The Book as Object

The Original Tablet

The transition from clay tablets to papyrus as a writing support, modes that coexisted for two thousand years, reveals the extent to which the book’s form and content influence one another’s development. The use of clay to record information arose in Sumer (modern Iraq), which shifted from a nomadic to an urban culture between 8500 and 3000 BCE. As people settled in villages and a system of kingships formed, Sumerians needed a way to track trade and record information about their governance. Cuneiform writing developed in Southern Mesopotamia around 2800 BCE thanks to a confluence of material availability, linguistic development, and utility.² Sumerians had long relied on clay, an abundant and renewable material supplied by the Tigris and Euphrates, the two rivers that give Mesopotamia (which means “between two rivers” in Greek) its name, in their architecture and crafts. While their region did not provide stone or wood in any great quantities, they had highly developed techniques for sifting and working with clay to create durable and lasting artifacts, making it a natural fit as a support for writing.

Initially, Sumerians used clay tokens in various shapes for their accounting, in some cases tying together groupings with string. Around 3500 BCE, they began containing these tokens in spherical clay envelopes, or bullae, impressed with each token’s shape to demonstrate the sealed pod’s hidden contents (see figure 1).³ Rather than associating three conical tokens with three sheep in the world, for instance, this system associated three impressions of tokens with the tokens themselves, a level of abstraction necessary to make the representational leap from spoken language to writing. With impressions standing in for objects, they no longer needed the receipt to serve as a container, so around 3200 bullae became solid, a shape that gradually flattened to create the clay tablet that would serve as a portable recording device for several millennia. Around 3100, scribes began to add designs inscribed with a stylus depicting the goods these token impressions represented, and a pictographic writing system in clay was born.⁴

Figure 1 (a) Clay bulla, based on Louvre Museum SB1940 (ca. 3300 BCE); (b) cuneiform tablet, based on Metropolitan Museum of Art 11.217.19 (ca. 2041 BCE); (c) scroll (ca. 2500 BCE); (d) *jiance,* based on The Museum of the Institute of History and Philology, Academia Sinica 128.1 (ca. 95 CE); (e) palm leaf manuscript/*pothī* (ca. 200 BCE); (f) *khipu*/*quipu* (ca. 1500 CE). Illustration by Mike Force for Lightboard.

While impression in clay worked well, the medium was not really suited to drawing, given the resistance of wet clay to the stylus tip and the challenge of standardizing drawn forms—your sheep and my sheep will likely look quite different unless we have a shared approach to drawing it. Working with the affordances of the clay, Sumerians developed a special wedge-shaped stylus (hence cuneiform, from the Latin cuneus, or “wedge”), also drawn from the material they had in abundance: reeds, which could be easily split and peeled to generate one of these beveled implements ready-made.

With the stylus in one hand and the damp tablet in the other, a scribe impressed a corner of the reed into the clay at an oblique angle, using combinations of wedge shapes to make characters, thus transitioning from pictographic to syllabic writing. This shift from shapes depicting words to signs representing sounds had the additional benefit of reducing the number of characters required to convey information. Rather than a one-to-one correspondence of drawing to object or idea, language could be abstracted from the things it represented, and this reduced phonetic character set could be adapted to represent other spoken languages in the region during the second millennium BCE, facilitating the spread of writing across the Near East.

This early medium, the tablet, takes on epic proportions in popular imagination. The term conjures up images of Charlton Heston as Moses, bearing the Ten Commandments on two stone slabs the size and shape of small headstones. While the Mesopotamians did carve texts into stone, because it was rare in the region it is thought to have been reserved for recording important events.⁵ The reality of the clay tablet was far more modest—most were small enough to fit comfortably in the palm of the scribe’s hand, and their shape, generally rectangular with a slightly convex bulge, suggests the cupped palm that formed them (see figure 1). Ranging from the size of a matchbook to that of a large cell phone, cuneiform tablets were highly portable, could be inscribed on multiple sides, and could rest stably on a flat surface for storage or consultation. Some were cured in ovens, but most were simply allowed to dry in the sun.

Although the history of writing might have to include anything from the cave walls of Lascaux to ancient stellae, to a computer disk or sky-writing, our definition of the book must be narrowed to records in portable form, going back at least as far as the Sumerian clay tablets, and following the availability of materials and techniques of material manipulation.

—Thomas Vogler, “When a Book Is Not a Book”

As the need for written documentation of law, commerce, religion, and cultural history increased, so too did the need for specialists who could both read and write. Thus scribes were born, though in this early period they were seen more as transcriptionists than authors. One notable exception, the Sumerian princess and high priestess Enheduanna composed poems to the moon goddess Inanna in which she addresses her, not on behalf of a king, but as herself:

my Lady

what day will you have mercy

how long will I cry a moaning prayer

I am yours

why do you slay me ⁶

Her name appears both in these poems and on tablets of temple hymns she composed, making Enheduanna the first named author.⁷ Drawn from upper-class families, the young men who predominantly served as scribes were accorded important status for their skill and received an education befitting their task.⁸ Scribal schools provide many of our surviving tablets: copies made by students on clay discs whose rounded form kept them distinct from official documents.

The increase in writing also led to the development of archives to store these texts. The most impressive such collection, that of seventh-century Assyrian King Ashurbanipal of Nineveh, included more than 30,000 tablets and used a topic-based indexing system, laying the foundation for contemporary libraries. Thanks to their clay form, these tablets survived the great fire that destroyed the city in 612 BCE, though countless scrolls and wax tablets were likely lost.⁹ Among the collection were letters and government documents, but also astrological, mathematical, medical, and scientific texts, as well as proverbs, songs, epics, and myths. These were dispersed among a series of rooms, each with a tablet by the door describing the subject matter within. Unlike today’s libraries, the collection was developed not as a public good, but as a symbol of King Ashurbanipal’s stature and scholarly achievement. Evidence suggests the library was also consulted by priests, professionals, and members of the learned class—some tablets are inscribed with threats to would-be thieves demanding borrowed tablets be returned the same day.¹⁰

Perhaps the best-known cuneiform tablet in King Ashurbanipal’s collection is a fragment of an Assyrian translation of the epic of Gilgamesh. Known as the “flood tablet,” it made waves when it was translated in the 1860s because of the similarity of its narrative to the biblical flood story in the book of Genesis.¹¹ Gilgamesh originated as a series of Sumerian praise poems about the king of Uruk (ca. 2700 BCE), raising him to myth-like status. These were combined to create a long epic poem detailing the adventures of King Gilgamesh and his bellicose sidekick Enkidu. Their escapades include war, rivalry, romance, and brotherhood—all the makings of a great road movie.

The version of Gilgamesh found in Nineveh, spanning twelve tablets, forms the basis for most contemporary translations and reveals much about the ways cuneiform literature was shaped by the tablet structure. To facilitate reading, scribes sometimes marked off sections with horizontal lines (easily made by impressing the side of the stylus into the clay) and indented the opening line of the subsequent section. They used special marks to separate words and to indicate a word was a name, and they developed determinative marks that categorized the words to which they were appended (as, for instance, related to people, place names, divinities, or specific materials). These latter are a particularly fascinating case in that they would not be voiced, but rather served as a kind of readerly metadata for disambiguation of words with multiple possible meanings.

Scribes developed finding aids for works spanning multiple tablets like this one, using the tablet’s reverse to provide, in some cases, summaries, a colophon with the scribe and/or owner’s name, the work’s title (generally its first line), and the opening line of the subsequent tablet (known as an incipit) if the text continued. Incidentally, these terms, colophon and incipit, come from manuscript studies, and while they are useful to us in mapping the shifting conventions of the book, they would not have been applied to such texts by their creators. Colophon, Greek for “finishing stroke,” suggests the use of pen and ink to close a text with information about its production on the final page. Incipit, Latin for “here begins,” comes from the scribal tradition of beginning a text with this term to name what follows.

The Papyrus Scroll

While the Sumerians were developing a book from the materials at hand, the Egyptians reached to their own river for a support to writing: papyrus, which only grows in the Nile Valley. Egyptians used the plant widely: for building materials, clothing, and even food. The earliest Egyptian writing appears on stone faces inscribed with hieroglyphics that date from the fourth millennium BCE.¹² Hieroglyphics are sure to be familiar to many readers as a system in which drawings of figures and objects are combined to represent things (pictogram), ideas (ideogram), and sounds (phonogram).¹³ Hieroglyphs were inscribed on temple walls and obelisks, providing religious and historical records, but they also appear on potsherds as more ephemeral notes. As the need for documentation increased and Egyptians sought a more portable surface for writing, they developed an ideal material from papyrus: a paper both smooth and flexible that could be sized to the needs of a given document.

To write on this surface, they developed a water-soluble, charcoal-based black ink and a red ink from oxidized iron, as well as a brush-like rush pen that allowed for smooth and rapid transcription, which gradually transformed hieroglyphics into a simplified script known as hieratic. One of the many ways the material form influenced its content was the indistinguishability of up-and-down strokes in hieratic, which some scholars associate with concern for piercing the papyrus.¹⁴ The uniform thickness of the line suggests even pressure, unlike the thick downstrokes and hairline upstrokes associated with calligraphy on paper and parchment. It might also be attributed to the reed pen itself, a soft brush with a fine tip. Whatever the case, in the fifteen hundred years in which papyrus prevailed, scribes took great advantage of their chosen medium.

Pliny the Elder (23–79 CE) provides a useful, if limited, explanation of Egyptian papermaking (a description he copped directly from Theophrastus) as part of his Natural History.¹⁵ The cyperus papyrus plant, which was used so extensively in ancient Egypt that it was nearly eradicated by the first millennium CE, consists of clusters of long, triangular stems (up to eighteen feet tall) with tasseled heads. Papyrus was made by cutting the stalks into uniform lengths, removing the outer green rind, and making strips from the plant’s pith using one of two methods. According to Pliny, papermakers sliced the pith into strips that gradually decreased in size (given the triangular shape of the stalk), discarding the smallest section. Contemporary research, however, suggests that in some cases the triangular stems were carefully peeled, working inward in a spiral fashion.¹⁶ This would have resulted in a wider continuous sheet and less waste. In both methods, these strips were laid out in two layers—one vertical, the other horizontal—and beaten until the fibers fused, using the plant’s natural sap as an adhesive. The resulting sheets averaged between eight to thirteen inches high and eight to ten inches wide, much like contemporary office paper. These were dried and bleached by the sun, then burnished with a piece of stone or shell, leaving behind a smooth white surface with natural flecks.

Egyptians glued these sheets end to end using starch paste to make rolls of twenty, which could be trimmed into shorter widths as needed and for easier handling.¹⁷ Such rolls were generally inscribed on only one side and in columns so they could be held open to reveal a narrow portion of text, much like a newspaper. The two layers of pith created a natural paper grain that dictated how papyrus could be inscribed and rolled: with the horizontal grain on the inside and the vertical on the outside to prevent cracking as the sheet curled.¹⁸ Once a scroll dried, its curvature would set, making curling it the other way difficult, so in those rare instances in which scrolls have been found with writing on both sides, the reverse generally contains a second text, suggesting reuse rather than continuation (see figure 1).

A book might be best understood as the material support for inscribed language, a category that includes rolls and codices and even monumental inscription, both written by hand and printed by many different mechanisms, and also a wide variety of digital media.

—Jessica Brantley, “The Prehistory of the Book”

Among its affordances, papyrus was durable, could be extended by adhering additional sheets, and allowed texts written on it to be amended, unlike hardened clay. The smooth surface made possible the development of curvy hieratic script, and the use of the brush facilitated the development of colorful illustration, of which Egyptian papyri offer some beautiful examples. Among the best known is a collection of texts to facilitate passage into the afterlife, referred to by ancient Egyptians as the “book of coming forth by day.” Known colloquially as the Egyptian Book of the Dead, this collection of two hundred spells, originally written on burial chamber walls and sarcophagi, was codified around 1700 BCE, when it began to be composed on scrolls for interment with the deceased.¹⁹ The order and number of spells varied from roll to roll, and its design reflected the status of its owner: from elaborately illustrated custom versions for the wealthy, who both selected their preferred spells and were depicted within them, to more anonymous prefabricated templates with gaps for the name of the deceased.²⁰

The papyrus scroll contains precursors to both the codex book and contemporary digital reading devices. The scroll, after all, which allowed continuous writing in columns on a surface that could be thirty to forty feet long, provides the verb we use for horizontal or vertical movement in a text that extends beyond the screen’s bounds. Many finding aids developed for the scroll persisted into codex form. The work’s contents or first words and the name of its creator were written on the outside edge of the scroll, providing an early title page, though the unfortunate placement at the scroll’s vulnerable edge meant that most of these fragile bits were lost over time.²¹ Egyptian scribes took advantage of ink’s variety and incorporated rubrication into their work, using red ink to highlight important words and ideas, as well as to indicate the start of new paragraphs. Such contrast was not possible with cuneiform impression. Rubrication would be adapted by Greek and Roman scribes in their manuscripts, and the tactic persisted into early printed books. Headings, glosses, and titles might be written in red, as would dots and dashes used to separate sections and sentences. In every case, scribes developed techniques to facilitate the reading of written work, one of the hallmarks of the book as not only a storage, but also a retrieval, device.

While papyrus facilitated the development of writing and illustration, it was not an ideal archival material because it becomes fragile as it dries and is susceptible to moisture and insects, particularly in European climates. We have very few intact papyrus scrolls as a result. Another drawback to the scroll form, its tendency to snap shut due to its curled shape, required readers to use both hands to hold it open or lay it on a flat surface and place objects on it, tactics that sounds cumbersome to us today, but that scholars point out would have become second nature among Egyptian readers.²² This normalization of reading practices bears remembering, since from the vantage point of the twenty-first century, our own codex book has been normalized to such a degree that we question the “book-ness” of anything that challenges our expected reading experience, with little regard for the fact that reading in one direction rather than another, scanning text silently, and putting a title and author’s name on a book’s cover are all learned behaviors.

Mineral, Vegetable, Animal

Despite their drawbacks, scrolls persisted for more than two millennia as the dominant book form in Egyptian and Greek culture, to which it was exported. Parchment, developed around 1600 BCE, provided a durable alternative to papyrus and ensured a long life for the scroll in Greece and Rome, where the Latin name for such a roll, volumen, gives us an important foundational term for the book.²³ The term parchment itself comes to us from the Latin pergamum, for Pergamon (in modern Turkey), a key center of its production in the fourth century BCE.²⁴ Made from animal hide rather than plant fibers, parchment was flexible, strong, could be cut in larger sizes than papyrus, provided an exceptionally smooth writing surface, and was opaque enough to allow clear writing on both sides. These features helped it eventually replace papyrus, but its greatest asset was mobility: parchment could be made wherever there was land to raise cattle, goats, and sheep—unlike papyrus, whose manufacture and export Egypt had cornered.

Like Egyptian papyri, Greek and Roman scrolls were inscribed in columns, which the Greeks called paginae, providing both the foundational term and the concept of the page, establishing what scholar Bonnie Mak calls its “cognitive architecture.”²⁵ While we associate the word page with trimmed sheets sold by the ream or bound in volumes, we might, instead, think of it as ancient scribes did, as a means of guiding the reader’s eye and containing information for easy access. These paginae, after all, were similar in size to the content of an 8.5 × 11 in. page, though scroll dimensions varied with content and quality. Works of Greek poetry, for instance, were trimmed to around five inches in height, and epigrams appeared on short, two-inch scrolls suited to their pithy texts.²⁶ With the development of concertina scrolls, discussed in this chapter, the shape of the page as we know it emerges. Folding the parchment between each pagina emphasizes its discrete edges while still allowing continuous reading.

Some scrolls were wound around rods that extended beyond the top and bottom of the roll to facilitate opening and closing. This umbilicus²⁷ (a term that points to the rollers’ centrality but also suggests a Cronenbergian connection between the hand at one end of this cord and the text at the other) could act as a weight if allowed to drape over a table’s edge, holding the scroll open.²⁸ Generally, readers would unroll a scroll with the right hand while rolling it with the left, an active process revealing only a column or two at a time, which meant that to read it again one had to rewind it, much like a reel-to-reel, cassette, or VHS tape. This process takes a ceremonial form in Judaism, where the Torah scroll is publicly rewound on a holiday known as Simchat Torah, or “rejoicing of the Torah.” After the final portion is read, the scroll is paraded around the congregation before being returned to the start so the opening portion can be read as well, symbolizing the cyclical nature of both the year and the text.

While one might expect parchment to have quickly superseded its more fragile counterpart, scrolls of both kinds existed side by side for centuries, much as tablets and scrolls did. Some scholars attribute this parity to the difficulty of systematizing and scaling its production. Parchment and vellum, its highest-quality exemplar (typically of calfskin), were made by skinning an animal, removing the hair from its pelt, bathing the skin in lime, stretching and drying it slowly, then treating the surface to make it hard and smooth. In addition to requiring utmost care, the process necessitated the slaughter of great quantities of livestock, a costly prospect. Much as in our own technological moment, where print books and e-readers continue to be used despite staunch proclamations in favor of the portability, durability, and cost-effectiveness of one or the other, established systems of production and use take time and resources to change.

The Alphabet

The material text, as we have seen, arose largely for administrative purposes: it helped burgeoning cities keep records and accounts, established the power of rulers, and codified ceremonial practice. The movement from orality to literacy plays a central role in the further development of writing to produce literature and its necessary audience: readers. It is through the Greek development of the alphabet that writing gained enough of a foothold to foster the book in the West, so we’ll take a short diversion here to establish how Greeks revolutionized the written word.

Much like Egyptians, early Greeks initially relied on pictographic writing (ca. 2200 BCE), from which a syllabic system (peppered with a few logograms: symbols representing a word, phrase, or concept) developed in the seventeenth century BCE. Syllabics, in which characters represent sounds rather than objects or ideas, both simplify writing and enable a vast increase in vocabulary. Their reliance on consonant-vowel combinations, however, means such alphabets require large character sets to express that vocabulary: upward of ninety in the Minoan script scholars call Linear A and seventy-five in its fourteenth-century Cretan successor, Linear B. This made writing so ponderous only a select group could master it. Such linear written forms, suitable for inscription in clay, were thus used primarily for administrative records.

The term “book,” then, is a kind of shorthand that stands for many forms of written textual communication adopted in past societies, using a wide variety of materials.

—Martyn Lyons, Books: A Living History

The alphabet as we know it arose not from a syllabic source, but a consonantal one that developed in the Sinai Peninsula around 1700 BCE through the mutual influence of Egyptian and Semitic languages. The Phoenicians developed this system of consonants into a twenty-two-character alphabet during the tenth century BCE.²⁹ The Greeks, in turn, adapted these letters to their own spoken language, swapping consonants they didn’t have for vowels and adding new letters for missing Greek sounds. Establishing a writing system of consonants and vowels with one symbol for each sound, the Greeks exploded the capacity of language, magnifying the number of words and ideas it could represent. Not only could these letterforms transcribe spoken Greek, they could adapt to any number of regional dialects and languages, ensuring their spread and development in the West. The Romans adapted Latin from an Etruscan variant of Greek around the seventh century BCE, and as their writing developed over the next two centuries, they established the familiar letterforms that would become the most widespread in the world.

The Greek invention of the consonant-vowel alphabet assured the development of literacy and the shift from tablet to scroll. This alphabet proved easier to learn and write than its predecessors, since it involved far fewer signs. And, like Egyptian hieratic script, it was better suited to papyrus than clay, leading to widespread adoption of scrolls. This reliance upon papyrus led to another important Greek invention: the pen. Cut from a reed and possessing a sharp split tip, the ink pen ensured clarity and speed—it was a vast improvement over the Egyptian reed brush and was eventually adopted, along with the Greek alphabet, by the Egyptians themselves in the fourth century.³⁰ The enhanced speed of writing in turn influenced the alphabet. Initially written in straight-sided majuscule only, Latin grew curves in the fourth century CE, and minuscule letterforms arose in the fifth century, perhaps because of the increased production of Christian codices at that point. We’ll return to upper- and lowercase letters when we look at the invention of movable type, but it bears noting that writing’s form and materials developed in dialogue with one another, even in this early stage, shaping the form of the book to both writers and readers.

Scrolls and the Advent of Paper

Like Mesopotamians with clay and Egyptians with papyrus, the Chinese developed their earliest book form, jiance, from the plant fiber they had in abundance: bamboo. A versatile material, it supported architecture, agriculture, and the arts before its adoption as a writing substrate. At around the same time King Ashurbanipal was amassing his library, Chinese artisans were cutting bamboo stalks at their natural joints, splitting them into strips about half an inch wide, cutting these into slips of equal length, and curing them over a fire. They polished these slips smooth on one side (known as the yellow side), and tied them together with cords of silk, hemp, or leather so they could be rolled up like a mat for storage and transport (see figure 1).³¹ Though in most cases their cords have disintegrated, some jiance include gaps in the text or notches in the bamboo that indicate where such bindings would have been placed. The very name for these Chinese scrolls, scholar Liu Guozhong points out, “ce,” which translates loosely to “volumes [of strips]” (as in jian/bamboo strip + ce/volume), is a pictogram representing uneven strips of bamboo encircled with string: 册.³² Incidentally, this form provides an excellent model for the idea of grain: the direction in which a sheet of paper’s fibers lay. One folds paper parallel to this grain to get the smoothest crease and ensure pages will turn easily. As book artist Scott McCarney demonstrates to his students with a sushi mat, folding a jiance against the grain will snap it like a bundle of twigs and likewise crack the paper’s surface.³³

This preliminary form influenced the very shape of Chinese writing. Jiance were inscribed in ink using a fine, stiff-bristled brush. As with Egyptian and Greek scrolls, a knife was used to scrape away mistakes, though saliva and water also did the trick. The traditional Chinese style of writing from top to bottom arises directly from the book’s materiality—a bamboo slip was too thin to permit more than one character per line. They were thus inscribed from top to bottom in a column of single characters, and the text continued to the left. One would expect scribes could just as easily have developed a top-to-bottom, left-to-right orientation, but here the form again impacts the content. Because scribes wrote with their right hands, blank slips were held in the left. Moving the painted strip to the right to dry and adding a blank slip to the left was the most expedient approach, and thus they were bound in this fashion.³⁴ In addition to this columnar orientation, scribes established other characteristics that would continue into manuscript and printed books at this point, numbering their slips either at the bottom of the yellow side or on the back, including chapter numbers and titles on the outermost slips of a roll, and inscribing titles or colophons on the reverse (much as we saw with clay tablets).

The narrowness of the strips contributed not only to the way texts were laid out but also to the evolution of the characters themselves. Because the strips were so thin, scribes developed vertical ideograms that could be written more easily on them. Guozhong gives the example of the characters for horse and pig, both of which appear to be standing on their hind legs, rather than on the ground, as we might expect.³⁵ In addition to fitting comfortably on their supports, figures for humans and animals also face to the left, indicating the direction in which writing and reading will proceed.³⁶

While writing on other surfaces, including tortoise shells, pottery, bronze artifacts, and seals, appears as early as 1400 BCE, jiance provide the first portable method for distributing information in China, including funeral lists, divinations, and civic records early on, before expanding to include medical books, philosophical and scientific treatises, and literature (a pattern that repeats in nearly every culture to develop writing). The best-known example of a book composed of slips might be the Yijing or I-Ching (Book of Changes, ca. 1000 BCE), a divinatory text consulted by casting lots to generate hexagrams whose meaning must be interpreted. The oldest complete jiance, whose seventy-seven intact strips are threaded together with hemp cord, dates from between 93 and 95 CE and consists of a monthly weapons inventory for the platoon to which it belonged.³⁷ While bamboo slip scrolls were durable and portable, in lengths over two feet the rolls would have been cumbersome to transport and read. Light, durable silk, another fiber whose production the Chinese perfected, was also used for writing and illustration, particularly between the third and fifth centuries, but it was far more expensive to produce, and those silk scrolls that have been found are thus thought to have been luxury goods.³⁸

For the book to take hold, a cheaper, lightweight alternative to bamboo and silk had to be developed: paper, attributed to Cai Lun, a eunuch of the Han imperial court, though examples precede him by two centuries. Cai Lun reportedly presented a method to the emperor in 105 CE whereby a mash of hemp, mulberry bark, fishing nets, and rags was suspended in water (known in papermaking as a slurry) and then sifted through a fine screen to lift out the matted fibers, which were then dried and bleached.³⁹ The random orientation and interlocking of the fibers made paper durable and flexible. These sheets were adhered into scrolls much like those in use in Egypt and would be inscribed in the same method as jiance. Once again, while we might expect this new technology to have quickly superseded its predecessor, scrolls of paper and bamboo continued to be used alongside one another through the fifth century, when Emperor Huan Xuan felt the need to issue an imperial decree commanding bookmakers to use paper and stop relying on what he believed were antiquated techniques.⁴⁰

We should keep in mind that no text exists outside of the physical support that offers it for reading (or hearing) or outside of the circumstance in which it was read (or heard). Authors do not write books: they write texts that become written objects—manuscripts, inscriptions, print matter or, today, material in a computer file.

—Guglielmo Cavallo and Roger Chartier, A History of Reading in the West

Paper and Islam’s Golden Age

Despite its superiority as a material, paper did not reach the West until 751. Its arrival was fortuitous, circuitous, and linked to the spread of the Islamic Empire. When Muslim soldiers captured a group of Chinese sailors in battle, they immediately recognized the utility of the paper found aboard their vessel.⁴¹ These Chinese prisoners of war built the first paper mill in Samarkand, and the papermaking techniques they taught their captors spread rapidly throughout the Arab world, replacing papyrus entirely by the tenth century.

As the Islamic Empire expanded, bookmaking flourished, fed, in part, by the availability of paper. Islamic artisans developed the process they inherited, and the essentials of their paper milling system continue to this day: the replacement of mulberry fibers, unavailable in the region, with linen rags; employing trip hammers to mechanize the slurrying process, rather than beating the fibers by hand; and the use of laid-wire screens that leave faint lines on the finished sheet. The high value placed on writing and intellectual pursuits in Muslim culture, coupled with the expansion of the empire, led to a boom in bookmaking between the eighth and thirteenth centuries. Some calculations suggest Islamic book production during this period of expansion was greater than that of Greece, Rome, Byzantium, and Christian Europe combined.⁴²

This “golden age” of Islamic culture, marked by significant advances in philosophy, mathematics, science, and medicine, also saw extensive translation and annotation of ancient Greek texts inherited through the conquest of Egypt and Syria. Caliphs collected millions of books in libraries in Baghdad, Cairo, Lebanon, Córdoba, and elsewhere, furthering scholarship and book production throughout the empire. It was ultimately through Islamic Spain that Europe would receive paper in the twelfth century, at which point the codex had become the dominant book structure. The founding book of Islam, the Qur’an, had embraced this form from the outset. The text was initially transmitted orally by the prophet Muhammad and his followers in the first decades of the seventh century, but because they were soldiers actively engaged in conquest, the narrative was threatened by their own bodily precarity. To prevent its loss, Caliph Abu Bakr had the text committed to the page soon after its appearance.⁴³ After Muhammad’s death, Caliph Uthman compiled an authoritative version, and the Qur’an took shape as a parchment codex.

A Networked Text

While we are considering the deep history of the book’s material forms, we should look briefly at an example that arises elsewhere, outside the lineage of the codex. The South American khipu (or quipu in the Spanish spelling) provides an entirely different method of recording—knots (khipu in the Quechua language) in string. While our earliest samples come from the first millennium, most archeological specimens represent the Inka Empire (roughly 1400–1532).⁴⁴ At least one proto-khipu, which includes twigs knotted into the strings, likely dates to the third millennium BCE, making it concurrent with cuneiform tablets.⁴⁵ These devices consist of a main cord of cotton or wool with pendant strings affixed to it, most descending, while a few are attached in the other direction (see figure 1). Some of these pendants include further, subsidiary strings dyed in a range of colors and spun in multiple plies. When displayed in museums, khipu are often mounted linearly, like a timeline on the wall, or radially, like a clock face or dramatic rope necklace. While some are relatively simple, a handful include over fifteen hundred pendants, making them intricate and beautiful objects of study.

Scholars are still deciphering these dense systems of knots, based on around 750 specimens found in burial chambers in the Peruvian Andes and on descriptions of their use by the Spanish colonists who first encountered them. The opacity of the khipu is a legacy of colonization: Spanish authorities outlawed their use in the sixteenth century, destroying many of them in the process. The only khipu to have remained within their community are those of Rapaz, Peru, where they were kept until the 1930s.⁴⁶ While they are still maintained and used in ritual, they are no longer updated. In 2012, Chilean artist Cecilia Vicuña published a contemporary, conceptual artist’s khipu. Produced in an edition of thirty-two, her Chanccani Quipu (2012) takes the knot-book as metaphor, in her words, for “the clash of two cultures and worldviews: the Andean oral universe and the Western world of print.”⁴⁷ Knotted to a bamboo spine, four-foot cords of unspun wool cascade like soft white hair to the ground when the piece is hung. Stenciled in rust-colored ink onto their surface, fragmented Quechua phonemes extend across and down each strand, inviting eye and ear to sound the chords of language.

Khipu were maintained by a specialized bureaucratic class of knot-makers, khipukamayuq, who understood the conventionalized signs represented by different configurations of knots, colors, and types of string. Their role was to breathe life into the threads, conducting the census, documenting taxation and labor, and maintaining calendars of ritual practice. The chief question, whether their knots served as mnemonics or as a written system that could be read, remains unanswered. These devices were certainly portable, like a book, and they could contain both accounting records and narratives. Scholar Gary Urton, founder of Harvard’s Khipu Database Project, contends that the system represents a form of binary coding that helped keepers recount narratives based on relational pairs.⁴⁸ On some quantitative khipu, knots are arranged in a decimal hierarchy based on distance from the primary string, with the 1s being furthest, then 10s, then 100s, and so on. Other khipu have a more complex arrangement, which suggests they were used to record some other kind of information, like images, ideas, or sounds.

While the khipu represents a distinct departure from the kinds of record keeping we have so far traced, it is knitted to the history of the book’s changing forms by its materiality. Like the clay tablet, papyrus sheet, and bamboo scroll, it was made from a material both abundant and highly refined within the culture in which it arose: cloth. The cultivation of alpacas, llamas, and other camelids in the Andes provided raw material for the perfection of weaving. Incan cloth was pliable, three-dimensional, visual, and already embedded in the culture as a marker of status.⁴⁹ Colored embroidery and geometric designs had clear symbolic significance, and knotted tapestry patterns in royal clothing were used to establish authority. As in each of the other early book forms we have looked at, the chosen material formed part of the fabric of existence for its users and likely shaped not only how information was transmitted but also the very nature of thought itself. Our challenge, as students of the book, is to think about the way its materiality is both a product and constituent of its historic moment.

In China, an entire mountain can be a book, a calligraphic catalogue of historical events. For example, in Quanzhou, there is a mountain whose cliffs are inscribed with maritime stories, such as historical accounts of Zheng He’s travels to the Western Ocean. In this way, even nature can be considered a medium for writing books.

—Cai Guo-Qiang, Statement in Shu: Reinventing Books in Contemporary Chinese Art

From Roll to Accordion to Codex

Given this range of antecedents, where does the codex, that ubiquitous structure we all recognize, come from? How do bookmakers transition from continuous scrolls to bound volumes, and why? There are a few different ways to answer this question, depending on which material you used to make your scrolls. In China, paper’s strength and malleability led to the development in the eighth century of sutra-folded books, named for their use in producing copies of the Buddhist sutras. Scrolls were folded back and forth in even widths to create an accordion (see figure 2). Such folded books facilitated reading by allowing access to any part of text and would play a key role in both spreading Buddhism throughout Asia and establishing the codex in China. Incidentally, the term “role” itself is a legacy of the scroll, a reference to the rolls on which actors’ parts were written during the Renaissance.⁵⁰

Figure 2 (a) Wax tablet diptych (ca. 800 BCE); (b) quire/parchment notebook (ca. 55 BCE); (c) accordion/concertina (ca. 700 CE); (d) butterfly binding (ca. 800 CE); (e) wrapped back binding (ca. 1200 CE); (f) stab binding (ca. 1300 CE). Illustration by Mike Force for Lightboard.

The folded book, also known as an accordion or concertina, was a flat, rectangular volume, and its height came directly from the scrolls that preceded it, since the paper sheets used were generally the length of the papermaker’s arm. The format may also have developed from the palm leaf manuscript, or pothī, carried by monks between India and China starting at the beginning of the Common Era. Produced in many scripts and languages, such collections of Hindu and Buddhist teachings were copied and memorized as a form of devotion by scribes across Asia. Using dried palm leaves that had been flattened, polished, and trimmed into rectangular strips, they wrote these sutras with an early intaglio technique: text was inscribed with a stylus, ink or soot was applied to it, and the excess was wiped away, leaving behind darkened impressions.⁵¹ The leaves were bound in a style reminiscent of the venetian blind: encased in wooden boards of the same size and shape as the leaves, one or several holes were drilled through the lot and a cord passed through to hold them together (see figure 1). The name sutra itself relates to this form: it derives from the proto Indo-European root syū, “to bind, sew,” which yields the Sanskrit sūtram, meaning “thread [or] string.”⁵² The Chinese accordion book turns this volume on its side, adapting it to the vertical writing method established in the jiance.

The earliest example of a woodblock-printed book comes from China as well. The Diamond Sutra, a fourteen-foot scroll commissioned in 868 CE by Wang Jie, was made from a series of carved wood blocks containing both illustration and text.⁵³ These were rolled with ink, and a sheet of paper placed on top was rubbed to transfer the ink to the page. This type of printing, known as xylography, was highly portable and could be completed by a single person (unlike early movable type, which would require a heavy press and several artisans to operate). It had the added benefit of enabling the reproduction of multiple sequential pages simultaneously.

The rubbing technique, coupled with the thinness of mulberry paper, meant that only one side of a page could be printed—rubbing the inked side to print on the reverse would have smudged it. This restriction likely led to the development of glued bindings from the concertina.⁵⁴ For the butterfly binding (ninth to thirteenth century), individual sheets were printed with two facing pages, then folded in half, stacked, and glued along their crease. This results in a codex whose open, printed folio perches like a spread-winged butterfly on the closed folios on either side (see figure 2). Unfortunately, it also leaves two blank pages between each printed pair, a trait that both breaks readers’ concentration and encourages skimming.⁵⁵ To create an uninterrupted reading experience, printers developed the wrapped back binding (thirteenth to seventeenth century).⁵⁶ Sheets were folded outward, so text appeared on both sides, and piled, and holes were drilled through the open edges of the stack (see figure 2). Bookmakers threaded these holes with thick paper pegs to keep the pages together, then glued the spine and wrapped a stout paper cover around it, creating a stronger book block and hiding the blank sides of the page while also producing the first glue-based or perfect binding.

Because glued bindings attracted insects, they eventually gave way to a thread-based system: the stab binding most associated with Chinese and Japanese books. This sewn binding used the same folded-folio technique as the wrapped back binding, but rather than gluing covers on, a single thread was passed through a series of holes along the spine to create a decorative geometric pattern. In addition to being less appealing to bugs than glue, the sewn binding was easier to repair, since it could be taken apart without damage to the pages (see figure 2).⁵⁷ The stab binding remained popular until the early twentieth century, when the Chinese book began to more closely resemble the codex.

The Greeks and Romans had an accordion-like structure of their own before the introduction of papyrus or paper: wax writing tablets. Employed as early as the eighth century BCE in ancient Greece (and borrowed from the Assyrians, who used them from the fourteenth century BCE on),⁵⁸ they were made by carving a rectangular depression in a wood slate and filling it with wax. Known as pugillares, their name derives from pugnus, or fist, suggesting these “handbooks” offered sturdy and portable writing surfaces that could be held with one hand, much like a contemporary e-reader.⁵⁹ They also clearly demarcated the boundaries of the pagina. A similar device can still be found in novelty shops—the magic writing pad that so intrigued Sigmund Freud, a gummed surface protected by a layer of cellophane. Using a plastic stylus or the tip of a retracted ballpoint pen, one can write secret messages on the surface of the plastic that are easily “erased” upon reading by lifting the cover away. Early wax tablets were used for secret messages as well. According to Herodotus, when the exiled king Demaratus wanted to warn Sparta of an impending Persian attack (ca. 480 BCE), he carved his message in the wood of a tablet and covered it with wax to avoid interception. The warning reached its readers who were, at first, perplexed by the blank slate. Herodotus credits Queen Gorgo with the idea to scrape the wax away, revealing the secret message in time to warn the people.⁶⁰

Wax tablets were inscribed with a stylus that was sharp on one end and flat on the other, allowing one to erase one’s mistakes. Writers might warm the stylus with their lips for easier inscription, as some ancient artworks suggest. Tablets could be used singly but were often threaded together with leather cord into pairs or polyptychs, gatherings of as many as ten tablets. When more than two tablets were threaded together, either end to end or along one edge, those in the middle could be inscribed on both sides.⁶¹ The bound group of tablets, known to Romans as a codex, is suggestive of the form’s wooden supports, though the term was also used for tablets carved from other materials, including ivory and bone.⁶² Representations of such bound tablets on Greek vases suggest they were held not like our codex, with the hinge running up the middle, but more like a laptop, with a horizontal hinge between two writing surfaces (see figure 2).

A book is never simply a remarkable object. Like every other technology, it is invariably the product of human agency in complex and highly volatile contexts.

—D. F. McKenzie, Bibliography and the Sociology of Texts

The Greeks initially used wax tablets for important documents like wills and birth announcements,⁶³ but with the introduction of papyrus and parchment rolls, they began to use them in education and for more ephemeral documents like notes and household accounts. Romans used pugillares and codices extensively, particularly for legal documents (giving us the term codicil). They are credited by most scholars with shifting the material of the codex from wood to skin in the first century CE—a process that likely began with the use of small parchment notebooks, or membranae.⁶⁴ Horace (ca. 65–8 BCE) mentions their use for drafting in both his Satires and Ars Poetica, suggesting groups of folded pages provided a lightweight alternative to the wax tablet for composing longer texts before committing them to papyrus rolls.⁶⁵

These bundles of folded pages provide the essential concept from which the codex developed in the Common Era, though the inventor of the form remains unknown. Folded papyrus and parchment sheets from the first to the fourth century CE, wrapped with leather and sewn with a loop of thread along the fold—a sewing now known as the saddle stitch—provide some of our earliest codices, each using a slightly different method to produce a quire, or gathering of folded pages (see figure 2).⁶⁶ A single sheet can be folded into a variety of gatherings named for the number of smaller sheets (known in bookmaking as leaves) they generate: folio (two), quarto (four), octavo (eight), sextodecimo (sixteen), etc. (see figure 3), and book historians use these names to describe a work’s size. Papyrus gatherings consisted of nested folios, sheets folded in half with the grain, and are thus the same height as papyrus rolls. Parchment gatherings, however, were made from a range of pelt sizes and, given the malleability of the material, could be folded and cut several times before binding. The most common such folding, in octavo, yields the gathering of four nested folios that give the quire its name.⁶⁷

Figure 3 Single sheet folding and resulting book sizes. Illustration by Mike Force for Lightboard based on A. W. Lewis, *Basic Bookbinding* (New York: Dover, 1957), 9.

When a large sheet is folded down, the resulting pages are called conjugates because they are still attached. When the conjugates have been sliced along their folds, the pages become leaves, each with two sides: recto (front) and verso (back). These terms come from the names Romans gave the sides of a papyrus scroll: recto for the horizontally ruled interior that was the right place to write, and verso folium, for the “turned leaf” of the scroll’s back.⁶⁸ When looking at a two-page spread or opening in a codex, the left-hand page is always a verso and the right always a recto.

This easy-to-assemble format, in which a single sheet is folded down into a quire, played an important role in European publishing from the sixteenth to the nineteenth century, with itinerant peddlers known as chapmen hawking such cheap pamphlets to the masses at a time when books were luxury goods.⁶⁹ These chapbooks, ranging in length from four to twenty-four pages, continue to play an important role in small press publishing today, where they are used for compact, inexpensive collections of poetry.

It bears repeating that the advent of the codex doesn’t mean scrolls disappeared—scrolls and codex books made of both parchment and papyrus existed side by side within Roman culture for centuries. Archaeological evidence shows a gradual decline in the number of rolls from the third century onward, with codex books increasing, until the two reach parity between the third and fourth centuries.⁷⁰ The extant codex books from this time include works by Homer and Plato as well as treatises on medicine and grammar, which suggests they were used in education. Each tends to be only as long one or two scrolls, evidence that the predominant format had established in readers’ minds some concept of a book’s duration. To make the leap from the pamphlet-like notebook to the codex as we know it required more complex bindings that could link several quires tightly together. Our earliest complete example, the Mudil Codex, a fourth-century psalter discovered in an Egyptian Graveyard in 1984, reveals the bound book’s Roman roots: its thirty-two quires are enclosed in wooden covers and stitched with leather.⁷¹

The Manuscript Tradition

The side-bound book as we know it is well suited to the needs we have seen so far. It is portable and durable, it is easy to reference, and its economical pages allow writing on both sides. Unlike the scroll, it does not require both hands to pry apart, and like the tablet, it can rest open on a surface for consultation. But perhaps its spread is most attributable not to what it is, but to what it isn’t: a scroll like that used for the Hebrew Scriptures—the Torah—and pagan religious texts. Early Christians, the essayist and book historian Alberto Manguel suggests, embraced the codex as a means of clandestinely transporting texts banned by the Romans.⁷² This differentiation would serve an important purpose with the rise of Christianity in the Common Era, as Christians and Jews selected bindings for their religious tracts to reinforce their distinction—monks even bound the Septaguint into codex form for inclusion in monastic libraries, suggesting the extent to which such differentiation was internalized.⁷³

It is through the rise of Christendom that codex book production developed in the West—in the form of monastic manuscripts. In the sixth century, as the first monasteries were establishing Catholicism in Italy, St. Benedict of Nursia issued a rule requiring Benedictine monks to read daily, complete a book during Lent, and carry books when traveling to peruse each time they stopped.⁷⁴ The emphasis placed on literacy led to a boom in book production within monasteries, each of which had its own library and a scriptorium for copying texts by hand. Staffed by scribes, correctors, calligraphers, and rubricators, it produced codex books for trade or sale to other monasteries predominantly for the consultation of other monks. Monasteries monopolized book production until the thirteenth century, and when you picture early books, it is likely monastic illuminated manuscripts that come to mind. These ornate productions included both Christian texts and ancient Greek and Roman works, copied and recopied by scribes who hadn’t exactly signed up for the task of preserving and disseminating literature.

The monks who served as scribes did not, in fact, relish the task. While their brothers worked in the fields or traveled, they spent six hours a day hunched before the page in a cold scriptorium, incurring back-aches, headaches, eye strain, and cramps, all while wasting away the daylight hours, since candles were costly and fire was a great risk to their highly flammable materials.⁷⁵ They labored in silence, communicating by hand signal when they needed materials or simply wanted to commiserate about their lot. This they sometimes did in the margins of their pages, leaving complaints in ink among the text’s glosses like this one, reminiscent of a schoolchild’s plea: “St. Patrick of Armagh, deliver me from writing.”⁷⁶ Scribes need not have understood the Latin, Greek, and Hebrew texts they copied, since literate correctors and proofreaders ensured the quality of their work.⁷⁷ Some roles in the scriptorium could be filled by laity, including calligraphy, rubrication, and illumination. Manuscript production thus employed a great number of people in the manufacture of a single book through a time-consuming process. Bound in wood and leather covers, sometimes encrusted with jewels and filigree, and decorated with bright colors and gold leaf, the labor and expense that went into each codex manuscript shows tangibly on every page.

Scribes worked on gatherings of four parchment folios, known among book historians as quaternions (the source of our term quire), whose sixteen pages were arranged with “hair” sides facing “hair” sides and “flesh” to “flesh,” an effect of the octavo fold described above.⁷⁸ The two sides of a sheet of parchment, though treated, scraped, and stretched, bear the history of the animal from whom it was made: the “hair” side is visibly darker and freckled with tiny follicle holes.⁷⁹ Dividing manuscripts into these smaller sections enabled several scribes to work on the same volume simultaneously, speeding up the process, but multiplying the potential for error, since each quaternion had to align in appearance and content with its neighbors in the finished book. To keep their writing clear and uniform, scribes pierced or marked lines and margins on their pages. Their layout was drawn from a familiar source—Roman scrolls and codices: a series of justified columns, two to a page, both efficient and pleasing to the eye.

Scribes copied the text before them with a goose quill in one hand and a knife in the other, the latter allowing them to hold pages flat, sharpen their pens, and scrape away mistakes. They wrote in a majuscule style, or hand, known as uncial, developed from Roman letterforms, but with curved edges.⁸⁰ This hand persisted for religious texts through the eighth century, when Charlemagne commissioned a more legible, lowercase hand to standardize manuscript writing. Known as Carolingian for its association with his reign, it was much faster to write, so scribes could produce more quickly with this small affordance and keep up with an increasing demand for books.⁸¹ By the fourteenth century, Gothic scripts in a variety of regional styles had developed across Europe, and these would provide the basis for our earliest typefaces, a subject we will return to in chapter 2.

When a quaternion was completed and proofread by a corrector, the rubricator took over, continuing the Egyptian tradition of embellishing important passages in red ink. At this point, titles, chapter initials, and headings were added. If the manuscript was important, or if a wealthy patron commissioned it, the rubricated text was next turned over to the illuminator, whose marginal paintings illustrated and beautified it. Working primarily in red and blue pigment and gold leaf, the illuminator added flourishes to the initial letters of passages (sometimes painting little scenes around the capital to create decorative letters known as historiated initials), decorated the page’s borders, and introduced illustrations depicting the text’s themes (see figure 4). This method persisted through the fifteenth century, when it reached its height with about 10 percent of all manuscripts receiving illumination.⁸²

Figure 4 Page from an Italian breviary by an unknown artist. The freckles in the upper left corner are hair follicles. Initial V: The Descent of the Holy Spirit, 1153. Tempera colors, gold leaf, gold paint, and ink on parchment. Leaf: 19.2 × 13.2 cm ( × ). Image courtesy The J. Paul Getty Museum, Los Angeles, via the Open Content Program.

One of the most well-known illuminated manuscripts, the Irish Book of Kells (ca. 800 CE), demonstrates both virtuosic technique and the innovations introduced by scribes of the British Isles: an elegant uncial script, richly decorated initial letters, and, most importantly, word spacing.⁸³ The four gospels and prefatory texts in Latin are written in several colored inks (in addition to the standard black), and are more densely illuminated than any surviving Gospel book of the period. Likely the work of at least three scribes, the Book of Kells reflects the intricate and laborious process that made manuscripts such costly and tightly controlled products.⁸⁴ Lavishly illuminated throughout its 340 vellum folios, with ten full-page illustrations, it includes interlaced motifs that mimic Celtic metalwork, red dots evoking Roman tradition, and illustrations influenced by Byzantine, Armenian, and Mediterranean iconography.⁸⁵

With its large scale (unfortunately cut down to around 13 × 9 in. for rebinding in the nineteenth century) and sumptuous appearance, The Book of Kells would have been an altar-book reserved for use during special services by a reader who had memorized the text. It was meant more to be seen and heard than studied, as evinced by a number of errors in the text including missing words, repeated passages, and illustrations added to cover up mistakes.⁸⁶ Such large volumes required an important seventh-century development in codex binding: the use of stout cords or leather thongs, threaded through the cover and run across the book’s spine, to which each quire was sewn. These supported bindings, perfected in the seventeenth century, continue to be standard in both the repair and production of fine books.

Reading and Writing’s Shifts

The shape and style of these early manuscripts reflect the reading practices of their day and the needs they were designed to meet. Reading was, in the manuscript era, a practice fundamentally different from the kind of private, meditative engagement we now experience. A monk did not sit silently at a desk or reclining in bed or while in transit from one place to another. He might be read to in assembly by a fluent reader among his brothers, or he might mumble to himself as he learned the Latin text. When he traveled to other monasteries to consult their volumes, he found codices chained to lecterns at pews, ensuring these valuable documents did not walk off. Copying the book would require the significant effort of sliding the neighboring volumes off the metal rod to which they were affixed before he could take his own to a desk that would accommodate his labor. Such chained bindings persisted through the eighteenth century and are emblematic of the different place the codex held in cultural and religious life. Each book was a unique and hard-wrought object to be enjoyed by a limited audience.

Reading had been, since the Hellenic era, an oral practice—one reflected in writing itself. Greek bookrolls were written in a continuous script, or scriptio continua, withoutspacesbetweenwords or changes of case and minimal punctuation. They both required and rewarded sounding aloud. As Paul Saenger, curator of rare books at the Newberry Library in Chicago and a scholar of medieval reading practices, explains, continuous script could not have developed without the Greek introduction of vowels, which allowed readers to parse syllables and hold them in memory as the eye traversed the text.⁸⁷ While Greek was composed from right to left, like Phoenician, the Romans developed a method of speeding up reading by reversing direction from line to line, allowing continuous reading as one scanned from right to left, then left to right, and back again. This boustrophedon form, named for the method of plowing, allowed farmer, writer, and reader to crisscross the field without lifting their instruments, suggesting that it was far less cumbersome than we might assume.

Though it seems awkward to us now, continuous script was not a naïve construction, but a choice, as evinced by the fact that the Romans discarded their own punctuation in the first century in favor of the Greek model. It established literacy as the purview of a cultured elite, who either studied from a young age to master the identification and inflection proper to each text or employed a professional reader, or lector, for the task. It also facilitated a culture of shared inquiry, in which challenging texts were read aloud in groups as a springboard for debate.⁸⁸ In ancient Greece, literature was primarily a social activity, with audiences gathering for performances of epic poetry and drama. The epic bears the hallmarks of this orality: it relies on repetition, formulaic images, meter, and rhyme as mnemonic aids to the performer.⁸⁹ The term used to describe performances of such works, rhapsody, means “to stitch together”—suggesting the extent to which oral composition relies on weaving familiar lines.

The great thinkers of ancient Greece, in fact, mistrusted writing as a technology that would destroy the oral arts of debate and storytelling on which they based their sense of the world, of philosophy, of time and space. In Plato’s Phaedrus, Socrates disdains the written word for separating ideas from their source, citing Egyptian king Thamus as the first to voice this concern when he received the gift of writing from the god Thoth.⁹⁰ Transcription, Socrates fears, is a crutch that will both hamper memory and mire philosophical thought in ambiguity, leaving interpretation in the hands of the reader. Texts, after all, can circulate without their author, thus preventing one from explaining or defending them. Despite these fears, the very writing Plato used to record his dialogues proved instrumental in the development of ancient Greek oratory. As scholar Walter Ong points out in Orality and Literacy, his study of the ways writing technologies restructure consciousness, the written word enabled Greek scholars to transcribe and codify effective rhetorical strategies.⁹¹ It also vastly increased human vocabulary, since we no longer had to rely on memory to hold all of language at the ready. Writing, in fact, allowed rhetoric to flourish.

For the kind of silent reading we now experience to take hold, reading would have to change its context and text its form. It would have to become a more private experience, which means literacy would have to extend beyond the elite and monastic communities. Texts, too, would need to become more legible, with standardized punctuation and word spaces so that the mumbling of readers sounding out text, common through the sixth century, could dissipate. And libraries designed for quiet, contemplative reading could then develop to serve this new readership.

Insular scribes, like those who crafted The Book of Kells, played a key role in making text more accessible. Because Latin was a second language, and one more challenging to sound out in scriptio continua, they introduced several changes to improve its legibility, including word separation (around 675 CE), additional punctuation, and simplified letterforms. Still, it took nearly four hundred years for these small innovations to spread.⁹² The translation of Arabic scientific writing in tenth-century Europe likely played a vital role in solidifying word separation, since it was inherent in the language (because unlike Greek and Latin, it is written in consonants). Translators kept Arabic word separation when rendering these texts in Latin, in part because it made the complex technical prose significantly more comprehensible—an important case where content had a direct influence on form.⁹³

What is a book? A book is an experience. … A book starts with an idea. And ends with a reader.

—Julie Chen and Clifton Meador, How Books Work

Ultimately changes in who read and what they accessed were essential in reshaping writing and the look of the page. The population boom in Europe during the late Middle Ages meant a middle- and upper-class laity needed to be educated, and therefore needed access to books. Universities developed in the thirteenth century to provide education for both clerics and the aristocracy beyond rudimentary language and rhetoric. These institutions drove a new market for the exchange of ideas: academia, an almost entirely male province.⁹⁴ To meet the rising demand for and production of texts, guilds of stationers, acting as copy services, binders, booksellers, and book lenders cropped up to serve both faculty and students. Using exemplars approved by the faculty, the stationers sent quires out to copyists and illustrators who completed the work piecemeal and sent it back, then arranged and bound the finished book in-house.⁹⁵ The piecemeal system meant that many more copies could be made, with identical sections interchangeable between them. For a small fee, students could make their own copies from these exemplars, section by section—a highly manual reading process. They then brought these books to class with them, reading along silently with the professor’s lecture,⁹⁶ thus learning multimodally, with hand, eye, and ear.

Here again we have the Arab world to thank for the spread of ideas and shifts in reading. Scholars translated ancient works preserved in Arabic into Latin, and works of science and mathematics began to spread as well. As a scholastic audience for books developed, so too did the structure of the page and of the codex designed for individual, silent consultation and annotation.⁹⁷ This moment swiftly ushered in that period of artistic and intellectual activity we know as the Renaissance, which changed everything for books. Indispensable in the exchange of ideas by thinkers far removed from one another, the easily transported written codex allowed exactly that asynchronous development of thought Socrates and Plato feared.

Their concerns echo contemporary anxieties about the ways digitally mediated reading and writing shortens our attention spans and ability to engage deeply with texts. The thing we fear is precisely what worried the ancients: mediation. At the root of Socrates’s accusation lies a vision of writing as a technology that interposes between thinker and thought, severing the two and allowing them to travel independently of one another. While Socrates believed this would make it impossible to defend oneself against others or clarify one’s ideas, this separation proved essential to the development and dissemination of knowledge in a rapidly growing world. It bears emphasizing that writing itself fundamentally changed human consciousness, much as our reliance on networked digital devices has altered us at the core.

1 The Book as Object