The Adaptation of Cuneiform to Akkadian
University of Michigan, Ann Arbor
No one knows exactly how many languages there are in the world today. The most generous estimates suggest that five thousand may be the upper limit, but this is only a remnant of a much richer human heritage. Thousands more languages have been lost to history, and most of them are known to us by name only. Some of these dead languages have survived through the medium of writing in a few parts of the globe, and the largest such cluster comes to us from Western Asia and North Africa, where millennia ago a variety of scripts were used for preserving messages on clay, papyrus, and stone. The first among them was the cuneiform writing system.
Cuneiform was used for 3300 years and perhaps even longer, starting around 3200 BC. What began as a complex but systematic device for registering administrative transactions eventually became a vehicle for the permanent notation of utterances in the Sumerian language. Sumerian writing went through many changes, and in due course became flexible enough to be used as a vehicle for other tongues from many different families, including Eblaite, Elamite, Hurrian, Hittite, Urartaean, and, most important of all, Akkadian. Akkadian belonged to the Semitic language family, which includes, among many others, Aramaic, Arabic, Phoenician, and Hebrew, and itself was divided into dialects, primarily Assyrian and Babylonian. The story of the development of Akkadian writing is complex and involves many issues, only some of which are directly related to technical epigraphic matters.
Cuneiform, or rather proto-cuneiform, was invented around 3200 BC in southern Mesopotamia, perhaps within the walls of the city of Uruk during the latter part of the period that archaeologists call Late Uruk. The origins of the writing system and the historical background of the times has been described above by Hans Nissen and need not be recapitulated here; in what follows I discuss only those matters that have direct bearing on matters of language and writing in early Mesopotamia.
The Late Uruk period was a unique epoch in human history. Within a mere two hundred years or so the social and physical landscape of Southern Mesopotamia was irrevocably altered. The focal point of this process was the city of Uruk and its environs. The city grew into a veritable metropolis, encompassing an area of 250 hectares and housing, according to some estimates, as many as 50,000 people; Uruk at the time was larger than Athens during its glory. The surrounding regions grew accordingly, and surveys have shown that in Late Uruk times the number of surrounding towns and villages grew from 11 to 110, an astonishing increase by any standards. As impressive as these numbers are, they are merely signs of complex social changes. We do not know where the bulk of the new population came from, and we can only surmise the disruptions as well as new opportunities that these upheavals brought to peoples' lives. The novel urban environment, where the intimacy of small village life was exchanged for a more complex fabric of relationships, brought with it new allegiances and new social hierarchies. Individuals now had many different roles, and at different times had to juggle multiple identities as members of families, wards, as well as of the city as a whole. The very size of Uruk created new problems and new opportunities, as individuals were drawn into relationships with strangers in other parts of the city, and a whole new bureaucracy had to be created to manage such affairs. Communication at a distance became a key social fact and this must have had a profound effect on the languages of the time.
We are speaking here of the period immediately preceding the invention of writing and therefore all discussions of language must be speculative. If we do wish to continue in this manner, however, then one other historical circumstance must be considered in any discussion of language behavior in late Uruk times. Just prior to this period, a remarkable expansion of southern Mesopotamian culture can be traced in the archeological record of Western Asia, and even as far away as the Nile Delta. On the trade routes that connected Sumer with Iran, Syria, and Anatolia, there appeared new settlements that were clearly outposts of the Uruk culture. Judging by the excavated material remains, some of them, such as Habuba Kebira in Syria, were made up entirely of Mesopotamian colonists. In other places, such as further up on the Euphrates in Anatolia, one can clearly distinguish distinct quarters for the Mesopotamians and the members of the local culture. A secondary offshoot of this expansion is represented by Buto in the Nile Delta, where the material culture is heavily influenced by Uruk traditions from Syria, but is not identical with that of the other colonies. The true functions of this expansion have been vigorously debated but these issues are of little concern to us here. What does matter is that during the Late Uruk almost all of these places disappear from the archeological record. Whatever the reasons for this apparent collapse of the trade system, it is reasonable to assume that at least some of the occupants of these settlements returned to the heartland, to Sumer. One can reasonably speculate that their own language would have preserved certain archaisms that had disappeared, or been altered in the less conservative environment of the city; but also that they brought with them loan words that they had picked up from the indigenous populations that surrounded the colonies.
There can be no doubt that the Late Uruk period was a time of rapid linguistic change. As a rule, the rates of language change are more rapid in urban environments, and the complex social and historical forces outlined above must have increased the rate of these developments. We must also suspect a large degree of bilingualism, even monolingualism, within much of the population. Even today, in a much different world, long after the rise of the nation-state, it is still estimated that at least half the people of the world are bilingual. All of this may appear self-evident, but in fact many of these socio-linguistic factors have rarely been taken into account in discussions of language change and the development of writing in Sumer. The primary model for language change has been, for many years, the replacement model. Briefly stated, it was believed by many that in far prehistory southern Mesopotamia was occupied by an unknown language group or groups, sometimes referred to as Proto-Tigridians. These people were replaced by a incoming group, the Sumerians, and they, in turn, made way for speakers of Akkadian. This historical process was thought to have been mirrored in writing: the cuneiform system was invented by the first group, quickly applied to Sumerian, and then, in turn, adapted to Old Akkadian.
Attractive as this scheme might seem, it simply will not do. As we have already noted, it would be a mistake to assume that areas such as Sumer were linguistically homogeneous. More important, there is no reason to think that the language of writing would be indicative of the one hypothetical language spoken in a specific area. To the contrary, human history is filled with examples of people writing one language and speaking another. We must therefore be cautious not to equate language, especially written language, with cultures or with specific ethnic groups.
To complicate matters even further, the linguistic identity of the first tablets--which we should not confuse with the vernacular of their writers--is a matter of some controversy. Some scholars have claimed that cuneiform was not well suited for Sumerian, and therefore had to have been invented by speakers of an unknown, earlier language. Others, more cautiously, have suspended judgment on the matter, arguing that we simply do not have enough information to go on at the present time. In recent years, the theory that Sumerian was the underlying language has been gaining ground. The main clue for this argument is found in the nascent phonetization of the script. A good example of this is the sign AMA, "mother," which is written as PISAN, "box," inscribed inside with a smaller sign that has the value AM6. In this combination, which we conventionally render as PISANxAM6, the smaller sign is a phonetic complement, helping us to read it as Sumerian AMA. This does not necessarily mean, however, that the whole Uruk system was created to represent Sumerian utterances. The first script was not linguistic in nature; it was an independent communication system that was parallel to language, was closely allied with it, but was essentially independent. Only essential, formalized bureaucratic notations were registered, and part of the message was expressed by means other than writing itself. Thus the type of commodity, or the textual genre could be noted by the format of the text or even by the shape of a tablet. The subsequent history of cuneiform shows a gradual merger of the two--language and the new invention--until they reach a point of unstable merger. Then, in turn, the opposite took place. Natural language continued to change and even die, while written language remained relatively stable.
This process of merger between language and writing is not well documented. By 2600 BC, when the first literary texts appear in large numbers, the process had taken place, and there is no doubt about the linguistic identity of the contents of clay tablets. But while the language was still alive, the Sumerian writing system was never intended to express fully any linguistic segment. It was only in the second millennium, when the language was no longer anyone's mother tongue, surviving only as an archaic artifact in schools and temples, that it was fully written out. We can understand some parts of those early literary texts, but we cannot fully reconstruct them from the so-called nuclear writing that was used at the time. Only some of the grammatical and phonological information that was needed for comprehension was written down. The reader received only a skeletal version of a text and was expected to provide the missing elements. Here is a classic example of this kind of nuclear writing from an Early Dynastic Sumerian literary text; the original text is followed by the full spelling that would have been used in an 18th century version. One should note that the accents and number indexes in transliterations are not markers of pronunciation, but only serve to identify signs and the raised d is an abbreviation of dingir, the classifier for divine beings.
den-ki isimud gù dé
den-ki-ke4 disimud-ra gù mu-un-na-dé-e
"(The god) Enki addresses (his vizier) Isimud"
Although the reading of full texts of this kind is a difficult, and sometimes impossible task for us, we must assume that it was effortless for those for whom the writings were actually intended. First, they actually knew the language and we can only approximate it, and, second, but equally important, they also were intimately familiar with the texts. We know that in later periods scribes memorized literary compositions, even in languages that were no longer spoken, and that the oral component accompanied the written form. There is no reason to assume that things were very much different in the third millennium.
The manner in which an utterance can be expressed by such a nuclear method of writing differs from language to language. Sumerian word roots were primarily monosyllabic, and thus one could easily express one word with one sign. This does not mean, however, that signs were developed for each and every word in the language, or even for the average vocabulary used in everyday speech.. This was neither practical nor necessary. Since the script was developed for a specific purpose--the recording of a limited range of bureaucratic activities--the number of discrete symbols was limited. The Uruk IV scribes utilized approximately 1200 signs, but this number was rapidly reduced to a more manageable repertoire.
So far we have discussed the development of cuneiform writing used for the expression of the Sumerian language. The adaptation of the script for Akkadian, or, more broadly speaking, for Semitic languages, is a complex problem and one that is not yet well understood. A few decades ago the issue was relatively simple. In the twenty-fourth century, a ruler from the Kish area by the name of Sargon (2334-2279 BC), a throne name that means "True King" in Akkadian, succeeded in conquering all of southern and northern Babylonia, established his capital in the new city of Agade, and, among other major reforms, introduced Akkadian as an official language of statecraft, bureaucracy, and schooling alongside Sumerian. The period of his reign and that of his successors is called the Sargonic, or Old Akkadian dynasty (2334-2154 BC). Although Sumerian would be brought back as the main official language for a mere century of rule under the succeeding Third Dynasty of Ur (2112-2004 BC), this was but a linguistic holding action. Akkadian eventually gained the upper hand as the sole administrative language of Mesopotamia, while Sumerian remained in use only as a literary and liturgical tongue. One assumed that cuneiform was adapted to Old Akkadian--the dialect of Sargon's empire--somewhere in the north in the era preceding the new king's rise to power. Today this matter is no longer that simple. Before we describe the complicating factors, it might be useful to describe the features that make Semitic languages different from Sumerian and provide a general description of the principles of Old Akkadian cuneiform writing.
The basic structural principles of cuneiform were relatively simple. There were, in essence, only three kinds of signs: word signs, also called logograms, syllabic signs, and classifiers, that is signs that provided clues concerning the semantic class of a word. Since most signs had more than one use, there were graphs that could take on each one of these functions, depending on context. The best example of this started life as a picture of a star.
LOGOGRAM SYLLABLE SIGN CLASSIFIER
dingir, "god" dingir, "god"
an, "heavens" an
An, "Sky God"
While this may seem unreasonably difficult, in practice there was little ambiguity in the script. It is clear how the syllabic values came to be: the large number of monosyllabic words in Sumerian provided the material. The classifiers were usually unpronounced. Some, such as dingir, which specified that what followed was the name of a deity, or gi , "wood," were placed before nouns, while others, such as ki, "geographical name," came after.
Perhaps the one salient difference between Sumerian and Akkadian, or any other Semitic language for that matter, lies in the structure of word roots. In Sumerian, roots were primarily monosyllabic and, more important, were internally unalterable. They had no grammatical gender, and the only way in which they could be modified was by prefixing or postfixing particles. Noun roots were distinct from verbal roots and adjectives were essentially simplified verbs. Words such as en, "ruler," kur, "mountain, foreign land," kalam, "(native) land," gal, "large," or pad, "to call," could be modified with particles, but not by means of infixes or internal changes; they could be strung together in chains with the addition of affixes that expressed grammatical relationships. Thus en was "ruler," but nam-en was "rule," kur was "foreign land," and kur-kur meant "all the foreign lands." In combination they would provide clauses and sentences so that in grammatical terms the utterance "lord of the land" would be en kalam.ak, that is lord+land+possessive particle(of). Linguistic as well as orthographic rules required that this be written as en kalam-ma, since /k/ was dropped in final position and vocalic endings were written with a syllable that included the last consonant of the root, in this case /m/.
The Semitic root is an abstraction, usually consisting of a three-consonant skeleton which could be filled with vowels and even with other consonants, altered by reduplication of one of the consonants, as well as modified by prefixes and suffixes. Thus the root KRB in Akkadian could be used to create a wide variety of words with the general meaning of "blessing, prayer:" karabu, "to bless," takrub, "you blessed," takarrab, "you will bless," or ikribu, "blessing, prayer." In order to accommodate the different structural characteristics of this language, the Old Akkadian scribes used a system that was primarily syllabic. Thus to write the verb takrub, "you blessed," they would use four syllabic signs: ta-ak-ru-ub. All these values were already available in Sumerian: ta was a grammatical particle, ak was the verb "to do," ru was a verbal root, and ub was "corner." Additional syllabic values were derived by abbreviating Akkadian translations of words. Thus Sumerian gi , "wood," was isum in Akkadian and from this came the syllable /is/, which could be used for /iz/, /is/, as well as /is/ (s is an emphatic sound similar to ts). In this manner, the Mesopotamian scribes developed a full repertoire of syllabic signs: simple vowels such as a, u, and i, as well as consonant-vowel and vowel-consonant units such as ad/t or da/ta. Thus, from the middle of the third millennium on, one could write Akkadian, and indeed even Sumerian, in a purely syllabic manner, using a reduced repertoire of approximately 150 signs. Eventually the sign repertoire was increased, and many signs acquired multiple logographic and syllabic values, but letters and documents continued to be written in a relatively simple, primarily syllabic manner.
Discoveries made in the last few decades have made it clear that Semitic languages, including Akkadian, were written down before the advent of fully syllabic writing. There are texts from the period preceding the time of Sargon that are clearly in Akkadian, albeit they were written exclusively, or almost exclusively, with Sumerian logograms. In theory, one could say that there are only so many possible ways of using a system such as cuneiform for writing Sumerian and Akkadian: A text could be:
1. written and read in Sumerian,
2. written and read in Akkadian,
3. written in Sumerian and read in Akkadian, or
4. written in Akkadian and read in Sumerian.
An early tablet written only with word signs would seem to be linguistically impenetrable; if we knew all the signs then we could understand it, but we could not be sure of the underlying language. In practice, there are a number of clues that one can use to establish the probable linguistic identity of a given text. The origin of the object provides some information: a text from northern Babylonia is more likely to be Semitic, while one from Sumer should be Sumerian. This is only a rule of thumb, however, since there are demonstrably Sumerian texts from the north and Akkadian ones from the south. Another indicator might be found in one or more syllabic signs that provide readings from a specific language. Thus, if a royal inscription seems to be written in Sumerian, but contains, before a place name, a syllable /in/ that can only be interpreted as the Akkadian particle in, we may assume that the whole text is to be read in Semitic. Other indicators are less obvious. There were a few logograms and syllabic signs that were used exclusively in Akkadian and other Semitic languages such as Eblaite, but never in Sumerian. A good example of this is the word for "witness" which was written in Sumerian as ki inim-ma, corresponding to Akkadian bum, which means "elder," as well as "witness." The standard Sumerian signs were not used as logograms in third millennium Akkadian texts. Rather, this was expressed by means of the logogram that we transliterate as AB+Á , and which corresponds to the Sumerian word for "elder." These kinds of rules would seem to provide sufficient criteria for the linguistic identification of texts, but certain texts defy easy classification. Take, for example, the following fragmentary tablet:
1. lugal-á-zi-da Mr. Lugalazida
2. árad Lugal-ki-gal-la the slave of Lugalkigal
3. énsi-da from Mr. Ensi
4. in-da-zàh escaped.
5. ki zàh-a-na His hiding place
6. géme ur-ki-ke4 the slave girl of Urki
7. ba-dug4 revealed:
8. in ma -kà-ni- "In (the town of) Mashkan-
9. [ SHA]BRA shapir..."
Armed with the criteria enumerated above, how are we to establish the language of this legal text? On the basis of the sign forms we can date it to the time of the dynasty of Sargon. The provenance of the tablet is uncertain; it may have come from Nippur, but that is not a decisive feature since we have Sumerian as well as Akkadian texts from that city. One must add that the text is broken and that better information may have been contained in the missing lines, but the investigator must confront, more often than not, such incomplete tablets. On first glance, everything points to Sumerian, with one classic exception--the particle in contained in line 8. Suddenly we find that the neat series of possibilities enumerated above may prove to be insufficient. The possibility does exist that the whole text was Akkadian and that lines 1-7 are simply Sumerian logograms that were designed to be read in the Semitic tongue. More interestingly, however, one could claim that the two languages were intermingled; the main protocol of the juridical proceedings was redacted in Sumerian, and the testimony of the main witness, registered as direct speech, was written down in the language in which it was given, that is, in Akkadian.
During the Sargonic period Akkadian was fast becoming the dominant written language. It co-existed with Sumerian, primarily in the south, as the language of letters and administration, but soon began to dominate in royal inscriptions and other monumental display texts. The Sargonic state disintegrated in the decades after 2200, and after a short period of local rule, Mesopotamia was once again reunited under one banner, this time centered on the southern city of Ur. The Third Dynasty of Ur (2112-2004) is one of the best documented periods of Mesopotamian history, leaving behind countless administrative and economic records. We currently have at our disposal over 30,000 published cuneiform tablets from the time of this dynasty and countless more lie unread in the drawers of museums and in the mounds of modern Iraq. Only a very small percentage of this vast documentation was written in Akkadian, and this is one of the reasons that some scholars have dubbed this period "Neo-Sumerian," even going so far as to refer to it as the "Neo-Sumerian Renaissance." This is meant to convey the idea that the introduction of Akkadian by Sargon and his successors was an act of ethnic and linguistic politics that was then countered by a nationalistic reaction on the part of the kings of Ur, who reestablished the venerable Sumerian traditions, including their language, after the fall of the Akkad state. This is a highly anachronistic interpretation of the facts, which projects into the past language policy controversies of modern times. The fact that most texts from the time of the kings of Ur were written in Sumerian is partly a consequence of the centralization of bureaucratic power and an imposition of a relatively uniform set of accounting procedures throughout the core areas of the state, and partly due to sampling problems. Simply stated, although we have an unprecedented mass of documents from this period, the vast majority of tablets comes from only a few cities. We know that Akkadian was used alongside Sumerian in a few places, primarily in the north. When the empire crumbled and Ur fell in 2004 BC, the fate of Sumerian as a language for everyday use was sealed. The small successor states in Sumer continued to use Sumerian for a few generations, but in the north Akkadian was taking over, soon to become the sole written language of administration, business and commerce. The last Sumerian letter is dated to approximately 1930 BC.
The syllabic principle which was developed for writing Semitic languages was never rigidly formalized, and remained fluid throughout the history of cuneiform. The kind of syllabic signs that were used, the way they were combined, as well as the number of signs, changed over time and space. Nevertheless, one can state unequivocally that the principles of Old Akkadian writing created the foundations for all subsequent developments. The syllabic principle allowed one to spell out any word in a relatively simple manner. This does not mean, however, that the older logographic way of writing words became obsolete. To the contrary, the two principles were used together for efficient communication. The relative percentage of syllabic and logographic writings fluctuated synchronically as well as diachronically, and at any given time cuneiform could be used differently depending on the type of text one was dealing with. A scribe writing in Babylon around 1800 BC would use one hundred and fifty or less syllabic signs as well as a few logograms to write a letter on behalf of a businessman, but would need a larger repertoire of logograms to write a routine account. Scholars who copied literary works at Assyrian courts a millennium later would have to know a much larger inventory of signs, perhaps more than six hundred, in order to read and write a wide variety of textual genres. The main principle that dictated the use of logograms was predictability and routine: words that occurred often were written with word signs. Thus a scribe who had to regularly account for a limited repertoire of items that were being brought in, stored, and disbursed from a central bureau, would regularly use the same few signs to notate the goods that were being moved. An overseer of animals could thus write one sign UDU for the sheep in his charge rather than to spell out im-me-ru in syllabic Akkadian. In much the same manner, scribes writing medical or scientific texts would develop specific logographic sets to express the specific technical vocabulary of their work.
This broad principle goes only so far towards explaining local and historical differences for aesthetic, ideological, and psychological forces were at work as well. As a general rule, Akkadian literary texts from the second millennium are primarily syllabic, but in the first millennium we find an increased use of logographic writing as well as the development of new syllabic repertoires, so that the number of signs used increased dramatically. The reasons for this are many, and some of them have nothing to do with the technical aspects of writing. The Akkadian literary language by now was quite distinct from the vernacular, which, in turn was being rapidly replaced by another Semitic language, Aramaic, in the streets, markets, and courts of Babylonia and Assyria. The small number of people who could read and write now had to learn two dead languages--Akkadian as well as Sumerian. For some scribes, the literary tradition became a celebration of the roots of civilization and complexity acquired a special prestige sanctioned by the past.
The last dated cuneiform tablet is an astronomical almanac from Babylon dated 74/74 A.D. By this time the cities of Assyria and Babylonia lay in ruins and the exploits of their scholars and kings were just faded memories that were the stuff of rumor and legend. A small number of scholars kept the writing alive but the alphabet had already taken the place of the venerable old cuneiform. We do not know how long this knowledge lasted; some evidence suggests that the ability to read the wedge-shaped signs persisted as late as the second century or perhaps even longer. A typical short passage from this text will serve to demonstrate the manner in which a technical vocabulary could be expressed with word signs that can be read even if one is not always certain of the pronunciation of the underlying words:
14 gud ma -ma kur 14 na 27 kur
"On the fourteenth day Mercury will rise in Gemini. On the fourteenth, moonset after sunrise. On the twenty-seventh, last lunar visibility before sunrise."
Here gud stands for the planet Mercury and ma -ma for the constellation Gemini. The sign kur, which originally meant "mountain" in Sumerian, was used in late period astronomical texts in two distinct ways. It could be a logogram for the Akkadian verb napahu, which described the rising and first visibility of celestial phenomena, but it could also serve as a term for the period between moonrise and sunrise and, at the end of the month, denoted the last time that the moon was visible before the rising of the sun. The sign na, the reading of which is not known, was used to register the first time in the month that the moon set after sunrise.
And so in the end it all came full circle, for the
last dated tablet that we have was written almost exclusively with logograms,
unconsciously recalling the third millennium origins of Akkadian writing.