If Paleolithic cave paintings are taken as the earliest records of human communication, it can be seen that notation began with mineral substances: cavern walls and ground-earth pigments. Vegetal and animal substances successively came into use.

Mineral Substances.

The oldest surviving documents, the proto-cuneiform corpus from Uruk, are written on clay, “a natural, earthy, fine-grained material that develops plasticity when mixed with a limited amount of water” (Grim, 1979, p. 700), “formed by the decomposition of … rock-forming minerals, notably the feldspars”; they are “hydrated aluminium silicates with subordinate amounts of alkalis, alkaline earths, iron oxides, etc.” (Scott, 1954, p. 379). [See Cuneiform; Uruk-Warka.] Clay used for cuneiform tablets (as opposed to pottery clay) has apparently never been physically analyzed; Denise Schmandt-Besserat (1992, p. 29) reports that the tokens used in preliterate accounting are most commonly made of a clay called montmorillonite (not mentioned by Grim, 1979). Lindsay Scott (1954, pp. 379–380) distinguishes primary from secondary clays. The former (bedded where they form, in marshes or bogs) include the kaolins or china clays; the latter (transported by water, wind, or ice from their primary source, acquiring additional impurities) are virtually ubiquitous, and it is these that seem to be used in cuneiform tablets. Mixed with an appropriate amount of water, clay is plastic; a mass of it suitable to accommodate the length of the text to be written would be modeled by a scribe into a shape dictated by tradition or convenience (usually a rectangular parallelopiped—called a tablet—for most uses; a disk for a practice text; or a hexagonal prism for annalistic texts). [See Tablet.] The obverse surface—the first to be inscribed—was flattened (Jerrold S. Cooper [1996, p. 38] claims that the flattening occurred when the tablet was set down on a flat surface for drying after the obverse was written, but is this not precluded by the high viscosity of the material?), the reverse being somewhat convex. Air drying was the normal way for inscribed objects to harden; very exceptionally, they could be baked, like pottery. Charles E. Jones (personal communication) observes that far fewer, if any, tablets were intentionally baked (as opposed to being accidentally cooked in conflagrations) in antiquity than is generally thought. Jones has also noted that at least some tablets have a mud core with a clay veneer to receive an inscription; and that when such a veneer chips off, pieces can be mistaken for fragments of clay envelopes. The India House Inscription (a very large slab inscribed on five sides recovered from Babylon at the beginning of the nineteenth century, one of the key documents used in the decipherment of cuneiform) seems to be hollow. [See Babylon.]

The wedge-shaped impressions that comprise cuneiform characters were made by lightly touching the surface of the clay tablet with a stylus (see below), or by stamping a complete commemorative inscription onto the faces of building bricks; clay is, thus, the primary medium in the Mesopotamian sphere. Mycenaean texts, in Linear A and Linear B, were also written on small clay tablets, but the characters were formed of incised lines rather than impressed wedges. Cuneiform (and other) writing could be imitated on non-yielding surfaces, including cliffsides, quarried or freestanding stones, or portable objects. Stone could be incised using mallet and chisel; the gypsum of which Mesopotamian palace reliefs were often fashioned is a very soft stone and easily worked, whereas boundary stones and other objects for unsheltered display needed to be more durable and must have presented a greater challenge to the engraver. The limestone of statues and other objects is of intermediate hardness. The Hammurabi stela, on which the wedges of the characters are outlined rather than hollowed out, is said to be of diorite. The Rosetta stone, originally described as basalt, is identified in a not-yet-published analysis as a form of granite. In Egypt, hieroglyphic inscriptions (originally written with a pen or brush, see below) could be carved at great length into limestone walls. [See Hieroglyphs.] Some monuments in various peripheral areas—among them some Luwian hieroglyphic and Aramaic and South Arabian consonantal inscriptions—exhibit characters that are not incised but are in relief, with the background cut away. [See Luwians; Aramaic; South Arabian.] Another material on which writing is sometimes incised is metal: a gold tablet of Ashurbanipal, bronze plates both at Byblos and in early Italy, a copper scroll from Qumran, strips of lead with Gnostic Mandaic inscriptions. [See Byblos; Copper Scroll; Mandaic.]

Naphtali Lewis's statement “The need to adapt the cuneiform script of the Fertile Crescent to the pen-and-ink technique of papyrus undoubtedly was a factor in the Phoenicians' development of the world's first alphabet” (1974, pp. 84–85) cannot be defended. [See Papyrus.] Assyrian scribes had no difficulty in imitating impressed wedges on a flat surface (see Baer, 1960). Conversely, the Northwest Semitic linear consonantary was successfully adapted to cuneiform technique at Ugarit. [See Ugarit.]

Writing also appears on pottery (baked clay): when a vessel shatters, no longer serving its original purpose, the sherds can be written on with ink—they are then called ostraca. Ostraca are often the only kind of ink inscription that manages to survive the stresses of the physical environment; they have been found throughout the Fertile Crescent, from Aššur to Aswan. [See Ostracon; Aššur.] Examples of pottery created especially for the purpose of being written on are the apotropaic bowls of Jewish, pagan, and Mandaean tradition. They bear incantations written in a spiral on the bowl's inner surface and were buried upside down under the floors of houses to ward off demons, such as Lilith.

Such ink from the ancient world that has been analyzed generally proves to consist of carbon. Lampblack or incense soot was mixed with water and a small amount of gum arabic to produce a nonfading writing liquid. Iron-based black is not found until about the seventh century CE, and ink made from plant sources (e.g., oak galls) found later, in Europe, is not reported from the Near East. Rubrication could be done in red ink made with ocher, an iron-containing earth (Lucas, 1962, pp. 362–364).

Vegetal Substances.

A blank roll of papyrus has been found in an Egyptian tomb at Saqqara dated to 3100 BCE (Černý, 1952, p. 11). [See Saqqara.] The Egyptians wrote on papyrus as long as their civilization endured; the earliest testimony to papyrus elsewhere in the ancient Near East comes from the Tale of Wenamun (c. 1100 BCE), but it must have been in use well before that. Only the exceptional desiccation of the Egyptian and Transjordanian deserts allowed the survival of papyrus over the millennia.) Papyrus was ubiquitous in classical Greece and Rome and spread with the expansion of the Roman Empire. The Merovingian chancery ceased using it in 677 CE, but elsewhere in France it continued for at least another century. Dated papyri come from Ravenna in 967, Spain in 1017, and the papal chancery in 1057; not until the twelfth century was papyrus completely superseded in Europe (Lewis, 1974, pp. 84–94).

Papyrus “paper” is formed from layers of the pith of Cyperus papyrus, a sedge that formerly grew to 3 meters high in marshy ground throughout the Nile Valley from the Delta as far south as Lake Tana in Kenya, as far west as Morocco, and north into Palestine. [See Delta.] In the Seleucid era it was introduced to Babylonia, and by the tenth century CE to Sicily (Lewis, 1974, pp. 3–20). [See Seleucids.] The chief, if not completely reliable, source for the preparation of the writing material is Pliny the Elder's Natural History (13.74–82). Fresh strips of pith were laid side by side on a flat surface, with their edges touching; more strips were laid on top of them, perpendicularly, again with their edges touching. They were then pressed together (and perhaps pounded with a mallet), and the plants' sap bonded them into a very smooth, white, flexible sheet that could be inscribed on both sides. Usually, the sheets were glued into rolls of twenty, with the horizontal fibers on the inside. The joins were nearly imperceptible, and when a scroll was written in columns, the column breaks did not need to correspond to the joins. [See Scroll.]

True paper is “a matted or felted sheet of fibres formed on a fine screen from a water suspension” (Tsien, 1985, p. 35). It was invented in China, perhaps around the beginning of the common era (rather than in the second century CE by Tshai Lun, the traditional account). The earliest such fibers seem to have been silk, though paper-mulberry bark was used later. Only textile fibers were used outside China (not bark, and not, until the nineteenth century, wood pulp). The manufacture of paper was not a closely held secret monopoly of China, as is often stated. It was made in neighboring areas as soon as they came into contact with the Middle Kingdom. (The Qur'anic word qirtas is believed to be a Chinese loanword in Arabic.) Paper was manufactured in Muslim Samarkand from 751 onward, and in Baghdad by the end of the eighth century. [See Baghdad.] From there it spread to Damascus and across Africa in the ninth century; it was introduced from Africa separately to Spain and Italy in the twelfth and thirteenth centuries, and then spread across Europe (Tsien, 1985, pp. 296–303, citing and correcting Hunter, 1947). The oldest dated paper book in the West is a fragment of the Thousand and One Nights reused by a legal scribe in 879 (Oriental Institute, 1982, p. 130).

Writing Materials

WRITING MATERIALS. Three inkwells from Qumran. (Courtesy ASOR Archives)

view larger image

Writing implements fall under the heading of vegetal substances. Not surprisingly—because it is assumed that the stylus was made of wood or reed—no stylus used for impressing wedges on clay has ever been identified in excavations in Mesopotamia. The literature unanimously assumed that the stylus had to be cut to some angle in order to produce wedge-shaped impressions (Driver, 1976). H. W. F. Saggs (1981) plausibly suggests that a sedge widespread in Iraq, with a stem that is triangular in cross section (like the related papyrus plant), could be used to produce the needed shapes without modification beyond stripping off its leaves. (Unfortunately, he does not supply the botanical identification of the species.) This account also explains why the impressions are wedges rather than dots, rectangles, stars, or anything else—the script did not go through stages of experimentation as to suitable shapes of the components of the characters.

Robert M. Whiting (personal communication) has observed that wall paintings from Tell Barsip show the stylus held palm downward, passing under the palm at a low angle to the clay [See Til Barsip.]; he has determined by microscopic observation that the stylus was rotated through 90 degrees to make horizontal and vertical wedges, noting that the stylus leaves a mark with one smooth side and one side that shows the grain. Horizontal wedges have the grain on the bottom, whereas vertical wedges have the grain on the left side. Thus, the stylus was held in the direction of the writing to make a horizontal wedge and perpendicular to the direction of writing to make a vertical wedge; in positioning the stylus to make wedges of different orientations, the scribe could rotate both the stylus and the medium (which was held in the other hand), to make them meet in the proper relative position.

The notion that the pens used to write Egyptian hieroglyphs were flat-tipped reeds chewed into brushes apparently goes no further back than an unsupported assertion by James H. Breasted of the University of Chicago in 1916. Rather, the sharpness of line even in very early papyri indicates that the pen differed little from that discernible in fifth-century BCE Aramaic papyri or that in current use: it was likely a hollow reed trimmed so that a shallow arch was what made contact with the writing surface, split for a centimeter or so to facilitate the flow of ink (Daniels, 1984).

Animal Substances.

The transition to the category of animal substances is represented by wood tablets covered with beeswax that takes the wedge impression of the stylus. They are known from contemporary accounts, but very few have survived (Driver, 1976, pp. 225–226).

The three terms for prepared animal skins used as a writing ground—leather, parchment, and vellum—are not clearly differentiated. Driver (1957, pp. 1–3) apparently uses the word leather for any writing on skin, including parchment; Alfred Lucas and J. R. Harris discuss leather and parchment in separate sections (1962, pp. 33–37, 38–39); Dard Hunter (1947, pp. 13–16) distinguishes parchment and vellum by both animal of origin and technique of manufacture; and David Diringer (1953, p. 170) agrees with current usage in calling vellum merely “the finer quality” of parchment, “finer in grain, whiter, and smoother.”

Egyptian leather was prepared by treating hides and skins with a number of different substances; true tanning (i.e., with substances containing tannin derived from acacia pods or oak bark) was known in Egypt in the predynastic period. Other materials used include ocherous earth, salt, fat, urine, dung, brain substance, and alum.

For Lucas and Harris (1962), parchment is prepared merely by “first removing the hair and then rubbing the skin smooth with some abrasive material, such as pumice stone.” Other authorities insist that the preparation of parchment involves baths in lime (details in Saxl, 1956). For Hunter (1947), parchment is made from the inner layer of a split skin (the outer layer is used for leather), while vellum is made from the unsplit skin of a calf only. Thus, the two sides of the product differ—one being the hair and the other the flesh side. In gathering folded sheets into signatures for codices, care is taken to face hair side to hair side and flesh side to flesh side, to preserve the uniformity of appearance of each opening. [See Codex.]

It is difficult to date the earliest use of skin for writing. The “Driver letters” (1954, rev. ed., 1957), from the fifth century BCE, are the earliest surviving corpus, but scribes writing on flexible material—either papyrus or leather—are depicted in eighth-century BCE Assyrian reliefs. [See Assyrians.] The Akkadian textual evidence for leather and parchment is even later (Driver, 1976, pls. 23/2, 24; pp. 16–17, 228). By the time of the writing of the Dead Sea Scrolls, it seems that the preferred material for biblical texts was leather (as continues to be the case to this day for sacred Torah scrolls). [See Dead Sea Scrolls.]

It is often said that parchment replaced papyrus because papyrus is inferior in various ways (e.g., Diringer, 1953, p. 165). Lewis refutes this notion, attributing the gradual supersession of papyrus by parchment in the West (and by paper in the East) to the diminution of the supply from Egypt, which was the result not of manipulation of the market or of withholding of the supply, but of the gradual draining of papyrus marshes for agriculture (Lewis, 1974, pp. 57–61; 1989, pp. 27–28). An advantage of both papyrus and parchment over paper is that they can be reused by washing or scraping off existing writing. Fortunately for later scholarship, the underwriting does not entirely disappear from such palimpsests. Many an otherwise lost ancient text has been partially preserved beneath a later text that to modern eyes is of less value.


See also Libraries and Archives; Literacy; Scribes and Scribal Techniques; and Writing and Writing Systems



  • Baer, A. “Un cylindre d'offrande à Sennachérib.” Revue d'Assyriologie et d'Archéologie Orientale 54 (1960): 155–158.
  • Černý, Jaroslav. Paper and Books in Ancient Egypt (1952). Rpt., Chicago, 1985.
  • Cooper, Jerrold S. “Mesopotamian Cuneiform: Sumerian and Akkadian.” In The World's Writing Systems, edited by Peter T. Daniels and William Bright, pp. 37–57. New York, 1996.
  • Daniels, Peter T. “A Calligraphic Approach to Aramaic Paleography.” Journal of Near Eastern Studies 43 (1984): 55–68.
  • Diringer, David. The Hand-Produced Book. London, 1953. Reprinted as The Book before Printing: Ancient, Medieval, and Oriental. New York, 1982.
  • Driver, Godfrey R. Aramaic Documents of the Fifth Century B.C. (1954). Abr. and rev. ed. Oxford, 1957.
  • Driver, Godfrey R. Semitic Writing: From Pictograph to Alphabet. 3d ed., rev. by S. A. Hopkins. London, 1976.
  • Grim, Ralph E. “Clay Minerals.” In Encyclopaedia Britannica, Macropaedia, vol. 4, pp. 700–706. Chicago, 1979.
  • Hunter, Dard. Papermaking: The History and Technique of an Ancient Craft. 2d ed. New York, 1947.
  • Lewis, Naphtali. Papyrus in Classical Antiquity. Oxford, 1974.
  • Lewis, Naphtali. Papyrus in Classical Antiquity: A Supplement. Papyrological Bruxellensia, 23. Brussels, 1989.
  • Lucas, Alfred, and J. R. Harris. Ancient Egyptian Materials and Industries. 4th ed. London, 1962.
  • Saggs, H. W. F. “The Reed Stylus.” Sumer 37 (1981): 127–128.
  • Saxl, H. “A Note on Parchment.” In A History of Technology, vol. 2, The Mediterranean Civilizations and the Middle Ages, c. 700 B.C. to c. 1500 A.D., edited by Charles Singer et al., pp. 187–190. London, 1956.
  • Schmandt-Besserat, Denise. Before Writing, vol. 1, From Counting to Cuneiform. Austin, 1992.
  • Scott, Lindsay. “Pottery.” In A History of Technology, vol. 1, From Early Times to the Fall of Ancient Empires, edited by Charles Singer et al., pp. 376–412. London, 1954.
  • Singer, Charles, et al., eds. A History of Technology. 5 vols. London, 1954–1958.
  • Tsien Tsuen-hsuin. Paper and Printing. (Joseph Needham, Science and Civilization in China, vol. 5, Chemistry and Chemical Technology, part 1.) Cambridge, 1985.
  • Waterer, John W. “Leather.” In A History of Technology, vol. 2, The Mediterranean Civilizations and the Middle Ages, c. 700 B.C. to c. 1500 A.D., edited by Charles Singer et al., pp. 147–187. London, 1956.

Oriental Institute, University of Chicago. A Guide to the Oriental Institute Museum. Chicago, 1982.

Peter T. Daniels