The Genesis story of the Flood is a crucial text for those wishing to develop a non-Darwinian understanding of Earth history. The catastrophe was the biggest geological event of all time, purportedly destroying the land and annihilating every creature that lived on the land. However, many scholars see discrepancies in the internal chronology of the story and take these as evidence that the text was composed from different, mutually conflicting sources (the ‘documentary hypothesis’). If they are right, it follows that the text cannot be accepted as a reliable record of an ancient and faithfully preserved tradition of a historical event; either the tradition was garbled or it was completely factitious. On the other hand, if the problems can be resolved, the credibility and historical importance of the Genesis story will only be enhanced. The allegations need to be addressed, for the stakes are high.

Perceived inconsistencies and previously proposed solutions

The normally understood sequence of events may be summarised as follows (Kidner 1967).

Set out like this, the narrative may seem straightforward enough, but to some readers it is complicated by the following inconsistencies:

1. The flood is said in 7:17 to have lasted 40 days, in keeping with the warning to that effect in 7:4, whereas the recorded calendar dates (column three above) suggest a duration of up to 370 days (assuming months of 30 days and a year of 360 days).
2. In 7:12 the rain is said to have lasted 40 days and nights, whereas 8:2 suggests that the rain was not restrained until after 150 days.
3. 8:13 says that the earth became dry on the 1st day of the 1st month, whereas 8:14 states that it was dry on the 27th day of the 2nd month.

To these may be added a fourth possible inconsistency which usually passes unnoticed:

4. The birds were sent out ‘to see if the waters had abated from the face of the ground’ after the hill tops had already become visible.

It may seem justified, then, to conclude that the discrepancies are evidence of the story’s having been cobbled together from sources with at least two different accounts of how events unfolded.

This is not a conclusion we should jump to with any alacrity, however. To suppose that the discrepancies are real is to impute a high degree of incompetence to the supposed compositor and, by the same token, to run a high risk of the imputation bouncing back on oneself. For it might not be the text that lacks sophistication so much as our reading of it. Is it plausible that any writer wishing to convey the impression of an historical event, or any subsequent editor, would have tolerated the imputed inconsistencies? Could he not make up his mind about which was the more reliable of his sources, sources for whose existence we have no independent evidence? As noted below, the narrative is the product of immense elaboration – in some ways more elaborate, more worked, than literary story-telling is today. Is this sophistication compatible with a failure to ensure that the final product constituted a seamless whole?

Adolf Dillmann, Niels Lemche and Frederick Cryer have all attempted to explain the discrepancies from within the documentary hypothesis, and they, in turn, have been critiqued by Lloyd Barré, who also operates from within it. All explanations have involved radical departures from a natural reading of the text.

In Dillmann’s solution, the statement (8:3) that the waters continually receded and were abated after 150 days is taken to refer to a second 150-day period, not the 150 days which terminate with the date ‘17th day of the 7th month’. Thus in the first period, the waters rise; in the second, they fall. Unfortunately for this assumption of literary symmetry (and not a symmetry which real events could be considered likely to follow), the text omits the date which terminated this second period. However, it does, in Dillmann’s view, specify the midpoint, namely the 1st day of the 10th month. This too is open to question, since the date is one day short of the midpoint, and one might suppose that the phenomenon associated with it – the first appearance of the hilltops – would be conceived as occurring earlier than midway through the recession period. Indeed, according to the text itself it isn’t midway. The two dates given for the drying of the land are not 76 days but 90 and 146 days later, so that Dillmann must conclude that these are editorial additions, or alterations, to the story. A final difficulty is that the supposed midpoint is not balanced by any similar midpoint in the first period (the final disappearance of the hilltops as they became submerged). All in all, Dillmann’s solution only replaces one set of perceived inconsistencies with another set.

Lemche’s argument relies on a contention that Genesis 7:6 (’Noah was 600 years old when the flood of waters came on the earth’) conceals an original tradition (’P') that the flood began on the 1st day of the 1st month. The flood, ending on the same day in the 601st year, then lasted exactly one year – assumed to be a solar year of 365 days. Lemche further postulates that, according to another tradition (’J'), the ark – a vessel measuring 135 metres long, 22 metres wide and 13 metres deep – was built in just 40 days. Evidence for this chronological detail is lacking because it was edited out. In the J tradition the flood lasted 148 days and began on the 17th day of the 2nd month. Later, a redactor blended the two traditions together and, wishing to translate P’s solar year to his own 354-day lunar year, added 10 days so that the flood ended on the 27th day of the 2nd month. Unfortunately, he could not count, since, as Barré observes, the difference between 365 and 354 is 11 days, not 10.

Cryer’s principal contribution to the complexities is to postulate that one tradition had the flood lasting 360 days, another 370 days, and that the sum of these, being exactly equal to two solar years, accounts for the ‘perplexing’ reference to two years after the flood in 11:10.

For most of the way Barré follows Dillmann’s analysis. Agreeing that there are two 150-day periods, he considers the problem of how to account for the 14-day gap between Dillmann’s missing date, i.e. the 17th day of the 12th month when the second period came to an end, and the date preserved in the text at Genesis 8:13. His solution is to assign the sending forth of the raven to J and the two sendings forth of the dove to P, so that the latter sequence immediately follows P’s second 150 days. Since the word ‘another’ in 8:10 implies that there was a 7-day interval before the first sending forth of the dove (after the release of the raven), and this does not accord with the solution, he proposes that this word was not part of the original P tradition. It then follows that two intervals of 7 days passed between the first sending forth of the dove and the third, with the latter coinciding with the end of the flood on the 1st day of the 1st month, in Noah’s 601st year. It may be doubted whether this multiplying of untestable hypotheses amounts to a solution. Perhaps the biggest problem is that it loses sight of its own premise, namely that according to P the flood ended on the 17th day of the 12th month. It solves a problem that never existed.

The literary approach

Another contributor to the discussion is Gordon Wenham (1987). Wenham points out that parallels with Mesopotamian flood stories from the early second millennium make it unlikely that the Genesis narrative was cobbled together from independent Hebrew versions. Either the biblical account was a rewriting of the Mesopotamian account or each went back to a tradition common to both. Literary analysis indicates that the dates and periods of the Genesis text fit neatly together without the need to postulate different Hebrew sources.

Amongst scholars who take this literary approach, the prevailing view is that the text exhibits some kind of chiastic (mirror-like) structure, in which corresponding verses fan out from the hinge phrase ‘And God remembered Noah’. Wenham’s scheme (1978) is probably the most elaborate of those proposed, and all have been criticised by J. A. Emerton. Emerton’s principal criticisms are: (a) there is no symmetry in the length of the paired units, which varies from half a verse to several verses; (b) the schemes are selective, omitting phrases and sentences which are at least as significant as those which are included; and (c) some of the pairings are forced. He therefore rejects the literary analysis as too subjective: although such analysis deals with the text in its final form rather than with its hypothetical antecedents, it does not, in practice, deal with, and account for, the whole text.

Both the ‘documentary hypothesis’ and the ‘literary analysis’ schools place a high premium on formal symmetry and seek to establish their arguments by selecting parts of the text at the expense of the whole. The one concerns itself with the evolution of the text, the other with the final form of the text. Despite Wenham’s raising the possibility that the biblical and Mesopotamian stories might derive from a common oral tradition – in which case neither a ‘documentary hypothesis’ nor a ‘literary analysis’ approach does the text justice – neither perceives much relevance in asking whether Genesis is describing a historical occurrence: whether the flood actually happened.

That, nonetheless, is the all-important question, and an appreciation of a text’s literary qualities need not be incompatible with a recognition of its historiographic nature (Younger 1990). History writing has been regarded as a literary art without any sense of incompatibility until quite recent times. If our outlook is not to be limited to a peculiarly modern concept of historiography, we need to consider the possibility that a text’s formal and rhetorical qualities – chiasmus, repetition, variations of vocabulary and so on – may simply be devices of a skilful story-teller, aspects of the art of capturing memorable events memorably.

That a basic form of chiasmus has been employed to enhance the intrinsic symmetry of the events seems undeniable. Wenham sets out this structure as it relates to the chronological intervals as follows:

The reality of a deliberate pattern seems confirmed by the suppression of the first 7-day period, obliquely referred to in 8:10, in which the dove searched for a resting place. To have overtly mentioned this period in addition to the others would have disturbed the symmetry. But even here the symmetry is not perfect. Whereas the initial pair of sevens refer to the same period, the corresponding pair in 8:10-12 are consecutive. Similarly, as Wenham recognises (1987, p 187), there is in fact only one 150-day period.

The chronology makes it clear that the 150 days mentioned in both scenes 4 and 5 [comprising 7:5-16 and 7:17-24] fall between the 17.2.600th day of Noah’s life and 17.7.600th day, i.e., a period of about five months. In others words, although the waters appear to triumph for 150 days, they were actually falling well before the period elapsed, or else the ark would not have grounded on 17.7.600. Presumably, then, we are meant to understand that God remembered Noah and blew this wind long before 17.7.600. Yet to an ordinary observer, the waters appeared to be triumphing throughout this time.

That the wind must have begun to blow long before 17.7.600, despite its being mentioned after the 150 days are first mentioned, is a point to which we shall return. For the moment, the crucial thing to note is that the literary structure is not identical with the chronological structure. As shown by the artful but ‘imperfect’ symmetry of the pairs of sevens, the author/editor has resisted the temptation to amend the true chronology in the interests of an exact chiasmus. That is especially evident in the fact that the first 40 days of the flood are not balanced by a concluding 40 days.

Flood distinct from ‘cataclysm’

The alleged discrepancy between the chronology of 7:17, which allocates 40 days to the ‘flood’, and the chronology of the whole narrative, which shows that it lasted some 314 or 370 days, is an artefact of translation. In English a flood is an inundation which ends only when the land is again above water. The word may therefore properly designate the entire period from 17.ii.600 to 27.ii.601. In Hebrew, however, the word translated ‘flood’ – mabbul – designates the initial 40 days of destruction only, when the mountains were shattered and water engulfed the earth and every creature on the earth was obliterated.

The word is used thirteen times in the Old Testament: twelve times in Genesis 6-11, and once in Psalm 29:10. In every case the word refers exclusively to the event which blotted out all living things in the days of Noah. Other words were available if an author wished to refer to floods within the compass of ordinary experience, such as nachal (Job 28:4, Ps 74:15, Jer 47:2) and ye’or (Jer 46:7f, Amos 8:8, 9:5). Mabbul was never used in a general sense.

The twelve occurrences of mabbul in Genesis may be analysed as follows:

After verse 7:17, stating that the mabbul was upon the earth 40 days, the term is not used again until 9:11, when the narrative looks back, and it is not used at all in describing events after the 40th day. Instead the author chooses the term ‘waters’: the waters prevailed, were assuaged, decreased, were abated, were dried up. To translate it by the word ‘flood’ is thus misleading, since the latter designates an inundation which ends with the draining away of the waters and is not reserved for events of extreme violence and destruction. A better translation would be ‘deluge’, or ‘cataclysm’, the equivalent of the Greek term κατακλυσμος by which mabbul is rendered in the Septuagint and the New Testament. ‘Cataclysm’ connotes a sudden and violent outbreak of waters, and this was largely spent by Day 40. The continual rain, which was the most perceptible expression of the mabbul (7:4, 12), had ceased.

The range of the cognate Akkadian word, abubu, is similarly restricted, but the situation is not quite so straightfoward, since the Sumerians conflated the primordial cataclysm with a violent flood that overwhelmed much of southern Mesopotamia at the end of the Jemdet Nasr period (c. 2900 BC in radiocarbon dating). Consequently, when they referred to the abubu, they did so in terms which were a mixture of what they knew from tradition and what they had themselves experienced, and the event as portrayed took on something of the non-global character of the later inundation. Nonetheless, they still looked back on it as a unique event, as the Deluge, and the word referred uniquely to it. Time was divided into lam abubu and arki abubu, before the Deluge and after it. The word came from the verb ebebu, ‘to purify or clean,’ a reflection of the fact that the effect and purpose of the cataclysm had been to purge the Earth of its moral pollution, not, as in later myth, to solve a problem of overpopulation.

Even as a metaphor, it denoted power of extraordinary violence, with specific reference to the cataclysm. In the Gilgamesh epic there is a demon called Humbaba who guards the forest of the gods (a Sumerian version of the garden of Eden), and ‘his voice is the Deluge (abubu), his speech is fire, and his breath is death’. When Marduk in Enuma elish is appointed king of the gods and the champion who will fight the monstrous Tiamat (representing the Sea), the gods give him the Deluge for his most fearsome weapon, and the ensuing battle with the dragon Tiamat and her demonic allies is a dramatisation of the Deluge, after which the Earth is recreated. Occasionally gods were pictured standing over the dragon and the abubu, the latter represented as a lion with raptor-like legs and wings (denoting the ferocity and suddenness of the onslaught). And Assyrian armies, similarly, boasted of overwhelming their enemies like the Deluge. The abubu had been a cosmic battle, a day when God had confronted and punished his enemies and had shown his overwhelming power in its waters.

Enuma elish is the Mesopotamian counterpart of a similar Canaanite myth, the Epic of Baal, where El (’God’) appoints his son Yam (’Sea’, the equivalent of Tiamat) as king of the gods in place of his son Hadad (or Baal, the equivalent of Marduk). But in the ensuing battle Hadad defeats Yam and obliges El to accept him as king of heaven. As in Enuma elish, the battle is a polytheistic dramatisation of the Deluge, where Hadad in effect usurps the kingship of heaven. Psalm 29 is a call to remember who God really is and who is really king. Whereas Israel was continually tempted to follow the heresies of the nations around her and worship the sons of God instead of God himself, David exhorts the host of heaven to ascribe glory to Yahweh (not Hadad), for it is he who as king of the universe thunders over the waters, flattens the forests and makes the mountains quake. It is he who is enthroned above the abubu beast, the ‘many waters’ that once became the instrument of his wrath.

The narrative structure

In narrative the basic unit is the paragraph. Where dates and periods are recorded, these typically begin the paragraph, and the remainder of the paragraph then describes what occurred during the period specified. For example, in I Kings 15-16, which summarises the reigns of successive kings of Judah and Israel, each king is introduced with the date in which he came to the throne and the duration of his reign, and the succeeding sentences describe significant events that occurred during his reign. The opening statement outlines the chronology of what is to follow. Another example is the first verse of Genesis. The statement ‘In the beginning God created the heavens and the earth’ is a summary of the events which the rest of the chapter then proceeds to narrate in detail. Much the same structure governs the Flood narrative.

The first paragraph begins at 7:11. “In the 600th year of Noah’s life, in the 2nd month, the 17th day of the month, on that day all the fountains of the great deep burst forth and the windows of the heavens were opened, and rain fell upon the earth 40 days and 40 nights.” The succeeding sentences then focus on the first of those days, describing the completion of the task of herding all the animals into the ark on that same day, until Noah and his family boarded the vessel themselves and Yahweh shut the door behind them.

The second paragraph (to some extent expanding on the first) begins at 7:17. “The cataclysm was 40 days upon the earth; and the waters increased and bore up the ark, and it rose high above the earth.” The succeeding sentences elaborate on this summary. Verse 18 emphasises that the waters increased greatly during the period and the ark began to float. Verses 19-20 state that the waters increased so greatly that every mountain under the entire heaven was covered. And verses 21-23 describe the consequences for the creatures on the earth: they perished, without exception, and were totally obliterated. Only those in the ark survived.

The third paragraph begins at 7:24. “And the waters prevailed upon the earth 150 days.” As verse 8:4 makes clear, this period is inclusive of the initial 40 days. In other contexts the word ‘prevail’ (gabar) means to have mastery over, have the upper hand, be stronger than (cf. Ex 17:11, II Sam 11:23), so that ‘prevail’ here (AV, RSV) is a good translation. The Hebrew does not signify a passive covering of the earth, but active dominance over it. The waters prevail from the very first day, and cease to do so when the land re-asserts itself by emerging above the waters. During these 150 days, but by implication not until after the first 40 days, a wind blows over the earth, the flood waters are no longer multiplied from below, and the rain abates. The waters flow back and forth, subsiding. On the 151st day (this date ending the paragraph) the ark runs aground.

The fourth paragraph begins at 8:5. “And the waters continued to abate until the 10th month; in the 10th month, on the 1st day of the month, the tops of the mountains were seen.” The first 40 days of this period, beginning from the 151st day, end with Noah sending forth a raven. At this point no land is visible. Noah watches the raven flying back and forth but never venturing far enough, or high enough, to detect whatever land might lie beyond the horizon. After a week he sends forth a dove, which looks for land but does not find it, and so returns. The second time, however, she brings back a fresh olive leaf: a sign that the waters were abated (since she was unlikely to have picked the leaf from a branch bobbing in the waves). After leaving the ark a third time, she does not return. Thirteen days later land is seen from the ark.

The fifth paragraph, at 8:13, is a short note recording that by the 1st day of the new year the waters had receded to the point where all land within view appeared dry.

Finally, the sixth paragraph, beginning at 8:14, records that by the 17th day of the 2nd month of the new year, the earth – not merely the surface of the ground – was dry. It was dry enough to walk on, and Noah at last receives the order to leave the ark.

Implications for the historicity of the event

In this analysis there are no longer chronological discrepancies. The chronology of the events works out as follows:

The first 40 days are the period in which all land becomes submerged and all terrestrial life perishes; it is distinct from the total duration of the flood, which is 370 days. The cataclysm is accompanied by 40 days of unceasing rain, and since verse 8:2 is within the paragraph relating what happens up to Day 150, it no longer appears to say that the rain abated only after Day 150. Finally, the birds are sent out before, not after, land becomes visible.

The chronology, moreover, is consistent with the events being related. (Contrast the exotic idea that eight adults could have built a wooden vessel 130 metres long in 6 weeks.) In particular the following points should be noted.

The entire earth was overwhelmed with water and destroyed (6:13) in just 40 days – a comparatively brief period which began when, all on the same day, the earth convulsed with the eruption of the fountains of the deep. These were terrestrial springs supplied by the waters underneath the earth (the dry land of 1:10) (Tsumura 1989, Seely 1997), their eruption releasing the waters in the same way as a volcano releases magma. A cataclysm of unimaginable ferocity is implied, simultaneously affecting every part of the earth and shattering its whole fabric, possibly destroying even mountains (as foretold for the corresponding day of wrath at the other end of history, e.g. in Isa 40:4). The narrative describes an event quite different from floods within normal experience, both in origin and in scale. It was not a tsunami hitting low-lying coastlands, or a river bursting its banks, or the undamming of a lake – disasters in which there are always some survivors and which leave other regions and lands untouched. The inundation came, first and foremost, through the eruption of water from below, affecting ‘all the high mountains under the whole heaven’. Accompanied by a continual downpour of water from the heavens, it resulted in the death and obliteration of every animal living on the earth, every flying creature, and all mankind, without exception.

Whereas the period of destruction was concentrated into just 40 days, the period in which the earth was under water was possibly as much as 200 days, since the first definite mention of emergence is the appearance of the hilltops on Day 225 (1.x.600). Again, it was no ‘normal’ flood that could have sustained total inundation for over six months. Such a duration is, however, consistent with the magnitude of the initial cataclysm. The water level falls gradually: the ark runs aground on Day 151, land first emerges above water after a further 74 days, and is completely above water only after a further 90 days.

The narrative is also credible in its portrayal of human actions. Noah, it seems, was told beforehand only the length of the period in which the earth and its inhabitants would be destroyed, not how long the waters would persist on the earth after the rains had subsided. After 6 months in the ark he was becoming restless. He had received no word from Yahweh since Day 1, and although the ark had run aground, land was still not visible. Anxious to find evidence that the inundation was coming to an end – further off if not in his immediate vicinity – he sent out a raven. But the bird frustrated the intention of the inexperienced seafarer, as animals often do. So he tried again, this time with a dove.

Finally, the report in 8:13 that the earth is dry is repeated in 8:14, but with a different date. Followers of the documentary hypothesis see the repetition as evidence that two traditions have been ineptly pasted together – ineptly, because the traditions have not been fully amalgamated and the same moment is dated as if it were two different ones. However, if an author gives different dates for two similar events, the presumption must be that he intends the reader to understand that the events were not the same, albeit similar. In verse 13 what is being recorded is apparent dryness: hence the term ‘behold’, and the reference to the surface of the ground. In the next verse we are told simply that the earth was dry, with the implication that the water table was now low enough and the sediments sufficiently dewatered for animals to walk on the surface without getting bemired. Because of the enormous scale of the flooding a further 56 days had to pass before the animals could disembark with safety.

Summary

Source analysis of the flood story depends on the perception of discrepancies in its chronology. To a lesser extent discrepancies also appear in readings which interpret the sequence of events in relation to a chiastic structure. However, attempts to explain the inconsistencies lead to explanations that are more problematic than the text itself, and fail to do justice to the text as ostensibly recording the tradition of a historical event. That the world had been destroyed in a cataclysm of waters early in its history was the tradition of Hebrews and Sumerians alike. Their word for it was reserved uniquely for that event.

The story is divided into paragraphs, each of which begins with a synoptic statement that sets out the chronology of the events about to be described. Once this is recognised, the narrative makes good sense as the work of essentially one writer, describing an event which, although unlike anything either a Hebrew or a Sumerian could have experienced, is devoid of inconsistencies. The whole earth was inundated and destroyed in 40 days, but almost a year passed before it was safe enough for the survivors to step onto what was once more dry land. Entirely consistent with these overall dimensions, the carefully recorded dates of the Flood’s various stages underscore the magnitude of the disaster.