Abstract
This essay argues that the vocalization of the very first word of Gen. 1.1 in the Masoretic Text (MT), bĕrēʾšȋt, which is often thought to be in error in some way, may instead be the result of exegetical activity. Specifically, in light of the well-attested tradition that links Wisdom with creation both within the Bible and without, it is possible that bĕrēʾšȋt in MT Gen. 1.1 participates in the line of interpretation that ciphers Wisdom as “Beginning” (rēʾšȋt) in light of Prov. 8.22. If so, the MT of Gen. 1.1 is not a grammatical error to be corrected, but an exegetical cross-reference, referring readers to Wisdom’s role in creation as known, inter alia, in Proverbs 8.
“With wisdom [bĕḥokmāh] Yhwh founded the earth; with understanding [bitbûnāh] he established the heavens” (Prov. 3.19)
1. Introduction
It is not without a decent measure of irony that the very first word of the Bible is debated. On the one hand, this may be because “the text’s account of the very beginning is shrouded in mystery” (Kass 2003:27). On the other hand, it may be just further proof of what Farrar wrote in the nineteenth century—namely, that “we need not go farther than the interpretation of the first chapter, and even the first verse of the Bible without being forced to confess that exegesis has stamped even its initial labours with the impress of its own incompetency” (1961:34). 1 That said, according to the vast majority of interpreters reaching back to ancient times, there is a problem with the pointing of the first word in the vocalized Masoretic Text (MT) of Gen. 1.1, or, if that word is correctly vocalized, then with the second word. 2 Today, the problem is typically stated in grammatical and syntactical terms, that is, the temporal clause introduced by bĕ- should be followed by an infinitive, so that the second word should properly have been (or should now be repointed to) bĕrōʾ (qal inf. cst.), especially in the light of comparable constructions in Gen. 5.1 (bĕyȏm bĕrōʾ); 2.4b (bĕyȏm ʿăśȏt); and 2.4a (bĕhibbārĕʾām; niphal). 3 Alternatively, if the verb as presently vocalized (bārāʾ; qal pf. 3ms) is given precedence and deemed correct, then the first word of the Bible should have been (or should now be repointed to) bārēʾšît—that is, with the definite article 4 —not the anarthrous form that is presently found in Codex Leningradensis (Codex L or MS Firkovitch B19A), 5 our most important vocalized exemplar for the MT of Genesis 1 and the base text used for Biblia Hebraica Stuttgartensia and Biblia Hebraica Quinta. 6
The specification of the precise manuscript in the previous sentence is far from pedantic because, as will be argued here, it is of considerable importance to see how other versions and witnesses thereto—not to mention other Hebrew manuscripts (the Masoretic tradition is hardly a singular entity, see Orlinsky 1966:xviii; Ginsburg 1966)—treat the vocalization of the first word of the Bible. 7 According to von Rad, after all, “[n]othing is here [in Genesis 1] by chance; everything must be considered carefully, deliberately, and precisely” (1972:47). Before proceeding further, however, two important points must be mentioned.
First, the treatment of grammar is as much a descriptive exercise as it is a prescriptive one. Indeed, when the grammar in question is historical grammar, it is safe to say that it is primarily, if not exclusively, a descriptive endeavor. In bĕrēʾšît, that is—if it is correct—we may have a grammatical “problem” that was not perceived as such by earlier speakers and/or writers of what we now call “Biblical” or “Classical” Hebrew. 8 This first point is a rather obvious one, but it is not infrequently forgotten from the privileged vantage point afforded by the major archaeological and linguistic discoveries of the past century and a half that have shed enormous light on comparative Semitic philology. But, to paraphrase Hamilton (1990:225; commenting on a different verse), textual emendations suggested by scholars armed with such knowledge sometimes stem from a desire to make the text conform to our own contemporary understandings of Biblical Hebrew and its syntax. Such understandings may, but also may not, be accurate—historically or idiolectically—and the unwarranted, if not at times arrogant, tendency to emend at will in prior generations of scholarship has been rightly critiqued. 9
Second, as most vocalized Masoretic texts are of late provenance, one must take seriously the possibility—if not, in fact, the actual reality—that later, secondary interpretive traditions have crept into the manuscripts by means of such vocalization and other para-textual elements. Indeed, with reference to certain well-known scribal phenomena such as the Kethib-Qere, the tiqqunê sopherim, or even the ʿiṭṭur sopherim, the point is widely acknowledged (see McCarthy 1981). Unfortunately, this possibility seems to be only rarely applied in thoroughgoing fashion by biblical scholars, who often operate with an abstract and reified notion of “text(s).” But if one does reckon with such a possibility, then the text(s) with which one is working—and here the point is the actual, physical manuscript(s)—is not just a window onto the past or onto other texts and textual traditions but is itself an artifact and product of that past, even and especially of its own, most recent past, testifying as much, if not more, to its own time of production as to times far removed. 10 It seems that, in the quest for historical knowledge behind the text and for text-critical knowledge about manuscript families, affiliation(s), and so forth, biblical scholars have sometimes underestimated or neglected the significance of the text/manuscript as a cultural and material artifact in and of itself. That an alternative approach can be both productive and highly insightful is well demonstrated by work done on various Versions and in medieval codicology. 11
Each of these two points has relevance for the present study, though it is the second that receives special emphasis. In what follows, I will argue that the vocalization of the first word in Gen. 1.1 MT may not be a problem after all—whether that be a scribe’s (or tradition’s) idiolect or a scribe’s (or tradition’s) poor understanding (!) of proper Hebrew grammar—but may, instead, be understood as a reflex of a rather common exegetical tradition. That tradition, found widely in both Christian and Jewish sources, understood rēʾšȋt to be a cipher for personified Wisdom, who, according to Proverbs 8, was present with God in or at the “beginning” (rēʾšȋt, Prov. 8.22; cf. mērōʾš, 8.23) and who assisted God with the creation of the world (8.30). If this suggestion is correct, the calculus of these data would indicate that the vocalization of the first word of the Hebrew Bible in the MT is exegetically motivated, signaling a particular (sapiential) interpretation of Gen. 1.1, which, among other things, directs readers to other passages about creation found elsewhere. The MT of Gen. 1.1 as it now stands can (and no doubt will) still be understood by scholars as either a temporal clause, subordinate to v. 2 or v. 3, or as an independent sentence, which might function as the introduction, title, or “motto verse” of the chapter (cf. Stipp 2007). 12 But the vocalization preserved in the MT (bĕrēʾšît) can also, and perhaps better, be seen as a subtle preservation of and index to an exegetical position – namely, that it was “with [bĕ-] ‘Wisdom’”—here ciphered as rēʾšȋt—that God created the heavens and the earth.
Establishing this case involves correlating two datasets: (i) the interpretive tradition that connected Wisdom to creation and, additionally, frequently used rēʾšȋt as a cipher for Wisdom (§2); and (ii) the text-critical evidence that suggests that the vocalization bĕrēʾšît is not invariable but may be by design (§3). Both datasets are well-known but, as Wittgenstein once said, it is often the case that “[t]he problems are solved, not by reporting new experience, but by arranging what we have always known” (2001:40, no. 109). In the present study, the (re)arrangement in question leads to several implications for studies of Gen. 1.1, inner-biblical exegesis, and the textual criticism of the Hebrew Bible (§§4-5).
2. rēʾšȋt in Proverbs 8, Genesis 1, and Subsequent Interpretation
rēʾšȋt in Early Jewish Interpretation
Kugel, among others, has demonstrated the existence and endurance of the notion that Wisdom “came first” in God’s creative activity. In his compendium on the history of the interpretation of the Torah, Kugel locates the beginning of this tradition—at least in Scripture—in Proverbs 8, especially vv. 22-27: These words clearly state that God had created wisdom even before the heavens and the earth were made….There was thus every reason to believe that the creation of wisdom had come at the very beginning of things; this fact was plainly stated in the book of Proverbs. (Kugel 1998:44)
To be precise, Prov. 8.22 states that God “created” (√q-n-h) Wisdom rēʾšȋt darkȏ, “as [or: at] the beginning of his work(s).” 13 The first word of that adverbial clause is, of course, the same as the first word of Genesis, excepting the preposition bĕ-, though it is noteworthy that some Versions of Prov. 8.22 do reflect the preposition, perhaps in a harmonizing move with Gen. 1.1. 14 Beyond this lexical linkage, the specific content of the Proverbs passage is obviously what gave rise to the numerous interpretive discussions regarding the prime position of Wisdom in God’s creative activity and created order. Indeed, it seems clear that several of these discussions—especially by recourse to texts like Jer. 10.12; Ps. 104.24; and Prov. 3.19, which indicate that God established (√k-w-n), made (√ʿ-ś-h), or founded (√y-s-d) creation “with (his) Wisdom” (bḥkmh, bḥkmtw)—drew the logical conclusion that Wisdom was instrumental in all subsequent creation (Kugel 1998:45). 15 A brief listing may be offered here. 16
Aristobulus (mid-2d c. B.C.E.): “But Solomon, one of our ancestors, said more clearly and more eloquently that it [Wisdom] was there before heaven and earth” (Frg. 5). 17
Ben Sira (early 2d c. B.C.E.): “Before the ages, from the beginning [ἀπ’ ἀρχῆς], he [God] created me [Wisdom]” (Sir. 24.9a; cf. 1.4, 9). 18
Wisdom of Solomon (late-1st c. B.C.E. or early-1st c. C.E.): “And with you [God] is Wisdom who knows your works and was present when you created the world, and knows what is pleasing in your eyes and what is right according to your ordinances” (Wisd. 9.9; cf. 9.1-2; Winston 1979:203).
Hymn to the Creator (11QPsa = 11Q5) (first half of 1st c. C.E.): “Blessed is he who created the earth with his power, who established the world with his wisdom [בחוכמתו]” (11Q5 xxvi 13-14; Sanders 1965:47 and pl. XVI). 19
Philo (1st c. C.E.): “Wisdom is older than the creation…of the whole universe” (On the Virtues, 62; Kugel 1998:45; cf. Colson 1939:200-201). “And who is to be considered the daughter of God but Wisdom, who is the first-born mother of all things” (Questions in Genesis 4:97; Marcus 1953:381). 20
Within this interpretive trajectory, Wisdom came to be intimately associated with the creation of humanity on the sixth day. Note, for example, 2 En. 30:8 (late 1st c. C.E.): “And on the sixth day I commanded my wisdom to create man.”
21
What is of most interest to the present discussion, however, are those materials that demonstrate how the early tradition of Wisdom as first-creation and/or co-creator with God was conjoined with a move that ciphered Wisdom by means of the term “beginning” (rēʾšȋt).
22
At this point, the lexical linkage between Prov. 8.22 and Gen. 1.1, noted above, returns as a critical element in the exegetical process. Kugel explains: But if wisdom was the first thing that God had created, and if God had in fact used it to create the rest of the world, then biblical interpreters had to wonder: why did the book of Genesis leave out this crucial detail? Why didn’t the first verse in the Bible read: “In the beginning God created wisdom, and afterwards, the heavens and the earth”? In looking for an answer, interpreters noticed a striking coincidence. In Prov. 8.22, wisdom says, “The Lord made me the beginning of his work,” while the Genesis account opens, “In the beginning God created the heavens and the earth.” Perhaps this was not just a coincidence. Perhaps the word “beginning” in the Genesis verse was in fact a subtle hint, an allusion, to wisdom. For, if wisdom is called the beginning of God’s work in Proverbs, then (one might argue) the word “beginning” itself might be used elsewhere in the Bible as a kind of nickname for wisdom, a shorthand reference to the first thing that God created. If so, then the first verse of Genesis could now be understood as meaning not “At the start God created the heavens and the earth” but “In [or “with”] wisdom God created the heavens and the earth.” (1998:46; his emphases)
As primary examples of this exegetical move, Kugel (1998:46) cites the Targums and a passage from Philo:
Targum Neophyti:
23
“From the beginning, with wisdom [mlqdmyn bḥkmh],
24
the <Word> of the L
Fragment Targums:
27
“Brʾšyt:
28
With wisdom [bḥkmh]
29
the L
Philo (1st c. C.E.): “By using many words for it Moses has already made it manifest that the sublime and heavenly wisdom is of many names; for he calls it ‘beginning’ [ἀρχὴν] and ‘image’ and ‘vision of God’” (Allegorical Interpretations, 1.43; Colson and Whitaker 1929:174-75).
“And so,” Kugel summarizes, “interpreters came to the conclusion that not only was wisdom the first thing God created, but the phrase ‘In the beginning’ in Gen. 1.1 was intended to imply that it was by means of, or with the help of, wisdom that God had created the world” (1998:46; his emphasis).
rēʾšît in the New Testament and Beyond
The interpretive moves just described are attested elsewhere in related but less-overt ways. A parade example is John 1.1 (late 1st c. C.E.), which states that “in the beginning” (ἐν ἀρχῇ) was “the Word” (ὁ λόγος). The consensus among New Testament scholars is that the Johannine text is evoking Gen. 1.1. But which version (or interpretation) of Genesis? Several scholars have argued that the evocation is primarily of the Targumic tradition of creation by means of the word (mȇmrȃ) of the Lord.
31
G. Anderson (1990:27-28), however, has argued that the background of John 1.1 is not to be found solely in those Targums that explicitly attribute creation to (or by) the “Word,” but rather in the more widely attested interpretive tradition that saw Wisdom present at the creation. If so, then, the understanding of bĕrēʾšît in Tg. Neofiti would provide a remarkable parallel to John 1.1. Like Tg. Neofiti, John offers a twofold exegesis of bĕrēʾšît: on the one hand it is a simple prepositional phrase (“in the beginning”), while on the other it is a witness to the preexistent figure of Wisdom (“…was the word”). Moreover, John emphasized the fact that all creation was superintended (“all things were made through him…”) by this preexistent figure just as the targumist had done. (Anderson 1990:28)
32
Anderson goes on to consider the possible relationships—direct/genetic or indirect/mediated—between Neophyti and John. But again, given the wide attestation of Wisdom’s presence at creation and, therefore, in Genesis 1 (esp. 1.1), it seems that the latter option is most likely. What Anderson does not make explicit, however, is that these data can be reversed: John 1.1 may be seen as additional evidence of the interpretive tradition about Gen. 1.1 that associated Wisdom with (bĕ)rēʾšȋt. The Fourth Gospel belongs, therefore, in the listing of texts provided above.
If this is correct for John 1.1, it raises questions about similar language used elsewhere in the Johannine corpus. Space precludes a full analysis of all the texts here, but 1 John 1.1, a text that evokes the first verse of the Fourth Gospel, deserves particular attention. The verse begins with a piece of unusual and convoluted syntax, 33 speaking of “that which was from the beginning” (ὃ ἦν ἀπ’ ἀρχῆς). Because the unexpressed antecedent of this and the following phrases “concerns the word of life” (περὶ τοῦ λόγου τῆς ζωῆς), and given that some of the expressions subsequently employed speak of seeing, hearing, and touching, it seems clear that “that which was from the beginning” is really—despite the neuter construction—“the One who was from the beginning” (cf. 1 John 2.13-14), namely, Jesus Christ (1 John 1.3). 34 If this is correct, it raises interesting possibilities about a number of other passages in the Johannine epistles that employ the same prepositional phrase (ἀπ’ ἀρχῆς) and do so in a fashion that may not only evoke that verse (and John 1.1 behind it and Gen. 1.1 still further behind John 1.1) but also the referent of that phrase. That is, in several of these passages, one is tempted to think that ἀπ’ ἀρχῆς is being used in substantival, not just temporal, fashion. This seems especially possible for passages that refer to hearing something—either a commandment (ἐντολή) or a message (ἀγγελία)—“from the beginning” (ἀπ’ ἀρχῆς; see 1 John 2.24; 3.11; 2 John 1.5-6; cf. 1 John 2.7). To be sure, translating the prepositional phrase ἀπ’ ἀρχῆς as “from the beginning” and understanding it as a temporal reference to either (a) the ministry of Jesus, especially as found in the Fourth Gospel (cf. John 15.27; 16.4; see esp. R. Brown 1982:158), or (b) the point at which the Johannine community received and believed the Gospel are both widespread and perfectly acceptable positions. 35 But given the interpretive traditions outlined earlier, one might pause to ask whether ἀπ’ ἀρχῆς might also, perhaps even simultaneously, be a cipher for Jesus—“the Beginning”—who is associated with the Word, but also (and because of that) with Wisdom, both of which are associated in various ways with the “b/Beginning(s)” of Prov. 8.22 and Gen. 1.1. 36
Though such an understanding of the Johannine data is suggestive, it is far from certain, even though it is significant that LXX Prov. 8.22-23 twice uses ἀρχή with reference to Wisdom.
37
What is certain, in any case, is the wide attestation of the logos-wisdom-creation connection elsewhere. These other examples may, in turn, lend further support to the notion that the same idea is at work in ἀπ’ ἀρχῆς in the Johannine corpus. The first text to be mentioned is Colossians (second half of 1st c. C.E.), which calls the Son (Col. 1.13) “the image of the invisible God, the firstborn of all creation [πρωτότοκος πάσης κτίσεως]; for in him [or: “for by him”; ὅτι ἐν αὐτῷ] all things in heaven and on earth were created” (1.15-16; NRSV). Several scholars have argued that Col. 1.15-18 is, in fact, “dependent on the exegetical history of bĕrēʾšît.”
38
The passage goes on to state that the Son “is before all things” (καὶ αὐτός ἐστιν πρὸ πάντων; 1.17) and that “he is (the) beginning” (ὃς ἐστιν ἀρχή; 1.18).
39
Sentiments like these receive further support, and resoundingly so, in some of the early patristic writings. Justin Martyr (ca. 100-165 C.E.), for example: Yet another testimony from Scripture will I give you, my friends, namely that God has begotten as a beginning [ἀρχὴν], before all his creatures, a kind of power endowed with reason from himself. [Elsewhere in Scripture this power] is also called the “glory of God,” or “the Son,” or “wisdom” [σοϕία], or an “angel,” and sometimes “God,” “lord,” or “word” [λόγος]. (Dialogue with Trypho 61.1; cf. 62.4)
40
Or, even more to the point, Origen’s (ca. 185-254 C.E.) commentary on John: For Christ is creator as (the) “beginning” [ὡς ἀρχή], since he is wisdom. It is because of his being wisdom that he is called (the) “beginning” [ἀρχή].
41
To underscore the argument further, two additional texts might be considered. The first is from Jerome (ca. 347-420 C.E.), who, in his Questions on Genesis, rejected the notion that Gen. 1.1 referred to creation through the Son, appealing to the LXX, Symmachus, and Theodotion to support his case (Hayward 1995:30; cf. Kugel 1998:67). Jerome’s opinion shows that not everyone, not even every Christian, agreed with the beginning-wisdom-logos-Christ interpretive equation, at least in Gen. 1.1—especially not a Christian who espoused Hebraica veritas! 42 At the same time, however, Jerome’s comment shows that the tradition he opposes did, in fact, exist, which further supports the trajectory outlined above. 43 The second piece of evidence comes from the Nag Hammadi codices that also make the connection, although from an admittedly far different perspective. Both The Sophia of Jesus Christ (104.16-18) and Eugnostos the Blessed (9.3-5) indicate that the first-begotten (or begetter) Son of God’s (or Christ’s) “female name [or “aspect”] [is] ‘First Begettress [or “First-begotten”] Sophia, Mother of the Universe’” (Parrott 1990:231; 1991:104-7). If space permitted, still other examples could be adduced from within the sources already cited and from others not yet mentioned. 44
To summarize to this point: there is a long-standing (and early) exegetical tradition in both Christian and Jewish circles that connects Proverbs 8 and Genesis 1. This tradition is attested to in texts that stem from different periods, locations, and languages. It is also found in a wide variety of genres, including translations and interpretive literature. The specific connection between Proverbs 8 and Genesis 1 (sometimes with the help of other texts like Psalm 104 or John 1) concerns Wisdom and her presence at the start (rēʾšȋt) of God’s creative activity and how this connection (and thus Wisdom herself) can be found in Gen. 1.1, especially by means of the “cipherization” of Wisdom as “(the) Beginning” (rēʾšȋt)—a move facilitated by the lexical linkage provided by Prov. 8.22 and Gen. 1.1. With the Wisdom-as-rēʾšȋt tradition sufficiently established, we may turn to the second dataset, namely the text-critical information surrounding Gen. 1.1.
3. brʾšyt in the Ancient Versions
Brʾšyt in Septuagintal, Latin, and Hexaplaric Texts
The data pertaining to (b)rʾšyt in the history of interpretation are both well documented and fairly well known. Equally well documented, but rather less well known—or at least underutilized with reference to the particular point under discussion—is the text-critical data regarding the first word of Genesis. Prima facie, the majority of it appears to support the text as vocalized in MT (e.g., in B19A) by reflecting an anarthous form of bĕrēʾšît, though it must be immediately stated that, in some of the Versions, this may have as much (or more) to do with the nature of the target language than the putative Hebrew original. Moreover, given the frequent (double-)use of the first word as the book’s title, brʾšyt is often written decoratively and without vowels, leaving us without any indication of its precise vocalization, even in Hebrew MSS that are otherwise vocalized. 45 With these caveats duly noted, the data from the Greek and Latin versions deserve discussion because of the preservation in both of transliterations of the first Hebrew word—again, sometimes as the title of the book (see Rüterswörden and Warmuth 1993).
Greek. The main Septuagintal tradition translates brʾšyt with ἐν ἀρχῇ (ΕΝ ΑΡΧΗ) for MT’s brʾšyt (see Wevers 1974:75; Brooke and McLean 1917, 1:1). Origen, Symmachus, and Theodotion are identical, 46 with the important exception of Aquila, who translated with ἐν κεϕαλαίῳ. 47 Transliterations that seem to reflect the anarthrous form of bĕrēʾšît in Gen. 1.1 also exist. The second column of the Hexapla ( Ἑβρ.), for example, reads βρησιθ or βρασιθ. 48 Elsewhere, Origen gives the title of the book as Βρησιθ (Origen 1862:1084b [§529 (54)]).
Latin. The Old Latin and Vulgate translate brʾšyt with in principio (see Gasquet 1926:141). The title of the book in most MSS is (or includes) bresit or bresith (followed by id est genesis or the like) with the exception of five MSS, which read beresit (SM; 11th c.) or beresith (relv; all 16th c., from 1532-1583 CE). The former two titles do not reflect the simple vocal shewa in the received form (bĕrēʾšît), while the latter two evidently do. Either way, these vocalizations are again confirmed in other sources. In his Questions on Genesis, for instance, Jerome writes that the name of the book “in hebraeo scriptum” is “bresith” (Hayward 1995:30; Antin 1959:3; cf. Field 1875, 1:7 nn. 1-2).
In the cases of both the Latin and Greek traditions, therefore, the representation of the first word of Gen. 1.1 tends to reflect MT as we now have it, though the fact that the Hebrew Vorlagen of these Versions—or, at least, the first word therein, given its (double-)use as the title of the book—may have been unvocalized is not to be missed and certainly complicates matters, preventing absolute certainty on the point.
That said, alongside this predominant, apparently anarthrous tradition, other and contrary evidence exists that must not be neglected. Note, first, the transliterations of Ἑβρ found in four manuscripts (Rüterswörden and Warmuth 1993:169): MS 73 (11th/12th c.): βαρησηθ
49
MSS 57, 413, and 78 (11th-13th c.): βαρησειθ.
50
By means of α in the first syllable, these renderings seem to reflect an articular form bārēʾšȋt (see Skinner 1963:13; Hasel 1971:159). Of course, one must tread carefully when dealing with Greek and Latin transliterations of Hebrew, not only because of phonological differences between the languages, but also especially because (again) the Hebrew Vorlagen at the time of initial translation were almost certainly unvocalized. 51 But that same judgment also holds true for transliterations that seem to reflect the anarthrous form, bĕrēʾšît: they too were likely based on unvocalized Vorlagen. Whatever the case, the fact that anarthrous vocalizations exist without any indication of a vowel (Ø) or by use of ε (or e in Latin MSS), in apparent emulation of the simple vocal shewa, makes it at least possible to see in the spellings with α renditions of an articular form. 52 It is possible, that is—especially given other available options—that the α-vowel in βαρησηθ or βαρησειθ may not represent the simple vocal shewa. To put the matter simplistically: βαρησηθ/βαρησειθ is not the same as βρησιθ/βρασιθ, bresit/bresith, or beresit/beresith.
If this point be granted, the question that immediately presents itself is whether the α-vowel (possibly) representing a determined form reflects a preexisting tradition, whether oral or written, or is simply an alternative transliteration of the unvocalized Vorlage(n). 53 This question seems impossible to answer because so much depends on the date of the specific traditions, the date of the specific MSS in question, and speculations concerning the data and shape (especially regarding vocalization) of the Hebrew Vorlage(n) or—in the absence of vocalized versions—in the reading tradition(s) pertaining to the consonantal text. 54 For the purposes of the present study, it is enough to note that, although the anarthrous vocalization of bĕrēʾšît in B19A receives support from other textual traditions, there is nevertheless some, admittedly slim, evidence that may reflect an articular vocalization, 55 thus indicating that the MT’s vocalization as we now have it was not the only option available or known. It is especially important that two of these readings stem from the second column of the Hexapla. And should one object that the MSS containing these readings are late, it should be recalled that their date is actually in the same horizon as the date of B19A itself (1010 C.E.), not to mention several of the interpretive traditions (specifically their extant MSS witnesses) discussed earlier. 56
brʾšyt in the Samaritan Tradition
That the articular vocalization of brʾšyt was a live possibility may receive further support from the Samaritan tradition. In 1917, Ritter recorded a reading of the Pentateuch in Nāblus, transcribing the first word of Gen. 1.1 as bā̀rā́šit; Ben-Ḥayyim transcribed the same realization as bā̊rā̊šǝt. 57 Either way, this specific form is anarthrous, as Samaritan Hebrew would double the r if the definite article were present; 58 the long a-vowel is the result of lengthening in an open syllable (see Ben-Ḥayyim 2000:45, 325). If the form were articular, the expected reading would be barrā̊šǝt. 59 In short, this tradition “does not provide firm testimony for the determinative state” (Tal 2015:77*; Rüterswörden and Warmuth 1993:175).
Two caveats may be entered, however. The first is to note, once again, that the double use of brʾšyt as both first word of and title for Genesis, meant that it was often left unvocalized even in much later, vocalized (and printed) traditions—Samaritan or otherwise. brʾšyt in such environments may well have intended bĕrēʾšît, but the unvocalized form must be acknowledged, as it contains at least some uncertainty with regard to precise realization. 60
Second, according to Schorch, “the Samaritan Targum and some manuscripts of the Samaritan Arabic translation contain readings of brʾšyt that seem to have been based on a reading with determination, which in turn seems to be a clear indication that, besides the reading without the article, as heard today and documented by Kahle and Ben-Ḥayyim, the older Samaritan tradition knew also this alternative.” 61
Such evidence, though not as widespread or dominant as the anarthrous reading, nevertheless demonstrates that alternative vocalizations, with the article, were known and available. It would seem, therefore, that one assumes the equation brʾšyt = bĕrēʾšît at one’s own peril, a point further underscored, perhaps, by those traditions witnessing a double translation of brʾšyt. 62
brʾšyt in the Masorah
A final item to consider is the note to בְּרֵאשִׁית found in the Masorah parva, collated in the first two lists of the Masorah magna (Mm 1-2). 63 The note reads: “בְּרֵאשִׁית: five times (total); three times at the beginning of a verse, twice in the midst of a verse.” Mm 1-2 conveniently list the five instances:
At the beginning of a verse (Mm 1): Gen. 1.1; Jer. 26.1; 27.1.
In the midst of a verse (Mm 2): Jer. 28.1; 49.34.
One might well ponder the purpose of such a note. Does it, perhaps, anticipate and attempt to circumvent a possible scribal mistake or problem? To be sure, not every comment in the Masorah performs such a function, but there can be little doubt that many do, and if that is not the case in this instance, it is hard to understand why it exists other than to simply collate data about various words and phrases in the Hebrew Bible. 64 The latter scenario is far from impossible, but even if correct, the question remains: why this particular word in this particular location?
One possible explanation is suggested when it is observed that in the five instances of בראשׁית in the Hebrew Bible, each is indefinite (i.e., בְּ־)—none is articular. Perhaps it is conceivable, therefore, that the note serves to remind scribes that Gen. 1.1 should be pointed just as all the other instances of בראשׁית are—that is, anarthrously—and not some other way. If so, that “other way” would likely be with the article (i.e., בָּ־), as that is the only alternative. If this scenario is plausible, the Masoretic note could be seen as evidence that (a) the articular alternative was, in fact, known; (b) certain scribes (or scribal traditions) existed who might have tended toward it; but (c) it was to be avoided in favor of the anarthrous vocalization. Whether the articular option was automatic (default?) in some way or simply a temptation of sorts, the Masoretic note would help protect against it by referring the scribe to the other Jeremianic attestations of בראשׁית, all of which are anarthrous. An anarthrous vocalization of Gen. 1.1 would thus be preserved and protected. Though suggestive, this explanation of the Masoretic note cannot be definitely proven. And yet (again), if it is correct, it would constitute one more piece of evidence that the articular vocalization of brʾšyt was known in the Masoretic tradition, perhaps even widely, and—or perhaps better, but—that it was to be avoided, perhaps even studiously. 65
This raises another issue that has been studiously avoided at least to this point: what was the original reading of brʾšyt—was it articular or anarthrous? In this case, as in so many others, recovering the original (or archetypal) text seems impossible given the existence of differing readings, each of which can lay claim to at least some antiquity and support from other sources (though, to be sure, not equally so). Moreover, in the case of brʾšyt, it may be unnecessary to make a decision with regard to the putative original. Good sense can be made of both articular and anarthrous readings, even while it is true that the adoption of either has significant bearing on subsequent reading and interpretation. 66 Furthermore, with regard to the connections with Proverbs 8 and Wisdom, both readings may work, perhaps even equally well. 67 It is to those connections I now turn.
4. Arranging the Data: “Sapiential Exegesis” in Genesis 1:1 (MT)
When the two datasets discussed above are correlated, a new, but rather old, interpretation of bĕrēʾšît in Gen. 1.1 becomes possible. It can be stated simply in two sequential points:
The existence of textual traditions that attest to an articular understanding of brʾšyt indicates that an anarthrous vocalization may be the result of a particular decision-making process (see §3).
The existence of interpretive traditions in premodern biblical interpretation that witness to the use of rēʾšȋt as a cipher for Wisdom, God’s first work of creation (see §2), suggests that the anarthrous understanding may be the result, not only of a particular decision-making process, but of an exegetical decision-making process.
These two datasets, that is, add up to the possibility that bĕrēʾšît in Gen. 1.1 is not a case of mistaken grammar or syntax, nor an example of a different, perhaps older or more native idiolect that no longer corresponds to the vast majority of similar constructions in Classical Hebrew prose. 68 Instead, the particular vocalization of brʾšyt in Gen. 1.1 MT as bĕrēʾšît may be understood as an exegetical and interpretive—perhaps even (semi-)midrashic—reading of the first verse of Genesis. If so, it would mean that, already in the very first word of the Bible, we have evidence of Jewish exegetical practice. 69 The phonetic realization /bĕrēʾšît/ is not, in this view, indicative of a temporal or circumstantial construction, though it can certainly still be read and understood as such. 70 It is also (or instead) an exegetical reading: bĕ-“rēʾšȋt”—that is, “With ‘Beginning’ [i.e., with Wisdom],” God created the heavens and the earth. 71 The connection between Proverbs 8 and Genesis 1 then may not be coincidental (Kugel 1998:46) so much as created – and not only by secondary interpretative sources but by the very vocalization itself. 72
If so, the most important conclusion to be drawn is that Gen. 1.1 is no longer a text in a vacuum, so to speak, with no associated antecedent literary context—even if the verse is judged an independent sentence. 73 Instead, the first verse, indeed the first word, as now present in the MT provides a kind of footnote or inner-biblical exegesis referring the knowledgeable reader to Proverbs 8. 74 One might build off Fishbane’s work at this point and call this an example of sapiential exegesis: an exegetical move concerned with Wisdom, especially as that is known from other texts elsewhere in the Hebrew Bible. 75 Such a move not only ciphers Wisdom and refers learned readers to the book of Proverbs and its discussion of Wisdom and creation, it also helps bind together, if not actually unite, what might otherwise be seen as two discrete and divergent creation theologies in Genesis 1 and Proverbs 8. 76 But when one sees that God creates “with ‘Beginning’”—that is, “with ‘Wisdom’”—in Genesis 1, there is no longer rough juxtaposition, contrast, or contradiction between these two creation theologies. Instead, there is connection, relationship, intertextuality (cf. Brown 2010; Strawn 2009; Keel and Schroer 2015). Such an outcome, suggested and prompted already by the lexical linkage provided by the use of the term rēʾšȋt in both passages, should, in turn, probably be seen as a (if not the) primary motivating factor in the attempt to connect the two passages in the first place—whether that is by interpretive art or by scribal practice (e.g., vocalization). 77 Whatever the case, G. Anderson’s conclusions regarding the Targumic renderings—namely, that “the targumist is not free to speak discursively about the general nature of ḥokmȃ in the creation process,” instead, “[t]his idea has to be woven subtly into the fabric of the text” (1990:27), seems to apply equally well also to MT Gen. 1.1 and those responsible for it. Given the existence and availability of other vocalizations in the reading and written traditions, it seems quite possible, that is, that the present vocalization found in MT Gen. 1.1 is the work of scribes familiar with the widespread exegetical tradition of Wisdom’s presence in and at the creation.
5. Implications for Textual Criticism
In closing, the significance of this argument for larger issues remains to be stated, especially in terms of the textual criticism of the Hebrew Bible. If the (re)arrangement of the data that has been offered here is at all instructive, it would constitute further proof that within the texts we now have—not just in the history or traditions that they relay but which lie at considerable remove behind them—we are faced with the end result of a longue durée. It is perhaps best to make this assertion more specifically about the MSS presently at our disposal. Texts are not pristine artifacts; the exact opposite is the case, however, when we are speaking of particular MSS instead of (reified) “text(s).” 78
These are hardly novel insights; they have been abundantly demonstrated in the study of the Dead Sea Scrolls and other Versions, on the one hand, 79 and, on the other, in Christian interpolations within Jewish writings like the Pseudepigrapha (see Davila 2005) or the Targums. 80 Unfortunately, biblical scholars often seem to operate de facto without these insights, afforded to us with force since 1949, when the Scrolls were first discovered, but certainly available earlier, and widely employed in codicological studies. That said, the point to be emphasized is that texts—any and all texts, even excavated ones—stand at the end (though never a complete and definitive one) of a long process and at the same time remain irreducibly the cultural-material products of specific religious communities and tradents, including also their exegetical traditions. 81 These latter may very well be—indeed, most certainly are—embedded within the very texts that those communities and tradents preserved and passed along for posterity, which become, in turn and in time, so much grist for the mills of subsequent exegetical reflection. In brief, text and exegesis, so often separated, are in reality inextricably conjoined. 82 Of course, this, too, is hardly a novel insight (see Breed 2014).
To their credit, scholars have long admitted that one biblical text, passage, or author is influenced by another (e.g., Kugel 1994:261-66). But in practice, this seminal insight has often been thought to operate in preexisting levels—in the traditions that underlie and gave rise to the texts as they now stand. Thus, for example, it might be granted that the author of Proverbs 8 had Genesis 1 in mind or that Gen. 1.1 somehow influenced the writer of Proverbs 8. The possibility being entertained here, however, is that the phenomenon of exegetical influence is at work within the MT itself, even in the first word of the MT, at least as that has come down to us in our main diplomatic exemplar. 83 This should come as no great surprise given the evidence of traditioning and related textual/scribal phenomena found elsewhere—most notably, perhaps, at Qumran. That we have similar phenomena in the MT is, once again, more than ably demonstrated by the tiqqunê sopherim or other theological changes to the text. 84
In sum, the case of bĕrēʾšît in MT Gen. 1.1 may be yet one more example of the same. The particular vocalization, without the article, may simply be the result of exegetical influence (cf. Heckl 2001; Maori 1992). Given the antiquity of the exegetical tradition conjoining Wisdom and creation, it is impossible to determine when this move might have taken place; we cannot be certain that it is the product of later, medieval exegesis alone, even if much of our manuscript evidence dates to that horizon. Whatever the precise case, the present study has argued that it is not simply text that produces exegesis. Exegesis also produces text. Or, said differently: it is not just vocalization that produces interpretation; interpretation produces vocalization. 85 In either formulation, it is clear that traditional models of text → exegesis, production → interpretation are typically too simplistic and must be revised to more accurately reflect the grittier and messier realities of “texts” and MSS. 86
6. Conclusion
In summary, bĕrēʾšît in MT Gen. 1.1 may be an example of exegetical influence within the MT itself. When all the data regarding bĕrēʾšît are considered, they give one pause. With all due respect to the vast secondary literature that it has spawned, perhaps the first word in the MT of Gen. 1.1 is correctly pointed after all, and this for exegetical, not (necessarily or simply) grammatical and syntactical, reasons. 87 If so, we have a simple and elegant solution to a crux interpretum that is as old as Rashi. It is a solution, moreover, that has not been proposed heretofore and yet is one that is constructed from preexisting knowledge—a rearrangement, as it were, of “what we have always known” (Wittgenstein 2001:40). The important work of rearranging, in turn, can bring to light still more questions and “new” data in our ongoing search for answers in the texts that continue to intrigue and elude us.
Footnotes
Acknowledgements
Thanks to several colleagues who commented on earlier drafts of this paper or otherwise assisted in important ways: Bill T. Arnold, Lewis Ayres, Brennan W. Breed, Steve Delamarter, Christopher B. Hays, Gail O’Day†, David L. Petersen, Stefan Schorch, J. Ross Wagner, and Christine Roy Yoder. I am also indebted to Doug Watson and Michael Chan for editorial support and also wish to express my gratitude to the editors and anonymous reviewers of JSOT.
1.
Cf. the sentiments of Barr (1998:55) and 
:93).
2.
The problem was noted already by Rashi (1040–1105 C.E.), who was the first to make famous the temporal understanding of brʾšyt that is now fashionable among modern translations of Genesis: “If you wish to interpret it according to its clear grammatical sense then interpret it this way: ‘At the beginning of God’s creating heaven and earth…then God said, ‘let there be light.’ Scripture does not intend to teach the order of creation, that heaven and earth were first. If it had it would have written ‘in the beginning (bārīʾšōnȃ)…’” (G. Anderson 1990:21; see also Orlinsky 1983:208). But cf. Barr (1998:58) who makes the case that Rashi’s temporal interpretation is the result, not solely of the linguistic and grammatical elements of the text, but “a semi-allegorical attempt to deal with a theological problem.” For precursors to Rashi, see Schäfer (1971:161-66, esp. 163) and :62-63).
3.
Some scholars leave bārāʾ intact but treat the words that follow bĕrēʾšît as an extended genitive clause. An objection often raised to understanding v. 1 as an independent sentence is the assertion that v. 2 should then have begun with wattĕhȋ hāʾāreṣ. See, e.g., Heidel (1951:92); Powis Smith (1928:108-15, esp. 108-10); Speiser (1964:12); and Orlinsky (1983:208). Although Heidel raises this point, he disputes it, as does Hasel (1971:159-60). Interestingly, , vol. 1:19-20) argues that it is the subordinate clause interpretation of v. 1, not the independent clause interpretation, that requires wattĕhȋ hāʾāreṣ at the start of v. 2.
4.
5.
But note the similar anarthrous construction that is found in Hos. 1.2 (tĕḥillat dibber-yhwh bĕhȏšēaʿ; cf. LXX), indicating that emendation of Gen. 1.1 may be unnecessary. See, among others, Rattray and Milgrom (2004:270); Westermann (1994:78); Arbez and Weisengoff (1948:140); and, on the Hosea text, Andersen and Freedman (1980:153). It is often argued that terms like rēʾšȋt, rōʾš, qedem, and ʿȏlām, when used adverbially sometimes appear without the article and yet in the absolute state. See, e.g., Heidel (1951:92 and n. 41); Arnold (1997:1025); Wenham (1987:12); Westermann (1994:95-96); Hasel (1971:158-59); and Skinner (1963:13). But as Barr (1998:57-58) notes, the absolute and construct forms of “beginning” are identical, so that the main question remains the presence or absence of the article—a problem that Schäfer (1971:161-62 and n. 1) rightly states is simply one of vocalization. For an extensive argument for understanding rēʾšȋt as temporal, see Humbert (1955:85-96, esp. 85-88; :121-31).
6.
Unfortunately, Genesis is not extant in the Aleppo Codex. For a facsimile of the folio containing Gen. 26.34-27.30 in Aleppo—the only part (picture) of Genesis in Aleppo that survives—see the frontispiece in Wickes (1887). The Damascus Pentateuch Codex (formerly MS Sassoon 507) is only extant from Gen. 9.26 on. See further .
7.
8.
Cf., e.g., G. Anderson (1990:22) on the absence of asyndetic relative clauses in postbiblical Hebrew; note also :6).
9.
As, for example, in the important but oft-neglected volume by :52-55, 103-104, and passim) who speaks of the need to revise our (preconceived) understanding of biblical Hebrew grammar in light of the actual behavior (often idiosyncratic and not always regular or regulated) of the Hebrew found in the Bible. Note, esp., the following remark: “It is high time that Bible scholars outgrow this attitude of superiority, and approach the Bible not as schoolmasters teaching the prophets how Hebrew sentences should be formed and Hebrew words spelled; but as humble students of these great masters of Hebrew, anxious to learn from them” (104). He is careful to indicate that such an attitude need not eschew a critical approach, however (103).
10.
There are points of contact between this statement and New Historicism, but the insight that texts speak of their own time as much as the time they putatively report is as early and fundamental as Wellhausen (see also :6). That this insight applies equally well to the physical manuscript’s time, that is, the time of production, and not just the time of composition, is where New Historicism as well as contemporary codicological work helpfully inform the discussion. See further §5 below.
11.
Note esp., e.g., M. Brown (2003) and Crown (2001); see also, for Ezekiel, Lilly (2012). I am grateful to Steve Delamarter for discussions on medieval codicology. See also : esp. 12-13) on scholarly (over-)attention to textual content at the expense of para-textual elements.
12.
See, among others, Fretheim (1994:342) for the three main grammatical interpretations: (1) v. 1 is temporal, subordinate to v. 3, with v. 2 as a parenthetical aside; (2) v. 1 is temporal, subordinate to v. 2; or (3) v. 1 is an independent sentence. Westermann (1994:78) gives four options, two of which subdivide Fretheim’s third depending on whether or not v. 1 is taken as a summary statement (almost a title) or as a description of the first act of creation (see further below). Option 1 was advocated by Rashi (see note 2 above) and is present in the NJPSV (Tanakh). Option 2 seems to have been held by Ibn Ezra (see Skinner 1963:12, 14; Hasel 1971:157)—though he may have later abandoned it (Cheyne 1885:50)—and is present in NEB, NAB, NRSV (also Gross 1981:131-45, esp. 142-45). Option 3 is as early as Targum Neophyti and the Fragment Targums (see G. Anderson 1990:24-25 and further below), and, as Arnold (1997) notes, is the traditional interpretation supported by (virtually) all ancient versions (also found in KJV, NASB, JB). Despite the historical weight of Option 3, many exegetes and modern translations follow one of the first two options (see Eichrodt 1962:1-10, esp. 1). For further discussion, in addition to the standard commentaries, see W. Brown (1993:62-72). For the first of the subdivisions of the third option, see van Wolde (2009:200: “‘In a beginning God differentiated the heaven and the earth’….a caption to what will subsequently be narrated”); for the second, see, e.g., Oswald (2008), who argues that bĕ- in brʾšyt is the beth essentiae and so the sense of the verse is “As his first work God created heaven and earth”; also Chambers (2019) and Poythress (2017). Bauks (1997:146) offers something of a hybrid approach that is not fully convincing: “Im Anfang, als God die Welt schuf.” See Keel and Schroer (2015:139-40); also .
13.
The literature on the meaning of √q-n-h in Prov. 8.22 is immense. See briefly Fox (2000:279-80, 411-12), and, far more extensively (with bibliography), Thomas (2017, esp. 473-88), who argues strongly against “create” as the meaning of the root. Even so, Thomas must admit that the creational sense is present already in the LXX, Targums, and Peshitta of Proverbs 8 (2017:473; see further below) and that “the weight of exegetical and philological tradition lies with understanding qny in this context to mean ‘to create’” (2017:474). Schipper argues in favor of “create” (2019:308). For drk (“way”; perhaps a collective singular, parallel to mipʾālāyw) as “work(s), task(s),” see Clifford (1999:91-92) and Fox (2000:280-81, esp. 281: “Derek means deed or deeds as a collectivity”). Contrast , who understands Prov. 8.22 to be analogous to Greek notions of preexistence.
14.
So Syriac (mryʾ brny bryS bryth), Targum (אלהא בראני ברישׁ בריתיה), and several MSS of the Vulgate, all of which may, of course, go back to a common tradition. See Gasquet (1957:52); J. de Waard in BHQ; J. Fichtner in BHS; G. Beer in BHK; Toy (1902:173); and Burney (1926:169); cf. G. Anderson (1990:24). Jerome knew of this “prepositional” tradition himself: see Epistle 140 where he transliterated Prov. 8.22a as Adonai canani bresith dercho. Burney (1926:167) thinks that Jerome’s citation is a mistake resulting from his quoting from memory, and goes on to list other passages from Jerome’s commentaries on Micah and Isaiah that also refer to Prov. 8.22 (the former uses in principio viarum suarum, the latter initum viarum suarum [173]). :156), however, thinks that the b in Jerome’s transliteration “is probably influenced by Gen 1:1.” The differences between Gen. 1.1 and Prov. 8.22 should not be neglected, of course. They include different verbs and divine names, among other things. These differences did not stop early interpreters, however, from drawing connections between the texts.
16.
For the entries that follow, see Kugel (1998:44-47) and . Translations usually follow Kugel unless otherwise specified, though I have had recourse to the original or critical editions and have occasionally made slight modifications toward these. The dates of composition also typically follow Kugel, though other sources have been consulted. Emphases, throughout, have been added.
17.
Preserved in Eusebius, Praep. ev., 13.12.11 (text and translation in :180-81). Note also Frg. 5a (preserved in Clement, Strom., 6.16.138.4a): “Wherefore Solomon also says, that before heaven and earth, and all existences, Wisdom had arisen in the Almighty” (Holladay 1995:180-81).
18.
All Greek citations of Ben Sira are taken from Ziegler (1965:237). The Hebrew of chap. 24 is not extant, but cf. the Syriac: mn qdm ʿlmʾ ʾtbryt (see Vattioni 1968:127; cf. also :145).
20.
21.
MS J (:45; F. Andersen 1983:150). MS E of 2 Enoch is nearly identical though it lacks the reference to the sixth day (see Andersen 1983:151). Cf. also the Hellenistic Synagogal Prayer (2d-3rd c. C.E.) preserved in Apostolic Constitutions 7.34.6: “having given order by your Wisdom, you created, saying, Let us make man according to our image and likeness” (Fiensy and Darnell 1985:679 [3:19], emphasis there; cf. Kugel 1998:45). Note also Wisd. 9.1-2.
22.
Cf. Schäfer (2008:269) on Genesis Rabbah 1.1: “the Torah is the uman…of God, and the beginning of the verse Gen 1:1 (bereshit bara elohim) must be understood, in the light of Prov 8:20, not as ‘In the beginning God created heaven and earth’ but ‘By means of reshit = Torah did God create heaven and earth.’” The ciphering allows one to speak of or evoke Wisdom as “(the) Beginning” even though rēʾšȋt has no direct semantic-lexical connection with √ḥ-k-m. Note the cognate epithet used of Apsu in the Enūma Eliš (I 3): apsû-ma reš-tu-u za-ru-šu-un, “Primeval Apsu, their begetter” (Labat 1935:76; :384).
23.
The date of Neophyti is debated; its traditions may be as early as the late-1st c. C.E. The date of the extant copy, however, is given in the colophon: 1504 C.E.
24.
Note the marginal variant (with definite article): bḥwkmtʾ (Díez Macho 1968:3). Mention should also be made of the articular reading in Targum Pseudo-Jonathan: mn ʾwwlʾ (see G. Anderson 1990:22; :16). Pseudo-Jonathan dates no earlier than the 7th or 8th c. with the MS dating to 1598 C.E. (ibid., 11-12).
25.
There seems to be a corruption here. See G. Anderson (1990:23 n. 7); McNamara (1992:52 nn. b, 2); and Díez Macho (1968:2-3; 1964:173-74) for discussions and for the insertion of mymrh (signaled by the < > brackets above). See also Schwarz (1982:136-37, esp. 136 n. 3) who makes a good case for the alternative spelling mmrʾ (cf. mmrʾ in vv. 3, 4, 5, 6, 8, 10, 11, 16, 20, 22 vs. mmr[y]h in vv. 9, 24, 25, 27, 28). The text as it now stands in Neophyti reads bḥkmh brʾ dYYY (“with wisdom, the Son of the L:102 and n. 7; signaled by the {} brackets above). Compare, at this point, the reading in the Fragment Targums below.
26.
The translation (mine) is close to that of Kugel (1998:46) and McNamara (1990:52). Kugel (1998:46 n. 2), McNamara (52 n. 1), and G. Anderson (1990:27) all concur that the reading “From the beginning, with wisdom” is an instance of double translation “in which the original word [brʾšyt] is translated twice…to fit two different understandings of the text” (:46 n. 2). See also note 28.
27.
The dates of the Fragment Targums are debated. , 1:25) states that they certainly cannot be earlier than the 2d c. C.E. and in actuality are probably substantially later than the Palestinian Targums. The specific manuscripts are dated as follows: MS P = 15th c.; MS V = 13th c.; MS N = 1291 C.E. (see ibid., 1:26 and n. 50; 1:29; and 1:31, respectively).
28.
, 2:4) translates “In the beginning,” but it is hard to know if brʾšyt here is intended simply as the title/incipit of the book or if it is the first word of the text itself with bḥkmh forming a double translation (cf. note 26). Given the large writing of brʾšyt in, e.g., MS V (see ibid., 2: plate 3) and the different Aramaic translations of brʾšyt found in the other Targums, it seems likely that the word is functioning as the book’s title.
29.
After bḥkmh, mn lqdmyn is written as a marginal gloss in MS P. See Klein (1980, 1:43); :23).
30.
31.
See Schwarz (1982:136-37); and, further, :27 n. 14) for bibliography.
32.
Similarly Kugel (1998:66). See also Schwarz (1982) and :431-32), both of whom discuss additional parallels in John 1.3-4.
34.
Cf. Rev. 22.13 where Jesus (see 22.16) states that he is “the beginning [ἡ ἀρχὴ] and the end” (cf. 21.6). R. Brown notes that some versions (especially Latin MSS) translated the difficult Greek of John 8.25 as “(I am) the beginning who also speaks to you” (1982:156; see also 1966-1970, 1:347-48). Note Augustine, Two Books on Genesis against the Manichees, I.2.3: “We answer them that God made heaven and earth in the beginning, not in the beginning of time, but in Christ. For he was the Word with the Father, through whom and in whom all things were made. For, when the Jews asked him who he was, our Lord Jesus Christ answered, ‘The beginning; that is why I am speaking to you’” (:49).
35.
Cf. Lieu 1986:75-77, 143-44, 173-74, 181, who thinks the first clause in 1 John 1.1 refers to the Gospel message, not Jesus. R. Brown (1982:155-58) actually summarizes six possible interpretations of ἀρχή in 1 John 1.1. Bultmann speaks of a double meaning (“the origin and the historic fulfillment of the proclamation”) for ἀρχή (1973:9 n. 10). See further Conzelmann (1954:194-201); La Potterie (1977:379-403); and .
36.
Cf. La Potterie’s conclusion (1977:402) and note 34 above. Elsewhere in the NT, ἀπ’ ἀρχῆς is used to refer to creation or the creation stories. See Matt. 19.4, 8; 24.2; Mark 13.19; 2 Peter 3.4; cf. Mark 10.6; Matt. 24.21; Heb. 1.10; and Rev. 3.14. Note also Sir. 15.14, “God created humankind from the beginning” (Gk: αὐτὸς ἐξ ἀρχῆς ἐποίσεν ἄνθρωπον; Heb: אלהים מבראשׁית ברא אדם). See Vattioni (1968:77) and Beentjes (2006:44, 52, 142) for variants and corrections in the different Hebrew copies (the one given here is MS A; MS B reads הוא מראש ברא אדם, but in the margin has מבראשׁית א]ל[הי]ם]) and for the Syriac (ʾlhʾ mn bršyt brʾ bny ʾnšʾ ). The use of מבראשׁית, with doubled preposition (prefixed/infixed בּ־ and prefixed -מן), in the Hebrew version is odd and may be an indication that the term is frozen, if not actually a cipher, even at this point. See further W. Brown (1993:64); Lella (1966:121); but cf. Bauer (1964:2-3). Note that the Gospel of John uses ἐν and ἐκ with ἀρχή, not ἀπό as in 1 and 2 John, but the variation may not be significant (:155-56). Cf. also LXX Hab. 1.12 and esp. Sir. 24.9a (cited above).
37.
Prov. 8.22: κύριος ἔκτισέν με άρχὴν; 8.23: ἐθεμελίωσέν με ἐν ἀρχῇ.
38.
G. Anderson (1990:29 n. 19), appealing to the work of Burney (1926: esp. 175-76). Burney is also followed by Davies (1967:150-53, cf. 172). See also :66) who compares Heb. 1.2.
40.
See Marcovich (1997:174-75, 177) and :66-67) for the text and translation, respectively.
41.
On John 1.22. See Blanc (1996:116-17; XIX.109[22]) and :67).
42.
Perhaps Aquila’s translation of brʾšyt with ἐν κεϕαλαίῳ is significant here (MS 78-413, which dates to the 12-13th c., reads ἐν τω κεφαλαίῳ; Wevers 1974:15, 17, 75). Could Aquila’s revision be contra Christian uses of ἀρχή as a cipher for Christ? Wevers (1993:1) finds Aquila’s translation “preposterous” and indicative of “the lengths to which Aq[uila] would go to maintain etymological equivalents.” Regardless of Wever’s assessment of Aquila’s technique, it is important to note that Aquila translated ראשׁית in Prov. 8.22 with κεϕάλαιον. Contrast LXX, Symmachus, and Theodotion, all of which read ἀρχή (see :169).
43.
44.
Note, e.g., OdeSol (1st-2d c. C.E.) 41.9, where Christ says, “the Father of Truth…possessed me from the beginning [but note bryšyt not articular bryšytʾ]” (Charlesworth 1985:770). A full review of the literature would have to make recourse to the debates over the meanings of Gen. 1.1 and John 1.1, the use of ἀρχή in these and related passages, and the logos-sophia Christology in early Church writings. See, among others, Alexander (2009); Simonetti (1965:9-87), Parys (1970), Winden (1997:3-48, 61-93), Logan (1985:123-29), Vignaux (1973), and the data from Clement of Alexandria, Theophilus of Antioch, Irenaeus, Tertullian, Hilary, Origen, Justin, and Tatian collected by Daniélou (1958:219-22)—all of which connect creation with the Son. As yet another example, see Augustine’s discussion of the Word as Wisdom of God by means of Ps. 104.24 in his Sermon 117, which is on John 1.1-3 (see :210). See also note 34 above.
45.
Unvocalized MSS of the Samaritan Pentateuch are, similarly, of little help (see von Gall 1918:1; :145). But see below on the oral (reading) tradition of the Samaritan Pentateuch.
46.
The latter two are confirmed by Jerome who states that they read in principio (see Hayward 1995:30; :3).
47.
See note 42 above, also there for the variant with τω.
48.
See R. Kittel in BHK; O. Eissfeldt in BHS. The βρασιθ reading is from MS 135 (10th c.; Wevers 1974:16, 75; cf. Brooke and McLean 1917:1). For βρησιθ, see Field (1875, 1:7), Montfaucon (1713, 1:1), and PG 15:143. Heidel (1951:93), relying on Lagarde (1882:5), also notes the spelling βρησίδ. Lagarde adds βρισηθ (ibid.). See further Wevers (1974:75) and :1) for the numerous MSS that have superscriptions employing ἀρχή.
49.
Cf. also Field (1875, 1:7 n. 2), Lagarde (1882:5), and Montfaucon (1713, 1:14), all of whom note βαρησηθ in the margin at the beginning of Cod. Reg. 1825 (see also :170).
50.
Cf. Eissfeldt in BHS; Kittel in BHK; and Lagarde (1882:5), who also give βαρησεθ. See also Skinner (1963:13), Heidel (1951:93), Westermann (1994:94), Hasel (1971:159), Shäfer (1971:162 n. 1), Ridderbos (1968:229), and :103 [= 3rd German ed., 1910:101]), all of whom cite these alternative Greek transliterations.
51.
See Kahle (1959:160 n. 3) and also, more specifically, Brønno (1966:209, 318-56) and Janssens (1982:67-75, 92-96, 107-10) on the various realizations of Hebrew shewa and what might lie behind Greek α in Secunda. See also, most recently, Kantor (2017). (Much) later copies—or, rather, copyists—would have had access to vocalized manuscripts—certainly by the medieval period from which so many of our MSS (in all traditions) stem. So it is possible that (the scribes of) MS 73 and MSS 57, 413, and 78 of the Hexapla might have had access to vocalized forms of the Hebrew text. Even so, the first word of Gen. 1.1 was often left unvocalized due to its double use as the title of the book. Whatever the case, it seems safe to say that translators had at least some familiarity with the vocalized tradition, even if only in oral (reading) form(s; these may have several), in order to do their work. The second column of the Hexapla proves at least this much (see Grabbe 1977:185-88, esp. 185; Brønno 1966; Janssens 1982; ). For the oral (reading) tradition of the Samaritan Pentateuch, see below.
52.
Contra, e.g, Powis Smith (1928:110), who calls the forms with α “stray” transliterations, and contra Heidel (1951:93), who thinks it may have been possible to say bĕrēʾšît or bārēʾšȋt “without any difference in meaning.” Similarly, Gunkel (1997:103 [= 101]), Hasel, (1971:159), and others. But note Waltke’s (1975:223) response: “It is more likely…that the change [in vocalization to articular forms] shows that those responsible for this reading were under the impression that the absolute sense demands the use of the article and accordingly altered the oral tradition. It does supply us with additional evidence…that traditionally the word was understood as an absolute.” What Waltke does not realize or allow for is the possibility that the articular vocalization may be proof that the oral tradition at this point was not altered but preserved. Cf. Heidel, who wants it both ways: while granting that βαρησήθ and βαρησέθ “support the old and generally accepted translation and interpretation of verse 1 [as an independent sentence]…the absence of the definite article in bĕrȇshȋth cannot be used as a point against it” (1951:93). On the latter assertion, see, among others, Wenham (1987:3) and Barr (1998:58); see also note 5 above. Note Stipp (2007), who thinks the interpretation calls for a determined form, without requiring that it be indicated morphologically. See the previous note, esp. the work of Janssens (1982:92-96, 107-10) on the realization of simple vocal shewa as Ø, α, ε, and ο. According to Brønno (1966:329), the number of instances for these vocalizations in Secunda are 126, 43, 33, and 5, respectively. See more recently for the default realization of shewa in Tiberian: “The quality of vocalic shewa in the Tiberian reading tradition was generally the same as that of the pataḥ vowel sign, i.e., the maximally low vowel [a]” (544).
53.
Or an interpretation thereof? Cf. Barr (1967) and :188-91). But even if interpretive, transliterations have an important part to play; they are, after all, witnesses to the text’s reception.
54.
Note Kahle’s theory that Origen used preexisting Jewish sources for the second column of the Hexapla (1959:158-63; 1960). Contrast Jellicoe (1968:106-10), with presentation of other perspectives on Secunda. Of these, note Emerton (1956; 1971), who argued that Secunda functioned to allow those who knew Hebrew to vocalize the consonantal text (presumably for reading). Jellicoe’s rebuttal (1968:110)—that those who knew Hebrew would already know enough of the language to supply the vowels—loses some force in the face of scenarios in which multiple vocalizations were known and possible, as, e.g., in Kethib/Qere variance. Emerton’s theory is positively evaluated in :216-17).
55.
See note 42 on ἐν τω κεϕαλαίῳ in MS 78-413, though such a reading could simply be stylistic, not the result of an articular understanding of brʾšyt.
56.
For the date of B19A, see the comments of Lebedev (1998:xxii) and :xxix). The Hexaplaric MSS range from the 11-13th c.; Neophyti is dated to 1504 C.E.; Pseudo-Jonathan dates to 1598 C.E.; the Fragment Targums date to the 13th-15th c. (see notes 23-24, 27 above).
57.
See Murtonen (1959:318); so also Eissfeldt in BHS; Kittel in BHK gives barašēt (cf. Kittel 1977:353). For the antiquity of the Samaritan pronunciation, see :153-57).
58.
See Tal (2015:77*); and Ben-Ḥayyim (2000:238; 316, 324-26; also 53-60, 90-92) and the following forms derived from √r-ʾ-š as recorded by Ben-Ḥayyim (see :257-58): barrāʾīšon (Gen. 8.3; Exod. 12.18; Num. 9.5), barrāʾīšūna (Gen. 13.4; Num. 10.13-14; Deut. 13.10; 17.7), karrāʾīšon (Lev. 9.15), karrāʾīšūna (Deut. 9.18), karrāʾīšūnem (Exod. 34.1, 4; Deut. 10.1, 3), and larrāʾīšūna (Gen. 28.19). In each of these cases, the MT has an articular form of ראשׁון (with preposition and/or suffix). Contrast, e.g., Exod. 24.17; Lev. 13.29; Num. 20.28; Deut. 20.9 (abrēʾoš)—instances in which the MT is anarthrous (בְּראשׁ).
59.
60.
The vast majority of MSS do not vocalize בראשׁית, largely because they set the word or at least the first letter (or two letters) in larger print (see Tal 2015:15*). When the entire word is large, it is typically centered, often decorated, and functions doubly as the title of the book. See Kennicott (1776-80, 1:1) for a few odd MSS that lack ב (592) or even the entirety of בראשׁית (8, 69, 264, 335). It is noteworthy that it is not until the 5th edition of the Rabbinic Bible (1617 C.E.) that בראשׁית is vocalized despite its larger setting/printing. The next two editions (1618 and 1723 C.E.), however, return to the unvocalized form. The Parma Pentateuch is another vocalized example of בראשׁית, despite the term’s use as the title to Genesis. But contrast the Parma Bible, which does not vocalize. (Both of the latter MSS can be found in -1958.)
61.
Stefan Schorch, personal communication (email of July 9, 2020). I thank Prof. Schorch for discussions on this point and for sharing with me the pertinent pages in Schorch (2021) prior to publication. See also :175, who evaluate the evidence differently.
62.
See above at notes 26 and 28, the discussions of Neophyti and the Fragment Targums. Cf. also the discussions of the Gospel of John and the double preposition in the Hebrew of Sir. 15.14 (note 36 above). This is contra :175, who deem any reading of the MT (at least) with the article as “eine freie Konjektur, die sich weder auf griechische Transkriptionen der Väter noch auf das samaritanische Material stützen kann.”
63.
See BHS and Weil (1971:1); also :54*).
64.
For the Masorah and explanations of its function(s), see Revell (1992), Yeivin (1980), Morrow (1992:24-30), and Kelley, Mynatt, and Crawford (1998).
65.
But see :210-11) for how even rejected readings/traditions may still exercise power and influence on readers and on the readings/traditions they deem authentic. This situation, which Parker describes as indicative of the contemporary context, was likely operative in antiquity as well; the double translation of brʾšyt (see above) seems apropos at this point.
66.
The dating of different vocalization traditions deserves brief comment. Tov (2012:47) is of the opinion that the Tiberian vocalization system reflects the 8th–9th centuries, whereas the “Samaritan tradition, the transliterations in :102) notes that “[s]ynagogues, where Torah scrolls were read, may have come under rabbinic control [only] as late as the seventh century C.E.”
67.
Anarthrous bĕrēʾšît as a later, intentional reading against an earlier, articular bārēʾšȋt might be the ideal scenario for a Proverbs/Wisdom connection, given anarthrous rēʾšȋt in Prov. 8.22. The explanation of the Masoretic note offered above fits such a reconstruction well. Articular bārēʾšȋt could also be seen as a plausible “Wisdom reading,” esp. if Wisdom could appear with a generic definite article (cf. forms like האלהים in Qoheleth), which seems at least possible in light of 1 Kgs. 7.14; 2 Chr. 1.12; and Eccl. 7.12, 19. Even so, the first scenario seems most attractive. Regardless, the fact that two vocalizations for brʾšyt exist must be accounted for in some way. These differing vocalizations are also what have focused scholarly attention on the first word of Gen. 1.1 and possible changes to it, not the second word, because there is no comparable discrepancy in the Versions or interpretive traditions with regard to the verb.
68.
Note the judgment of those scholars who believe it is impossible solely on the basis of grammatical considerations to determine the meaning of Gen. 1.1. E.g., Eichrodt (1962:6); Lane (1963:64, 69, 73); Hasel (1971:167); Westermann (1994:78, 96). Contrast on Gen. 1.1 as containing “an unmarked, restrictive relative clause.”
69.
It is noteworthy that, after citing, “In the beginning God created,” Genesis Rabbah begins with R. Hoshaiah citing Prov. 8.30-31. See Neusner (1985, 1:1; Parashah 1; §I:I); also Burney (1926:174-75); more generally, Schäfer (2008). Note also that the Mekhilta states that the reading “God created in the beginning” (nb: ʾĕlōhîm bārāʾ bĕrēʾšît), was one of the places in the Torah changed for King Ptolemy. See W. Brown (1993:63); Schäfer (1971:163). Cf. further Hendel (1998:18-19) on how LXX Gen. 1.2 reflects “Platonic cosmology in biblical guise.” Similarly, and further, Winden (1997: esp. 78-93). :100) thinks that Col. 1.13-20 “may, indeed, have arisen from the author’s meditation on the first words of Gen. I:1, and his elaboration of them in the manner of a Jew expounding scripture according to midrashic rules of exegesis.”
70.
Witness the double translation in Neophyti at Gen. 1.1 (see note 26 above; cf. also notes 28, 36). The doubled nature of the reading is part of its subtlety and brilliance. Such moves are common in cruxes and in exegetical tendencies elsewhere (see Fishbane 1985). See :222 n. 37) for the possibility that Gen. 1.1 MT is an instance wherein two traditions—a subordinate clause and independent sentence—were preserved.
71.
See Burney (1926:175-76) for possible meanings of the preposition bĕ- in this construction, esp. vis-à-vis Colossians 1. Burney lists spatial (“in”), instrumental (“by/through”), teleological (“into”), substantive (“before/beginning”), sum-total, head, and first-fruits interpretations of the phrase bĕrēʾšît. The latter is also found in Genesis Rabbah, where R. Huna ties bĕrēʾšît to first fruits: “For it is said, ‘On account of [the merit of] what is first, God created…’ (Gen. 1:1)” (, 1:11-12).
72.
Of course, relatively speaking, it is entirely possible that in its fullest (latest) form, the MT might be said to have “learned” of this exegetical connection from earlier interpretive traditions (see further §5 below).
73.
This would render irrelevant most if not all of the debate regarding whether (b)rʾšyt in Gen. 1.1 signifies an “absolute” beginning or not. The secondary literature on this question is enormous (see, e.g., Wenham 1987:12; Müller 1997; Arnold 1997:1025; cf. Jenni 1989)—with a good bit of it motivated by doctrinal concerns about creatio ex nihilo (e.g., Moskala 2011). But see G. Anderson (2017:41-58) and :29-39) for more subtle treatments.
74.
The low frequency of brʾšyt (only five times in the Hebrew Bible) and the remarkable mnemonic capacities of the rabbis should be kept in mind.
75.
Fishbane does not treat “sapiential exegesis,” but some of his material is analogous (see esp. 1985:522-24 and, more broadly, the section on scribal exegesis, 23-88). Whatever the case, Gen. 1.1 would certainly not be the only example of sapiential interpretation. Note, e.g., the earlier tendency to find “Wisdom influence” virtually everywhere. :129-42) helped control (if not debunk) this movement.
76.
See W. Brown (1999:35-132, 271-316), respectively, for theologies of creation in Genesis 1(-2:4a) and Proverbs. Note also his larger conclusions (1999:381-410) and his earlier work (1993), which argues for ideological differences underlying and motivating the different versions in MT and LXX of the first creation account. Landes (1974) argues for a closer relationship between the texts than is sometimes realized (so, similarly, ).
77.
See above at note 14 for the harmonizing preposition in the Syriac, Targums, and Vulgate MSS at Prov. 8.22. Note also Aquila’s translation of Prov. 8.22 and Gen. 1.1 (see notes 42 and 55 above).
78.
The notion of “text” is often an artificial one, and too frequently separated from the realia of specific MSS. The MT is as much if not more a witness to the MT tradition than it is simply (and certainly not facilely) to an original “text.” See Delamarter (2004a:7-8) and Parker (1997); also :4-5).
79.
One thinks especially of the Samaritan Pentateuch (see Crown 2001) and the LXX, the latter of which can be viewed as a testimony to Hellenistic Judaism and Jewish interpretation as much as a witness to the Hebrew text in antiquity. Of course, such an approach (similar in some ways to that of Kahle or Rösel) might be contested by those of a more genetic (Lagardian) perspective. For a convenient overview of LXX origins and interpretations thereof, see : esp. 53-66).
80.
81.
Again, there are some methodological similarities here to trends in New Historicism (see note 10 above). For convenient overviews, see Carroll (1998: esp. 52-57) and Hens-Piazza (2002), both of whom speak of the historicity of texts and the textuality of history, as well as the cultural materiality of texts about history and texts/textuality in general (on the latter, see also : esp. 795).
82.
What is argued here bears much similarity to Fishbane 1985, esp. on how the text was manipulated according to the needs of the community, prior to final stabilization, in a fashion akin to later midrash and exegesis (see 5-6); or on how later usage could have been “retrojected” into earlier sources (526). I concur with Fishbane’s judgment, that “our received traditions are complex blends of traditum and traditio in dynamic interaction, dynamic interpenetration, and dynamic interdependence” (542-43). And yet it is revealing that Fishbane speaks here of received “traditions” and not texts nor, esp., MSS. It seems that even Fishbane’s work is still somehow behind the text (or behind the MSS) to some degree. Fishbane speaks rightly of “exegesis as action and exegesis as literary form” (ibid.), but what might be added is a notion of exegesis as text/MS. For works that reflect on these latter factors—one on a MS from Qumran, the other on the Gospels—see Pulikottil (2001) and Parker (1997). Note also the Hebrew codicological work by Beit-Arié (1992:79-103; 1993; 1981) and Sirat (2002), and, more generally, : esp. 71, 193, 200-71, 397-402).
83.
It is possible, of course, that the vocalization issue has as much to do with the second word as the first. If the second word of Gen. 1.1 was originally bĕrōʾ, it might have been intentionally (re)vocalized as bārāʾ to prevent a temporal understanding of bĕrēʾšît, requiring, instead, the instrumental reading of bĕ- and securing (or at least suggesting) the cypherization of Wisdom as “Beginning.” Ps. 104.24 is important here: “in wisdom you have made them all” (kullām bĕḥokmāh ʿāšȋtā). I thank Bill T. Arnold for discussing this matter with me. Whatever the case, instead of the “back-referencing” discussed by :261-66, this would be a kind of “forward-referencing.”
84.
One example may be the ס (sĕtûmâ) after 2 Sam. 12.13a in many MSS, which some scholars have argued is meant as a pause to facilitate cross-reference to Psalm 51 (see Arnold 2003:534 n. 53; Ackroyd 1977:111; A. Anderson 1989:163; cf. Talmon 1966:19; Kugel 1994:261-66). In late-medieval codices, the text is a synchronic entity: scribes have the whole of the tradition before them to copy and comment on as they please. In such a setting, it is entirely possible that a chronologically or literarily later text may have influenced the writing, copying, and interpreting of an earlier-in-the-canon (codex) composition. This possibility also holds true if what is chronologically later is not a text at all but an exegetical tradition. See Delamarter (2004b:4) on the “sociological matrix that binds manuscripts and all of their details to the identity of the community that created them.” He argues that textual explanation, not just description, must be located in efforts to trace how scribal practices are rooted fundamentally in the ontology of the community and its texts (so also Delamarter 2004a). For a fascinating study of Qumranic material that is not unrelated at this point, see .
85.
Cf. Noegel (1996), who demonstrates that pointing a word removes what might have been intended ambiguity (14-15) and notes that non-normative orthography allowed a writer to achieve polysemy (146-47). See also .
86.
See note 78 above and cf. Levinson (1997), who has argued that Deuteronomy “breaks down any facile bifurcation between text and interpretation or between text composition and text reception” (17 and n. 46). From the perspective of the Qumran material and its impact on standard text-critical positions, see also Schniedewind (1999:252) and :31-40, esp. 34-35: “In one sense every copy of an authoritative scriptural book made in the late Second Temple period is a rewritten scriptural manuscript”).