The genealogies of the Persian kings were preserved by the Greeks. Given the general time period, the names must have been Xerxes, his father Darius, and Darius’ father Hystaspes, who was not a king. With an informed guess as to what these names would have sounded like in Persian (rather than Greek), Grotefend was able to assign phonetic values to some of the signs. The name of the last king, Xerxes, began with the same letters as the hypothesized word for “king,” and indeed in Persian they both began with the sounds [x∫]. So the language was Persian, a language that (in other versions and written in another script) was fortunately becoming known to Western scholars at the time. Decipherers who followed up on Grotefend’s initial work could thus check their guesses and deductions against what was known about the language: if the text could be made to make sense in Old Persian, the decipherment was likely to be correct. Here the budding field of historical linguistics played a key role, enabling scholars to make systematic use of different varieties of Persian and of related languages such as Sanskrit to reconstruct the properties of Old Persian.
The most famous and longest of the trilingual inscriptions covers a large part of a cliff face at Behistun (also known as Bsitun or Bisotun). The British army officer Henry Rawlinson (1810–95) copied the inscription at some personal risk and made it available to Western scholars.
With the publication of the Behistun inscriptions, further progress could be made. The Irishman Edward Hincks was able to make significant headway, demonstrating that the Old Persian script was at least partly syllabic. He also studied the other versions of the trilingual text, recognizing the second script (now known to be Elamite) as a simpler version of the third (now known to be Babylonian Akkadian), and the third as being the same (generally speaking) as that of tablets being found in great number at Assyrian and Babylonian sites in neighboring Mesopotamia (see figure 2.3).
Hincks applied Grotefend’s name formula to the Elamite and Babylonian inscriptions, beginning the decipherment of these scripts. He deduced that the third language was Semitic, and used his knowledge of common Semitic properties to work on Akkadian. Akkadian presented a significant challenge, owing to the complicated nature of the writing system, with its polyvalence of signs and combination of logograms and syllabograms. Elamite, by contrast, had a much simpler script, but the language itself was more difficult to understand because it had no recorded relatives.
At the time, much of the credit for the decipherment of cuneiform after Grotefend’s initial progress went to Rawlinson, with Hincks understood as having achieved similar results working independently but a little more slowly. It is not unlikely, however, that it was Rawlinson who trailed Hincks, making use of Hincks’s results as they became available.
Hincks was a clergyman and a prolific scholar who also did much in the decipherment of Urartian. His was also the honor of bringing the decipherment of Akkadian cuneiform to a close in 1852 by correctly identifying an Akkadian tablet that listed a number of logographic signs and gave their pronunciation spelled out in syllables. He had found the Akkadians’ own ABC. Such a sign list would have been used in the training of ancient Akkadian scribes; it and the many other sign lists that have since been found have also proved invaluable to the study of Akkadian by modern scholars.
Figure 2.3 The first sentence of Darius the Great’s trilingual cuneiform inscription at Behistun in (a) Old Persian, (b) Elamite, and (c) Babylonian Akkadian. Note that the Old Persian, being partly alphabetic, is the longest, while the Babylonian version, being logosyllabic, contains more complex signs. The sentence reads, “I am Darius, the great king, the king of kings, the king of Persia, the king of the provinces, the son of Hystaspes, the grandson of Arsams, the Achaemenian.” The Babylonian version is more succinct than the other two (and, where in brackets, partly restored through comparison with other inscriptions). For those interested in trying their hand at decipherment, the first words of the Old Persian can be transliterated as a-da-m da-a-ra-ja-va-u-∫ x-Sa-a-ja-θ-i-ja va-z-ra-ka (I, Darius, great king), with signs belonging to the same word separated by hyphens. The lone angular sign with which the sentence begins is a word divider.
But where in all this were the Sumerians? The decipherment of cuneiform proceeded largely backward, from the relatively late (and streamlined) Old Persian to the ancient and elaborate (one might even say baroque) Akkadian. The Persians were known to history, and so were the Akkadian-speaking Assyrians and Babylonians. The Sumerians had been lost entirely. The Sumerian language first re-emerged at the site of the royal library at Nineveh. As Akkadian became fully readable, it was obvious that some of the texts at Nineveh were written bilingually in Akkadian and something else. For a while some scholars believed that the “something else” might not even represent a real language – it could be some sort of code version of Akkadian, or a scribe’s game.
But other evidence was accumulating that pointed to the existence of a form of cuneiform older than that of Akkadian. For one thing, the syllabic and logographic versions of Akkadian signs did not sound the same. Why, for example, was one sign pronounced [gal] when it denoted just a phonological syllable sound, but [rabu: m] when it was functioning as a logogram meaning “great”? Wouldn’t it make more sense if the syllabic use of the sign were based on the logographic use and pronounced like it? Where had the syllabic values of the signs come from, if not from Akkadian words? Assyriologists began to suspect that a different language underlay the Akkadians’ use of cuneiform. But what was it? The answer came from early Akkadian records in which the ancient kings called themselves “king of Sumer and Akkad.” The unknown language must have been that of this hitherto unknown land, Sumer. And sure enough, tablets from the Early Dynastic period in southern Mesopotamia began to come to light. These were written in the same language as the mysterious texts assembled much later at Nineveh.
And thus the words of the Sumerians were restored to us. Once people were willing to believe that the bilingual tablets contained a real language – Sumerian – much about that ancient language could be deduced. The disparity between the syllabic and logographic readings of the signs disappeared: in Sumerian, the word for “great” was gal. The use of the same sign for “reed” and “render” and the syllable [gi] became obvious: in Sumerian, they were all pronounced [gi]. Sumerian, it turned out, explained a lot about how Akkadian was written.
Nevertheless, to this day Sumerologists are plagued by the fact that we can only read Sumerian because we can read Akkadian, an unrelated language. Thus we do not know precisely how the Sumerians pronounced their language. We know that the Akkadians thought that the Sumerian word for “great” sounded like [gal], and that the Sumerian words for “reed” and “render” sounded like [gi], but whether the Sumerians would have agreed is a somewhat different matter. Conversely, the writing of Akkadian was somewhat hampered by the traditions of writing Sumerian. The number of different sounds represented in Sumerian influenced which sounds were distinguished in writing Akkadian. Thus Akkadian was always written with what might be considered a Sumerian accent. But when we look for evidence as to what Sumerian sounded like, we have to take the Akkadians’ word for it, and so we now read Sumerian with an Akkadian accent!
The spoken words of the Sumerians and Akkadians disappeared long ago. Yet through the efforts of decipherers we can read words 5,000 years old, even if we do so in an accent the ancient writers would never recognize. Most other languages of the time – the vast majority of languages around the ancient world – were not so lucky. They perished without a trace or mutated into new forms. Language is inherently ephemeral; writing lasts, particularly if it is written on baked clay tablets.
3
Egyptian Hieroglyphs and the Quest for Eternity
A Sumerian or Akkadian scribe keeping records in cuneiform would have been astounded to learn that his text would survive for millennia. An Egyptian scribe incising hieroglyphs for a commemorative or funerary text would merely have considered such permanence a job well done. The
Egyptians were probably the second people to develop a writing system, but they trailed the Mesopotamians only slightly if at all, and they soon caught up to apply their unique script to the full range of bureaucratic, religious, and literary uses. They also made writing into an art form that still inspires fascination and admiration today. The lasting popularity of hieroglyphs is quite apt: writing both encouraged and expressed the Egyptians’ devotion to permanence and the pursuit of eternal life.
The Egyptians’ preoccupation with eternity was fostered by their physical surroundings, where life flourished in the narrow valley of the Nile River, surrounded by changeless deserts and a changeless sky. The Nile defined the Egyptian country and culture. The river creates a slim band of cultivable land extending the length of the country, widening out into the fan-shaped delta in the north as it flows through its own deposits of sediment to reach the Mediterranean Sea. On either side of the narrow line of green are inhospitable deserts. These deserts were formed in prehistoric times, driving their inhabitants to settle in the Nile Valley where water continued to be available. There the migrants found plenty of game in the marshlands fed by the river. The yearly flooding of the Nile brought fertile sediments and moisture to the riverbanks, allowing edible plants to flourish. The prehistoric Egyptians discovered that they could control the growth of plants by sowing, irrigating, and harvesting crops. With the development of agriculture, the population grew, and as the population grew, swamps were drained and put into cultivation. As the cultivated land expanded, so did the need for irrigation. As in Mesopotamia, the collective labor required for irrigation appears to have played a part in the increasing complexity of society and the birth of civilization in Egypt.
Since the habitable land was very narrow, the population did not tend to cluster into towns and cities as it had in Mesopotamia, but remained spread out along the river in villages. As civilization began and the political power of chieftains grew, their domains therefore were not city-states but stretches of the Nile. By the end of the prehistoric period, Upper Egypt was a cohesive cultural and political region in contrast to Lower Egypt, which contained the delta. (Upper Egypt was southern Egypt, as the Nile runs downhill from south to north. The Egyptians always thought of south as “up” and north as “down.” Similarly, “left” could mean “east” and “right” “west.”)
Toward the end of Egypt’s prehistoric period, the Egyptians must have come into contact with the Mesopotamians. They borrowed the cylinder seal (which originated in the Uruk period of Sumer), and certain artistic and architectural styles. They soon rejected these borrowings for their own developing native styles. More importantly, the Egyptians may have learned at this time of the recent breakthrough in Sumerian record keeping – the ability to record words by stylized markings on a durable surface.
It is likely that the Egyptians at least heard about Sumerian writing, but it is unclear exactly how much the Egyptians learned about protocuneiform, and it is even possible that somehow they did not hear of it at all, for they certainly went about creating their writing system in their own way. Egyptian hieroglyphs make their first appearance in the archaeological record around 3150 BC, as Egypt was making the transition from its pre-dynastic to dynastic period.
According to later Egyptian tradition, the nation’s history began around 3100 BC with the unification of Upper and Lower Egypt into a single kingdom under the first pharaoh of the first dynasty, Menes. Menes has never been identified historically, but he was said to have been a king of Upper Egypt who conquered the delta land of Lower Egypt. Despite uncertainty as to the precise date (it may have been a little earlier than 3100 BC) or the actual name of the pharaoh, by 3100 BC Egypt had become the first nation-state in the world, unified under the command of divine kings, with all the administrative requirements that this entailed. The success of this new political arrangement depended on their new invention, writing.
For all we know, the Egyptian writing system may have been the invention of a single person, or it may have required the collaboration of many over a considerable period of time. If the latter, then the early stages of development have been lost to archaeology. Although the oldest surviving inscriptions – labels on ceramic vessels and stelae (commemorative stones) – are terse and hard to read, they appear to contain the same essential characteristics as later Egyptian writing: logograms supplemented with three different kinds of phonograms. In Mesopotamia, by contrast, the maturation of the script was much slower. While changes and elaborations in the Egyptian system did occur along the way, these characteristic elements were present from its earliest surviving uses.
The first and longest-lived form of Egyptian writing was what we now call hieroglyphic after the Greek term for “sacred carving.” While hieroglyphs were often carved into stone, they could also be written with a brush and ink on pottery or on papyrus, a heavy, paper-like material made from the papyrus reed. Unlike cuneiform, whose original pictograms eventually lost all visual connection to the real world, Egyptian hieroglyphs always remained pictorial (see plate 2). Individual hieroglyphs were pictures of actual beings or objects. They could be elegantly carved and painted, as they often were in commemorative and sacred inscriptions, or they could be drawn in outline, a style known as cursive hieroglyphs and usually used for writing with brush and ink. The two styles are attested from the earliest stages of hieroglyphic writing. The fancier style may be an elaboration of the simpler, or the simpler style may be a reduced form of the fancier. In either case, the two styles were adapted to their different uses and the different surfaces on which they were written.
An individual hieroglyphic sign could have one or more of three possible functions: logogram, phonogram (phonological sign), or determinative. Probably the first signs to be developed were logograms, as in Mesopotamia. Such a sign would stand for the word that named the object pictured. Oft-cited examples include (pr), a schematic of a house plan meaning “house,” and (rc), a depiction of the sun’s disk meaning “sun” and also, by extension, hrw, “day.” A number of other easily depicted things, such as animals, body parts, and everyday objects, could be written with logograms.
The inventors of Egyptian writing would soon have realized, however, that designing a logogram for every word in the language would be a tall order. Thus in practice, words written with merely a logogram were relatively rare; when they did occur, they were usually followed by a stroke below or beside them to indicate their logographic use:
For a fully developed writing system, something else was needed. The Egyptians may have heard of the rebus principle from the Sumerians, or they may have discovered it independently. At the heart of rebus writing is the idea that the sound of a word is separable from its meaning, so that a pictorial logogram for one word can be transferred to another word of the same sound but a different meaning. The rebus principle eventually led the Sumerians and Akkadians to dissociate sound and meaning enough to create syllabograms – phonologically based signs with no inherent semantic meaning. In Egypt the rebus principle was applied differently because of some crucial differences between the Egyptian and Sumerian languages.
The Sumerians had begun by developing symbols for simple monomorphemic words – a practice known as “nuclear writing” as it represents just the uninflected core of the words. From there they used rebus writing to represent homonyms, to add phonetic complements, and eventually to spell out words and names syllabically.
A crucial aspect of the Sumerian nuclear writing, however, was that the cores of words were themselves words. The same is true in English. The word breath can function alone, or it can form the core of the plural breaths, the verb breathe, the adjective breathless, or the derived noun breathalyzer. The logograms of early Sumerian nuclear writing sometimes represented full words (like breath) and sometimes just the core of longer words.
The logograms of the Egyptians represented full words. So was “house” and “sun,” etc. Unlike in Sumerian or English, however, the cores of the words wer
e not themselves words. This takes a little linguistic theory to explain.
The Egyptian language belonged to the Afro-Asiatic language family, to which also belong the Semitic languages of western Asia (such as Hebrew, Arabic, and Akkadian) and a number of languages of North Africa such as Berber. Egyptian and Akkadian are the earliest attested Afro-Asiatic languages.
Like Akkadian and other Semitic languages, Egyptian generally built its words around a core of three consonants. The derivation of one word from a related word was usually done by changing the vowels that were intermingled with those consonants, as well as by the repetition of consonants or the addition of affixes. Thus related words shared consonants, just as English drink, drank, and drunk do, with vowels carrying the difference in meaning.
The core three consonants of an Egyptian word constituted the root morpheme of the word, while the vowels that were interspersed with these consonants carried the inflection, varying with the grammatical form of the word and functioning like prefixes and suffixes do in other languages. For example, the consonant sequence n-f-r meant “good” or “complete.” The masculine form of this word was probably *nafir, where the asterisk shows that the word is a reconstruction made by historical linguists, not a form of the word that has ever been firmly attested. The feminine form of the word had a -t suffix, but it would also have had different vowels: it was probably *nafrat. In these two words not only are the vowels different (a and i versus a and a), but they are placed differently among the consonants. Both the particular vowels used and their placement affected the meaning of words.
In Egyptian, unlike in Sumerian, there was a difference between a simple unaffixed word (like *nafir) and the core of the word (n-f-r). The core of a word was not composed of a consecutive string of sounds pronounceable as a syllable or two. The feminine *nafrat, for example, could not be formed by the addition of -t to *nafir. Instead, the word core was an unpronounceable sequence of discontinuous consonants. A symbol that represented the core of a word therefore represented its consonants. In other words, the Egyptians realized that the common element of *nafir and *nafrat was the triconsonantal root n-f-r.
The Writing Revolution Page 5