The invention of writing in the strict sense of the word was preceded by a series of earlier developments that gradually led up to it. Soon after the invention of pottery in the neolithic period (ca. 8000 BCE) in the Near East, excavations reveal the existence of a system of clay counters and tokens that, in their earliest form, appear to have served to identify the number and kinds of goods traded, entrusted to second parties, or otherwise dealt with. By the fourth millennium BCE, such counters were being enclosed in hollow clay envelopes (bullae) which, after drying, constituted a sealed physical record of the transaction; but until and unless the envelope was broken open, this record could not be verified or retrieved. Before long it was realized that, like the seals, the counters could be impressed on the surface of the bulla before it dried, thus providing an ever‐present record of the counters enclosed in it; only if their number and character needed to be verified was it necessary actually to break the bulla open. Next it must have been seen that the counters could be dispensed with altogether, and thereafter it became logical to abandon the hollow bulla format as well, in favor of a more or less flat clay tablet with an only slightly rounded writing surface. Finally, the counters themselves were abandoned and their approximate shapes reproduced instead by means of a stylus made of reed. This final development seems to have occurred first in Sumer, where clay and reeds were both abundant; it signaled the emergence of writing in its full sense and may probably be dated about 3100 BCE.
Within a century or so, the system expanded from the depiction of concrete objects by means of pictograms to the expression of sounds or syllables by means of syllabograms, a first application of the rebus principle. An example is the name dEn‐lil‐ti, “the god Enlil lives (or gives life),” in which the first three signs may be described as logograms (word signs) but the last is a picture of an arrow whose pronunciation in Sumerian is ti, as is (more or less) the pronunciation of the Sumerian word for “life, live” (actually til).
No further significant innovations in the inner structure of the cuneiform script were needed beyond this point, nor indeed were they forthcoming. All subsequent changes involved only phonetic adjustments to the needs of other dialects and languages, or the external forms of the signs, whose total number was stabilized at about six hundred. From their pictographic origins, they rapidly evolved into stylized linear representations. Gradually, these were replaced in turn by the wedge‐shaped characters that result from the impression of a split reed (shaped like a prism) into wet clay and that justify the modern designation “cuneiform” first applied to them by Thomas Hyde in 1700 CE.
The newly invented cuneiform script remained in use in the Near East for three millennia (until ca. 100 CE), undergoing considerable changes in external appearance as it spread rapidly from its Sumerian base. In Elam (southwestern Persia) to the east and Egypt to the west, it stimulated native writing systems (“proto‐Elamite” and hieroglyphics respectively) that owed no more than their basic inspiration to the original invention. Elsewhere, the borrowing was more direct. The syllabic values of the cuneiform script were taken over in their entirety for writing Akkadian by Babylonians and Assyrians; the Sumerian logograms were given their Akkadian equivalents; some new syllabic values were assigned to existing cuneiform signs on the basis of such equivalents, others by convention to render sounds not found in the Sumerian phonemic roster. A little later in the third millennium BCE, the Elamites also adopted the Sumerian system.
In Anatolia (central Turkey) the first attested writing, early in the second millennium BCE, consists of Old Assyrian texts and a few native inscriptions modeled on them. But the conquest of the area by the Hittites about 1700 BCE led to the adoption of a Babylonian version of cuneiform, probably from Mari, and its adaptation to the special requirements of their Indo‐European language. A distinct script was developed by Luvian, a language (closely related to Hittite) that survived the fall of the Hittite empire about 1200 BCE, especially in northern Syria. This script is pictographic in form and essentially syllabic in structure. It is conventionally designated as Hieroglyphic Hittite, though it has nothing to do with Egyptian hieroglyphics (see below) and is not, strictly speaking, used for Hittite. Mesopotamian cuneiform also provided the vehicle for Urartian (see Ararat), the language of eastern Anatolia from ca. 1300–600 BCE.
While the Elamite, Hittite, and Urartian adaptations retained the essential forms and functions of Mesopotamian writing, a wholly new cuneiform script was devised about 520 BCE to record Old Persian. Its individual characters were wedge‐shaped, but beyond that they owed nothing to Mesopotamian cuneiform; they were nearly all syllabograms of the form consonant‐plus‐vowel (or simply vowel) and numbered no more than forty; only four of them were logograms. And while Persia thus continually experimented with new scripts inspired by Mesopotamian models, native factors predominated in the further development of writing in Egypt. From its origins before the end of the fourth millennium BCE, it quickly evolved into the elaborate logo‐syllabic system known by its Greek designation as hieroglyphics (literally “sacred carvings”). Its syllabograms were indifferent as to vowels; they distinguished only the consonants and must therefore be transliterated without vowels (or with the vowel e conventionally inserted between the consonants). Extensive use was made of logograms and (to a greater extent than in cuneiform) of determinatives or “semantic indicators,” that is, signs not pronounced in speech but alerting the reader to the meaning class of the ensuing word.
Like Mesopotamian cuneiform, Egyptian hieroglyphics survived for over three millennia, their last use dating from the Roman period. On monumental texts, the hieroglyphic signs retained their essentially pictographic character, and even elaborated on it secondarily by adding color or direction to the individual signs. But before the end of the third millennium BCE, a cursive adaptation of hieroglyphic, the hieratic script, began to evolve for use on papyrus and ostraca (potsherds). Early in the first millennium BCE, the Demotic dialect of Egyptian developed its own cursive script, while, at the end of the millennium, twenty‐three hieroglyphic signs and their Demotic equivalents were adopted for the unrelated Meroitic language of Nubia, south of Egypt. The various Coptic dialects of Egyptian were written in a modified Greek alphabet.
The internal development of hieroglyphic also continued. By the side of the traditional logo‐syllabic script, there developed in the second millennium a so‐called syllabic orthography, particularly for foreign names, which attempted to indicate vowel quality by means of semiconsonants. It is this form of syllabic hieroglyphic writing which is generally thought to have inspired the first attempts at the written recording of Northwest Semitic speech, beginning in the Sinai peninsula, where speakers of Canaanite (or proto‐Canaanite) came in contact with Egyptian culture in the middle of the second millennium BCE. At the Sinaitic site of Serabit el‐Khadim, where Canaanite laborers (or slaves) worked copper and turquoise mines under Egyptian overseers, graffiti and other inscriptions in the “proto‐Sinaitic” script employed clearly pictographic characters (which in some cases may have been conventionalized replicas of hieroglyphic signs) to represent the Northwest Semitic roster of phonemes. According to the acrophonic theory advocated by many scholars, the Canaanite word represented by the sign provided its name, and the first syllable of that word became the pronunciation of the sign. This principle works better for some signs than for others; it may therefore be best to regard the decipherment of proto‐Sinaitic as provisional and the assignment of letter names as a later development.
Meantime Northwest Semitic also came into contact with the cuneiform tradition of writing, knowledge of which had spread through the Near East in the Late Bronze Age (ca. 1550–1200 BCE), at least among scribes trained at scribal schools on the Babylonian model that are known from El‐Amarna in Egypt to Hattusha in Anatolia and along the entire Levantine littoral in between, especially at Ugarit on the north Syrian coast. Here, some time after 1400 BCE, thirty cuneiform signs were newly devised to represent the consonantal phonemes of Ugaritic (which represents another early form of Canaanite), its three basic vowels (when preceded by the glottal stop) and a Hurrian phoneme for words and names in that non‐Semitic language. Outside Ugarit itself, scattered examples of the new script have been found as far away as Israel. The spread of the invention was helped by the mnemonic device of arranging the signs in a conventional sequence and inscribing them in this sequence on practice tablets today referred to (like their later Latin counterparts) as abecedaries (from a‐b‐c‐d‐arium); allowing for the intervening reduction in the number of signs, this sequence already equals that of the subsequent Hebrew and Greek alphabet (from the latter's first two letters, alpha, beta), the former attested in the alphabetic acrostic‐poems of the Bible (Pss. 9–10; 25; 34; 37; 111; 112; 119; 145; Prov. 31.10–31; Lam. 1–4).
Both Ugaritic and proto‐Sinaitic served as prototypes and possibly as inspirations of the earliest forms of what is loosely called Phoenician script, the ancestor of all later alphabets. This script was not, however, a true alphabet, but more properly a syllabary, which continued to ignore vocalic quality in favor of syllables of the type consonant‐plus‐(any or no)‐vowel. It was adopted without essential modifications for the writing of such Northwest Semitic languages as Hebrew and Aramaic as well as Phoenician, the last probably responsible for spreading its knowledge westward as early as the beginning of the first millennium BCE to Greece and beyond, as Greek sources state. The Greek world had evolved an earlier system of writing in the Late Bronze Age, the Linear B known from archival texts found on Crete and the Greek mainland. These texts were written on clay and thus may represent a response to a Mesopotamian, or at least a cuneiform, stimulus. (The contemporary Linear A and the Phaistos Disc have not yet been satisfactorily deciphered.) But the greater simplicity of the new script recommended itself to the Greeks as it had to the Phoenicians.
The Greeks, moreover, turned the new invention into a true alphabet by employing some of the otiose consonantal phonemes into signs for vowels; they also took advantage of the fixed order of the letters to assign them numerical values. Both of these ideas were subsequently adopted, with modifications, for the original Northwest Semitic signs, notably in Hebrew. But the vowel signs (matres lectionis, or mothers of reading), never fully succeeded in distinguishing vowel quality, so that later Hebrew, Aramaic and, eventually, Arabic developed a system of (optional) diacritics for this purpose.
Via the Greeks, the alphabet was transmitted to the Romans and to the rest of the European world. Meantime the Aramaic version of the Northwest Semitic script inspired the Indic writing of Asoka and, indirectly, Sanskrit and other scripts of Asia to the east, and North Arabic, South Arabic, and Ethiopic to the south. The alphabet was thus truly ready to conquer the world, with the notable exception of China and Japan. But in antiquity, the older scripts of Mesopotamia and Egypt continued side by side with it. The last datable cuneiform text is an astronomical tablet for 75 CE. The last Egyptian texts are rock inscriptions from the island of Elephantine (Philae), below the Aswan Dam, and are dated 394 and 425 CE for hieroglyphic and hieratic respectively.
William W. Hallo