the long U in the biblical Hebrew form LUL. Doubled waw doesn't turn into long U. This explanation doesn't work either.

If one wants to save the form LWLM, the best approach would be to point to the Aramaic Tel Fakhariyah inscription from Syria, which dates to the ninth century or earlier and which has numerous medial vowel letters, especially in names and loanwords. One could then say that we don't know all that much about the development of the system of vowel letters and that it is not in principle impossible that we could find an internal vowel letter in an ninth century Judean inscription. I can't argue with this, but it's an obscurum per obscurius explanation, which makes me nervous. It makes me even more nervous that the word LUL means stairway in Modern Hebrew and is spelled with the waw. We already have reason to suspect a forger thinking in Modern Hebrew and this does nothing to allay our fears.

Bottom line: as above.