Copyright © 1999 by Westminster Theological Seminary, cited with permission.
AND THE INTERPRETATION OF GEN 2:4-7
The interpretation of Gen 2:4-7 is a traditional hard place for Biblical
studies. These verses are often cited as proof of discord between the creation
narratives of Genesis 1 and 2, and hence as evidence of disparate sources
of the originals. In response, many have sought to harmonize the two peric-
opes, but with widely differing conclusions. The purpose of this essay is to
employ the tools of discourse grammar to see if they can shed light on this
We begin by giving the Hebrew of Gen 2:4-8, with the RSV for a sample
English version. Our grammatical discussion will lead to an interpretation
that we can express by modifying the RSV
Mymwv Crx Myhlx hvhy tvWf Mvyb Mxrbhb Crxhv Mymwh tOdlOt hlx 4
Hmcy MrF hdWh bWf-lkv Crxb hyhy MrF hdWh HyW lkv 5
hmdxh-tx dbfl Nyx Mdxv Crxh-lf Myhlx hvhy ryFmh xl yk
hmdxh-ynp-lk-tx hqwhv Crxh-Nm hlfy dxv 6
MyyH tmwn vypxb Hpyv hmdxh-Nm rpf Mrxh-tx Myhlx hvhy rcyyv 7
hyH wpnl Mdxh yhyv
rcy rwx Mdxh-tx Mw MWyv Mdqm Ndfb-Ng Myhlx hvhy fFyv 8
(4) These are the generations of the heavens and the earth when they were created.
In the day that the Lord God made the earth and the heavens, (5) when no plant
of the field was yet in the earth and no herb of the field had yet sprung up--for
the Lord God had not caused it to rain upon the earth, and there was no man
to till the ground; (6) but a mist went up from the earth and watered the whole
face of the ground--(7) then the Lord God formed man of dust from the ground,
and breathed into his nostrils the breath of life; and man became a living being.
And the Lord God planted a garden in
the man whom he had formed. (RSV)
The way the RSV has divided verse 4 into two parts, with 4b as the
beginning of the sentence that continues on through verse 5, represents a
Jack Collins is associate professor of OT at Covenant Theological Seminary.
common analysis of the clause-to-clause relationships.1 Further, the inter-
pretation of the Hebrew ‘eres as "earth" in verses 5-6 is also common.
Indeed, it is this that leads to the declaration of contradiction between these
verses and the events of Genesis 1. S. R. Driver is typical:2
The words [of verses 4b-5], taken in connexion with the sequel (v. 7), are intended
to describe the condition of the earth at the time when man was created: no shrub
or herb--and a fortiori, no tree--had yet appeared upon it, for it was not suffi-
ciently watered to support vegetation. According to i. 11f., plant- and tree-life was
complete three ‘days’ before the creation of man: obviously the present writer
views the order of events differently.
Those who oppose source criticism but still accept this clause-to-clause analysis
typically contend that the two pericopes are better seen as complementary
rather than contradictory; and any successful harmonization between the
two pericopes diminishes the credibility of conventional source analysis.3
A recent example of such a complementary interpretation comes from
Mark Futato.4 He says,
I understand Gen 2:5 as having a global reference that would parallel the situa-
tion prior to Days 3b [Gen 1:11-12] and 6b [Gen 1:26-30], i.e., before God created
vegetation and people.... Rather than being a second creation account, Gen 2:4-25
is properly read as a resumption and expansion not of Day 6 but of Days 3b and
6b taken together as a unit.5
He uses this to support the conclusion that strict chronological sequence is
not a part of the communicative intent of either Genesis 1 or 2, and hence
to support the so-called "framework" interpretation of the Genesis days.
1 The RSV as it stands is non-committal on whether verse 4a goes with 1:1-2:3 as its
conclusion, or with 2:4b-25 as its heading. My discussion will address that question later.
2 S. R. Driver, The Book of Genesis (
Cf. Hermann Gunkel, Genesis (
1910]),4-5; John Skinner, Genesis
(International Critical Commentary;
51; Claus Westermann, Genesis 1-11 (Minneapolis: Augsburg, 1984 [German original, 1974]),
197. Terje Stordalen, "Genesis 2, 4: Restudying a locus classicus," ZAW 104:2 (1992) 163-177,
gives an overview of the process whereby this became dominant (at 163). The influence of this
approach is visible in many places, which it would be tedious to document; one example would
be Dianne Bergant and Carroll Stuhlmueller, "Creation according to the Old Testament,"
in Ernan McMullin, ed., Evolution
and Creation (Notre Dame:
1985), 153-175, at 155, who simply take it for granted.
3 Richard Hess, "Genesis 1-2 in its literary context," TynB 41:1 (1990), 143-53, gives a
rationale for this approach without discussing the particulars of Genesis 2:4-7. As he observes,
it is a feature of Genesis first to give the overall picture, and then to go back and focus on some
details. Derek Kidner, "Genesis 2:5, 6: Wet or Dry?" TynB 17 (1966) 109-14, attempts to
harmonize the two passages by taking 2:5-6 as describing the same conditions as 1:2, "the
unrelieved expanse of waters" (112).
4 Mark Futato, "Because it had rained: A study of Gen 2:5-7 with implications for Gen 2:4-25
and Gen 1:1-2:3," WTJ 60 (1998), 1-21. Although the analysis and conclusions of the present
paper originated independently of Futato's work, they have profited greatly from that work.
5 Futato, "Because it had rained," 12 n.41 and 14.
INTERPRETATION OF GEN 2:4-7 271
In my judgment Futato is probably right in supposing that, under this
analysis of Gen 2:4-8, the only way to avoid the declaration of incoherence
between the two pericopes is to do away with sequentiality. But this can lead
us to question whether the analysis is itself right. I am the more interested
in raising this question, because my own exegesis has convinced me of a view
of the Genesis days as "analogical days," namely they are God's work days:
they are analogous, and not identical, to ours, structured for the purpose of
setting a pattern for the human rhythm of work and rest. According to this
interpretation, the days are "broadly consecutive" (allowing for the possi-
bility that parts of the days may overlap, or that there may be logical rather
than chronological criteria for grouping some events in a particular day).6
II. Discourse Considerations and Literary Structure for Gen 2:4-25
Discourse analysis is the discipline that studies texts as acts of communi-
cation. Discourse grammar analyzes grammatical structures, such as verb
tense and aspect, to find patterns of usage related to communicative intent.
Described this way, its advantages for exegesis should be obvious and not
particularly controversial. Unfortunately, discourse grammarians often use
exotic vocabulary and make extravagant claims, and generally do not make
clear to the uninitiated just which parts of their position are common
ground among Hebrew grammarians, and which are not.7 I aim to make
use of those parts which are in fact common ground.
Our first task is to identify the genre of our text: is it narrative, exposition,
exhortation, eulogistic poetry, lament, or something else? There is no diffi-
culty in discerning that in this passage we are dealing with narrative prose.
Next we must delineate the boundaries and structures of the individual
pericopes. In this case, we must decide whether we should in fact divide
verse 4; and then whether any of its parts belong to the first pericope
(1:1-2:3) or to the second (2:5-25).
Many have noticed that in Gen 2:4 we have an elaborate chiasmus.8 In
general, the communicative function of a chiasmus is to unify its parts, with
6 C. John Collins, "Reading Genesis 1:1-2:3 as an act of communication: Discourse analy-
sis and literal interpretation," in Joseph Pipa, Jr. and David Hall, eds., Did God Create in Six
Days? (Taylors: Southern Presbyterian Press, 1999), 131-51. At 142-43 I offer some critique of
the framework scheme.
7 In this light it is understandable that Bruce Waltke and Michael O'Connor, An Intro-
duction to Biblical Hebrew Syntax (Winona Lake: Eisenbrauns, 1990), 55, "have resisted the
strong claims of the discourse grammarians"; but it is nonetheless an unfortunate decision on
their part. After all, the goal of discourse grammar is riot to replace the traditional grammar
(which seems to be Waltke and O'Connor's perception), but to incorporate that grammar into
a systematic description of what good readers in the receptor audience do when they receive
8 E.g., Yehudah
Kiel, Sefer Biresit (Genesis, Da’at Miqra;
1997), 43 (Hebrew page numbers); C. John Collins, "The wayyiqtol as ‘pluperfect': When and
the context allowing us to infer just what kind of unity the author has in
view.9 The chiasmus here can be seen thus: a heavens ... b earth ... c
when they were created c' in the day that the Lord God made b' earth and
a' heavens. As I observed in an earlier paper,
Such an elaborate chiasmus is evidence of art, not coincidence. Further, by this
means the author has tied the two accounts together: note how the word order
"the heavens and the earth" (a and b), as well as the verb bara "create" (c), point
us back to 1:1 (as well as 1:21, 27 for the verb); whereas the change in divine name
from ‘elohim, "God" (ch. 1) to yhwh elohim, "the Lord God" (ch. 2-3) is reflected
in the c' element. It is hard to escape the conclusion that the final editor wanted
his readers to read the two accounts as complementary, not contradictory.10
This further shows why the change in divine name from 4a to 4b does not of
itself indicate that the two parts are separable:11 instead, as Franz Delitzsch
put it, "The combination of the two names denotes ... the oneness of God
the Creator ['elohim, 1:1-2:4a] and the God of Israel, or the God of positive
revelation [yhwh, 2:4b-3:24]."12
Therefore the features of the text invite us to read verse 4 as a unit, and
to start a new sentence at the beginning of verse 5. However, is it a post-
script to the first pericope,13 or a heading to the second? The simplest
answer is that it introduces what follows: that is the function of the toledot
"generations" phrases throughout Genesis (cf. 5:1; 6:9; 10:1; 11:10, 27;
25:12, 19; 36:1, 9; 37:2),14 but so long as the communicative function is
observed (i.e., as an invitation to read the two narratives as complements)
it does not matter much.15
why," TynB 46:1 (1995) 117-140, at 138-40; Stordalen, "Genesis 2, 4," 169-75; Alviero Niccacci,
The Syntax of the Verb
in Classical Hebrew Prose (JSOTSS 86, Sheffield:
Press, 1990 [Italian original, 1986]), 200 n.26; Gordon Wenham, Genesis 1-15 (Word Commentary;
Magnes, 1961 [Hebrew original, 1944]), I: 98-99.
9 For example, the chiastic word order of Genesis 1:5 expresses simultaneous naming of the
light and darkness. A good discussion of the chiasmus appears in H. Van Dyke Parunak, "Oral
typesetting: Some uses of Biblical structure," Bib 62:2 (1981) 153-68.
10 Collins, "Wayyiqtol as pluperfect," 139.
Stephen Kempf, "Introducing the Garden of Eden: The structure and function of Gene-
sis 2:4b-7," JOTT 7:4 (1996) 33-53, acknowledges the chiasmus but supports the division of
the verse on the basis of, among other things, the change in divine name (at 39-41).
12 Franz Delitzsch, A New Commentary on Genesis (Edinburgh: T. & T. Clark, 1888), 114. Cf.
also Cassuto, Genesis, I: 86-88.
13 As preferred recently by Alviero Niccacci, "Analysis of Biblical narrative," in R. D. Bergen,
ed., Biblical Hebrew and Discourse Linguistics (Dallas: SIL, 1994), 175-98, at 184 (but without
14 Cf. Wenham, Genesis 1-15, 49; Delitzsch, Genesis, 110. This agrees with the paragraph
marker of the Masoretic text.
15 As Cassuto noted, Genesis, I: 99. Hence the suggestion that the difference in conclusion
between Niccacci and myself on this point is evidence of "the subjective nature of Discourse
Analysis" is a drastic overstatement (Joseph Pipa, Jr., "From chaos to cosmos: A critique of
the non-literal interpretations of Genesis 1: 1-2:3," in Did God Create in Six Days?, 153-98, at 179;
INTERPRETATION OF GEN 2:4-7 273
Next we must assess the structure of the whole pericope. Here is where
the discourse grammar of the verb can help us. In a Biblical Hebrew narra-
tive, the function of the wayyiqtol verb form (also improperly called "the
waw-consecutive with imperfect") is as "the backbone or storyline tense of
Biblical Hebrew narrative discourse."16 Hence, if we want to find the main
sequence of events in a narrator"s presentation,, we should begin by looking
for the wayyiqtol verbs.17 Other verb forms are used for supplying background
information: e.g., the "perfect" (qatal) is used to denote events off the
storyline, while the "imperfect" (yiqtol), "converted perfect" (weqatal), and
participle (qotel) denote background activities with process aspect ("some-
thing was happening").18
From this we can see that the storyline begins in verse 7 with the first
wayyiqtol verb (wayyiser, "and he formed"). Verses 5-6 are syntactically
background, or setting, for verse 7: with verbs describing what had "not
yet" happened in verse 5, and then verbs with process aspect in verse 6
(ya’aleh, RSV "went up," better "was going up"; wehisqa, RSV "watered,"
better "was watering") describing what was happening when the action of
verse 7 took place.19 This yields a structure as follows:
2:5-6 Background/setting--specific circumstances for following events
2:7-9 Events: formation of man, planting of garden, placing of man
2:10-14 Excursus: the four primeval rivers20
cf. Benjamin Shaw, "The literal day interpretation," in the same volume, 199-220, at 200 n.3).
The difference is simply one of literary judgment, and does not impinge on the validity of the
methodology (especially since Niccacci does not interact with the alternative).
16 R. E. Longacre, "Discourse perspective on the Hebrew verb: Affirmation and restate-
ment," in Walter Bodine, ed., Linguistics
and Biblical Hebrew (
1992), 177-89, at 178. Cf. also Randall Buth, "The Hebrew verb in current discussions,"
JOTT 5:2 (1992) 91-105.
17 Of course, since this verb form can be used for imbedded storylines, we cannot mechan-
ically identify the occurrence of the verb form with this function.
18 "Process aspect" has a number of contextually inferred nuances, such as habitual action,
repeated action, one action in process, inceptive action.
19 This analysis agrees with Niccacci;, "Analysis of Biblical narrative," 187; cf. Wenham,
Genesis 1-15, 46 n.5a. Futato, "Because it had rained," 2 n.5 and 5-6, argues that verse 7 is part
of the background with verses 5-6. However, this is unsatisfactory because (1) the wayyiqtol
sequence begins in verse 7 and runs through verse 9; and (2) it gives no indication of how the
discontinuity between verses 7 and 8 is detectable. He depends, not so much on the gram-
matical particulars as on his assessment of this section as having a "problem-resolution"
structure, as well as on his interest in supporting a version of the "framework" view of 1:1-2:3.
But if neither of those has independent support, his case loses its force.
20 The verb forms are participles, an imperfect (verse 10 yippared, RSV "it divided"), and
a converted perfect (verse 10 wehaya, RSV "and became"), which have the function of giving
process aspect background with past time reference. That is, these four rivers were flowing etc.,
though they might not flow the same way now (cf. John Munday, Jr., "
erodes flood geology," WTJ 58:1  123-54). The existence of the excursus explains why
verse 15 begins by re-stating the action of verse 8 (verse 8 "there he put the man" . . . verse
15 "the Lord God took the man and set him"): it resumes the narrative after a digression. This
2:15-17 Events: God establishes terms of relationship with man
2:18-25 Events: formation of complementary helper
Peak: verses 23-2421
III. Harmonizing with Gen 1:1-2:3
Since Gen 2:7 recounts the formation of the first human (cf. verse 6 which
says there was not a human up to this point), we cooperate with the author
by taking it as complementary to 1:27. In doing so we note that the forma-
tion of the woman, which is given in the same verse in the broad stroke
account of chapter 1, is in chapter 2 separated from the making of the man
by several events. The making of the woman is preceded by a declaration
of "not good" in 2:18, indicating that at that point we have not yet come
to the "very good" status of everything in 1:31. We note further that
Gen 2:19 describes the formation of the animals.22 All of this suggests that
the storyline events of 2:5-25 are events of the "sixth day" of 1:24-31.23
This being the case, it makes sense to see if we can interpret 2:5-6 in a
simple way as background to the events that begin in verse 7. Can we
cooperate with the invitation of verse 4, to read the two pericopes as com-
plementary, in a way that is consonant with the grammar and the lexicon?
We can if we take note of several factors. First, we note that discourse-
oriented exegesis shows that the "days" of Genesis 1 need not be the 24-
hour kind, and that hence the events of the sixth "day" could be some
number of weeks, years, or even longer after the beginning of the creation
week in 1:3.24
We note further that the semantic range of Hebrew 'eres in verse 5: is it
"earth," "land," or "region"? It is fairly common to take 'eres as "earth"
(cf. RSV, NASB, NIV text), and to find in this a description of the condition
of the whole earth.25 But the word quite often means simply "land" (cf.
explanation of verse 15 hardly supports Futato's contention that the narrative of Genesis 2 is
not governed by chronological concerns ("Because it had rained," 11-13; Futato takes verse
9-14 as an expansion of verse 8a).
21 The "peak" is the place of maximum interest in the narrative, and here it is marked by
(1) the poetic and rhetorical features of verse 23; (2) the fact that the project of verse 18 is now
complete; (3) the enduring consequences described in verse 24; and (4) the viewpoints of the
chief characters, man (verse 23) and God (verse 24) being clearly stated.
22 See Collins, "The wayyiqtol as pluperfect," 135-40, justifying the interpretation of 2:19,
"The Lord God had formed from the ground every animal of the field and every flying thing
of the sky," i.e., this formation actually took place before the making of the man, as recorded
in Genesis 1.
second pericope as an elaboration of the sixth day (see his note 7 for evidence).
24 See Collins, "
25 If we reject the view that there are two competing creation accounts, we have to decide
what to do about it not having rained: are we to suppose that rain did not fall until the flood?
INTERPRETATION OF GEN 2:4-7 275
NIV margin), either as dry land (its sense in 1:10-31) or as a specific region
(its sense in 2:11-13), where God made man prior to moving him into the
The discourse relation of verses 5-6 to verse 7, as the setting for the events
of verse 7, makes the latter line of interpretation the simplest: that is, in a
particular year, at the time of year before the rain fell to water the ground
when the "mist" (or perhaps "spring"?)28 was rising (possibly beginning to
rise),29 in some unspecified region, God formed the first human, planted a
garden, and then transplanted the man to this new place to enjoy it and
care for it. This interpretation has the advantages of (1) following directly
from the discourse relations; (2) using ordinary meanings of words; and
(3) being easily harmonious with Gen 1:1-2:3.30
IV. A Revised Translation of Gen 2:4-8 (with notes)
We may modify the RSV given above to reflect this analysis (I have
included several philological comments as annotations):
(4) These are the generations
of the heavens and the earth when they were created,
an argument for "region."
27 Cf. John Bimson et al., .New Bible Atlas (Downers Grove: InterVarsity Press, 1985), 14-15.
Futato, "Because it had rained," 3, notes this fact but does not apply it in the same way as
28 Hebrew ‘ed, no one knows for sure what this word means. I see no reason to dissent from
Delitzsch, Genesis, 117, who argues on the basis of Job 36:27 (its only other occurrence in the
Bible) and an Arabic cognate for the sense "mist" (i.e., "condensed vapor"). The Septuagint
rendered the Hebrew with phgh<, "spring," and comparative evidence may favor something
like "flood" (cf. Wenham, Genesis 1-15, 58; Victor Hamilton, Genesis 1-17 [New International
Commentary on the OT;
this evidence decisive. Futato, "Because it had rained," 5-9, argues strongly and, I think, success-
fully, against the "flood"/"stream" interpretation and in favor of something like "rain-
cloud," i.e., along the lines of Delitzsch.
that this is the Targum and Rabbinic tradition. (Futato does not like the rendering "mist"
because according to his definition the English word does not quite match the meaning
"rain-cloud"; but it seems clear that Delitzsch means something close to the sense for which
29 It is possible to infer from the context that the particular nuance of the process aspect is
inceptive action, "it was beginning to go up. .. it was beginning to water." Waltke and O'Connor,
Hebrew Syntax, §31.2c, support such a possibility, but their examples are not all persuasive
(2 Sam 15:37 is the best, cf. RSV).
30 This harmonization, by the way, which began with the possibility that the "days" were
not the 24-hour kind, seems actually to, favor the likelihood that they represent longer spans
of time. The passage itself supplies an explicit reason why the vegetation had not grown,
namely the absence of rain and man (verse 5). In order for this to hold there must be some
lapse of time longer than a few days.
in the day that31 the LORD God made earth and heaven.
(5) When no bush of the field32 was yet in the land and no small plant of the field
had yet sprung up--for the Lord God had not caused it to rain on the land, and
there was no man to work the ground, (6) and a mist was going up33 from the land
and was watering the whole face of the ground--(7) then the Lord God formed
the man34 of dust35 from the ground,36 and breathed into his nostrils the breath
of life, and the man became a living creature.37 (8) And the Lord God planted38
a garden in
31 I.e., "when the Lord God made." For this meaning of beyom followed by an infinitive
construct see Brown-Driver-Briggs, 400a; P Jouon and T. Muraoka, Grammar of Biblical Hebrew
(Rome: Editrice Pontifico Biblico, 1993), §129p A.2.
32 Futato, "Because it had rained," 4, argues that the only legitimate interpretation of siah
hassddeh "bush of the field" must be "wild shrubs of the steppe," in contrast to the seb hassadeh
"small plant of the field," which he takes to be cultivated (a possibility mentioned in Kidner,
"Genesis 2:5, 6," 109).
33 Taking the verb ya'aleh, as most do, as a simple Qal imperfect with 'ed as subject. Futato,
"Because it had rained," 8, argues that we should interpret the verb as a Hiphil imperfect
with the Lord God as subject of both it and the next verb wehisqa "and it was a mist that he
[i.e., the Lord God] was bringing up, and he was watering." This is possible, but the reasons
he offers do not settle the question. (I) When the first element of a clause is not the verb, as
is here the case, it is more expected for that element to be the subject. (2) A noun in the
semantic category "mist/cloud" can as easily be the subject of the Qal of the verb (cf.
1 Kgs 18:44) as the object of the Hiphil (cf. Ps 135:7). (3) Similarly, it is quite proper for a
source of water to be the subject of the next verb, wehisqa "and was watering" (as it is in verse
10). The credibility of his argument that "God would be the explicit solver of both the
problem of no rain and the problem of no cultivator" (8-9) depends in turn on the prior
acceptance of his literary reading for the text, and hence cannot establish that reading in
opposition to others. Hence I see no reason not to translate this in the usual way.
34 The Hebrew has a definite article ha'ddam, "the man," namely the first human. I would
take the article as anaphoric to the mention of "man" in verse 5, which does not have the
article: literally, "and as for man, there was none to work." Without the article it becomes the
proper name Adam in verse 20. In verse 23, using different terms, the "woman" ('issa) is taken
from the "man" (is).
35 That is, loose soil.
36 Many suppose that there is a play on words here: "human" is 'adam, while "ground"
is 'adama, from which man was made and now to which lie will returns because of sin (cf. 3:19).
Since, however, in verse 19 God also "forms" the animals "from the ground," we must not
push this too far.
37 It is difficult to give a good literal translation of this term (nepes hayya, traditionally
"living soul": cf. 1:21, 24, 30; 2:7, 19) and still have elegant English: "living animated being"
would be the closest. Delitzsch, Genesis, 94, points out that since a nepes (often rendered "soul")
animates a body, the expression denotes "animated material beings, bodies having souls." In
I Cor 15:45, Paul employs the LXX rendering of this expression, yuxh> zw?sa ("living soul")
in the sense of "living natural being," to contrast with the supernatural life he denotes by
nvefµa ("spirit") in verses 44-46.
38 In Collins, "The wayyiqtol as pluperfect," 140 n.75, I reject the NIV's making this
pluperfect: "the Lord God had planted." I think the end of the verse, 'aser yasar "whom he
had formed," places the formation of verse 7 prior to the planting of verse 8.
This material is cited with gracious permission from:
report any errors to Ted Hildebrandt at: