© Ian Lawton 2000

The texts can be spit into two segments: those written in the Sumerian language, and those written in the Akkadian language. Whilst both used cuneiform script, the former date to before the fall of the Sumerian civilisation around the start of the 2nd millennium BC, the latter after. Furthermore, many of the Akkadian texts which we will discuss are actually copies which were made in the late Assyro-Babylonian period, for example as housed in the library of Ashurbanipal at Nineveh; although these date to as late as the first half of the 1st millennium BC, they are still written in Akkadian, which remained the written and spoken language of trade and diplomacy throughout the Assyro-Babylonian era.

Although a few of the Akkadian texts appear to be almost perfect copies of Sumerian originals, most have no Sumerian original yet discovered. When attempting to trace their origins, the vast majority come under the category of 'mixed heritage', for a variety of reasons. First, many are thought to be older because of their frequent use of Sumerian loan-words. However after the fall of Sumerian civilisation its language survived in the fields of learning and culture, playing a similar role to that of Latin after the fall of the Roman Empire. This cannot therefore be regarded as an acid test. Second, many tablets state that they are copies of 'ancient texts', some even using the wording 'based on the Sumerian original'. However as with the Old Testament of the Bible - much of which is itself an edited version of older texts, including the ones under discussion here - the problem lies in unravelling the original aspects of documents which have been repeatedly copied and selectively edited over the course of centuries and even millennia. This is especially true if one is attempting to verify elements which might contain a grain of literal truth, such as the story of the Deluge.

The reasons for this editing are multiple. It may occur to promote a particular religious standpoint: the elevation of the Babylonian god Marduk from local to national status in later versions of the Epic of Creation is a case in point; similarly most non-Hebrew translations of the Bible use the singular term 'God' throughout to support a monotheistic viewpoint - which is fine as a translation of the Hebrew word Yahweh or Jehovah, but the word Elohim which is used repeatedly in the Old Testament actually means 'gods' plural.[i] Alternatively it may occur for political purposes, for example to elevate a ruler to a 'demi-god' status, or to elevate a particular city. Or the names of places and people may be altered to make the text more interesting and relevant to contemporary readers, just as one might update an old joke. For all these reasons and more, even in the Early Dynastic and Old Assyro-Babylonian periods, multiple copies of texts all dating to the same period can show significant differences - let alone those which stretch across periods of time.

The only mitigation against this widespread editing was the practice of copying ancient texts in order to form library collections which would preserve and respect Sumerian literary heritage. As far as we know this practice started with the Sumerian Shulgi, a ruler of the 3rd Dynasty of Ur c. the 2100 BC, and continued right through to the Assyrian ruler Ashurbanipal whose extension of the Royal Library at Nineveh c. 650 BC provided the many texts first uncovered by Layard. This copying of texts for official libraries involved no political or religious motive to edit and rewrite; rather it demanded accuracy and scholarship. This is attested by the fact that professional Assyrian scribes often inserted blanks, their own commentary, or 'I do not understand' in places where their own source tablets were incomplete. But this still does not mean we can take these 'official' versions as 'true originals', because of course they may have been copied from previously edited older texts.

Be that as it may, one could suggest that any original Sumerian texts are evidently older and therefore more likely to be accurate - both because they were written closer to the time of the events which they describe, and because they would have had less chance to be subjected to the distortion of the editing process. However there are considerable problems even then. The Sumerian written vocabulary was massive, with thousands of words with different signs, and its written grammar was similarly complex.[ii] In particular the same written word can have multiple meanings (providing ample chance for punning, a Sumerian favourite); grammatical prefixes and suffixes can be mistaken for word-syllables; the use of tense varies; conjunctions are rarely used; and nouns tend not to have a gender. In addition, their system of writing developed via a number of stages over several millennia.[iii] It is these facts which ensure that, even as late as 1987, the distinguished Sumerologist Thorkild Jacobsen could write in the introduction to his last book of translations:[iv]

It should not be left unsaid, however, that knowledge of Sumerian is still in a rudimentary, experimental stage where scholars differ on essential points, so that translations, even by highly competent scholars, may diverge so much that one would never guess that they rendered the same text. The reasons for this uncertainty are numerous. The writing is in many respects vague and leaves a broad margin for variant interpretation; meanings of words have not yet been exhaustively defined; and - worst of all, perhaps - scholars have not yet been able to agree on basic grammar and its restraints.

Although a number of Sumerian lexicons have now been published,[v] it is therefore undoubtedly the case that the Sumerian language is so complex that we should be extremely wary of anyone who pretends to be an expert on the subject who is not a dedicated, professional Sumerian linguist. This position contrasts somewhat with the translation of the later Akkadian texts. The understanding of the latter is now so advanced that at least two detailed modern dictionaries are available for the professional translator [vi] - which leaves some, but considerably less, room for debate.

The style of writing in Ancient Mesopotamia, as with any ancient texts, differs from that which we would use today, reflecting a fundamentally different society. Consequently certain nuances of the writing can only ever be guessed at, and some translations by necessity appear highly enigmatic if not downright unintelligible. Furthermore the fact that many epics were composed in a poetic style with much repetition of stanzas and verses, in order that they could be recited to the accompaniment of the lute and harp, does not always make them easy to read.

We can deduce that many of the texts recount epic tales which would have originally been preserved and spread by word of mouth over long periods of time. This practice would have continued even after the first written versions had been prepared. This means that there are many whole texts, and even more so passages within texts, which evidently come from a common (possibly verbal) source but in which details and characters get mixed up and placed in different contexts. A prime example of this is the various versions of the story of the Deluge. So although we must accept that many Mesopotamian epics are highly literary creations, it is perfectly possible that certain common elements of these stories carry grains of truth. This is even more likely to be the case if they appear to recount events which are repeated in the religious writings and legends of other ancient civilisations around the world.

Most orthodox commentators tend to use the term 'myths', especially for the Sumerian texts which principally involve deities, as we will see in the next paper. Although therein I use this term to assist the categorisation, it should be clear that I do not generally support its use, preferring instead the less assumptive word texts. This allows us to appreciate that there are undoubtedly some aspects of even the 'mythical' texts which may deserve a literal interpretation. Of course, working out which aspects is the area of greatest controversy and difficulty.

When I first came to examine the rich literary inheritance which the Ancient Mesopotamians have bequeathed to us, I found it particularly hard to establish the true extent thereof. Many authors tend to quote the odd major work but it can be difficult for the newcomer to place these in context. I have therefore prepared a proper catalogue of the most important works and the periods in which they were written, one each for the Sumerian and Akkadian texts respectively, in the papers that follow.

Source References

[i] For example see Stephanie Dalley, Myths from Mesopotamia (Oxford University Press, 1989), p. 177, Note 11.

[ii] In The Sumerians (University of Chicago Press, 1963), Appendix B, Samuel Kramer describes the Sumerian language: it is 'agglutinative', in that parts of words retain their individual meaning, so monosyllabic words are often combined to form polysyllabic ones (this is usually denoted in our script by a hyphen between the syllables). By contrast, in 'inflected' (for example, Indo-European and Semitic) languages, roots and vowels change - for example to reflect different tenses as in sing, sang, sung.

[iii] In ibid., Appendix A, Kramer describes how the development of writing in Mesopotamia went through 5 stages: 1) In the earliest 'pictographic' script dating to c. 3000 BC, each symbol corresponded to a pictorial representation of the object concerned; this required a large number of complicated symbols, although they were gradually simplified in style and reduced in number by substituting phonetic for ideographic values, i.e. , if one word sounded like another word, the same symbol was used. 2) The second stage involved lying the symbols on their sides, i.e., rotating them through 90 degrees, purely to facilitate the layout of the writing. 3) The third stage is referred to as 'Archaic' script dating to c. 2500 BC, which had begun to look like cuneiform proper in as much as the pictures had become stylised and wedge shapes were used throughout. 4) The fourth stage emerged c. 1800 BC, at which point the symbols were further simplified. 5) Ditto the fifth stage c. 1000 BC.

[iv] Thorkild Jacobsen, The Harps that Onceā€¦ (Yale University Press, 1987), Introduction, p. xv.

[v] Several of these of varying quality and detail are accessible on the Internet. A valuable page with links to Sumerian and Akkadian lexicons and a variety of other useful Mesopotamian sites can be found at

[vi] The Akkadisches Handworterbuch prepared by W. Von Soden, and the Chicago Assyrian Dictionary.