Creation/Evolution Journal

The Flood: Mesopotamian Archaeological Evidence

The assertion of some historians and archaeologists that a great flood devastated a region of Mesopotamia at the dawn of history and that this event was the origin of the biblical Flood story has become a curious backwater in the debate over creationism. The topic has not proved of major concern to either the advocates of recent-creationism or to their scientific opponents. It has, however, given considerable, if probably unwarranted, encouragement to day-age creationists, gap theorists, and those who hope to reconcile apparent contradictions between scripture and science.

Within a few months of one another during the 1928-1929 excavation season, archaeologists at two southern Mesopotamian sites, Ur and Kish, announced the discovery of flood deposits which they identified with the Flood described in the Hebrew scriptures and cuneiform sources. The famous and glamorous Sir Charles Leonard Woolley, after his deep excavations of the Early Dynastic royal tombs at Ur, had a small test shaft sunk into the underlying soil. He persisted through some eight feet of bare mud before finally coming to a layer bearing artifacts of late prehistoric date. It did not take Woolley long to arrive at an interpretation:

I . . . by the time I had written up my notes was quite convinced of what it all meant; but I wanted to see whether others would come to the same conclusion. So I brought up two of my staff and, after pointing out the facts, asked for their explanation. They did not know what to say. My wife came along and looked and was asked the same question, and she turned away remarking casually, "Well, of course, it's the Flood."

[1954, p. 27]


- page 15 -


Woolley's first test pit was very small, so during that and the next season he had dug a number of other test shafts, including an enormous pit, seventy-five feet by sixty feet and sixty-four feet deep. In this main pit, he encountered a deposit of clean, apparently water-laid soil up to eleven feet thick. Evidence of the Flood was absent from several shafts and uncertain or disturbed in a number of others. But in many, Woolley felt he had certain evidence of the Flood (1955).

Just slightly before Woolley's initial discovery, S. Langdon and L. Watelin encountered smaller flood levels at Kish (Watelin, 1934). Although the Kish discovery actually predated Woolley's find at Ur, Woolley published first (Woolley, 1929) and received the lion's share of the initial publicity. Woolley, moreover, produced a highly successful popularization of his work in which the Flood finds were recounted in a manner that is at once simple, authoritative, and filled with references to familiar biblical materials (Woolley, 1929, 1954, 1982). The finds from Ur achieved and maintain a predominant place in the public mind.

Initially, some assumed with great eagerness that the flood levels at Ur and Kish were identical and provided marvelous evidence for a historical kernel of the Genesis Flood story (Peake, 1930), but the enthusiasm could not be maintained. The level of the great flood at Ur was sandwiched between remains of the Al Ubaid cultural phase, the last purely prehistoric period of southern Mesopotamia, and a layer of debris from the early Protoliterate period. The great Ur flood, thus, can be dated with a high degree of certainty to about 3500 BCE. Kish, however, produced evidence of two floods at the end of the Early Dynastic I and beginning of the Early Dynastic II periods, around 3000 to 2900 BCE, and a still more impressive flood dating to the Early Dynastic III period, around 2600 BCE. All three of the Kish floods were much later than the great flood at Ur. Watelin argued that the earliest of these three was the deluge of the Bible and cuneiform literature.

Within a few years, excavations of a third Mesopotamian site, Shuruppak, also uncovered a flood stratum (Schmidt, 1931). It is of particular interest because, according to the Mesopotamian legend, Shuruppak was the home of Ziusudra, the Sumerian Noah. (The Sumerian Ziusudra means "life of long days." The Akkadian equivalent, Utnapishtim, is "he found life," while the alternative Atra-hasis means "exceedingly wise.") This flood level separated late Protoliterate and Early Dynastic I remains and dates from around 2950 to 2850 BCE. Perhaps, but not certainly, the Shuruppak flood may be equated with the earliest flood at Kish. No other Mesopotamian sites have produced flood remains of significance (Mallowan, 1964).

Which, if any, of these floods is to be equated with that recounted in the Bible? Despite the assurances of biblical literalists, no exact date or even close approximation can reasonably be derived from Genesis for the Flood or many other events. Simplistic compilation of patriarchal ages in the manner of the famous Bishop Usher is just not adequate. Crucial Hebrew concepts and terms, even those translated by explicit English words, such as generation, frequently carry in Hebrew a variety of meanings, some of which are neither commensurate with English nor immediately evident. Biblical genealogies, for instance, can and do sometimes contain omissions (Hyers, 1983, pp. 13-15). Biblical materials by themselves are inadequate to distinguish among the Mesopotamian flood strata.


- page 16 -


Mesopotamian flood tales are more useful. Similarities between the account of Noah's Flood in the Hebrew scriptures and the Mesopotamian flood tales are great and obvious. Despite some lesser differences, there is no reasoned body of opinion that claims they are unrelated. The accepted view is that the archetypal account originated in Mesopotamia. The earliest extant Mesopotamian version is far older than the biblical account, and the Flood story bears specifically Mesopotamian details that cannot reasonably be supposed to derive from a Hebrew original. Near Eastern scholars have consequently turned to the cuneiform sources.

The most well-known and detailed Mesopotamian account of the Flood is found in the Gilgamesh Epic (Tigay, 1982, pp. 214-240; for other accounts, see: Lambert and Millard, 1969; Kramer, 1967). Even this account, however, seems to have been somewhat abbreviated because of the literary role that it plays within the broader story of Gilgamesh's confrontation with mortality. Closely parallel are the lengthy but, in part, ill-preserved accounts in the Atra-hasis Epic and the shorter and incomplete Sumerian Deluge Myth. Briefer references to the Flood serve as prefaces to several other myths. Myths are frequently introduced by an abbreviated account of some monumental mythic event, such as the Flood or creation itself. There are other scattered fragments, and a version of the Mesopotamian Flood tale even survives in the sadly incomplete fragments of the writings of the Babylonian priest Berossus, who lived in the late fourth and early third centuries BCE (Lambert and Millard, 1969; Kramer, 1967).

The Sumerian King List also contains a reference to the Flood (Mallowan, 1964, pp. 67-69; Kramer, 1967, pp. 12-13). The King List is a complex document, existing in a number of different editions. Probably first composed about 2100 BCE and extant in an edition from about 1900 BCE, the King List purports to record the kings and dynasties of Mesopotamia from the time when first "kingship descended from heaven" until the time of composition. The list has many weaknesses. Early kings are credited with reigns of such fabulous length that Methuselah's span seems reduced to insignificance, and a number of early dynasties that were in fact contemporary are listed as if they were sequential. Despite these defects, the Sumerian King List appears to preserve the names and sequences of many early real rulers, a number of whom are independently attested elsewhere. The King List claims that, after a number of antediluvian rulers, the Flood swept over everything, after which kingship once again "descended from heaven" and the list of dynasties and rulers resumes. Gilgamesh, hero of the epic, is listed long after the Flood. Thus, the evidence of both the King List and the Gilgamesh Epic, which has Gilgamesh listening to an account of the Flood, agree that he lived well after the Flood.


- page 17 -


Although Gilgamesh appears as a mythic character in later Mesopotamian literature, he was, in fact, a real person, and references to contemporaries and near contemporaries allow the calculation of his date. Scholars generally agree with a high degree of certainty that Gilgamesh lived in the period of 2700 to 2600 BCE (Mallowan, 1964, pp. 67-68). How much earlier should the Mesopotamian Flood be placed? The Sumerian King List names twenty-three rulers of the city of Kish between the Flood and a contemporary of Gilgamesh, but there are good grounds for dividing this list into two nonsuccessive segments and reckoning only eleven generations of kings in the interval. Calculating on the basis of the average reign of Mesopotamian kings, no more than about two hundred years ought to be allocated to these kings, placing the Mesopotamian Flood around 2900 to 2800 BCE (Mallowan, 1964, pp. 68-70, particularly p. 69, n. 21a; Kramer, 1967, pp. 16-18).

The period 2900 to 2800 BCE is much too late to fit Woolley's impressive flood remains at Ur, which must be dated at about 3500 BCE. This period does, however, fit well for the two earliest floods at Kish and a flood level at Shuruppak, and many scholars specializing in the ancient Near East have concluded that the Flood stories of cuneiform literature and the Bible find their ultimate origin in the event attested to by the remains at Kish and Shuruppak (Mallowan, 1964, pp. 62-82; Kramer, 1967, pp. 12-18; Woolley, 1955, pp. 16-17. Woolley's findings were generally rejected by others, including his chief archaeological assistant, Mallowan).


What Role Has All This Played in the Creation-Evolution Debate?


Most recent-creationists simply ignore the entire matter. Presumably, the reason is the one set forth by John C. Whitcomb, Jr., and Henry M. Morris in The Genesis Flood: the Mesopotamian flood remains fail to agree with the literalist view of a universal flood survived only by Noah and family (1961, pp. 109-111). The Mesopotamian strata, whether at Ur or at Kish and Suruppak, testify only to a local flood which clearly left behind survivors and significant cultural continuity. The Ur flood apparently did not even cover the entire mound of Ur. Moreover, fundamentalists have generally demonstrated little interest in the investigation of possible nonliteral explanations of biblical material.

At the other end of the spectrum, the scientific critics of the recent-creationists also have ignored the Mesopotamian materials. They are concerned primarily with answering the arguments of the recent-creationists, who themselves have not emphasized these nonbiblical materials. Generally, the approach of the scientific critics has been to demonstrate the scientific impossibility of recent-creationist claims rather than to attempt to supply alternative explanations of biblical materials.


- page 18 -


A middle ground is held by a wide range of writers—from eccentric catastrophists, through the less-literal day-age, gap, and local-flood creationists, to nonliteralist theologians and secular historians. These groups often accept the equation of the Mesopotamian archaeological finds with the origin of the biblical Flood story. At first glance, this position may appear rational, but, in fact, it is usually based upon religious or other a priori assumptions and, thus, in essentials, is similar to the recent-creationist position. Many in these groups exhibit only superficial acquaintance with the Mesopotamian material—often just that from Ur and usually just through one of Woolley's popularized accounts. They frequently show no awareness of the problems surrounding Woolley's thesis or of alternative interpretations (Thomas, 1966, p. 15; Neil, 1962, p. 32; Hyers, 1983, p. 21, citing Daniel, 1968, pp. 39-47; Hyers, 1984, p. 102).

A few cite the full range of Mesopotamian flood discoveries as confirmation of the biblical Flood story. It is not apparent whether they simply fail to understand that these diverse archaeological discoveries do not pertain to a single event or if they are callously suppressing information that does not conform to their preconceptions (for example, Halley, 1978, pp. 77-80). Others who are primarily concerned with the Mesopotamian sources are well aware of the problems, but nonetheless presuppositions frequently seem to sap their critical abilities. The distinguished scholar Andre Parrot, for example, wrote: "It seems probable, a priori, that a disaster whose magnitude cannot be in doubt must have left traces in the soil of Mesopotamia" (1955, p. 45). The great Sumerologist Samuel Noah Kramer echoes a somewhat similar conviction: "And even among skeptics, there are some who feel that there must be at least a kernel of truth in the Flood-motif; it seems to have played too large a role in Mesopotamian myth and legend for it to have been nothing more than a total fabrication of fancy and fantasy" (Kramer, 1967, p. 13). Actually, there are no compelling reasons to identify any of the floods-at Ur, Kish, or Shuruppak-with the Flood of Mesopotamian literature and the Bible.

Woolley's popularization of his discoveries seems to account for much of the continuing visibility of the Ur flood thesis, but it has little actual claim to be the Flood of Mesopotamian and biblical literature. Despite the thickness of the deposit, it appears like the other Mesopotamian floods to have been a purely local event. Eridu, just seven miles distant, exhibited no sign of the Ur flood, although it was sought diligently there. On about the same or a slightly lower elevation than Ur, Eridu is separated from Ur by only a very low ridge. Equivalent strata at Eridu occupy a higher position on the mound that at Ur, yet no trace of the flood was found at all (Mallowan, 1964, pp. 75-77).

There is, moreover, question of whether memory of an event as early as 3500 BCE could have survived to historic times. The date is too early for a written account to have been made, and the Sumerians do not appear to have had a methodical oral technique that would have long preserved a record of the event. The experiences of other cultures indicate that even the most traumatic events tend to fade from memory after a few generations in the absence of either writing or a highly developed oral procedure, such as formulaic oral poetry.


- page 19 -


The hypothesis that the flood levels at Kish and Shuruppak represent the same event is no more than an assumption. Flood events occurred with frequency throughout southern Mesopotamia, as the two separate early flood levels at Kish indicate. Even more so than the Ur flood, the flood levels at Kish and Shuruppak fail to fulfill the biblical or even the Mesopotamian literary descriptions. In the degree to which those descriptions are "rationalized," any criteria for distinguishing between the biblical Flood and virtually any other flood disappear. The flood remains at Kish and Shuruppak are hardly imposing. The silt at Kish averages less than ten inches thick, and the deposit at Shuruppak is about fifteen inches-in comparison to up to eleven feet of material at Ur (Raikes, 1967, pp. 52-63). The severity of a flood cannot necessarily be deduced from the thickness of an isolated sample of the flood deposit. It is nonetheless suggestive that thicker, more impressive deposits from another flood have been discovered at Kish, dating too late to be identified with the innundation of the Bible and Mesopotamian literature, and yet that later flood left no record in history (Watelin, 1934, pp. 41-43; Mallowan, 1964, pp. 78-79 and plate XX). All that remains is the possibility that the Kish and Shuruppak materials do represent the same event and coincide chronologically with the date of about 2900 BCE for the Flood of Mesopotamian literary tradition.

The flood materials from Ur, Kish, and Shuruppak were excavated over half a century ago. Woolley's description of the flood level at Ur is far from scientific. It is not even possible to be sure of the exact number of sondages in which he found flood remains. While attempts to dismiss the remains of the Ur flood as merely windblown sand are unsubstantiated and probably unsubstantiatable, the two "scientific" examinations of materials from the Ur flood stratum are, by modern standards, vague and inconclusive. The same situation prevails at Kish and Shuruppak (Raikes, 1967, pp. 52-63). In all probability, the finds do represent floods, but the exact character of those events—fluvial or marine, rapid or slow deposition, unitary or episodic—remains unknown. The hydrology of southern Mesopotamia is very complex. Renewed excavation and modern scientific techniques could probably solve many of these questions, but current political and military conditions would seem to preclude any such activity in the near future. Until the situation changes, there are no compelling grounds on which to conclude that the Flood story found its ultimate beginning in an actual event that has been identified at Kish and Shuruppak or anywhere else in Mesopotamia.

The endemic character of flooding in southern Mesopotamia may well have been sufficient to generate the story about a supreme Flood, and the attachment of that story to a specific, long-passed, ill-known historical context may, in fact, be late and unreliable. The earliest edition of the Sumerian King List certainly includes no list of antediluvian kings, and the presence of reference to the Flood is in doubt. It may first have been added much later, during a period in which the Flood story was popular (Civil, 1969, p. 139). Ultimately, the search for a local Mesopotamian flood upon which a rationalization of the Bible story can be based may prove as illusionary as the search for Noah's ark.


- page 20 -


By David MacDonald
This version might differ slightly from the print publication.