1. IN THE BEGINNING (1:1–2)
2 And the earth—it was a desert and a wasteland; darkness was on the face of the deep; and the Spirit of God was hovering over the surface of the waters.
1 Among the most well-known passages of Scripture is its very first verse, traditionally translated as above. Nevertheless, no small controversy among biblical scholars has swirled around both the translation and the meaning of the verse.
The issues are at least twofold. First, should v. 1 be translated as an independent clause, which is the approach taken in this commentary? Or is the verse to be understood as a dependent clause, “When God began to create …,” and thus subordinated to some following main clause? Second, what is the relation of v. 1 to v. 2, and for that matter, what is its relation, chronologically, exegetically, and theologically, to the remainder of the chapter?
First, the proper translation of the verse. A number of options are available here: (1) The first word, berē ʾš îṯ, is in the absolute state (i.e., it functions independently of any other word) and all of v. 1 is an independent clause and a complete sentence. (2) The first word is an indeterminate noun, used as a relative temporal designation: “Initially (or first, to start with) God created.…” (3) The first word is in the construct state (i.e., it functions in close connection with another word, usually a noun) and the verse is a temporal clause subordinated to v. 2: “When God began to create … the earth was without form and void.” (4) The first word is in the construct state and the verse is a temporal clause subordinated to v. 3, with v. 2 taken as a parenthesis: “When God began to create the heavens and the earth—the earth being without form and void—God said.…”
How shall we decide among these possibilities? A knowledge of the Hebrew language will not be sufficient in itself to settle the matter, for all four positions have been advocated by competent Hebraists, both ancient and modern. A survey of the extensive literature on the subject reveals that interpretations (1) and (4) have by far the widest support.
The main lines of argument in support of taking v. 1 as a dependent clause which prepares for the main clause in v. 3 are as follows. (a) The vowels in the word berēʾšîṯ indicate the word to be construct, not absolute, and the phrase must thus translate as “In the beginning of,” not “In the beginning,” for which one would expect bārēʾšîṯ. (b) In the “second” Creation account (2:4bff.) the temporal construction is employed—“when the Lord God made the cosmos”—and is a structural parallel to 1:1. (c) The word rēʾšîṯ occurs some fifty times in the OT, and all of these, except possibly Isa. 46:10, are in the construct state. 3 Is it likely that Gen. 1:1 contains an exception? (d) Taking the first verse as a dependent clause provides further substantiation for the Babylonian background of this “Priestly” account of creation. That is, in the Babylonian Epic of Creation—the Enuma elish—the first nine lines parallel the first two verses of Genesis. 4 Thus:
(2) parenthetical clauses
Enuma elish, lines 1–2
Specifically then, Gen. 1:1—“When God began to create the heavens and the earth”—is the equivalent of the first two lines of Enuma elish: “When above, the heaven had not been named (and) below, the earth had not been called by name.”
Several of the more recent translations of the Bible have accepted this rendering: NEB, NAB, NJPS, RSV, and AB, but only in a footnote. Others, however, have retained the traditional translation; among them, NASB, NKJV, NIV, and JB.
The issue between these two options—“In the beginning when” and “In the beginning”—is not esoteric quibbling or an exercise in micrometry. The larger concern is this: Does Gen. 1:1 teach an absolute beginning of creation as a direct act of God? Or does it affirm the existence of matter before creation of the heavens and the earth? To put the question differently, does Gen. 1:1 suggest that in the beginning there was one—God; or does it suggest that in the beginning there were two—God and preexistent chaos? The latter approach separates itself from the former in that it dictates the existence of chaos prior to creation. But the concept of the creation of chaos would be a contradiction in terms.
In order to avoid this conclusion, several scholars (e.g., Westermann) have opted for the traditional translation, not on the basis of objective linguistic grounds—for they believe the Hebrew word itself to be ambiguous in form—but on the grounds of the wider context of the chapter. It is claimed, for instance, that the Creation story of Gen. 1 is a deliberate repudiation and demythologizing of a pagan cosmogony such as is found in Enuma elish.
If that be the case, is it possible to believe that the author would leave unchanged and unmolested, and thus endorse, one of the distinguishing concepts of the mythical worldview, viz., the creation of the world from preexistent matter which is outside the creator’s divine activity? Would such a vestigial motif be left undisturbed? Thus, speaking of v. 1, Brevard Childs says, “This verse can be interpreted grammatically in two different ways.… While there is a choice grammatically the theology of P excludes the latter possibility [viz., that 1:1 is a dependent, temporal clause subordinated to v. 3] … we have seen the effort of the Priestly writer to emphasize the absolute transcendence of God over his material.” 5
But one does not argue for the translation of 1:1 simply on the grounds of a biblical writer’s creation theology. While this is a legitimate criterion, if it is the only criterion the case for seeing absolute creation is seriously weakened. In our opinion valid lexical, grammatical, syntactical, comparative, and stylistic arguments have been advanced to substantiate the translation In the beginning. 6
They may be presented briefly as follows. Lexically, P. Humbert’s two studies are quite correct in their observation that rēʾšîṯ is almost always used in the OT in the construct state, the one departure being Isa. 46:9–10—“I am God … declaring the end [ʾaḥarîṯ] from the beginning [mērēʾšîṯ].” It cannot be denied that the prophet, in quoting God, is thinking in terms of God’s absolute disposition over beginning and end, with beginning and end indicating not “a specific period of time within history, but rather historical time as such.” 7 Now if one grants that, apart from the possibility of Gen. 1:1, Isa. 46:10 is the only bona fide illustration of this word in the absolute state, then this one example is sufficient to demonstrate that rēʾšîṯ may be used to express a temporal meaning by use of the absolute state construction.
The same word used here in Gen. 1:1, berēʾšîṯ (preposition plus noun), appears four other times in the OT (Jer. 26:1; 27:1; 28:1; 49:34, “in the beginning of the reign of X”). 8 Each time the noun is followed by another noun. Only in Gen. 1:1 is the noun followed by a verb in a finite form (specifically, a perfect form). This construction is not frequent, but it is known in most of the Semitic languages. 9 Here contrast needs to be made with Gen. 2:4b, literally, “in the day of Yahweh God’s making earth and heavens.” Everybody agrees that this is a relative sentence, “when Yahweh God made earth and heavens.” But the noun in this verse is followed not by a verb in a finite form, as in 1:1, but by a verb in a nonfinite form (specifically, an infinitive construct). Is it not plausible to suggest a different nuance in the two verses by virtue of their different verbal forms?
The absence of the article is not a fatal argument against construing the word as absolute. For one thing, if as we have argued Isa. 46:10 shows rēʾšîṯ used in an absolute sense, it also provides us with an illustration of this word used both absolutely and indeterminately, and thus an exact parallel to Gen. 1:1. Second, all the ancient versions translate the word as an absolute and the whole verse as an independent clause. 10 Third, the Masoretes understood the word to be absolute, for they accented the word with the disjunctive accent called a ṭip̄ḥá, which is normal for words in the absolute state, rather than with a conjunctive accent, which is normal for words in the construct state. 11
Syntactically, the argument in support of the traditional translation, or the translation that subordinates the first verse to the following, revolves around the interpretation of the verse’s relationship to vv. 2 and 3. Related to this matter is the comparative cosmogonic literature, that is, the alleged parallel in syntax between Gen. 1:1–3, 4bff. and lines 1–9 of Enuma elish. It is more accurate to say that there is a syntactical similarity between Enuma elish 1–9 and Gen. 2:4bff., but not between Enuma elish 1–9 and Gen. 1:1–3. If there is any parallel between Gen. 1:1–3 and Enuma elish 1–9 it is this: Gen. 1:2 parallels Enuma elish 1–8, and Gen. 1:3 parallels Enuma elish 9. Obviously Gen. 1:1 is unique. Gunkel was quite correct when he said, “The cosmogonies of other people contain no word which would come close to the first word of the Bible.” 12
On stylistic grounds the traditional translation conforms to the pattern of sentence lengths throughout the chapter. The rule is not long sentences combining subordinate and principal clauses, but rather a whole series of brief, terse sentences in paratactic style. Thus H. Shanks can say, “Why adapt a translation that has been aptly described as a verzweifelt geschmacklose [hopelessly tasteless] construction, one which destroys a sublime opening to the world’s greatest book?” 13
Finally, we may say a word about the interpretation that takes the first verse of the Bible with adverbial force, “initially, first,” zuerst as opposed to im Anfang. 14 Though this translation is possible, such a nuance would be expressed more directly in Hebrew by the phrase bāriʾšōná (Gen. 13:4; Num. 10:13–14; etc.), rather than by berēʾšîṯ.
2 And the earth—it was a desert and a wasteland. As we move beyond v. 1 and into v. 2, we do not, unfortunately, leave behind all problems of translation and interpretation. As will shortly become evident, v. 2 bristles with points of debate as much as does v. 1.
The first issue is the understanding of the two words tōhû wāḇōhû. It is unusual, but not unheard of, to have two juxtaposed words in Hebrew that rhyme. Rhyming could indicate, along with other factors, that the verse is poetry rather than prose. 15 It is less likely that the phrase is to be understood as farrago, that is, an expression made up of meaningless words but whose meaning may be determined from context. 16 No sure Semitic cognate for bōhû has yet been discovered, but tōhû may be safely equated with Ugar. thw, “desert.” 17
Both these words are nouns, and thus we have translated them a desert and a wasteland. The rendering “without form and void” (e.g., AV, RSV) might give the impression that the words are adjectives. But what do they imply and how does one arrive at a proper translation? On the one hand, the second word—bōhû—appears only three times in the OT and always in conjunction with tōhû (here; Isa. 34:11, “the line of confusion [tōhû] and the plummet of chaos [bōhû]”; and Jer. 4:23, “the earth, and lo it was waste [tōhû] and void [bōhû]”). 18
On the other hand, tōhû appears twenty times in the OT and unlike bōhû may stand on its own. Eleven of these occurrences are in Isaiah (24:10); 29:21; 34:11; 40:17, 23; 41:29; 44:9; 45:18, 19; 49:4; 59:4). In Deut. 32:10 the word is used in parallel with “desert” (miḏbār) and “wilderness” (yešimōn). The word also designates “desert” in Job 6:18, here a place of virtual death for any straying travelers. It is used to describe a deserted city in Isa. 24:10. The same concept of vastness and emptiness is illustrated by Job 26:7, “He stretches the north over the void [tōhû], suspends the earth on nothing [belîmá].”
Figuratively, the word describes that which is without substance or reality, something that is groundless, be that conversation (Isa. 29:21), the religious idols of the nations (“wind [rûaḥ] and emptiness are their images”) (Isa. 41:29); also the makers of these idols (Isa. 44:9); and apparent futility in labor, as expressed by the suffering servant (Isa. 49:4, parallel with heḇel).
For our purposes Isa. 45:18–19 is most interesting: “Yahweh … did not create it [the earth] a chaos … I did not say … ‘Look for me in chaos.’ ” As we shall see shortly, an issue is raised in the interpretation of Gen. 1:2 on the basis of whether Isa. 45:18 reads “Yahweh did not create the earth a chaos” or “Yahweh did not create the earth to be a chaos.” In sum, we observe that the nuance of tōhû is brought out most clearly by words with which it appears in parallel: desert, wilderness, wind, nothing, vanity. None of these obviously appeals to one’s sense of the pleasurable and the aesthetic. At the same time we shall need to discuss below whether the use of these two nouns in Gen. 1:2 designates a creation once pristine but now perverted, or whether 1:2 teaches a creation from a primordial chaos, or whether the expression is a generalization of which vv. 3ff. are a particularization. We reserve comment until we discuss the syntax of the entire verse.
darkness was on the face of the deep. Although Gen. 1 states that God created light (v. 3), it does not say that he created darkness. May we assume from this that darkness, unlike light, is not a part of God’s creation, but is independent of it? 19 Is day superior to night? Can one place spiritual meanings on physical phenomena? Other creation traditions within the OT do place darkness within the sphere of God’s creative acts. Compare Isa. 45:7, “I form light and I create darkness.” 20
Beginning with the discovery and publication of Enuma elish in the late 1800s, much attention has been given to the relationship between one deity in this epic, Tiamat, and the Hebrew word for deep, tehôm. In Babylonian lore, Tiamat is the belligerent and monstrous ocean goddess. As one who leads battle against the supreme god Anu, she is the personification of evil. Before she is able to win this battle, however, another deity—Marduk—defeats and kills her, then slits her corpse lengthwise “like a shellfish.” 21 From these two parts of her body Marduk forms heaven and earth.
Lending credence to the possible relationship between tehôm and Tiamat was the fact that the Hebrew word is feminine, and in all of its thirty-five occurrences it appears without the article except in Isa. 63:13 and Ps. 106:9. This fact suggests that tehôm may indeed be a proper name.
Further support for tehôm as a Hebraized form of Tiamat is found (a) in its association with verbs that can be applied only to human beings or animals; thus Gen. 49:25, “the deep that lies [couches or crouches] below,” and Hab. 3:10, “the deep gave forth its voice”; 22 (b) in several uses of tehôm, apart from Gen. 1:2, that occur in a paragraph dealing with Yahweh’s obliteration of superhuman monsters. The best example is Isa. 51:9–11, where a list of Yahweh’s conquests includes Rahab, the dragon, the sea, and the waters of the great “deep.”
Even if the etymological equivalence of tehôm and Tiamat be granted, this still does not demonstrate that the biblical Creation story has a Babylonian background. For one thing, many ancients believed in a primeval watery mass out of which the orders of creation emerged, whether these ancients were the Egyptians with their concept of the god of the primeval waters—Nu—who is the source of all things, or the Greek philosopher Thales. 23 Second, the deep of Gen. 1 is so far removed in function from the Tiamat of Enuma elish that any possible relationship is blurred beyond recognition. The deep of Gen. 1 is not personified, and in no way is it viewed as some turbulent, antagonistic force.
Strong negative arguments may be sounded regarding the linguistic relationship between Heb. tehôm and Babylonian Tiamat. 24 Much more likely is the correspondence between Heb. tehôm and Ugar. thm (dual, thmtm, plural thmt), “deep, depth(s),” or even earlier Eblaite ti-ʾa-matum, “ocean abyss.” 25
and the Spirit of God was hovering over the surface of the waters. The main issues in this phrase are the translation of the couplet werûaḥ ʾĕlōhîm (“the Spirit of God, the spirit of God, a wind from God, an awesome gale”) and the translation of the verb meraḥep̄eṯ.
As for werûaḥ ʾĕlōhîm, the translation “an awesome gale” may be disposed of most easily. This suggestion takes rûaḥ as “wind” (a valid translation for the word in many biblical passages) and ʾĕlōhîm not as a name for deity but as a way of expressing the superlative in Hebrew—hence “a powerful, awesome, tempestuous, raging wind.” To sustain this picture of intensity and a storm-like atmosphere the following participle, meraḥep̄eṯ, is rendered something like “was sweeping” or “was stirring.” 26
Several factors militate against this translation. First, none of the other eighteen occurrences of this phrase in the OT means anything like “mighty wind.” The next appearance of this phrase is Exod. 31:3, where Bezalel is filled with the rûaḥ ʾĕlōhîm in order to be equipped to build the tabernacle. Obviously a “tempestuous wind” did not come upon Bezalel. This key phrase unites, via an intertextual allusion, world building and tabernacle building, the creation of a world and the creation of a shrine. 27 It is most unlikely, therefore, that the phrase be read negatively in Gen. 1 and positively in Exod. 31.
Second, it is true that there are some plausible examples in the Hebrew Bible of ʾĕlōhîm used as a superlative, that is, as an attributive adjective rather than a noun. 28 But even these examples are ambiguous. Thus, in Gen. 23:6, is Abraham addressed as “a prince of God” or as “a mighty prince”? In 30:8 does Rachel wrestle with “wrestlings of God” or with “mighty wrestlings”? Is Nineveh “a great city of God” or “an exceedingly great city” (Jon. 3:3)? But even if the translation were transparent in these three references, this would not allow one to apply the same force to ʾĕlōhîm in Gen. 1:2c, for two reasons. First, how could the reader of the original or the translator be expected to differentiate the ʾĕlōhîm of v. 2c from all other occurrences of ʾĕlōhîm in the first chapter? Second, taking ʾĕlōhîm as superlative, and as a further descriptive part of the chaos of formlessness and darkness, places ʾĕlōhîm in v. 2c in opposition to the ʾĕlōhîm who in v. 1 creates the heaven and the earth, and who in v. 3 speaks. 29
As our third objection to this translation, we note that if the author had intended to say “a mighty wind” he could have used unambiguous expressions such as rûaḥ geḏôlá (1 K. 19:11; Job 1:19; Jon. 1:4, “a great wind”) or rûaḥ ṣeʿārá (Ps. 107:25; 148:8, “a stormy wind”).
The preferable translation, then, is either “S/spirit” or “wind” of God. In modern times H. M. Orlinsky has made the most cogent presentation of the arguments for “wind.” 30 For Orlinsky the translation “S/spirit” is an inauthentic “christianizing” of the Hebrew text, a tradition that is traceable to the philosophical interpretation of rûaḥ by Philo, the famous Jewish philosopher from Alexandria in the first half of the 1st century A.D. In Orlinsky’s judgment rûaḥ, which appears almost 400 times in the OT, does not translate as “spirit” in Genesis until 41:8, “in the morning his [Pharaoh’s] spirit was troubled.” By contrast, earlier uses of rûaḥ in Genesis clearly demand “wind” or “breath.” Thus, 3:8, “they heard the sound of the Lord God walking in the garden in the cool of the day [lerûaḥ hayyôm],” that is, in the windy or breezy time of the day. Similarly in 8:1, “God caused a wind [rûaḥ] to blow over the earth, and the waters subsided.” “Breath” seems most natural for rûaḥ in 6:17; 7:15, 22, “the breath of life.”
Some support for the translation “wind” is found in the Jewish targums, which are translations of the Hebrew Bible into Aramaic. So Targum Onqelos reads werûḥāʾ min-qoḏam-yeyā menaššeḇāʾ ʿal-ʾappê mayyāʾ, “and a wind from before the Lord was blowing over the face of the waters.” The other famous targums, Targum Jonathan and the Jerusalem Targum, both retain the verb “blowing” but, interestingly, qualify its subject as a “rûḥāʾ of mercy from before the Lord.”
The translations of the LXX and Vulg. do not clearly support either translation. The LXX kaí pneúma theoú epephéreto epánō toú hýdatos is exactly reproduced by the Vulg. spiritus Dei ferebatur.… Orlinsky translates the LXX “and the wind of God was sweeping over the water.” Note, however, that both epephéreto and ferebatur are passives; thus a literal translation would be “was brought” or “was carried.” 31 The passive form of the verbs reduces the necessity of rendering pneúma/spiritus as “wind.”
Probably the weakest part of Orlinsky’s argument is his contention that the translation “wind” provides another link between Gen. 1 and Enuma elish. In this myth Anu creates the four compass winds primarily as part of his arsenal to eliminate the antagonist Tiamat and then carry away her remains to a remote place. One would be hard-pressed to see any valid relationship between the rûaḥ of Gen. 1 and these storms or winds, called abubu and imhullu, which are Marduk’s weapons. 32
Further support for Orlinsky’s arguments comes from R. Luyster’s contention that the key to the phrase “the rûaḥ of God” is the entire clause: “the rûaḥ of God was hovering over the face of the waters.” 33 God’s ability to contain and rule over the cosmic waters is a sure indication of his power; for example, “you rule the raging of the sea” (Ps. 89:10 [Eng. 9]). Moreover, God’s wind or breath (as opposed to spirit) is a most potent manifestation of his presence. These two facts suggest that in Gen. 1:2 the antipode to the surging waters is God’s breath or wind, but not his spirit.
The arguments of both Orlinsky and Luyster seem inconsistent. First, one may safely say that the basic concept in rûaḥ is “breath.” Normally the context will indicate whether by “breath” is meant “wind” or “spirit.” What Luyster has not noticed is that in those passages he quotes about the wind as the most potent manifestation of God’s presence, the wind is destructive. Thus, Exod. 15:10, “you did blow with your wind, the sea covered them” (cf. also Isa. 11:15; 40:7; Hos. 13:15).
In those texts in which one has a legitimate choice between “breath” and “spirit” (cf. Gen. 6:3; Job 27:3; 33:4; 34:14; Ps. 104:30; Ezek. 37:14), the emphasis is one of energizing, giving life and vitality, creating and not uncreating. 34 If the emphasis that Gen. 1:2 wishes to make is that the rûaḥ is a destructive force, then we must opt for “wind.” If the emphasis that Gen. 1:2 wishes to make is that the rûaḥ is a beneficent force, then we must opt for “S/spirit.” It seems clear that the latter option is the preferable one. Even Luyster takes the participle meraḥep̄eṯ to mean “to hover” as a leader, a guide, a protector. Of course, the Hebrew alphabet, unlike modern alphabets, does not distinguish between upper case and lower case. Accordingly, there is no way to tell from the Hebrew whether one should read “spirit” or “Spirit.” To translate “Spirit” runs the risk of superimposing trinitarian concepts on Gen. 1 that are not necessarily present.
We turn our attention now to the force of the participle meraḥep̄eṯ, which is variously translated “was moving” (RSV), “moved” (AV), “was hovering” (NIV), “hovered” (JB), “swept” (NEB), “sweeping” (Speiser), “brooding” (Gunkel), “rushing” (Peters), 35 “swirled” (Fishbane). Obviously translations like “swept,” “sweeping,” “rushing,” “swirled” are dictated by the choice of “wind” for rûaḥ.
The verb is used infrequently in the OT. The Qal stem is used in Jer. 23:9: “my heart is broken within me [šāḇar]; all my bones shake [rāḥap̄].” The only other use is in Deut. 32:11 (in the Piel stem as in Gen. 1:2): “like an eagle that stirs up [ʿûr] its nest, that hovers [rāḥap̄] over its young.” Scholars have traditionally supposed that this verse concerns how a bird teaches its young to fly, specifically how the parent provokes the young to flight. The parent bird drives the young eagle from the perch by intimidation, by rushing at the young while vigorously flapping its wings.
But this interpretation may be called into question by the possibility that ʿûr in Deut. 32:11 does not mean “to stir up,” but rather “to watch over, to protect,” as in Ugar. ǵyr. 36 This parallel would indicate the likelihood of a similar meaning for rāḥap̄, at least in the Deuteronomy passage. In Gen. 1:2 is the rûaḥ “sweeping” over the waters or “watching over” the waters? If the latter, then “spirit” would be decidedly more accurate. Yes, there is a formlessness there, a forboding darkness, but all is kept in check and under control by the spirit of God.
Interestingly, in the Ugaritic texts this verb is always associated with eagles. For Ugar. rḥp C. H. Gordon suggests the meaning “soar.” 37 It has been found thus far only in the Epic of Aqhat: “Over him [Aqhat] eagles will soar, there will hover a [flight of b]irds” (3 Aqhat, obverse, line 20); “over him eagle[s] soar, there hovers a flight of bird[s. Among] the eagles soars ʿAnat” (3 Aqhat, obverse, lines 31–32); “Eagles so[ar] over the house of her father, there hovers a flight of birds” (1 Aqhat, line 32). Thus from the Ugaritic passages and from the Deuteronomy passage it appears that rāḥap̄ describes the actions of birds, not winds. 38
Finally, we must discuss the syntax of the entire verse. This verse has three circumstantial clauses with three different subjects and three different kinds of predication: a perfect verb (v. 2a), a nominal clause (v. 2b), and a participle (v. 2c). But do the contents of v. 2 describe something that came to be after God created an originally perfect universe? Or does v. 2 expand on and clarify the shape of the earth when God first created it? Or does v. 2 describe the situation before God begins his actual creation as introduced in v. 3?
The first of these suggestions is popularly known as the gap theory. In essence, this reconstruction suggests that v. 1 describes the original creation, which was flawless. Then something catastrophic happened (Satan’s fall from heaven?), throwing God’s perfect earth into turmoil and judgment so that it became (not “was”) without form and void. Subsequently God started a second creation, so that v. 3 describes not creation but re-creation. 39 The length of this gap between the first and second creation is impossible to determine.
For this analysis two points are essential. The first is that v. 2 be understood as describing something sequential to v. 1. This is accomplished by understanding the verb “was” in v. 2 as having an active rather than a stative force, and by reading the verb as a pluperfect: “the earth had become.…” The second necessary ingredient is that the phrase “a desert and a wasteland” be interpreted as a result of divine judgment, for it describes the exact opposite of a beneficent creation. Special appeal is made to Isa. 45:18—“he did not create it a chaos [tōhû]” as opposed to the translation “he did not create it to be a chaos.”
Now, at times the verb “to be” in the perfect tense can have an obvious active force. Certainly 3:22 says, “Behold, the man has become [hāyá] like one of us.” But for two reasons it cannot have this force in 1:2. First, if the writer had intended v. 2 to be read as a sequence to v. 1, he would never have used the construction he did: waw consecutive plus subject plus verb (in the perfect). Instead it would be: waw conversive attached to the verb (in the imperfect) plus subject. Thus, one would expect wattehî hāʾāreṣ rather than what we do have: wehāʾāreṣ hāyeṯá.
Second, in other circumstantial clauses the verb hāyá in the perfect tense normally carries its stative sense (3:1, “the serpent was wiser”; 29:16, “and Rachel was pretty; 34:5, “his sons had been [or were] in the field”; Exod. 1:5, “and Joseph was in Egypt”; Jon. 3:3, “now Nineveh was an exceedingly great city”). The burden of proof, then, is upon those who insist that here we have an instance of hāyá in a circumstantial clause with the meaning “became.”
We have already voiced our reasons for not interpreting tōhû wāḇōhû as a kind of early Sheol or Hades against which God’s wrath has been loosed. Instead, we see here a reference to the situation prior to specific creation, a situation of formlessness but over which God’s spirit superintends. 40
Syntactically, two possibilities remain in understanding v. 2. First, it may describe a condition concurrent with that described in v. 3, “the earth being without form and void, God said.… “This is the approach of Orlinsky, Speiser, and others. The most serious objection to this view is that contemporaneous circumstance is adequately handled by a verbless clause. 41 We would expect wehāʾāreṣ tōhû wāḇōhû. Hence, we opt for the second possibility, that on syntactical grounds v. 2 be understood as distinct from and prior to v. 3.
In sum, the position taken here is that v. 1 is an opening statement functioning both as a superscription and as a summary. As such, it is the functional equivalent to the colophon “these are the generations of,” which is the introductory sentence to each of the remaining major divisions of Genesis.
Verse 2 then describes the situation prior to the detailed creation that is spelled out in vv. 3ff. 42 It has long been observed that the creation days fall into the pattern of a movement from generalization to particularization. Days 1, 2, and 3 parallel days 4, 5, and 6. Thus day 1, the creation of light, goes with day 4, the creation of particular kinds of lights. 43 We suggest that this same movement occurs in v. 2 (generalization) and vv. 3–31 (particularization).
1 For the problems surrounding the vocalization of rōʾš, riʾšôn, and rēʾšîṯ, see C. H. Gordon, “Extensions of Barth’s Law of Vocalic Sequence,” Or 51 (1982) 395.
2 Lit., “the heavens and the earth,” which is to be taken as an illustration of hendiadys (an idea expressed by two nouns connected by “and”), or of merism (a means of expressing totality through two contrasting parts).
3 See P. Humbert, “Trois Notes sur Genèse 1,” in Interpretationes ad Vetus Testamentum pertinentes Sigmundo Mowinckel septuagenario missae, ed. N. A. Dahl and A. S. Kapelrud (Oslo: Forlaget Land og Kirche, 1955), pp. 85–96, esp. pp. 85–88; idem, “Encore le premier mot de la Bible,” ZAW 76 (1964) 121–31.
4 Among others, see E. A. Speiser, Genesis, p. 12; H. M. Orlinsky, “The New Jewish Version of the Torah,” JBL 82 (1963) 252–53.
5 B. S. Childs, Myth and Reality in the Old Testament, SBT 1/27 (London: SCM, 1960), p. 32. Von Rad takes essentially the same avenue: “Syntactically perhaps both translations are possible, but not theologically.… God, in the freedom of his will, creatively established for ‘heaven and earth,’ i.e., for absolutely everything, a beginning of its subsequent existence” (Genesis, p. 48). This attempt to translate the verse on the basis of the deciphered theology of the biblical writer by the biblical commentator is not without its detractors. See W. R. Lane, “The Initiation of Creation,” VT 13 (1963) 64–65.
6 Of the many studies, we may mention here: E. J. Young, “The Relation of the First Verse of Genesis One to Verses Two and Three,” in Studies in Genesis One (Philadelphia: Presbyterian and Reformed, 1976), pp. 1–14; W. Eichrodt, “In the Beginning,” in Israel’s Prophetic Heritage: Essays in Honor of James Muilenburg, ed. B. W. Anderson and W. Harrelson (New York: Harper and Brothers, 1962), pp. 1–10; G. F. Hasel, “Recent Translations of Genesis 1:1. A Critical Look,” BT 22 (1971) 154–68; H. Shanks, “How the Bible Begins,” Judaism 21 (1972) 51–58; B. K. Waltke, “The Creation Account in Genesis 1:1–3. Pt. III: The Initial Chaos Theory and the Precreation Chaos Theory,” BSac 132 (1975) 222–28.
8 W. Wifall (“God’s Accession Year According to P,” Bib 62  527–34) traces the source of the language and formula in Gen. 1:1 to the Deuteronomistic historian’s account of the reigns of the kings of Judah and Israel, royal chronicles that served as a model for the Priestly writer in his description of the reign of earth’s king. Creation, however, has replaced reigning.
11 Note, however, that the uses of berēʾšîṯ in Jeremiah are clearly in construct, yet they are also given a disjunctive accent, not a conjunctive one. See B. Waltke, BSac 132 (1975) 224. A. Sperber (A Historical Grammar of Biblical Hebrew [Leiden: Brill, 1966], pp. 463–64) also notes the appearance of the disjunctive ṭip̱ḥāʾ on words in a construct case, and concludes: “There exists no interrelation between accentuation and interpretation” (p. 465). At the same time Sperber makes the important observation (p. 627, § 100, and p. 637) that vocalization with a shewa may have a determinate meaning. Thus all agree that Gen. 38:25 has Tamar saying: “By the man [leʾîš] to whom these belong.” On p. 637 Sperber places berēʾšîṯ in this category.
16 This last passage from Jeremiah is not without significance, for it is used ironically by the prophet in portraying the reverse of the creation, God in judgment against his people undoing his creation. See M. Fishbane, “Jeremiah IV 23–26 and Job III 3–13: A Recovered Use of the Creation Pattern,” VT 21 (1971) 151–53.
17 D. T. Tsumura (“Nabalkutu, tu-a-bi[ú] and tōhû wābōhû,” UF 19  309–15) is inclined to see tōhû waḇōhû reflected in Ugar. tu-a-bi[ú], with the latter meaning “to be out of order, be unproductive.”
18 See J. S. Kselman, “The Recovery of Poetic Fragments from the Pentateuchal Priestly Source,” JBL 97 (1978) 163–64.
20 See J. L. McKenzie’s comment on Isa. 45:7: “In Israelite thought nothing, not even evil and darkness, could be removed from the dominion of Yahweh. Gen. 1:3–5 makes darkness the result of a work of division, not of creation in the sense in which the word is used elsewhere in Gen. 1” (Second Isaiah, AB [Garden City, NY: Doubleday, 1968], p. 77). If darkness is assumed in Scripture to be a threatening situation, what is one to do with a verse like Ps. 18:12 (Eng. 11), which states that God dwells in darkness, and darkness is his covering? Possibly darkness has both a benign and sinister nuance in the Bible. It is sinister in the sense of being the opposite of light, that out of which light evolves. But there is also a darkness that is protective, a darkness that conceals the location of a thief and acts as a veil for God, lest human eye behold him.
21 ANET, p. 67.
22 Note that the pronominal suffix on “voice” is 3rd masc. sing. (“his voice,” not “her voice”), indicating that tehôm was also understood as masculine. Also, the verb here is 3rd masc. sing. perfect. In fact, tehôm is masc. in form. The fem. is indicated when an adjective (tehôm rabbâ, Gen. 7:11) or a participle (tehôm rō ḇeṣeṯ, Gen. 49:25) follows, and in the pl. form (tehōmōṯ, tehōmôṯ, tehômōṯ).
23 See W. G. Lambert, “A New Look at the Babylonian Background of Genesis,” JTS 16 (1965) 293.
24 See R. L. Harris, “thm,” TWOT, 2:966.
25 See G. Pettinato, “The Royal Archives of Tell-Mardikh-Ebla,” BA 39 (1976) 50. M. J. Dahood (Psalms, AB, 3 vols. [Garden City, NY: Doubleday, 1965–1973], 2:231) wonders why scholars who know well the Ugaritic evidence continue to maintain an “incontestable connection” between Babylonian Tiamat and tehôm, both mythologically and philologically.
26 A modern commentator taking this approach is G. von Rad, Genesis, p. 49. His suggestion is “a terrible storm.” Cf. also B. Vawter, On Genesis: A New Reading, pp. 40–41. Referring to his own translation, “the mighty wind,” he says: “In this probably correct assessment of the expression, we have a final stroke added to the picture of initial disorder, with the world-to-be as yet inundated by the dark primordial waters and whipped into a vortex by a driving wind.” Similarly, the evangelical scholar R. K. Harrison opts for “an awesome gale” (“Genesis,” ISBE, 2:438). This approach assumes that the three parts of v. 2 all describe in substantially the same way the primordial chaos, instead of limiting such description to v. 2a–b, but not v. 2c. See further P. J. Smith, “A Semotactical Approach to the Meaning of the Term rûaḥ ʾĕlōhîm in Genesis 1:2,” JNWSL 81 (1980) 99–104.
28 See D. W. Thomas, “A Consideration of Some Unusual Ways of Expressing the Superlative in Hebrew,” VT 3 (1953) 209–24. Thomas’s own conclusion is expressed on p. 218: “In the Old Testament it is, I believe, difficult, if not impossible, to point to any unambiguous example of the use of the divine name as an intensifying epithet and nothing more.” The first scholar to suggest that ʾĕlōhîm is a superlative was J. M. P. Smith, “The Syntax and Meaning of Genesis 1:1–3,” AJSL 44 (1928) 111–12; idem, “The Use of Divine Names as Superlatives,” AJSL 45 (1929) 212–13.
32 Ibid., p. 40. Also, W. F. Albright, “Contributions to Biblical Archaeology and Philology,” JBL 43 (1924) 368.
33 R. Luyster, “Wind and Water: Cosmogonic Symbolism in the Old Testament,” ZAW 93 (1981) 1–10.
34 This important observation was made by W. H. McClellan, “The Meaning of ruaḥ∙ Elohim in Genesis 1, 2,” Bib 15 (1934) 523.
36 See H. N. Richardson, “A Ugaritic Letter of a King to His Mother,” JBL 66 (1947) 322; M. J. Dahood, Psalms, 1:56.
37 UT, p. 484, no. 2327.
39 The so-called gap theory has been given a wide hearing principally through the very popular Scofield Bible (particularly, The New Scofield Reference Bible [New York: Oxford University, 1907], p. 1 n. 5, p. 752 n. 2). See also A. C. Custance, Without Form and Void (Brockville, Ontario: Custance, 1970).
40 See G. M. Landes, “Creation Traditions in Proverbs 8:22–31 and Genesis 1,” in A Light unto My Path: Old Testament Studies in Honor of Jacob M. Myers, ed. H. N. Bream, et al. (Philadelphia: Temple University, 1974), p. 286.
41 See F. I. Andersen, The Sentence in Biblical Hebrew (The Hague/Paris: Mouton, 1974), p. 85: “It is more likely that Gen. 1:2a means the earth had become (or had come to be) … as a circumstance prior to the first fiat recorded in Gen. 13, than that it means the earth was … as a circumstance accompanying the first fiat.” This statement does not contradict our earlier contention about the translation of hāyâ in circumstantial clauses. Andersen is simply saying that the activity in v. 2 is prior to that in v. 3. He is not saying, if I understand him correctly, that the activity in v. 2 is also sequential to v. 1. On stylistic grounds B. W. Anderson argues that v. 3 cannot be the apodosis of a temporal clause whose protasis begins with v. 1 (“A Stylistic Study of the Priestly Creation Story,” in Canon and Authority: Essays in Old Testament Religion and Theology, ed. G. W. Coats and B. O. Long [Philadelphia: Fortress, 1977], p. 153). This would be further evidence against interpreting v. 2 as a parenthetical clause that is to be conjoined with v. 3.
42 Among those embracing this position see B. Waltke, “The Creation Account in Genesis 1:1–3,” BSac 132 (1975) 225–28.
43 “From form to fullness” or “preparation and accomplishment” is D. Kidner’s choice of words (Genesis, TOTC [Downers Grove, IL: Inter-Varsity, 1967], p. 46).