Abstract
Modern scholarship and medieval commentators have generally viewed the execution of the Tabernacle and the priestly ordination described in Exodus 35–40 and Leviticus 8 as the fulfilment of the earlier commands contained in Exodus 25–31, following the command-fulfilment pattern that is so common in priestly writings. We challenge this reading and conclude, on form, thematic, and narrative considerations, that we are in the presence of two alternate versions and that the skillful placement of the second version harmonizes these versions and creates a novel and largely ignored account. Priestly versions, when read in isolation, could have presented two competing (ideological) depictions of the Tabernacle.
Keywords
The Tabernacle accounts: a compiler’s narrative
Modern scholarship and medieval commentators have generally viewed the execution of the Tabernacle and the priestly ordination described in Exod. 35–40 and Lev. 8 as the fulfillment of the earlier commands contained in Exod. 25–31, following a “command-fulfillment pattern” that is common in priestly writings. However, the unusual verbosity of these sections, and in particular their tedious repetitiveness, confounded scholars who knew priestly texts to be generally parsimonious. 1 Driven principally by this problem, we challenge this reading and conclude, on form, thematic, and narrative considerations, that the prolixity is due to the fact that we are in the presence of two alternate versions, each with its own ideology though with many common structural details. We then suggest that the skillful placement of the second version by the putative Pentateuchal compiler-editor (or by a later redactor) harmonizes these versions and creates a novel and largely ignored account.
The indeterminacy of dependency
Exod. 25–31 and Exod. 35–40/Lev. 8 appear to be two similar accounts of the same event, the making of the Tabernacle, 2 the main difference being that one is expressed in the form of commands and the other in the form of their fulfillment. Many of the parallels are verbatim, particularly those dealing with the details of the tent, the furnishings, the priestly vestments, and the priestly ordination. The extent of the repetition attracted the attention of scholars, who concluded that one of the two sections had to be literarily dependent on the other. The dependency thesis led to a long but ultimately inconclusive search for the direction of dependency. Julius Wellhausen (1885) was convinced that Exod. 35–40 and Lev. 8 are dependent on Exod. 25–29 3 —and so were the majority of scholars after him, including Theodor Nöldeke (1869), 4 Abraham Kuenen (1886), 5 Heinrich Holzinger (1900), 6 Gerhard von Rad (1934), 7 Baruch Levine (1965), 8 and Helmut Utzschneider (1988), 9 although different reasons were offered for “proving” this direction of dependency. Noth (1962) would concede that “the pedantic repetition of the whole of chs. 25–31, given there in the form of a command and now in the form of a report on the execution of the work, would not in itself be a valid reason for holding the section to be secondary, for in view of the character of P as it can be observed elsewhere, such a long repetition in his work might be deemed quite possible”. 10 Still, he concluded that 35–39 is dependent on 25–31 and noted that even secondary elements in chapters 30–31 “have been incorporated [in 35–39] into a quite probable systematic order […]. At some points the narrative in chapters 35–39 has been extended even beyond that of chapters 25–31. But even here it is not so much a case of the inclusion of further independent traditions as of making narrative constructions obvious on the basis of chapters 25–31.” 11 The dependency discussions extended to the sections dealing with priestly ordination, some arguing for the dependency of Lev. 8 on Exod. 29, 12 some for the reverse, 13 and some for a common source for both. 14 It should be noted that a resolution of the direction of dependency does nothing to alleviate the main problem: the combined canonical account is tedious and unnecessarily repetitive.
Prolixity and the apologists
The problem of the prolixity of the accounts has been addressed by nearly every traditional and modern commentator of the book. Simply stated, the command could have been followed, for example, with a brief statement that Moses and/or the Israelites did everything that God had commanded. Conversely, the fulfillment could have been preceded by a short notice that a command had previously been given to Moses.
Propp pithily summarized the prolixity problem in a most restrained way, saying that “to the well ordered modern mind” the redundancy between chs. 25–31 and their execution in chs. 35–39 creates “great difficulty”. 15 By contrast, Wellhausen, whom he quotes, minced no words: Exodus 35–39, in his opinion, is “‘utterly meaningless in terms of content”, and it “would not be missed if it were absent.” 16 We find an assortment of answers to this “great difficulty” from traditional and modern scholars. Among the former, Hayyim ben Moses Attar (1696–1743), a famed Moroccan Kabbalist and biblical exegete, suggested that the detailed descriptions of the mishkan were repeated “because the mishkan was dear” (habib) to the author of the Torah, who wished to have it repeated. 17 S. R. Hirsch (1808–1888), rabbi and writer in Germany, commented that “all this prolixity is meant to tell us that both at the making as well as the delivering, and finally at the assembling and erecting, the holy and symbolic meaning and purpose of each and every part which made up the whole edifice was present and vivid in the minds of the workers and Moses, so that each part and the whole was made, delivered and set up in the spirit of its purpose, and so the symbolic character and meaning of the whole edifice and every part of it was vindicated.” 18 Among modern scholars, Richard Elliott Friedman, in his Commentary on the Torah, suggested that “it doubles the message that the Tabernacle is supremely important: the channel through which divine and human communicate.” 19 Gary A. Anderson suggested that “the furniture of the Tabernacle possessed something of the very being of the God of Israel. As such it bears careful repetition whenever the occasion arises, not unlike the piling up of divine epithets in a psalm of praise or description of the beloved from the pen of the lover.” 20 These all appear to be no more than justifications after the fact.
In a novel twist, James W. Watts has suggested (in an article dealing with the seemingly inexplicable editorial preservation of the anti-Aaronide story portrayed in Exod. 32–34 21 ) that the golden calf story “gives the P tabernacle story a creation (Exodus 25–31)/ fall (Exodus 32–33)/ restoration (Exodus 34–40) plot”. That is, these sections create “narrative suspense and intrigue” (p. 424); despite their tedious repetitions, Exod. 25–31 and Exod. 35–40 serve in toto as one single literary and rhetorical unit. But, of course, the same literary effect could be obtained without having to list every detail of the construction. In fact, the tediousness of the presentation takes away from the desired literary effect.
A more attractive solution to the problem of unnecessary repetition, with its excessive prolixity, is offered by the idea that the command and/or the fulfillment may be mirroring a sort of textual archival space. As suggested by Laura Carlson Hasler, late-Persian-period scribes felt a need to imitate ancient Near Eastern archival historiography and textually archived lists, decrees, and letters. Such archival texts were, for instance, embedded within Ezra-Nehemiah and possibly within a number of other late-Persian-period compositions. These textual archives tend to disrupt narratives. Archival history is carried out as a means of legitimation and, in the Ezra-Nehemiah case, restoration. 22 Assuming a late-Persian period date for the composition of the Tabernacle account, one might argue that Exod. 25–31 and/or Exod. 35–40 function as a sort of archival space and belong to “archival history”. We note that Laura Carlson Hasler has found one very convincing specimen of archival history—namely, the book Ezra-Nehemiah. However, is this form of writing pervasive in all the biblical books? One is hard put to find similar occurrences of prolixity through the rest of the priestly writings. Secondly, one of the two sections would have been sufficient for the purpose of archiving for posterity the details of the Tabernacle and its making. Unlike the letters and decrees of Ezra 4–6, one of the two Tabernacle account sections offers little that is newsworthy.
The prolixity problem is pronounced because commentators have largely failed to note significant differences between the two sections that may have made them worthy of being repeated. Only Benno Jacob (1997) identifies and focuses on the differences between the two versions and thinks therefore that the execution is not redundant. As he sees it, Exod. 35–40 narrates the proper execution, whereas Exod. 25–31 contains instructions to Moses about the purpose and significance of the works. 23 At the same time, Exod. 25–31 focuses on the content, whereas Exod. 35–40 is led by practicability; and again, Exod. 25–31 “provides Moses with instructions regarding meaning and purpose of the task and limits its technical instruction” while Exod. 35–40 “reports the actual execution and limited itself to that”. 24 According to Jacob, it is not the aim of the author to highlight the literal conformity and identity (“Gleichheit und Identität”) between godly order and its Mosaic execution. 25
The crux, Jacob adds, lies in how we understand כאשר: Jacob points out it does not mean “exactly as”, but rather “compared/analogue to”. Rather than the workers’ obedience, it is their sensible thought and implementation that is praised. Comparison and analogy are stressed as decisive features of the fulfilling of the godly orders through Moses. 26 To his great credit, Jacob focuses on the differences between the two and claims that these differences made it necessary for both accounts to exist. But as provocative as Jacob’s approach is, his answer is not satisfactory; other than for small details, many of which are attributable to stylistic differences and normal scribal variants, Jacob is hard put to find fulfillments that are not exact mirror images of the commands. Moreover, exactitude was always to be desired and aimed at as a sign of absolute and unconditional obedience. Why would Moses and the others depart in any way from the given instructions?
Having shown that prolixity is a matter to be reckoned with and having discussed dependency, it seems to us that it makes little sense for a redactor, whose aim is generally to improve a text, to deliberately create such a pointless expansion. In other words, it is inconsistent and unfair to assume that a dependency occurred and then accuse the borrowing scribe of being repetitive. Editor/redactors are far less likely to be repetitive than original authors. Solving the question of who borrowed from whom appears to us, therefore, to be an unproductive approach. Instead, it may be more logical to consider the possibility that we are dealing with an alternate second version or tradition. One literary clue pointing to alternate versions A and B (instead of the fulfillment of a command given prior) is the way the divine commandments are communicated in the respective Exodus chapters.
Dismissing the paradigm: a literary clue
Exod. 35–40 describes the making of the mishkan, its appurtenances, and the priestly vestments. The section is introduced by the announcement “This is what the Lord has commanded, saying […]” (Exod. 35.4). Lev. 8 describes the ordination of the priests, introduced by the announcement “This is what the Lord has commanded to be done” (Lev. 8.5). Moses then proceeds to ordain Aaron and his sons via a lengthy and complex set of rituals.
Divine instructions in priestly texts take one of two forms. One form is as a divine communication to Moses containing explicit instructions and a command to relay these to the Israelites (or to Aaron and his children as the case may be). The direct command formula takes the basic form “And the Lord spoke to Moses, saying: Speak to […] and say”. Through the use of this formula, the author informs the reader that Moses, and sometimes Moses and Aaron, are being asked to convey the latest set of divine laws to the Israelites (or just to the priests), or to perform a special deed. This is by far the most common form of divine instruction. The other form as described in the text is by way of an announcement that Moses makes to his audience that he has received a divine instruction to do so-and-so. For instance, “Thus says the Lord” or “This is what the Lord commanded”.
In the announcement modality, Moses introduces God’s wishes to his audience but does not disclose the content of the divine message. Instead, he is seen acting on the instructions. We find two such varieties of announcements that are not preceded in the priestly texts by an explicit instruction:
- “This is what the Lord commanded” (זה הדבר אשר צוה ה). This expression is found at Exod. 16.16, 16.32, 35.4, Lev. 8.5, 9.6, Num. 30.2, 36.6.
- “This is what the Lord meant [lit., said]” (הוא אשר דבר ה). This expression is found in only two places, Exod. 16.23 and Lev. 10.3.
The reader may automatically wonder where or when the Lord issued such instructions, but despite occasional attempts by commentators to identify them, these instructions are nowhere to be found. Quite possibly, the instructions are omitted for reasons of parsimony. 27 The (historical) audience of course is seen as accepting the divine nature of the instruction, though it is never privy to the oracle. It is not unlikely that Moses prefaces his address to the audience by informing the audience that the Lord had instructed him on what he is about to relay. If Moses does address his audience in this manner, however, we are not told about it. 28 The occurrence of a prior command remains implicit and understood; the reader is left to assume that Moses is carrying out exactly God’s will. The parsimonious style of our priestly writer is thus confirmed. In sum, prior commands are seen as wholly unnecessary. 29 The same, we suggest, should apply to the execution of the mishkan project which is introduced with “announcement formulas” at Exod. 35.4 and Lev. 8.5. The reader (and the audience?) need not be informed about the content of the divine communication. Instead, the instructions can be presumed to have been given. It is contrary to all precedent in priestly texts to find these announcements preceded by earlier oracles, as in this case at Exod. 25.1.
Seen in the light of the overall literary style of our priestly author, it is reasonable to believe that Exod. 25–31, containing God’s instructions for building the Tabernacle, was never meant to be the command governing Exod. 35–40 and Lev. 8. Instead, the latter sections are stand-alone units containing a complete pattern of command and fulfillment. 30 It should be noted that the converse is also true: Divine communications to Moses containing explicit instructions and a command to relay these to the Israelites (or to Aaron and his children as the case may be), like Exod. 25–31, are never followed by lengthy fulfillment sections, and certainly not by descriptions that do no more than repeat in toto the subject matter of the command. Instead, these types of commands are generally followed by short sentences informing the reader that Moses, Aaron, and/or the Israelites did as God commanded them. This consistent literary pattern, a mix of parsimony and aesthetic, is further evidence pointing to the lack of narrative affinity between Exod. 25–31 and Exod. 35–40/Lev. 8; the latter sections are just too long and too detailed to constitute a fulfillment of the earlier command.
Summary
We have argued, based on literary form, that Exod. 25–31 and Exod. 35–40/Lev. 8 ought not to be treated as command and fulfillment. Instead, we showed that Exod. 35–40/Lev. 8 plausibly represents a complete, self-contained unit possessing its own (implicit) command and fulfillment. The unit is introduced by Moses, who attests to an unspecified divine command, and then offers a lengthy description of its fulfillment. This structure follows a well-known pattern found in other priestly texts. From a formal, literary point of view, Exod. 35–40/Lev. 8 can stand entirely on its own and has no need of a preceding command section. The implications of our findings are that Exod. 25–31 is conceived not to serve as the command governing both the execution of the tabernacle described in Exod. 35–40 and the ordination ceremonies taking place in Lev. 8 but as an alternate account of it. 31 If there was a fulfillment statement to Exod. 25–31, it is no longer extant. We shall soon see that this conclusion based on form can be confirmed by noting the major ideological/theological differences between the versions.
Two alternate versions
B version
We commence with Exod. 35–40 because, as argued earlier, it is a self-contained unit. And we refer to it as the “B version”. It is worth noting here that the B version may have undergone a number of earlier iterations, some of them possibly visible in the Greek translations and exhibiting significant differences against the MT. We have no intention of dealing with the enormously complex relationship of the Old Greek and the MT, and neither with its second or third transmission. Our literary history begins with the document that gets incorporated into the canonical text (=MT). 32
The B version describes the making of the mishkan, its appurtenances, and the priestly vestments. The “mishkan project” (by this we mean the making of all the parts of the mishkan and its sacred furniture, the erection of the mishkan, the placement of the furniture, and the making of the priestly vestments as ordered by the Lord) is executed by specialized personnel, all under the supervision of two master craftsmen, Bezalel and Oholiob. Bezalel is filled with “divine spirit” and endowed with “wisdom” (hochmah), ability or technical know-how (tevunah, lit. “understanding”), and knowledge in every kind of craft (u’ve-daat u’vechol melakhah) […] (metal work) and to give directions (Exod. 31.2–5).” He and Oholiob “have been endowed with the skills to do any work”—that is, work with cloth. They and other skilled artisans are endowed by the Lord “with wisdom and understanding to perform expertly all the tasks connected with the work of the Holy […]” (Exod. 35.30–35, 36.1). Bezalel and his workmates proceed to make the various sacred objects; there is no indication in the narrative that the craftsmen seek specific instructions from Moses with regards to the design of the sacred objects. This seems reasonable for, after all, they are highly skilled and, to boot, they have been endowed by the Lord with great wisdom and understanding. Exod. 39.32–43 describe the completion of the Tabernacle and all of its furniture, the bringing of these components to Moses for inspection, Moses’ approbation of the work, and the blessing he bestows on the Israelites. The impression one gets is that the project is not only carried out by humans but also engineered and designed by them. This impression is reinforced when one reads that, on completion, the work is presented to Moses for approval, and, on seeing that “they had performed all the tasks—as the Lord had commanded, so they had done—Moses blessed them” (Exod. 39.43). Had Moses been involved in every step of the work, there would not have been a need to present the work for approbation. In short, there is neither a divine nor even Mosaic “plan”; there are no specific building directions. There is an affirmation at Exod. 38.22 that Bezalel (and Oholiob) “had made all that the Lord had commanded Moses”. 33
Exod. 39.32–43 is followed by Exod. 40.17–34, describing the assemblage and erection of the Tabernacle—presumably by Moses himself—and the completion of the work. These two sections are interrupted by Exod. 40.1–16, the command to erect the Tabernacle and to consecrate the furniture and the priests by anointing them with the special anointing oil. These commands are not carried out, as is evident from even a cursory reading of 40.17–34; there was no anointing of the mishkan, the furnishings, and the priests. 34 Aside from the lack of correspondence with what is stated in vv. 17–35, other problems make us doubt that vv. 1–16 belong where they were placed. The instructions to put the special vestments on Aaron and his sons and to anoint them so that “their anointing shall serve them for everlasting priesthood throughout the ages” (40.14–15) presuppose that Aaron and his sons have already been ordained and are ready to enter the priesthood. But, as we know, the priestly ordination does not occur until later, as described in Lev. 8.
If these instructions were indeed supposed to follow the priestly ordination, then one would have to assume that originally the command section contained a brief fulfillment section of all the instructions appearing in 25–29, including the priestly ordination. Only then can Exod. 40.1–16 coherently call for the final preparations of the construction and the priestly consecration. This assumption, incidentally, would further strengthen the case for two traditions or versions. If this hypothesis is correct, one must presume that the fulfillment section was no longer extant at the time of compilation or that it was consciously excised by a compiler in favour of Exod. 35–40/Lev. 8. 35 Alternatively, Exod. 40.1–16 could be a redactional addition, mindlessly and incoherently placed where it is for one purpose only—namely, recasting, reshaping, or reinterpreting the narrative presented in Exod. 40.17–34. The ideological divide between 40.1–16 and 40.17–34 appears to center around the ceremony of hinukh kelim, or ritual initiation of the holy objects—that is, making the mishkan and the holy objects fit for the cult. In rabbinic terminology, the priestly author of the B version holds that hanikhatam hi avodatam—ordination is achieved through use and practice. Stated differently, objects making their cultic debut have to be made to function or occupy a designated position in the primary way for which they were designed. By contrast, the 40.1–16 insertion and its sequel, Num. 7.1, 36 belong to the priestly school that held that hanikhatan hi meshikhatan—objects and priests had to be anointed before they could be made fit for cultic use. 37
Exod. 40.20–33 describes this ordination process. Moses placed the Edut in the Ark, placed the cover on the Ark, and screened off the Ark with the curtain (40.20–21), placed the twelve loaves of bread on the table (v. 23), lit the lamps of the lampstand (v. 25), burned aromatic incense on the golden altar (v. 27), offered up a burnt offering and a meal offering on the altar of burnt offering (v. 29), and filled the laver with water to wash (v. 30). With the erection of the mishkan and all its appurtenances and their consequent ordination, the mishkan becomes ready to serve its purpose: an earthly venue in which the Divine could communicate with Moses and his people. A cloud covers the mishkan, “and the Presence of the Lord filled the Tabernacle” (Exod. 40.34). 38
A version
Exod. 25–29 contains the core of the version introduced by the Lord’s direct command to build a Tabernacle and to ordain the priests—henceforth the “A version”. The core of A is bound by Exod. 25.1, where God speaks to Moses, instructing him to assemble material for the construction of the Tabernacle, and Exod. 29.44–46, where, following the instructions for the building of the Tabernacle and for the priestly ordinations detailed in the previous four chapters, God tells Moses that He will sanctify the Tent of Meeting, the Altar, Aaron, and his sons and that He will abide among the Israelites, and “they will know that I the Lord am their God who brought them out from the land of Egypt that I might abide among them”. These verses connect back to Exod. 25.8 (“And let them make Me a sanctuary that I may dwell among them”), forming an elegant inclusio (v. 45, “and I will abide among the Israelites”) while providing a fitting conclusion to the making of the Tabernacle. The core may have been followed by a fulfillment section that was either lost or excised by an editor, or it may simply have been written ab initio as an alternate or corrective to the (implicit) commands governing Exod. 35–40 and Lev. 8.
The core version is supplemented by a number of significant interpolations and some lengthy secondary additions (in particular, chs. 30–31; see appendix). 39 The core of A in Exod. 25–29 (minus five short interpolations 40 ), consists of (directions for) the making of the Ark followed by the making of the table, the lampstand, and the mishkan (curtains and tent over it) and instructions for the priestly ordination. A appears on the surface to be very similar to B, as the mishkan, its furniture, the priestly vestments, and the rites of ordination are described in nearly the exact same language. This, as we saw, led scholars to posit that A must either be borrowing from B or serve as a source for B. But, as we demonstrate below, there are fundamental differences between A and B.
The differences and their interpretations
Difference #1: In the A version, the mishkan is to be built after the Ark, table, and lampstand. B, however, begins with the making of the mishkan and then follows with the making of the holy furniture. The reverse sequence that the craftsmen followed in B has scholars, working under the assumption that B is the fulfillment of A, scrambling to find explanations.
Traditional exegetes, basing themselves on Exod. 25.9 (“Exactly as I show you—the pattern of the Tabernacle and the pattern of all its furnishings—so shall you make it”), which serves as an introduction and general statement, argue that A, too, intended to begin with the mishkan and follow with the furnishings. The specific instructions, however, begin where the last-named item left off—that is, the furnishings—a common feature in biblical language.
41
The talmudic rabbis are even more creative and argued that Bezalel is a better interpreter than Moses of God’s instructions:
Rabbi Shmuel bar Naḥmani said that Rabbi Yonatan said: Bezalel was called by that name on account of his wisdom. When the Holy One, Blessed be He, said to Moses: Go say to Bezalel, “Make a tabernacle, an ark, and vessels” (see Exod. 31:7–11), Moses went and reversed the order and told Bezalel: “Make an ark, and vessels, and a tabernacle” (see Exod. 25–26). He said to Moses: Moses, our teacher, the standard practice throughout the world is that a person builds a house and only afterward places the vessels in the house, and you say to me: Make an ark, and vessels, and a tabernacle. If I do so in the order you have commanded, the vessels that I make, where shall I put them? Perhaps God told you the following: “Make a tabernacle, ark, and vessels” (see Exod. 36). Moses said to Bezalel: Perhaps you were in God’s shadow [betzel El], and you knew precisely what He said. You intuited God’s commands just as He stated them, as if you were there. (Berakhot 55a, 12)
This is no more, of course, than a brilliant piece of homiletics. Bezalel’s justification for “disobeying” instructions on pragmatic considerations cannot seriously resolve the problem since, after all, the mishkan is not erected until all the furniture has been built. The sequence of construction would not greatly matter. Rather than being a justification for the countermanded execution, the fictive dialogue should be thought of as a creative midrashic flourish to explain Bezalel’s name: he was in the shadow (be-zel) of God (el); he knew best, or he understood better than Moses, his master, what needed to be done.
The source of the puzzlement, as already mentioned, is the belief that the execution countermanded the instructions given in Exod. 25–29. But it is clear that this question can only be asked if Exod. 25–29 and Exod. 35–40 are thought of as a command and its respective fulfillment and not as competing or alternate traditions portraying the same event. Since expediency suggests that all the tasks were likely undertaken simultaneously, it follows that the order of construction given was not describing a reality but rather an order that best suited the degree of sacredness immanent in each of the depicted Tabernacles. The more sacral tradition would list the construction of components in order of their theological importance, and the more “secular” tradition would describe the construction in the kind of order that comes naturally to builders.
While the A tradition considers the entire Tabernacle God’s dwelling place (Exod. 25.8) and the entrance to the Tent of Meeting the place at which God will “meet the Israelites” (Exod. 29.43), it is only from the Ark that God speaks directly to Moses and commands him with respect to them: “There I will meet with you, and I will impart to you—from above the cover, from between the two cherubim that are on top of the Ark of the Pact—all that I have commanded you concerning the Israelite people” (Exod. 25.22). Being the medium through which God speaks to Moses, the Ark is the principal structure and the most hallowed object of the Tabernacle. It’s little wonder, then, that A lists it as the first object to be built.
Difference #2: In the A version, the mishkan follows a divine design. The Tabernacle is built by Moses, who is instructed to follow a celestial blueprint (tabnit, Exod. 25.9, 25.40, 26.30; without tabnit, Exod. 27.8, “as you were shown on the mountain”). Moses follows a supernatural design conveyed by miraculous images projected by God on the mountain. There is no mention of this divine blueprint in B, nor do we find in B any mention that Moses hands over this blueprint to his craftsmen or that they follow any sort of blueprint.
Difference #3: In the A version, the mishkan is personally built by Moses, and not, as in B, by Bezalel, Oholiob, and their cohorts. All building instructions are given to Moses directly. He is addressed with the second person singular pronoun, leaving few doubts that God intends Moses to follow the instructions personally. 42 The exception that proves the (literal use of the) rule is Exod. 25.10, where the instruction to make the Ark is addressed in the plural form. The exception is noted by commentators, and reasons are given. 43 Further reading confirms this to be the case. In contrast to what we find in B, nowhere in the instructions to make the mishkan and its furniture do we find the Lord asking Moses to appoint Bezalel and/or Oholiob to carry on the task. Their absence is especially noted in chapter 28, where Moses is given instructions to make the priestly vestments. Exod. 28.3 states, “Next you shall instruct all who are skillful, whom I have endowed with the gift of skill, to make Aaron’s vestments.” The failure to name Bezalel and Oholiob in place of or with “all who are skilled” is notable and confirms the idea that the core of the A version was unaware of Bezalel’s and Oholiob’s participation or deliberately did not mention them. Compare this to Exod. 38.22–23 and Exod. 39.1, which discuss their active participation in the making of Aaron’s sacral vestments. We conclude that Bezalel and his craftsmen play no part in the construction according to the A version. 44 In the B version, Bezalel and Oholiob are given the task to build the mishkan immediately after all the materials needed for the construction have been collected (35.30ff.). As we stated earlier, the impression one gets is that the project seems to have been designed, engineered, and carried out by the chief architect and his people. It is eventually brought to Moses for approval, implying that Moses had little input in the ongoing work. 45 Summing up their contribution, the text affirms that Bezalel (and Oholiob) “had made all that the Lord had commanded Moses” (38.22).
Difference #4: The golden altar / altar of incense 46 is absent in version A. While B describes the making of the altar of incense or golden altar (Exod. 35.15, 37.25), its completion (Exod. 39.38), and its placement in the inner court (Exod. 40.5), A makes no mention of it whatsoever. It has been argued that 30.1–10, describing the instruction to make the golden altar, belongs to the core of A and that its absence is due to a scribal error, but the evidence is not persuasive (see appendix). We have therefore listed the absence of the golden altar from A as an important difference between the two versions.
Despite common elements, materials, and measures, the two versions described essentially two different tabernacles. A’s version was to follow a divine blueprint, a tabnit that was not inscribed on stone or papyrus, as common architectural plans might be, but that was instead miraculously projected on the mountain. One must further assume that the exact form of the plan was impossible to describe in writing or via a design since it was a divine design of a dwelling uniquely fit for God. This explains the aberrancy in the order of construction: this tabernacle did not follow human convention. The tabernacle of the B version, by way of contrast, was the creation of specialized craftsmen and artisans who followed as best as possible, even optimally, the general instructions they received. It was a humanly conceived dwelling, and it was built as all human structures are: first the outside surrounding structure, then the furniture. What is more, a truly divine dwelling can only be built by Moses because only Moses, a man just short of being a divine being, can understand God’s unfathomable design. By way of contrast, the second tabernacle is not a divine dwelling as much as a fitting temple for the worship of God. Such a tabernacle can and must be built by human beings, albeit only by those supremely endowed and inspired. The absence of the golden altar and the bronze laver in A, if indeed they are missing from the core, can also be explained along similar lines. The incense altar, lying inside the Holy, but with its primary function of becalming divine wrath when and if the Tabernacle’s sanctity (or a most critical divine instruction, see Lev. 4.1–21) is violated, and the bronze laver, lying in the outer court, with its sole role of ensuring the cleanliness and purity of those about to perform sacrifices, were, strictly speaking, not part of the divine abode. 47
The placement of the two versions in the canonical text
From a narrative point of view, the placement of B in the canonical text appears unobjectionable. It follows the Sinaitic theophany, the golden calf apostasy, and the Lord’s eventual reconciliation with the Israelites—all events taking place at Mount Sinai. And it’s at Mount Sinai, or at the foot of the mountain, that, according to various priestly texts, hukim, mishpatim, and torot were given (Lev. 26.46), and, in particular, most of the priestly sacrificial laws (Lev. 7.37–38). B not only represents a plausible sequence to those events but also forms a fitting link and connection to the book of Leviticus, since it is from this completed and erected mishkan that the cultic laws are issued.
The same cannot be said with regard to the placement of A. The insertion does not sit well in its literary context, as we shall see, and for other reasons as well has led even traditional commentaries to question the text’s implicit timeline. Some ancient and medieval commentators upheld the scriptural text’s timeline—namely, that the commands were given to Moses after he had ascended to the top of Mount Sinai for the first time. 48 Some disregarded the flow of the scriptural narrative and suggested that the command was given to Moses prior to his ascending the mountain of Sinai. 49 Finally, others have suggested that the commands were given to Moses after Moses descended for the second (third, as the rabbis read it) time and argued that the text was not being precise about the timeline. Underlying some of the confusion is the assumption that B represents the fulfillment of A.
The latter opinion represents perhaps the most dramatic reconfiguration of the biblical timeline and is worth mentioning. A well-known tradition, cited by Tanhuma (parashat Terumah siman 8) and followed by Rashi (at Exod. 31.18), asserts that, contrary to the biblical sequence of events, no time transpired between the command and its fulfillment—that the command was given to Moses on the Day of Atonement, just prior to the initiation of construction, to show the nations of the world that God had forgiven the Israelites for their sin. According to Tanhuma and Rashi, the command section is out of sequence, consistent with a well-known hermeneutic principle that Torah narratives do not necessarily follow a chronological order (eyn muqdam u-meuhar ba-torah, lit. “There is no prior or posterior in the Torah”). 50 Tanhuma and Rashi find it justifiable to relocate the command section away from its canonical setting. 51
Alternatively, Ramban (at Exod. 35.1–2) upholds the sequence of the parashyot, namely, that the command to build the tabernacle precedes the incident of the golden calf. Ramban takes as a given that the command relates to the later fulfillment and is not bothered by the distance separating these events nor by the lack of a new set of commands. Moses acts on his own cognizance and assumes that the original command is valid once he learns that God had granted full atonement to the Israelites. 52 Nor is Ramban troubled by the theological implications that God appeared to have failed to anticipate the violation of the covenant, raising, among other questions, the obvious one: why give Moses this command when it was not possible to carry it out immediately upon his descent from the mountain? A further difficulty is that, in the present canonical arrangement upheld by Ramban, Moses shatters the very same tablets that he was commanded to deposit in the Ark (Exod. 25.16). Did the omniscient God not know that these tablets would be shattered?
In sum, neither Tanhuma/Rashi nor Ramban provide a satisfactory rationale for the canonical arrangement, the former because it fails to provide a good reason for the out-of-order (gratuitous?) placement of the command section, and the latter because the narrative, as Ramban sees it, leaves us with a fallible God. Neither exegete entertains even the possibility that the commands of 25–29 speak of a different Tabernacle than the one actually built.
One traditional commentator who, almost uniquely, breaks the umbilical cord between the two versions is Joshua Heller (1814–1880), a Lithuanian rabbi and author. Heller, in Ohel Yehoshua, 53 recognizes a number of differences between the “command” and the “fulfillment”, among them the sequence in the construction, the role of Bezalel (appearing in ch. 31, well past the core set of commands), and alternate modes of expression in parallel passages. 54 Heller posits the existence of three stages between the first set of commands (Exod. 25–31) and the commencement of construction (Exod. 35:4ff.). 55 Heller explains that in the first stage, the relationship between God and the Israelite people stood on the highest level of intimacy, accounting for the command to build a divine dwelling patterned on the miraculous vision on the mountain. This intimacy was lost during stages two and three, resulting in the designation of a master designer who would oversee a team of craftsmen and a new set of commands to build a man-made Tabernacle.
While Heller’s exposition is a great deal more elaborate than described in just these brief lines, the three-stages exposition captures the essence of the argument. By recognizing the differences between the two versions, Heller affirms the notion that we are in the presence of two different commands and implicitly rejects the command-fulfillment structure that, heretofore, appeared so tedious and meaningless to a majority of scholars. At the same time, Heller draws profitably from the canonical placement of A, and even of a chapter (Exod. 31) which sits outside of the core, using it to build a highly plausible narrative. Heller’s overarching reading fits well the nuances and contours of the canonical text, even if some of his linguistic observations and interpretations are far from persuasive and some of his proposed resolutions are forced. As we shall soon see, Heller’s ingenious scheme may also help explain the literary-critical problem to be discussed next.
A critical approach
Attentive readers, though not necessarily Bible critics, cannot fail to note that Exod. 25–29 sits uncomfortably between Moses’ ascent to Mount Sinai for what appears to be the general purpose of receiving “stone tablets with the teachings and commandments” from the Lord (Exod. 24.12) and the receipt of these tablets just prior to his descent. How is one to reconcile stone tablets containing “teaching and commandments” with the specific command to build a Tabernacle? Was the command to build the Tabernacle the only substantive matter discussed at the top of the mountain to deserve so much space?
Without siding with any of the reigning literary-critical models, we could still agree that such rough edges and a general lack of coherence are the necessary consequences of compilation, the process by which the various texts and traditions were aggregated to form the Pentateuch. And this, we believe, is precisely what occurred in this instance. Exod. 19.3–24.15 is generally considered to be a non-priestly text—one that either belongs to an epic J/E source or, alternatively, to a large independent unit, or an aggregation itself of small units, perhaps denoting a special theme (e.g., a ‘covenant theology’). 56 This Sinaitic pericope includes the Sinaitic theophany, the giving of the Ten Commandments, the Book of Covenant, the ratification of the Covenant, and the recall of Moses to ascend to the highest level of the mountain, where God is to give him “the stone tablets with the teachings and commandments” (24.12). Immediately thereafter, the compiler has the distinctly priestly source, or P, resume its narrative, picking up from where it left off at 19.2 and, with the possible exception of 24.18b, continuing right through to 31.18. 57
A necessary corollary to these conclusions is that P is not aware of the events narrated in the Sinaitic pericope. 58 This may not be as shocking as it would seem. Priestly texts make it quite clear that divine legislation issues from (the foot of) Mount Sinai and directly from the Tent of Meeting; an out-of-the-tent revelation seems out of place in priestly thought. Determining that 24.15–31.18 is a priestly text, however, does not guarantee that the entire section is part of one priestly layer. A sensitive reading of the text might reasonably conclude that 25.1ff. does not follow 24.18 and that 25.1ff. probably belongs to a different priestly layer. Here’s why. After Moses ascended to the mountain and the Presence of the Lord descended on it, the Lord “called to Moses from the midst of the cloud” (24.16), and Moses responded to the call by ascending the mountain and going inside the cloud (24.18a). As a result, Moses now stands “right next” to the Lord, so to speak, in a spatially intimate relation. It’s at this point that the Lord addresses Moses and commands him to accept gifts for the construction of a tabernacle (25.1). Since Moses is already standing next to Him (as per 24.16 and 24.18a) the narrator need only say “the Lord spoke to him” rather than “the Lord called to Moses, saying”, which denotes that Moses is at a distance of time and/ or space. For an appropriate formula, under similar conditions, see Lev. 1.1: “The Lord called to Moses and spoke to him”, where Moses is already near the Lord. Exod. 25.1, and consequently, Exod. 25–31, which represents A and later secondary additions, does not appear, then, to be part of the layer that begins at 24.18a, skips 25–31.17, and resumes with 31.18. This latter version (B), by contrast, reads quite smoothly: During his stay on the top of the mountain, Moses is instructed, presumably on various matters, including no doubt the construction of a tabernacle. We are not told any of these details other than “when He finished speaking with him 59 on Mount Sinai, He gave Moses the two tablets of the Pact [luhot ha-edut], stone tablets inscribed with the finger of God” (31.18). Later, after Moses’ descent, we hear Moses say to the Israelites, “This is what the Lord had commanded” (35.4), instructing them to assemble material for the construction of the Tabernacle. Consequently, this announcement follows the exact same pattern as all other such types of announcements. As explained earlier, in the announcement modality, Moses introduces God’s wishes to his audience but does not disclose the content of the divine message. Instead, he is seen acting on the instructions.
In sum, Exod. 24.12–31.18, 35–40 is made up of two priestly layers. 60 One, version A (25.1–31.18), contains specific instructions for how to build the tabernacle; the other, version B (24.18b, 31.18, 35–40), remains silent in respect to what God had spoken to Moses on his ascent to the mountain but reveals those divine instructions when Moses announces that God had ordered the building of a tabernacle (35.4). The former does not cohere well with the context; it’s a forced insertion. The latter does, as it allows the reader to imagine that the construction of the Tabernacle was included among the “teaching and commandments” delivered to Moses during his stay on the top of the mountain.
A second assumption is that version A is not an authorial layer but a redactional one—that is, it postdates the compilation and was interpolated within an existing structure. This hypothesis does not depend on the compiler since, arguendo, it post-dates his work and offers an equally plausible explanation for the intruding pericope. If A is redactional rather than authorial, it stands to reason that A, at least at the outset, would borrow from B some of the material details of the tabernacle construction. This does not exclude the possibility that at an even later stage, A had a reverse literary influence on B. It is extremely difficult to determine the direction of dependency in the long literary history of works such as the Bible; mutual interdependence cannot be excluded. Whether authorial or redactional, the placement of A in its canonical setting infuses new meaning into this long but important account. Below, we summarize the possible interpretations.
Two readings
1. Reading as command and fulfillment
A and B form a command-fulfillment pattern. It is so interpreted by most, but not all, traditional and non-traditional commentators. Among non-traditional exegetes who note some material differences between the command and the execution, Jacob stands out, seeing the slight variations between command and fulfillment as indicating that Moses and his craftsmen followed their best understanding, that the execution did not always follow the instruction to the letter, but that this was acceptable and even laudable. Earlier on, we rejected his list of differences and interpretations but conceded that his approach had merit. Other scholars had nothing but contempt for the tiresome tediousness of the accounts.
Yet, as we saw, this command-fulfillment interpretation suffers from a number of major difficulties, the most obvious one being its unusual prolixity. We noted other difficulties: the fact that announcement formulas are never preceded by command sections; the awkward placement of A; a narrative that, among other things, diminished God’s omniscience; and, lastly, the major conceptual-theological differences between the two versions. While this canonical reading did not satisfy, it still remained the most widely accepted interpretation.
2. Reading as original intention versus implementation
Readers who, as a result of the above difficulties, rejected the command-fulfillment interpretation were offered an interpretive alternative made possible in the canonical text by the placement of A just before Israel’s apostasy. The A tradition described an intimate sanctuary, divinely conceived, miraculously described, and built by Moses, where God would meet and speak with Moses regularly. The B tradition, by way of contrast, described the cultic structure as man-made and recounted the actual implementation of the Tabernacle construction. In this scheme, A represents the original divine intention, and B represents what was eventually carried out. The obvious ideological/theological differences between the two versions could be attributed to changing circumstances. And these changing circumstances—that is, the golden calf incident, God’s anger, and His eventual reconciliation—were present in the narrative. It’s here that Heller’s provocative thesis nourishes the narrative logic of the compiler.
This brings us back, however, to the question of the origin of the A version. If the A version is an independent priestly tradition that was integrated into the book of Exodus by a pentateuchal compiler, one would have to note that this compiler did not produce a seamless account; the placement exposed a major discontinuity in the text—to wit, the instructions that Moses received to build the tabernacle are implausibly identified as “the stone tablets with the teachings and commandments” that Moses was to receive from the Lord (Exod. 24.12). To uphold this literary “sloppiness” one would need to subscribe to the idea that the compiler, being essentially an historian, interweaved his sources “in the only logical, chronological order possible”. 61 That is, the compiler is not a redactor or editor; he does not “fix” the text. Rather, he is “willing to allow significant contradictions, discontinuities, and chronological problems to exist in the combined text, intervening only on very rare occasions. His tolerance of these contradictions speaks to his desire to retain as much of his source material as possible”. 62 This seems to be a fair and reasonable supposition and appears consistent with other instances of aggregation of sources. 63
An alternative explanation, as indicated earlier, is that the insertion and placement of A was not the work of a compiler but rather the work of a later interpolator. This thesis, too, would explain the discontinuity in the text, but for a different reason. When intervening in a “formed” text, discontinuities and fractures are almost unavoidable. 64 The late redactor may have wished to indicate that, had the Israelites merited it, the original Tabernacle would have become a truly divine dwelling.
Concluding thoughts
Unless we are, some day, fortunate enough to physically find the original sources (or traditions that were aggregated to form these sources), or at least a series of earlier versions, as we did with the Gilgamesh account, 65 we are not likely to ever fully reconstruct the complex way by which the canonical text came to take its present form. But we can try to understand the obvious anomalies that are found almost everywhere in the canonical text—themselves the natural outcome of a long process of composition, collecting, compiling, editing, and redacting that preceded it. Verbatim repetitions are one such anomaly.
These repetitions have created almost a cottage industry of dependency studies, the aim of which has been to guess the direction of such dependency. To this task, scholars have applied their significant grammatical and linguistic skills and have sought to find hidden literary clues that could “betray” the borrower. For the most part, these efforts have not yielded definitive conclusions. As shown, the tabernacle account is one such case: much effort has been expended in determining the direction of dependency for each of the sections—and for more than 150 years—but no consensus has as yet been reached. Other than as interesting challenges and intellectual curiosities, dependency studies, we submit, are sterile if scholars are not able to offer reasonable justifications for such literary “inventions”. We ought not to be able to say, as Kuenen did in 1886, that “[t]o give such a diffuse account of how Moses executed the commands of Yahve, immediately after the commands themselves, and for the most part couched in identical terms, is an idea that would more readily occur to later readers and manipulators than to the author of 25–31”. Even if Kuenen were right, the question remains: Why would the “later reader or manipulator” borrow sections of the command to create a “diffuse account of how Moses executed the commands”? (Kuenen, 1886: 78). And if he did in order to show that Moses carried out his orders exactly as commanded, was the “later reader or manipulator” so careless as not to realize that the account had become impossibly long and tedious? Where else in the normally parsimonious priestly writings can scholars point to such unbearable prolixity? Why hold redactors to such a low literary standard? And even this is not always the full story. After concluding that Exod. 35–40 is dependent on Exod. 25–31, Noth says, “[T]he narrative in chapters 35–39 has been extended even beyond that of chapters 25–31. But even here it is not so much a case of the inclusion of further independent traditions as of creating obvious narrative structures on the basis of chapters 25–31 (Noth, 1962: 275).” In other words, chapters 35–40 are invented whole cloth or, more likely, he claims, copied and imaginatively extrapolated from 25–31. But how can Noth be certain that this was not another authorial composition or another “independent tradition,” and a competing one at that?
It is not enough for advocates of dependency to determine the direction of dependency. Their conclusions must include plausible justifications. 66 Methodologically, while similarities are easily spotted, efforts must be redoubled to discover differences. 67 It’s the latter that allows a reader to understand what the author or redactor of the “copy” had in mind. 68 The dependent section “speaks” to the source, reinterprets it, rewrites it, or polemicizes against it. Verbal parallels and verbatim repetitions are the subtle tools of ancient literary composers–deployed not for the purpose of adding useless information but rather to authorize their own versions and to offer us a new understanding.
Only when standing in a relation of command and fulfillment can these versions be called needlessly repetitive and tedious. But, as we have shown, these two versions do not relate to each other as command and fulfillment. Rather, these two versions reflect two independent and competing priestly traditions, containing many similarities but also important differences. These versions were creatively weaved into a narrative by a compiler, who treated them as representing original intention and final implementation. The compiler’s manoeuvre forces the reader to read through the similarities and find the differences. Thus, the two versions are no longer needlessly repetitive.
Footnotes
Appendix: golden altar
Exod. 30.1–10 and 30.17–21, which are not part of what we identified as the core of A, do contain the instructions to build the altar of incense and the brazen laver respectively. However, the altar of incense is disconnected from the making of the other furnishings that stood with it in the Holy, and both of these items are separated from the making of all the other furnishings—including those belonging to the outer court, as well as the priestly vestments—by 29.43–46, which, as we saw, forms a conclusion to the project. Their striking disconnection from the rest of the mishkan has led to some dramatic historical-critical conclusions.
Julius Wellhausen (1885) believed that Exod. 30, which mentions the command to build the golden incense altar, is a correction of Exod. 25–29 and therefore secondary. According to Wellhausen, the reason why Exod. 25–29 does not include the golden incense altar (cf. Exod. 30.1–10) is because the author simply did not know of it. 69 Kuenen took up Wellhausen’s argument, pointed out that Delitzsch and Dillmann, after reviewing the evidence, “remove the doubts urged by Wellhausen against Exod. 30.1–1,” but returned to Wellhausen’s conclusion, agreeing in the end that 30.1–10 is a “correction”. 70 There is little left of Wellhausen’s and Kuenen’s argument after the considerate rebuttal offered by Dillmann and Delitzsch 71 other than Kuenen’s observation “that the passage does not stand where the author of Exod. 25–29 must have placed it.” 72
Since the instructions to build the golden altar and the bronze laver appear just outside of the well-demarcated core of A, there are scholars who have argued that the golden altar pericope was misplaced due to the scribe’s carelessness. In Haran’s words, “one may explain the misplacement of the passage in question by assuming that the copyists or editors were not sufficiently careful about inserting where it belonged, and this was the natural result of an editorial method which gave rise to loose joints and rough edges.” 73 Haran provides little in the way of evidence, but some evidence does exist. The Samaritan version, for one, includes the command to make the altar of incense within the main command section, following Exod. 26.35 and nestled between the command to make the curtain (parokhet) and the screen for the entrance of the Tent. So does a Qumran Exodus scroll. Given that the editors of DJD conjectured that there is sufficient space left in the 4QpaleoExodm manuscript to accommodate the instructions to make the golden altar, offering, there is the possibility of a second textual witness to the more “correct” collocation. 74
We believe that the more difficult reading should stand (lectio difficilior potior). For one thing, the pre-Samaritan and Samaritan versions are known to contain a number of harmonizing entries; this one may be no exception. Moreover, the Samaritan text positions the golden altar “correctly” but fails to do the same with the bronze laver (Exod. 30.17–21 in the MT); we believe that it is more likely that a “harmonizing” scribe left something out—in this case, the bronze laver—than that the original author included the golden altar in his account but not the laver. Lastly, the MT text refers to the burnt offering altar (Exod. 27.1) as the altar, suggesting that this was the only altar known to the author and his audience. Had the golden altar been mentioned nearby, the author would likely have been forced to make the proper distinction and refer to these altars respectively as the burnt offering altar and the golden altar (cf. Exod. 38.1, 37.25).
We have argued that the golden altar is absent from the core of A. But is it possible that an earlier iteration of the B also omitted the golden altar and that its absence from A is not as noteworthy a matter as it is assumed to be. The canonical text mentions the golden altar four times, namely, Exod. 35.15, 37.25–29, 39.38, 40.26. By way of contrast, the LXX translation omits two of these mentions, Exod. 35.15 and, critically, Exod. 37.25–29, which represents the full description of the altar. Since the LXX version is thought to be based on an earlier Vorlage than the MT, a speculative case can be made that the golden altar “slipped in” redactionally and that an even earlier iteration did not know of the golden altar. Unfortunately, even the Old Greek version, known to us only from 4th–5th century AD manuscripts, is sure to have undergone some harmonization towards the MT during the many centuries that it traveled alongside the MT and therefore does not reflect faithfully the Hebrew Vorlage on which it was originally based. But an earlier version of the Greek translation may exist. In recent years. Pierre-Maurice Bogaert has studied a fragment of a Vetus Latina based, as these medieval Latin translations are, on the Old Greek and concluded that this well-translated manuscript reflected an Old Greek version of the narrative that predates the Old Greek of the 4th–5th century. The Monacensis (Munich, Bayerische Staatsbibliothek, lat. 6225, end of 5th century, sigla 104 in the vetus latina of Beuron), as it is called, allows us a better glimpse of a Hebrew Vorlage that may have preceded the version of the textus receptus. 75
This VL fragment represents a third omission, at Exod. 40.26. In effect, the only mention of the golden altar in the narrative version of Monacensis appears in the enumeration of the furnishings brought to Moses (=MT 39.38), the least consequential of the references. This might suggest that the golden altar was not part of the earlier iterations of the B version either and that it was introduced only at a later stage, perhaps to harmonize the narrative with the golden altar found in the First and Second Temples and having occurred at the same approximate time as 30.1–11 and 30.17–21, considered late redactional additions. Caution, however, should be exercised with respect to the Monacensis in view of the fact that the text varies significantly from other extant VL manuscripts and in view of J. W. Wevers’ comment, in his introduction to the Gottingen edition of the Old Greek of Exodus, that the Monacensis bears few similarities with the respective Greek text. The highly hypothetical supposition leaves us little choice but to consider the absence of the golden altar (and bronze laver) in the A version as notable and worth enumerating among the unique characteristics of the A version.
1.
Propp, 2006: 367. Even traditional exegetes like Gersonides (1288–1344) struggled with the repetitiousness and redundancies of the account. He was forced to conclude that “the prophet speaks in the accepted custom of his time” (Commentary to Exodus, Root 50).
2.
We will be using the terms “Tabernacle” and mishkan interchangeably, and occasionally “Tent of Meeting,” because our texts do so, though we recognize that the latter may represent a different tradition and a different tent, since it sat outside of the camp (Exod. 33.7).
4.
Theodor Nöldeke pointed out that the execution section follows a more “natural” sequence regarding building and setting up than the commandment section, indicating a later reworking: “Im Übrigen ist bei der Ausführung mehr die natürliche Ordnung berücksichtigt, in welcher die Anfertigung und Aufstellung erfolgen musste, als bei den Aufträgen dazu.” Nöldeke, 1869: 56f.
5.
Kuenen asserted Exod. 35–40 and Lev. 8’s dependency on Exod. 25–29 but failed to adduce evidence other than noting that the command section omits the altar of incense while the fulfillment section includes it. This suggested to him that Exod. 25–29 (followed by 30–31) is somehow earlier. Kuenen stated: “To give such a diffuse account of how Moses executed the commands of Yahve, immediately after the commands themselves, and for the most part couched in identical terms, is an idea that would more readily occur to later readers and manipulators than to the author of 25–31.” Kuenen, 1886: 78.
6.
Holzinger believed the execution section to be secondary because of its “better” sequence (Holzinger, 1900: 148).
7.
Von Rad considered Exod. 35–39 to be a copy of the prior instructions in which the material was adapted and rearranged according to Exod. 25–31. He highlighted the problematic sequence in Exod. 25ff. He believed the uneven order in the commandment to be the result of literary growth. The execution, on the other hand, follows impeccably the inner order of the tabernacle and is therefore secondary: “In 25ff. war die Reihenfolge nicht unproblematisch […]. Diese unebene Anordnung war das Produkt eines literarischen Wachstums. Ganz anders in Exod. 36 ff. Hier ist die Reihenfolge tadellos nach dem inneren Aufbau des Zeltheiligtums gegeben […]. Hier ist die gute Ordnung ein sicheres Zeichen der sekundären Weiterung” (von Rad, 1934: 68f.).
8.
Levine is not as categorical in relation to the dependency question. He agrees with the literary dependency model of von Rad and points out that chapters 25–27 provide evidence of originality and coherence (as they are unified by the recurrent theme of the tabnît Moses was shown). On the other hand, Levine questions whether the rigid order and system exhibited in Exod. 35–39 necessarily indicate artificiality (and, therefore, secondary character) by pointing out that palace records describing architectural and artistic projects reflect accepted forms, such as the description of Solomon’s temple in 1 Kgs 6–7, which is considered to be an authentic document based on actual archival records. Levine, 1965: 308.
9.
Utzschneider, 1988: 224.
10.
Noth, 1962: 274f. Contrarily, Ramban, a traditional exegete, notes a tendency on the part of the biblical author towards parsimony (see below, footnote 26). Gary A. Anderson better reflects the general scholarly consensus when he says that the priestly writer “is prolix only when a theme is being introduced for the first time; should the occasion warrant a return, the priestly writer is more than capable of abbreviation” (Anderson, 2009: 161).
11.
Noth, 1962: 275.
12.
One of them is Wellhausen, whose main argument is that the holy rites for people are now applied also to objects. (Lev. 8 adds the anointing of the Tabernacle and utensils and the purification of the altar, assuming that Exod. 29.36 is an addendum, and adds the “suet” (פדר) to head and sections in v. 20.) Similarly, he compares Exod. 29.13 (usage of המכסה) to Lev. 8.16 versus (secondary) Lev. 3.3, 9, 14 (המכסה) and points out that Lev. 8.26 gives a more detailed description by supplying “Mazzot” versus Exod. 29.23 “bread”. Christophe Nihan (2007) and Thomas Hieke (2014), too, argued that Exod. 29 served as the Vorlage of Lev. 8; 
provided a list of twelve reasons to reach the conclusion that Lev. 8 was dependent on Exod. 29.
13.
Karl Elliger (1966) reviewed many of the salient differences in formulation between the two chapters and argued that the author of Exod. 29, whom he calls the “Kopist”, is more precise, more expansive when needing to impart sense, clarifies ambivalent terms, uses more legal language, and so on—characteristics of a redactor working from a Vorlage. He concludes that most of Exod. 29 is secondary compared of Lev. 8: “[…] notwendig ist der Schluß, dass Exod. 29 gegenüber dem Gros von Lev 8 sekundärist.” Elliger, 1966: 107ff. devoted an entire monograph to the comparison between Exod. 29 and Lev. 8 and concluded that Lev. 8 has a traditional core that is older than Exod. 29, although some passages of Lev. 8 have been redacted and are therefore later than Exod. 29. Methodologically, he presumed that both Exod. 29 and Lev. 8 drew upon a common Vorlage, an ancient liturgical tradition.
14.
Nihan objected that even though many of the ritual acts described in Exod. 29 and Lev. 8 originated in traditional rites from the pre-exilic period, this does not necessarily mean that Exod. 29 and Lev. 8 were based on a distinct oral or written source. Rather, Nihan points out that both chapters are strongly permeated by the Sinai setting and are thus anchored in P’s narrative fiction (Nihan, 2007: 125). Graeme Auld, more recently, suggested that “[t]he relative priority of Exodus 29 and Leviticus 8 on the ordination of Aaron and his sons […] may continue to be debated backwards and forwards. And in each case it may be better to think of separate development from a common original than dependence of one present form of the text on the other” (Auld, 2003: 53). The dependency discussions of Exod. 29 and Lev. 8 will not be dealt with in this paper, in order to focus only on the construction of the mishkan.
15.
Propp, 2006: 367.
16.
Wellhausen, 21889: 142: “[E]ine blosse Wiederholung […] Tatsache ist jedoch, dass der fragliche Abschnitt inhaltlich ganz und gar bedeutungslos ist, dass er nicht vermisst würde, wenn er fehlte.”
17.
Ohr ha-Hayyim, Exodus 36.8.
18.
Hirsch, 1976: 682–683.
19.
Friedman, 2001: 299.
20.
Anderson, 2009: 161–162.
21.
Watts, 2011: 417–430.
22.
Hasler, 2020. We thank the anonymous reader for suggesting the application of this idea to our issue.
23.
“B berichtet die sachgemäße Ausführung und nur sie, A gibt dem Mose Belehrungen über die Zwecke und den Sinn der Arbeiten und nur insoweit technische Anweisungen” (Jacob, 1997: 993).
24.
For Jacob, “[t]his fundamental distinction explains all the differences in these accounts […]. These are sufficiently numerous to require a complete repetition of all the sections in the altered setting. This also presents an opportunity to add and to clarify nuances” (p. 1009 in English translation).
25.
In Exod. 35–40, it is frequently mentioned that the workers “did everything as He had commanded Moses.” Many would understand this to be a compliment, that everything was executed according to its command. This makes it the more surprising to find that this is not the case at all! In fact, the parallel sections deviate so much from each other that we have to ask ourselves where God had commanded such to Moses (e.g., the basin, the courtyard, the priestly gowns, the erection of the sanctuary, the dedication of the priests (Lev. 8)). Rather, “it would have been more correct to state that they did nothing as HE had commanded!” (“Sie machen nichts genau so wie ER dem Mose in A geboten hatte!”) Jacob, 1997: 995.
26.
“Vergleichung und Analogie”, in Jacob, 1997: 994–995.
27.
Ramban 16.4: “[W]e know that everything was ordered by God, but the text avoids repetitions in matters that are duplicated in the command and the fulfillment, as I explained to you many times”. See his comments on Exod. 9.18 and 10.2, though not with respect to these particular formulations. With the same formulation, see Rashi, Lev. 10.3, creating a confirming precedent through a highly midrashic interpretation; and see Ramban’s critique. Ibn Ezra, Lev. 10.3, cited also by Ramban, understands it as a communication from God that was never written down. On Exod. 16.23 (“This is what the Lord meant”), Saadya offers the following: “[…] not mentioned earlier. Though the messenger [Moses] had already heard it earlier from the sender [God] […] it was not written until now. And this is proof that the messenger heard many commandments from God […] and they were only written down later”. Saadya points out that this sort of thing—Moses giving instructions to the Israelites on a variety of commandments without us being apprised beforehand (as in the first four books) of God’s direct command to Moses in their respect—happens often in Deuteronomy. He then proceeds to give a number of examples and concludes by stating that “therefore if we find some commandment that the messenger relayed in the name of God, we do not need to find it written [i.e., commanded directly to Moses] in another, earlier, location in the Torah”. The continuation is lacking in the manuscript. Pirushe Saadya Gaon le-Sefer Shemot, p. 71. Saadya does not spell it out, but it is obvious from the biblical record that once Moses says that God had told him to do so-and-so, not only do we “not need to find it written”, but we never find the antecedent instruction. Moses’ testimony is sufficient evidence that God had previously commanded him.
28.
This is most probably the import of a rabbinic midrash at Lev. 17.2. The passage reads as follows: “The Lord spoke to Moses, saying: Speak to Aaron and his sons and to all the Israelite people and say to them: This is what the Lord has commanded […]”. It is immediately obvious that the expression “this is what the Lord has commanded” is unnecessary; Moses could have spoken to his intended audience without the special preface. Taking note of this redundancy, Sifra makes the following comment: “‘This is what the Lord has commanded’ teaches that the section was [originally] enunciated by way of ‘say thus’ [i.e., ‘saying,’ 17.1]. You might say that only this section was taught in this way [given the presence of the double formulas]. How do we know that this is the case everywhere [that whenever the expression ‘this is what the Lord had commanded’ appears, it is preceded by ‘the Lord spoke to Moses, saying’]? [Answer:] ‘this is what the Lord has commanded’ [the redundancy here] teaches [talmud lomar] that this is a general principle [Binyan av, lit. a father construct, one of the thirteen hermeneutic principles used by the Talmudic sages to interpret the Torah] that states that all such expressions were [always] preceded by ‘say thus’ [‘saying’]” (Sifra, Lev. 17.2). Different interpretations of this midrash are offered. See for example, Malbim, 1860: 327. We believe that this simple interpretation best interprets the problem and the answer. The midrash takes note of the awkward duplication and suggests that the exception comes to teach that all expressions formulated as “this is what the Lord had commanded” were preceded by a direct and detailed oracle in which Moses is asked to address others, just as we find here.
29.
Where Moses acts on divine instruction, the biblical author will confirm the instruction via a prior and, generally, immediately preceding command or will simply state that Moses acted as the Lord had commanded. Either of these two formulas was seen to suffice for the audience and reader to accept the divine origin of the instruction. A particularly interesting example is evident after Exod. 10.2. Verses 3–6 depict Moses and Aaron delivering the warning to Pharaoh about bringing locusts on his territory. Unlike other instances, there is no record of God commanding them to say this, but we do find Moses and Aaron prefacing their warning to Pharaoh with the expression “Thus says the Lord […]”. While for the MT scribe (even a non-priestly one), either one of the two formulas sufficed to indicate that indeed God had spoken to Moses and Aaron, this was not necessarily so for other biblical scribes. Dead Sea discoveries have revealed the existence of a circle of Palestinian scribes, operating in the Second Temple period, that sought to harmonize the biblical text further. In this respect, these texts resembled the expansionist/harmonistic tendencies of the Samaritan Bible, and for that reason they have been named pre-Samaritan texts, though they do not share the ideological variants of the latter. Both the pre-Samaritan and the Samaritan texts inserted between verses 2 and 3 an explicit divine command: “Say to Pharaoh […]”.
30.
This conclusion closely parallels the story of Solomon’s building project (1 Kgs 5–9), as Victor Hurowitz has noted. He points out further that “the description of the Temple in the Book of Kings is closer in structure and details to the Tabernacle description in the fulfillment section than to the description in the command section. Consequently, if the Temple story (1 Kings 5–9) is taken as a standard for the characteristics of building stories, the Tabernacle story is close to standard in the fulfillment section but very widely deviates from the standard in its command section” (Hurowitz, 1985: 25). He notes further that in standard literary patterns found in Mesopotamian and Ugaritic sources, the building of the temple “is not usually described within the context of the divine context”, even when the building is divinely sanctioned. This Hurowitz contrasts with the Tabernacle story, which describes the structure and furnishing twice. Such a deviation is rare, and even when it occurs, the two descriptions differ; one is about the god’s attributes, and the other one is more “nuts and bolts” (p. 26). However, Hurowitz would have seen the Tabernacle picture as being consistent with the literary pattern of the other temple projects had he been aware of the terse “content-less” announcement formula in the so-called fulfillment section.
31.
While, theoretically, one might be able argue that 39.32–43 serves as the fulfillment to A, making A also a complete, self-contained command-fulfillment unit, this is unlikely for various reasons. The summarizing statement mentions only that “the work of the Tabernacle” (v. 32) was completed; no mention is made of the Israelites’ worthy and generous contributions in material for construction. The execution or fulfillment account mentions this in great detail (35.21–29). Therefore, 39.32–43 can only be referring to the post-contribution phase, a natural coda to the execution account. The priestly bigdei haserad (v. 41) are unknown in the command section; they are only mentioned in the execution section. Lastly, it’s only in the fulfillment account that Moses can unequivocally be said to oversee and vet the work (v. 43); in the command section, Moses is seen as carrying out all the work himself. Therefore, 32–43 is far more likely to be part of and appended to the execution than to be the fulfillment of the commands. Wellhausen, who considered 25–31 to be the primary text on which 35–40 depended, thought that Exod. 39.42–43 served as the fulfillment of A (as cited by Hurowitz, 1985).
32.
33.
There is no reason to construe this affirmation as anything more than stating that these craftsmen had carried out what was demanded of them and specified in 35.11–19. The affirmations in chapter 39 can be read similarly. By way of contrast, the multiple affirmations in chapter 40 (vv. 19, 21, 23, 25, 27, 29, 32), all in connection to the arrangement of the sacred furniture inside and outside of the mishkan carried out of Moses himself, may be referring to actual divine directives. Much has been made of the seven repetitions of this affirmation and their mystical implications. However, the affirmations look suspiciously like the product of a later hand, and this is because though Moses is involved in the erection and placement of the furniture, the affirmations are phrased in the third person: “as the Lord had commanded Moses”!
34.
Ramban suggests that Moses was commanded a second time (Lev. 8) and that it’s only then that he performs the desired anointing. Similarly, he suggests that the instructions to light the lampstand, set the bread, place incense on the incense altar, and bring the olah and mincha offerings (v. 29) were not executed immediately but only after Moses got around to anointing the mishkan and the utensils. This is because no services can be performed before the mishkan and the utensils are consecrated with the anointing oil. To satisfy Ramban’s reading, however, the verbs must all be read as indicating an activity that will take place at a later time; so, for example, the execution “He placed the lampstand […] and he lit the lamps” (v. 25) must be read as “he placed the lampstand […] so that he could light the lamps”. His reading violates the plain sense of the text. Ramban’s suggestion suffers from two additional difficulties. Firstly, it posits that Moses waited for a second call before proceeding to anoint the mishkan. Would this delay not show a lack of respect? In his comments on Lev. 8.10–12, Ramban suggests that Moses wanted to bring the anointing of the mishkan closer to the consecration of the priests so that Aaron could be immediately ready to serve. But this same sequence—first the mishkan and its utensils and then Aaron and his children—is prescribed here too. Secondly, Exod. 40.15 demands that Aaron’s children be anointed “as you have anointed their father.” Nowhere in Lev. 8 do we find that the children were anointed in the same manner as their father. Therefore, Lev. 8 is unlikely to be the realization or fulfillment of Exod. 40.9–15.
35.
Whether the seven days of milluim (priestly ordination) occurred prior to the completion of the mishkan or subsequently (as a plain reading of the text implies) is the subject of a major rabbininc controversy. For the former view, see Mekhilta de-Milluim 96.32; Seder Olam, ch. 7; Bavli Shabbat 87b; Bamidbar Rabbah 12.15. For the latter view, see Bavli Sukkah 25b, where we find Rabbi Aqiba reading “the eighth day” of Lev. 9.1 as the eighth day of the month of Nisan. This is also Ibn Ezra’s position. (See his comments on Exod. 40.2 and Lev. 9.1, though he acknowledges the “traditional” former view in Lev. 9.1 and again in Num 7.1). The latter view is also held by Josephus in the Antiquities and is found in the Temple Scroll 14.9ff. This timing dispute may have more to do with the status of the priests and their identification with the mishkan than simply being a matter of antiquarian interest. We hope to discuss this matter in a subsequent article when dealing with the redactional layers in the Tabernacle accounts.
36.
In Num. 7:1, we read, “On the day that Moses finished setting up the Tabernacle, he anointed and consecrated it and all its furnishings, as well as the altar”. Num. 7.1 presupposes Exod. 40.1–16 and confirms, once again, as with Exod. 40.16, that the instructions were carried out, at least in respect to the mishkan and the furniture.
37.
The two initiation terms used here are drawn from a wide number of rabbinic sources, talmudic and post-talmudic. For a full discussion and the sources, see Encyclopedia Talmudit, s.v. Hinukh klei ha-miqdash, 16:202–214.
38.
Exod. 40.17–34 can in turn be connected to Lev. 8. The expression “just as the Lord had commanded Moses” appears with remarkable frequency in this pericope (40.19, 21, 23, 25, 27, 29, 32) as well as in Lev. 8 (8.4, 9, 13, 17, 21, 29, 36), suggesting that these two sections are of one piece.
39.
By interpolations, we refer to short insertions into a text, and by secondary additions, we refer to lengthy, stand-alone units. Generally, though not always, interpolations qualify, clarify, or add a detail but cannot stand on their own without reference to the context while secondary additions do stand on their own and can easily be understood without reference to the context. Interpolations and secondary additions are assumed here to be the work of redactors who work subsequently to the compilers of the work.
40.
We have identified the interpolations as Exod. 27.20–21, 28.30, 28.41, 29.29–30, and 29.38–42. These redactional additions can be detected with relative ease and have little or no bearing on the greater picture we are addressing here—namely, the ideological differences between the A and B versions on matters that pertain to the construction of the mishkan.
41.
Ibn Ezra ha-arukh, Exod. 25.22.
42.
Disagreeing with Propp, 2006: 378, who reads “you shall make” as “Moses is to commission the artificers”. One page earlier, and commenting on the tabnit, Propp makes the correct observation that “the text never indicates that Moses brings it down the mountain and shows it to anyone”. This is true and consistent with our literal reading; the instructions were solely for Moses, to carry them out himself. But then Propp surmises that “one rather gets the impression that he describes it to the craftsmen by memory”, an assumption lacking any textual basis. When Moses does turn to others to execute an instruction, the text says so clearly (see 28.2–3), as Propp himself notes (p. 378). Propp, following most scholars, was misled by a reading that saw the later execution as a fulfillment of A and missed the unique characteristics of the mishkan ordered in A.
43.
See Ramban and Ohr haHayim, Exod. 25.10.
44.
Bezalel and Oholiob are “reinstated” in 31.1–11, a later redactional layer. We intend to discuss the secondary additions of Exod. 30 and 31 in a future article.
45.
cf. Rashi, Exod. 39.33, based on Tanhuma siman 11.
46.
This altar is called by different names: in P, (1) the golden altar, מזבח הזהב, at Exod. 39.38, 40.5, 26, Num. 4.11, 1 Kgs 7.48, 2 Chron. 4.19; and (2) the incense altar, מזבח הקטרת at Exod. 30.27, 31.8, 35.15, 37.25, 1 Chron. 6.34, 2 Chron. 26.16. Or, מזבח מקטר הקטרת at Exod. 30.1. Or מזבח קטרת הסמים at Lev. 4.7, 3. Or simply המזבח אשר לפני יה at Lev. 4.18. In rabbinic literature, it is referred to as the “inner altar” due to its position in the Tabernacle in front of the Ark and in contrast to the outer or burnt-offering altar, at bYomma 25a (Mishnah), 58b (Mishnah), bZebahim 36b, and passim. We variously refer to it as either the golden altar or the incense altar.
47.
According to Ramban, at Exod. 30:1, the altar of incense was ordained as a supplement to the completed Tabernacle and follows the statement that the (completed) Tabernacle “shall be sanctified by My presence” (Exod. 29.43). Since the violation of this sanctity can express itself in a plague (as it did in Num. 17.11–14), Moses was taught the secret of the incense and its power to halt a plague. Hence the need to build an altar of incense as a response to a potential problem rather than by original design. In a somewhat similar vein, Elijah of Vilna, Aderet Eliyahu, at 30.10, notes that the altar of incense and the half-sheqel contributions (Exod. 30.11–16), were not indispensable for the in-dwelling of the Shekhinah (השראת השכינה) but were ordained solely for expiation purposes. Therefore, the altar cannot be listed among the inner furnishings like the Ark and the lampstand. Interestingly, rabbinic halakhah has it that, with the exception of the incense altar, services cannot be performed in the absence of the furnishings of the Tabernacle. Moreover, the burning of incense can be carried out in the place of the altar even in the actual altar’s absence. Also, priests could wash their hands and feet in any serving vessel if the bronze laver was not available (bZebahim 22a).
48.
Exodus 24.16–8, immediately preceding the command to build the Tabernacle. Views espoused, for example, by Exodus Rabbah 33.1, Ramban.
49.
Seder Eliyahu Rabbah 17. Ibn Ezra, at Exodus 25.1: “He spoke to him [Moses] on his way up the mountain [our emphasis] about the mishkan, namely, that he [Moses] should build a mishkan to the exalted Name and He will dwell in it, and address Moses from there without the need to go up the mountain.” Ibn Ezra’s terse and provocative comment appears to be saying that the command to build the mishkan at this point reflected a first intention. Might he be insinuating that the command issued with the first intention was different than the command given later? As an aside, Ibn Ezra seems to have anticipated some of the more modern ideas regarding the true function of the Tabernacle in priestly writings.
50.
Cf. Mizrahi supercommentary on Rashi, Exod. 31.18.
51.
While this reconstruction may be technically acceptable, the commentators fail to explain why, in the first instance, the Torah would want to place the command out of order. The principle of eyn muqdam u-meuhar ba-torah permits them to set aside the canonical order but does not explain why the compiler or editor chose to place the commands where he did. At a minimum, the fact that the command was not placed, as it should have been, immediately prior to the fulfillment underscores the logical weakness in a narrative that assumes the command to be organically related to the subsequent fulfillment.
52.
Ramban says, “Since God had reconciled Himself with them, gave him the second set of tablets and entered into a new covenant with him, it is evident that the relationship returned to the way it was during their youthful devotion. And it is clear that the Divine Presence would dwell among them in the same manner as He informed him at the beginning, saying: And let them make Me a sanctuary that I may dwell among them (25.8). Therefore, Moses commanded them now (35.4) in exactly the same manner as he was commanded earlier.”
53.
Heller, 1994: 1–50.
54.
E.g., “a woman to her sister” in 26.3, 6, 17 as against “one to the other” in 36.10, 12, 13, 22.
55.
These are (1) before the making of the golden calf, when the core commands (Exod. 25–29) are given to Moses; (2) on the fortieth day of Moses’ stay on the mountain, at the commencement of Israel’s apostasy and just as Moses is about to descend, accounting for chapter 31 and Bezalel’s designation as the master architect instead of Moses; and (3) after God is reconciled with the Israelites and forgives their sin, when a new set of commands are given.
56.
See Rendtorff, 1990: 186–187.
57.
Source critics are not exactly aligned with regard to the exact boundaries of each document. For Noth (1962: 200), P begins at 24.15b, being the sequel to 19.12, and continues to 31.18; for : 254–256) P begins at 24.15b, ends at 24.18a, allows E to interrupt with just one sentence (“and Moses was in the mount forty days and forty nights,” 24.18b), resumes with 25.1, and runs to 31.18a.
58.
This is not necessarily true; P may have had a very short mention of these events, and the compiler decided instead to use the fuller version offered by JE. However, the smooth transition between 19.2 and 24.15, two points firmly established as P, supports the scholarly consensus. According to Cross, however, it is inconceivable that P is unaware of the Sinaitic theophany, unless one assumes P to be a redactional layer (Cross, 1973: 312–313, 318).
59.
Note here the use of the pronoun “him”, consistent with the idea that Moses stood next to the Lord in an intimate relation and proving, as we argued earlier, that 25.1 could not be a sequel to 24.16.
60.
We avoid the designations Pg and Ps, as we do not wish to take a position when such a position is not necessary or relevant to our point.
61.
Baden, 2012: 225.
62.
Baden, 2012: 224.
63.
This is typical (e.g., following Carlson Hasler’s approach) of archival historiography, with its disrupting of the “narrative coherence”. See note 22.
64.
As Jeffrey H. Tigay noted, “the grossest disturbances in a composite text are those added in the latest redactional stage, when the traditional materials have lost plasticity” (“Conflation as a Redactional Technique”, in Tigay, 1985: 82).
65.
66.
That’s what we found in examining Deut. 14 and its relation to the Lev. 11 source. See, Friedberg/Hoppe, 2021.
67.
Similar issues arise in the analysis of biblical narratives. See, for example, Gen. 24.1–60, the long and seemingly highly repetitive account of Eliezer, Abraham’s servant, in his search for a wife for Isaac. Important and meaningful differences between the two “versions”, the telling and retelling of the mission, are glossed over because of what Meir Sternberg called, the “common fallacy of hindsight reading”, reading the beginning of a story in light of the end. Sternberg, 1985: 133.
68.
This, one should recall, is what Jacob and Heller did, each arriving at interesting and creative solutions even as they posited a single author.
69.
Wellhausen, 21889: 137f.: “Der Grund, weshalb der Verfasser von Kap. 25ss. […] [D]en goldenen Räucheraltar nicht mit aufführt, ist der, dass er von letzterem nichts weiß.”
70.
Kuenen, 1886: 75.
71.
See also the comments of Y. Isaac Halevi, who, in a perhaps excessively derisive and dismissive tone, rebuts categorically the inferences drawn by Wellhausen from the early prophets that had led him to conclude that the golden altar was not known in the pre-exilic period (Halevi, 1939: 216–228).
72.
“If the author of Exod. 25–29 had been acquainted with this ‘most holy’ object [= the altar of incense, Exod. 30.1–10], he would have described it in 25, and mentioned it in 26.31–37; and in 27.1–8 there would have been some indication that the altar there mentioned was not the only one. Lev. 16 likewise knows of but one altar, and does not mention the expiation prescribed in Exod. 30.10. On the other hand, it is urged that the golden altar, or altar of incense, appears repeatedly in P elsewhere […]; that the altar of Exod. 27.1–8; Lev. 16 is often called, as if for distinction, the altar of burnt offering […] or the brazen altar […], and that altars, in the plural, are mentioned […]; further, that the altar of incense is known not only to the chronicler […] and the authors of 1 Macc. […] and 2 Macc. […] but to the writer of the books of Kings […]: and probably also to Ezekiel, who speaks in 9.2 of the brazen altar, implying a knowledge of some other altar. All this evidence seems to Dillmann, 21880: 216, to remove the doubts urged by Wellhausen against Exod. 30.1–10. But this cannot be allowed. The fact remains that the passage does not stand where the author of Exod. 25–29 must have placed it, and the natural inference that this author had in his mind a sanctuary without an altar of incense is commended a posteriori by Ezek. 41.21, 22 and 44.16, which show that the prophet Ezekiel, likewise, left his proposed temple without any altar of incense distinct from the table of show-bread, inasmuch as he identified the two. Such then was the intention of the author of Exod. 25–29 (and Lev. 8) likewise.” Kuenen, 1886: 75. Likely, Kuenen intended to say that “the passage does not stand where the author of Exod. 25–29 should have placed it.” We interpret Kuenen to mean where he “must have placed it” had he been aware of the golden altar.
73.
Haran, 1978: 229.
74.
See Skehan, 1993:112–13.
75.
Bogaert, 1996: 399–427; : 62–76.