7

Writing

John Durham Peters

I

Next to the domestication of fire, plants, animals, and of humans themselves, the invention of writing constitutes probably the greatest transformation in human history. Like fire, writing has awesome powers. Its power to preserve meanings makes it explosive and dangerous. Like all other domesticates, writing has never been fully tamed, and can occasionally leap its bounds. (We have all been burned by fine print.) Things put in writing can come back to haunt us – or rescue us. Every document is potentially a “little sorcerer's apprentice” (Levy, 2001, p. 38). Written words have an oracular quality. They keep on saying the same thing over and over again, as Socrates complained in Plato's Phaedrus. They can determine our fates (“the writing on the wall”) and decide between life and death. “If only not to know how to write,” lamented the conquistador Hernan Cortés, “so as not to have to sign death sentences.” Writing indeed has a long-standing association with death (Ong, 1982, p. 81). Whenever one takes pen in hand, even in the most ordinary situations, one opens unforeseeable chains of events, large or small.

The introduction of the medium of writing in its diverse forms and practices has had enormous cultural consequences. Whatever group has possessed it has invariably outwitted those that did not. Writing is not simply a storage device for speech; it is also a power technology. It is a key part of civilization that goes together with such lasting facts as the division of labor, bureaucracy, and male dominance. (The history of writing is also the history of domination.) Writing enables remote control over people and property but also over time and space. A written contract binds the future; an archive of documents preserves the past; and writing has always been the best means of managing events and inventories at a distance. “Writing is magical, mysterious, aggressive, dangerous, not to be trifled with” (Powell, 2009, p. 11). It exercises remarkable leverage over communication, politics, and history and has exaggerated and amplified effects. We all understand Woody Guthrie's point that some men will rob you with a six-gun and some with a fountain pen.

Writing's implications are enormously diverse and impossible to treat fully in the short space of this chapter, which focuses on writing as a medium rather than as an array of practices. To use a distinction that the English term “writing” does not make, this chapter focuses more on l'écrit or Schrift (noun) than écrire or schreiben (verb). Despite its world-changing powers, writing has become so integrated into our lives that it can be hard fully to appreciate its historic and lasting importance. Like other givens of our existence such as eating, sleeping, thinking, dreaming, or wanting, writing's basicness both defies comprehension and invites neglect.

II

To some scholars, writing can seem the stepsister in modern media history. Its thunder was seemingly stolen by new media such as photography, the telegraph, and the phonograph, then by film, radio, and television, and finally by the computer and its digital spawn. Once some form of writing was the only way to store anything for transmission to future generations – whether poetry, law, religion, history, genealogy, property rights, or even music, dance, and cuisine. (Notation systems – such as musical scores, mathematics, chemical codes, and architectural plans – should be included as kinds of writing.) Incessant scribbling – inscribing, describing, prescribing – was the only bulwark against the mischief that time plays with memory. Of course, the very notion that memory is mischievous retroactively imposes an understanding of writing's capacity to fix texts permanently; an oral culture would never imagine invariant transmission as a desirable norm. Writing is itself the bias through which we read history; our access to the history of writing – and the history of almost anything else – comes from the medium of writing itself.

With the great analog explosion of the nineteenth century, writing's “monopoly” on storage was broken (Kittler, 1999). The new optical, acoustic, and electro-mechanical devices seemingly outdid writing in its historic job of making and sending records. By claiming to trace patterns directly from nature, photographs caught images, phonographs saved sounds, and movie cameras held movements much better than writing ever had, which always abstracted and codified events and sensations into abstract symbols. In sending symbolic cargo remotely, the telegraph and broadcasting outsped the post and other messenger systems. And yet, the nineteenth-century explosion of graphic methods that depict time-critical processes – from scientific instruments that inscribe heart beats and seismic activity to ones that record sounds and images – can be seen as deepening rather than subverting writing techniques. The new methods can be seen, to invoke Kittler again, as inscription systems or “discourse networks” (Aufschreibesysteme). The arts and crafts of writing proper did not, of course, sit idly by amid the whirlwind of sped-up pictures and sounds. Experimental texts by modernist authors mimicked the new audiovisual logics, and probably no century ever saw such a proliferation of letters, fonts, and calligraphies as the twentieth. The QWERTY keyboard remains one of the most pervasive communication devices well into its second century. By the early twenty-first century, telephone “texting” reminded us that writing as a medium of communication had never really gone away. Despite prophecies of its extinction, paper continues to be used in massive quantities. Indeed, our digital moment surely boasts the greatest abundance of writing material in all of earth's history. Photography, television, cinema, and video did not simply displace writing: their technologies rested upon its deep logics (inscription, storage, editing, transmission) and their products were full of written materials (labels, captions, subtitles, etc.). It is historically apt that you cannot buy a digital or electronic device without it coming wrapped in a package of paper covered with writing.

Even so, writing still generally plays a secondary role in our media histories and is insufficiently understood as what I will argue it is: the mother of all media. By the time scholars started to imagine a history of communication or media in the midtwentieth century, writing was often treated as an honored but aging medium compared to the audiovisual sounds and shows of the moment. For Harold Adams Innis, who has as good a claim as anyone to having invented the field of media history, the potential obsolescence of writing actually spoke in its favor. As a diagnostician, he saw the modern world tilting dangerously in the direction of what he called “space-biased” media that aimed to conquer minds and territories; certainly many of these were derivatives of writing (such as newspapers and popular print media). But writing was also the medium par excellence of “time-bias.” That is, writing enabled cultural transmission along the time axis. Just as the telegraph connected two points in space, so writing and other time-biased media could connect two points in time. Writing was the medium of remembrance and communication between past and future. Though his normative ideal was the oral tradition of ancient Greece, Innis thought writing could “reveal darkly” that vital way of life, and thus he prized writing (in some forms) as a counterweight to the spatial imbalance of modern life.

As a historian, Innis put writing at the heart of societal transformations. Each new inscription medium brought accompanying political and economic changes (Innis, 1951). In Egypt the shift from stone to papyrus empowered a new priestly caste and destabilized the grip of kings. In Sumeria the difficulty of writing on wet clay led to the more formal writing practice of cuneiform, giving incentives to abstraction and aiding administrative control over the Fertile Crescent. For the Jews, the divine law preserved as text did battle against the graven images of their idolatrous neighbors: part of their religious identity was a commitment to certain media forms over others. For the Greeks, the rise of alphabetic writing and the triumph of prose over poetry meant the decline of music and tragedy (here Innis follows Nietzsche). The Romans, in turn, managed their empire with roads, troops, and a postal system for transmitting written messages and commands, the cursus publicus. later giving way to networks of correspondence among monks, rabbis, and Muslim scholars in the Middle Ages. The printing press starting in the fifteenth century worked all kinds of changes in European and world history, changes still being debated and felt; printing, in various ways, was a conditio sine qua non for the Protestant reformation, the rise of modern science, the French and American revolutions, and modern imperialism. For Innis, the story of civilization is largely the unanticipated consequences of changes in word-media, especially writing.

If Innis died too soon, in 1952, to analyze the new audiovisual maelstrom, his Toronto colleague and disciple (of sorts), Marshall McLuhan, jumped in feet first. For McLuhan (1962), the new modes enabled a restorative kind of sensory fusion that writing, more specifically typography, had ruined. (Of course printing is but one branch – a huge one – on the family tree of writing, but to get McLuhan's arguments we must delve briefly into printing.) With a tone that sometimes verged on the gleeful, McLuhan argued that the social ensemble of the printing press, with its linear, assembly-line logic, was vanishing. Working by analogy from the micro to the macro, from cognition to culture, he argued that typographic reading had regimented the human sensorium into a line-by-line discipline and given birth to a social order – the Gutenberg Galaxy he called it – obsessed by segmentation, causality, and abstraction. McLuhan, though himself a man of letters to the core, saw writing as an endangered species in the new electronic habitat. Audiovisual media such as photography, phonography, and cinematography largely meant the outmoding of writing and literacy – despite the hidden tribute to writing in many of their names (“-graphy” comes from the Greek word graphē, meaning writing). In some of his later books McLuhan engaged in modernist literary experimentation himself, with type sometimes bursting the confines of the line to become image. For him, electronic media beckoned us back into an older mode of being. At the cognitive level, this was syn-aesthesia, i.e., harmony among the sense organs, in contrast to print culture's “hypertrophy” of the eye; at the social level, this was “the global village” of a renewed oral culture, in contrast to the lonely crowd of walled-in individuals produced by centuries of book culture. Other thinkers in this intellectual tradition, though differing in detail, likewise tend to praise orality for a special interactivity and vivacity compared to writing. For Walter Ong (1982) and Eric Havelock (1986), for instance, writing can never fully represent the warmth and communal solidarity of the oral word.

This normative preference for orality was hotly contested by Jacques Derrida, surely the most influential philosophical thinker about writing in the last half century. His blockbuster book Of Grammatology (1976), first published in French in 1967, argued that Western thought's view of writing as secondary, as a mere means of recording speech by visible marks, had a number of deleterious effects. Though his arguments could be flamboyant, Derrida was very well informed about the history of writing. He took the term “grammatology” from I. J. Gelb, a University of Chicago Assyriologist whose book A Study of Writing (1952) was long regarded as the best theoretical treatment of the history of writing, and which also played a strong underground role in postwar French thought about signs and language (de Moraes Rego, 2006), though it has since been criticized for its teleological narrative and its failure to recognize Mayan glyphs as writing. Derrida's larger points about history, politics, and ethics – that writing's downgrading was in league with “the violence of Western metaphysics” and helped to foster the illusory belief that the voice was somehow more revealing of or more “present” to the self's interior than the text – were grounded in the solid philological truth that no script or writing system in history has ever fully captured the voice and that many writing practices have no clear vocal counterpart. “A purely phonetic writing is impossible and has never finished reducing the nonphonetic” (Derrida, 1976, p. 88). He showed that writing was more than a mere ancilla to speech, more than “the painting of the voice” (Voltaire) or “the art of speaking to the eyes.” Derrida helped a wide range of subsequent thinkers see the artful fertility of writing as a medium in its own right. For him, writing did not pervert the supposed purity of speech; rather, writing intensified and revealed the symbolic structures inherent in language generally. Phonetic writing based on the Greek alphabet is not writing in general; it is a peculiar, historically specific form. “In its obsession with phonetic accuracy, the Greek alphabet was a great anomaly in the history of writing” (Powell, 1997, p. 4). Derrida provides a crucial opening for media theory especially, showing that writing's distance from speech and presence is a gift, not a curse. Absence is writing's genius, and all media to one degree or another trade in absence (Winkler, 2008, pp. 235, 237). By criticizing the dream of immediacy, Derrida makes room for the medium of writing and also invites us to see, more generally, media as variations on writing's logics. Writing can thus be understood as primary rather than derivative in our stock of media forms.

That writing has its own logic apart from its vocal accompaniment can be illustrated by one of Derrida's favorite examples. Consider a riddle: what is shared in common by a birth certificate, a wedding license, a coroner's report, a diploma, a birthday card, a declaration of war, a house sale, a legislative bill, a will, and a death sentence? A signature, of course. This cultural form – somewhere between a word, image, act, and gesture – still rules over life and death in our world. A signature transforms a piece of paper into a legal act (Fraenkel, 1992). Nothing binding can be done without one. Not all cultures, of course, work this way, and the signature has been a device for the systematic exploitation of native peoples who guilelessly signed written “contracts” with Europeans. But there is no vocal speech act that accompanies a signature (Harris, 1995, p. 80). It is a pure act of writing, a kind of legal glue in the world of binding documents without need of vocalization. To perform the concomitant speech act with each signature would require an enormous list: I thee wed, I understand the contract, I authorize withdrawal of $50, etc. The signature is thus a particular instance of the larger principle, discussed more fully below, that images cannot unambiguously designate linguistic utterances. To say a picture is worth a thousand words may be to praise its eloquence, but it is also to condemn its lack of precision.

Derrida's work was part of a wave of scholarly interest in writing in the last five or so decades. The apparent threat of extinction often reveals outlines sharply – this is the point of Hegel's dictum that “Minerva's owl only takes flight in the gathering dusk” (Innis, 1951, p. 9). Interest in orality seems to have been inspired in part by the massive new voice-based media of the twentieth century such as radio, sound film, and telephony (Havelock, 1986. pp. 30–33), and studies of oral poetry would have been impossible without tape recording (itself a neglected writing medium of sorts). The computer – another kind of reading and writing machine – has also inspired much interest in the philosophy, history, and cultural meaning of writing. Scholars who have tackled the subject include classicists, especially those interested in the rise of the alphabet and the “Homer question” (i.e., who wrote the Iliad and Odyssey, when and how); ancient historians; linguists and anthropologists interested in nonliterate cultures; literary scholars and historians of the book; the occasional computer scientist; and historians of science interested in writing as the first among many forms of epistemic instrumentation. Perhaps most importantly for present purposes, media theorists and historians have embraced writing, partly under the encouragement of poststructuralist thought associated with figures such as Derrida, Foucault, and Lacan, all of whom offer elaborate, sometimes baroque, theories of writing. Especially in German media studies, a vital and fascinating intellectual scene since the 1980s, writing is the starting point for studying media, which are always considered to be “paper machines” of one sort or another, however intense disagreements on other matters may be. This chapter does not review German-language media scholarship in any detail but builds upon its invitation to recognize the historical centrality of writing in media history. “Traditionally writing was regarded as derivative from oral language, as secondary; with the growing importance of technical storage media it becomes central” (Winkler, 2008, p. 197). New media not only reshape current social practices; they also alter our approach to the past. Our era's incessant texting, posting, tweeting, blogging, tagging, and clicking on two-dimensional surfaces freshly underscores the world-historical importance of writing.

III

Writing is unquestionably the most “momentous” (Ong, 1982, p. 85) of all technical innovations in human history. In contrast to spoken language, which seems “hard wired” and a universal part of the human equipment, writing is a software package that requires installation and constant updates. A toddler with something to say will not ask for pen and paper but will use gesture, facial expression, and the interactive array of oral language. Writing, in contrast, is a difficult, unnatural, and even painful medium, and anyone who teaches university students knows that 15 or 16 years of intense training in the art is often not enough to master it. Speech is a more or less “natural” endowment of all humans; writing is a technical accomplishment that requires a long apprenticeship. (And, unlike writing, almost everybody enjoys talking.) Of course, defining what is “natural” for homo sapiens is a highly vexed problem; as Ong (1982) nicely puts it, “artificiality is natural to human beings” (pp. 82–83). Everyone who writes knows that it can be an intensely physical process involving sore backs, bitten fingernails, headaches, and pulled hair. Writing is work. It exacts a fierce bodily and cognitive discipline. It has a crucial biomechanical dimension, requiring integration of multiple skills such as eye tracking, vocalization, auditory perception, manual dexterity, posture, and mental exertion (Liu, 2010).

And writing is always material and worldly. Though speech can have enduring consequences, its medium of record – memory and lived practice – is comparatively ephemeral and immaterial. Writing takes advantage of the conditions and resources available such as stone, clay, bones, shell, silk, bamboo, jade, brass, wood, animal or human skin, papyrus, paper, and silicon chips. Even dust has been a writing material – abacus descends from the Hebrew word for dust, which once covered tablets used for figuring. The same is true for oil vapor, the short-lived medium of skywriting. There may be speech without tools or tangible materials, but writing always requires both. The history of writing tools, styles, arts, and practices has become a fascinating scholarly subspecialty of its own. An entire material-cultural infrastructure supports writing, from tables, chairs, lighting, and ink to forestry, paper-milling, education, and optometry. Writing is always material and technical, however automated or naturalized it may feel to its users.

Just what is writing? Writing is not only a difficult practice to master, it is even a harder concept to define. Its practices are so overwhelmingly diverse and technically complex as to defy a single overarching understanding. As general definitions, Gelb t 1952) offers “a system of human intercommunication” (p. 12) or “a device for expressing linguistic elements” (p. 24) both “by means of conventional visible marks.” Powell's (2009) definition is similar: “a system of markings with a conventional reference that communicate information” (p. 18). For both scholars, whose analysis I follow here, writing must have visual markings and be governed by systematic rules of decoding, but it need not necessarily stand directly for speech, though most forms do. Like Derrida, Gelb and Powell warn against generalizing too glibly from the alphabetic system passed down from the ancient Greeks, in which discrete letters stand more or less exactly for the acoustic elements of words (more exactly, for instance, in Spanish, and less exactly in English). A two-dimensional plane capable of receiving visual–spatial marks also seems definitive for almost all forms of writing. Words used for writing in many languages imply actions such as incising, scratching, or scraping on some surface – the Latin scribere, German schreiben, and French écrire are all cognates to the English scribble, and the Greek graphō (“I write”) is a cognate to the English engrave and the Spanish grabar (to record or to tape).

Permanence of record, though essential to writing's power, is not always a failsafe criterion, given the crucial part played in the history of writing by erasable surfaces such as the wax tablet, the slate, the blotter, correctible typewriters, and the computer screen. The history of the pencil, for instance, shows a vast realm of experimental, erasable, and fruitfully non-durable writing – what Sybille Krämer (2005) calls “operative writing.” First drafts, architectural drawings, mathematical calculations, and engineering plans require disposable, correctable writing. “Ink is the cosmetic that ideas will wear when they go out in public. Graphite is their dirty truth” (Petroski, 1989, p. 6). Not all writing, fortunately, is permanent – the humble pencil eraser is one in a long line of highly useful methods of deletion. Writing entails not only the power to store and transmit meanings, but to process them as well. Like the abacus and tablet, the pencil shows that writing's revocability is an important counterweight to its finality. Cancellation is a prerogative of power (Vismann, 2008, pp. 25–29). Editing is a sign of majesty. Amnesty has long been figured as the blotting out of a written record. Without deletion, we would always be stuck with Pontius Pilate's dour decree: “I have written what I have written” (John 19:22). The eraser is a small symbol of forgiveness.

The history of writing is itself full of deletions. The preservation of written records (perhaps like all historical records) often depends on happenstance. Linear B, an ancient script that offered essential clues to the history of ancient Greek, was preserved by a palace fire in Knossos, Crete that baked clay tablets that otherwise would not have endured; Aramaic, the language chosen by King Darius to administer the far-flung Persian empire, is best documented in Egypt, not because Egypt was the geographic center of the language's use, but because a desert climate preserves documents on papyrus, parchment, leather, even stone or metal (Ostler, 2005, pp. 78, 83, 131). If the overall history of writing is patchy, the history of human communication is even more so. Humans have been writing for over 5,000 years, but they have been drawing and using language to speak for perhaps ten times that long. Sexual pair-bonding seems to be about 1.7 million years old, and to sustain life-long attachments and cooperative activities, a rich repertoire of non-verbal, that is, meaningful but non-linguistic, signs and gestures must have developed; basic facial expressions of emotion such as fear, joy, and surprise, for instance, seem more or less universal across humans. Much of this history is still unclear but archaeological and genetic evidence offers indirect material for conjectural reconstructions. “Prehistorians,” writes the French paleoanthropologist André Leroi-Gourhan (1993), “must resign themselves to doing without the evidence that would have been most significant – gestures, sounds, arrangements of objects” (p. 107). Writing gives us almost all of what little we have of the long history of communication among humans.

One of the most compelling approaches to the early history of writing comes from Leroi-Gourhan, who is increasingly seen as one of the twentieth century's great media theorists and exercised an important influence on philosophers of writing such as Derrida. Using a broad comparative evolutionary framework, Leroi-Gourhan argues that writing put language under the control of the hand as well as the face. Before writing, the instruments of language were voice, tongue, and ears, but writing puts the two great centers of human interaction, the face and the hand, on an equal footing. The roots of writing are found in drawing or “graphism,” to use Leroi-Gourhan's compromise formulation. Cave paintings were, he claims, an accompaniment to an oral context, not an effort at realistic depiction of animals or hunts, and he looked for language-like rhythms and syntax in their design. But drawing and writing, he notes, do not operate according to the same logic. Drawing has a multidimensional freedom, while writing is more or less linear, subjugating the image to the rule of phonetic sequence. The admiring eye is free to rove over an image as it wills, but the reading eye must track a text in serial order. The image may have structure, but it does not have syntax. Drawing is figurative, but not linguistic, but it is clearly the primal source from which writing practices emerge.

Gelb gave this historically early stage of writing the rather ungainly name “semasiography” and the label stuck. From Greek terms meaning “the inscription of significance,” semasiography is detached from speech. It communicates a more or less clear meaning via visual representations but without designating an unambiguous vocal or linguistic correlate. Consider, for example, the “not allowed” sign – a red circle with a diagonal slash superimposed on an image of the banned activity in question (Powell, 2009, pp. 19–20). Figure 7.1 shows a sign widely understood by the eye, hugely divergent to the ear: not only could it be translated as variably as interdit de fumer, ikke røyke, or μη καπνíζετε, it could also be stated in one language as variably as no smoking, don't smoke, smoking is notpermitted, etc. Semasiographic signs convey meaning without providing instructions for decoding the underlying language. Like visual art, with which they share a deep tie, such signs do not specify propositional content. A picture of a cat on a mat can never distinguish among crucial shades of meaning that are so easy for language to indicate such as “the cat sits on the mat,” “the cat sat on the mat yesterday,” or “cats should sit on mats” (Gombrich, 1972). Other modes allow for much tighter verbal specification – alphabetic writing allowing us to dictate what words we mean letter by letter, for instance – but semasiography is not a precision instrument for encoding speech. Even though it lacks a feature that we alphabet-users might think essential for writing – exact recreation of discourse – we should not think of semasiography as outmoded. Semasiographic conventions are used mundanely in road signage and do-it-yourself instructions for assembling furniture and electronics, and sublimely in mathematical and musical notation. Nor should we view semasiography as deficient in deep meanings. Images, like music, remain some of the most moving things we know, even if words fail us to say why. One thinks of Felix Mendelssohn's answer to the question of why he wrote songs without words. “What I have to say,” he said, “is too precise for words.”

Figure 7.1

The key breakthrough occurred in Mesopotamia around 3400 BCE with the rise of logographic encoding, or word-writing. Goody and Watt (1963) put the innovation well: “The notion of representing a sound by a graphic symbol is itself so stupefying a leap of the imagination that what is remarkable is not so much that it happened relatively late in human history, but rather that it ever happened at all” (p. 315). (Whether the history of writing is a history with a single origin is debated by historians, but it is clear that diverse writing practices have independently found ways to indicate sound; the case of Mayan glyphs would seem to argue for polygenesis.) The key device seems to have been the rebus, a picture sign which evokes a sound which then evokes a word. The rebus uses images to call up names which are exact or near homophones of other words. Like all historic modes of writing, the rebus is still in use; it is found in children's picture puzzles, and if we follow Freud, in our dreams as well. (Freud, the amateur Egyptologist, liked to think of dreams as hieroglyphs and himself as their Champollion.) The rebus may seem a roundabout way of representing speech, combining image, sound, homophone, and word, but all writing systems involve complex contrivances to some degree. Uniting diverse sorts of data is fundamental to the work that writing does.

Egyptian and Chinese are two leading examples of logographic writing systems. Since each word in a logographic script has a unique sign, learning to read and write is an enormous investment of time and effort. To read a modern Chinese newspaper, one must know about 3,000 characters. (Chinese is unique in that it is the only major ancient script to have survived to this day – Sumerian cuneiform, the first of all logo-graphic systems, Egyptian hieroglyphs, and Mayan glyphs have all gone extinct.) Historically most priests, scribes, scholars, and bureaucrats have had little incentive to make writing technology more accessible. Barriers to entry, as Innis would remind us, aided monopoly control, and the history of writing is never only a succession of scripts, but a struggle between groups for access to the means of communication. Indeed, many pre-alphabetic scripts are full of maddeningly labyrinthine rules that seem almost intentionally designed to be obscure and difficult. The sociological norm for ancient scripts seems to be what some scholars call “oligoliteracy” - literacy of the few. Modernity had a very different vision of literacy as a requisite for all citizens.

Logographic practices are by no means dead. We still use logograms such as $, £, €, ©, Σ, √, ∞, †, ♥, 1, 2, 3, 4, or 5, etc. (The letters of the alphabet can be used as logograms for some purposes as well.) Logograms may have an implied sound, but phonetic rendering is often open-ended. Though speakers of mutually unintelligible languages such as Mandarin, Min, Cantonese, Shanghainese, and Hunanese can-understand each other's speech, they can read the same Chinese characters. (This fact gives rise to the false idea that Chinese characters are purely ideational and without phonetic instructions.) We can read “Rx” as “prescription” or “are ex” and it won't matter. The meaning of “I ♥ NY” is clear whether one says “I love New York” or “I heart New York.” If I write the sign “1” a majority of earthlings will know what it means but not agree on how it should be read aloud; different speakers may say “one,” “eins,” “satu,” or “yksi.” But additional signs can clarify pronunciation: “1st” = first, “1°” = primero (Spanish), and “1^er” = premier (French). The meaning remains clear to the eye, but the additional sign hints the sound to the voice and ear. (The additional sign complementing a logogram is often called a “determinative.”) This ability to imply the sounds of language distinguishes logography from semasiography. All scripts give indications as to both meaning and sound; there is no such thing as a script that is nothing but sounds or nothing but meanings (Coulmas, 2003, p. 18).

Thus we should not think that logographic scripts do not provide phonetic instructions. Both Chinese and Egyptian hieroglyphs are phoneticized logographic systems. The great majority of Chinese characters are compounds of radicals that indicate both meaning and sound. Figuring out that Egyptian hieroglyphs were full of phonetic indications – in contrast to the old romance of a purely visual language – was the key step for Champollion's decipherment of the Rosetta Stone in the 1820s. (Chinese and Egyptian have been subject to a long history of fantasy projections by European thinkers – Chinese as a purely ideational language and Egyptian as sensuous and allegorical [Derrida, 1976, pp. 76–81]). The combination of signs that represent concepts and signs that represent sounds is not foreign to English. The cross, a very well-known logogram in English, can be spun in many different phonetic and semantic directions: “†s” = cross and “†d” = dead (Gelb, 1952, p. 118); “xmas” = Christmas and “xing” = crossing. “Twenty” was the name of a British spy ring, since that number in Roman numerals is a “double-cross”: XX. Such far-fetched ingenuity is found in writing both ancient and modern. Medieval heraldry compressed names into designs on shields. Voltaire supposedly responded to a dinner invitation with the message “Ga.” That is, “Gé grand A petit” (big G, little A), or “j'ai grand appétit” (I am very hungry). Today one must read logographically to make sense of phenomena as diverse as vanity license plates and hip hop names and titles. Telephone texting also reactivates ancient logographic strategies (Crystal, 2009): “l8r” in English means “later”; “7ac” in French means “c'est assez” (that's enough) and “je x” means “je crois” (“I believe” – “croix,” cross, is a homophone with “crois”); in Spanish “s3” yields “es” (S) plus “tres” (three) or “estrés” (stress). To make sense of a text message, you have to know which letters and numbers should be read as logograms with a distinct name rather than as phonetic markers. “7ac” is not read as “set-ack”; rather, the “a” and “c” are treated as logograms with a distinct name and sound. The “l” and “r” in “l8r” work alphabetically as phonemic elements, while the “8” is to be pronounced as a free-standing logogram. Deciphering text messages can be quite a jumble for the non-initiate who is not privy to the code – very much the status quo in the history of writing.

It is thus an alphabetic prejudice to think that alphabets are the only writing systems that instruct the reader in how the underlying speech sounds. Before the alphabet, Sumerian cuneiform, Chinese and Egyptian, and proto-alphabetic syllabaries (such as Aramaic and Hebrew) were able to suggest the sounds of speech. Sound encoding has been a feature of writing since the discovery of the “phonetic principle” (Gelb, 1952, p. 67ff.) in the fourth millennium BCE in the Fertile Crescent, even if rules for sounding out were obscure and not always reliable. By any standard the alphabet was a huge transformation, but it is difficult to get an adequate fix on its historic significance, since there is so much accompanying ideological freight, claims about writing history being mixed up with claims about cultural superiority. For Powell (2002, p. 120), the writing of vowels does nothing less than demarcate the historic divide between East and West, a view enthusiastically seconded by Kittler (2006). Havelock (1986) clearly sees Greek writing, and therefore culture, as superior to Hebrew, a valuation Derrida contests.

The invention of “a short atomic table learnable in childhood” (Havelock, 1986, p. 9) did have all kinds of consequences, but it is easy to miss the continuities with earlier graphic systems for representing speech. For Gelb (1952), the “greatness of the Greek innovation lies [...] not in the invention of a new method of indicating vowels but in a methodical application of a device which the early Semites used only in an irregular and sporadic fashion” (p. 182). (Several Hebrew letters can do double duty as vowel and consonant – alef, he, vav, and yod.) As we have seen, the Greek alphabet was not the first writing method to record the sounds of speech in a decodable form; it was, rather, the first to represent speech by nothing but signs for the phonemes of the voice – and thus in a form relatively accessible to all. Only a competent speaker of Egyptian or Hebrew could read those scripts aloud: the visual marks yielded sounds only if you could divine what the underlying words were. With the Greek (vowel) alphabet, a reader could figure out the sounds without knowing the language. The ambition of the Greek alphabet – never perfectly realized – was to provide unambiguous codes of sound to the eyes, that is, to create a one-to-one correspondence of grapheme and phoneme. Despite what some of its devotees such as Havelock (1986, pp. 59–61) seem to claim, the Greek alphabet was no tape recorder capturing vocal sounds in all their particularity. It didn't represent prosody, pitch, pacing, or the infinite variety of tone-color. It represented phonetics graphically, in a brilliant, innovative, and enduring way, but it did not create an exact medium for sound production. There are near-infinite shadings of how to pronounce any given vowel but the alphabet crunches them all into one sign. No writing system is ever purely phonetic – and we should be grateful for the alphabet's astonishing data reduction. This abstraction is not a loss of, say, the fullness of the voice; it is a service. Backgrounding the voice is writing's gift.

The history of scripts resembles more a punctuated equilibrium – sudden lurches of innovation followed by long plateaus of use – than an ever-progressive spiral. The principles that govern alphabetic writing today are essentially the same as those laid down by the invention of the “vowel-alphabet” (as German scholars call it) somewhere in Greece about 800 years BCE. But some alphabetically written languages such as English are rather far removed from systematic vocal representation. “Ghotiugh” can famously be pronounced as fish, taking the “gh” as in rough, “o” as in women, “ti” as in nation, and “ugh” as in bough (Gelb, 1952, p. 225). English arguably even has some features of syllabic writing systems such as Arabic or Hebrew (in which vowels are implied) or Japanese hiragana (which indicate sounds through fixed syllables). The same “sh” sound in the English words suspicion, shut, sure, station, crustacean, session, for instance, can only be recognized in context; this is no atomic table where each free-standing letter possesses a unique sound but a much more volatile and interactive chemistry!

The written form of a language can contain intelligence that its spoken version does not. This is true of both English and Chinese, though much more intensely in the latter's case (Powell, 2009, p. 48). Anciently, many Mandarin words were phonetically more differentiated, but many have since merged in pronunciation, leaving the language with many monosyllabic homophones that stand for distinct characters. This leads to a situation in which a single story of 92 characters about a poet who eats lions can make perfect sense to the eye but, when read aloud, consists of nothing but variations on the single syllable “chih” (Fallows, 2010, pp. 39–43). Here the written form takes authority over the spoken. The visual or mental “reading” of Chinese ideograms disambiguates homophones more quickly than speech, and though Chinese speakers usually rely on syntactic context to specify the implied character, they will sometimes draw the character in the air or on paper in line with Chinese's important manual-graphic dimension (Liu, 2010). In English, the priority of the written form is weaker but still evident. English's erratic spelling, though often a target for reformers, holds firm because it is a treasure-trove of the language's history, and even more, an indicator of the identity of words. In English, word differentiation depends on the graphic sign: such terms as but and butt, threw and through, sight, site, and cite, are phonetically identical. Context normally distinguishes them, but recourse to the written form, either as seen or spelled, decisively clarifies which word is meant. The written form – here is the point Derrida would underscore – is the decisive way to tell “but” and “butt” apart. (The very ideas of an isolated word and of spelling are derivative of specific writing technologies.) Writing as much as speech defines the word.

In studying writing, we come face-to-face with the bigger mystery of how signs mean at all. We may find it odd or amusing that a single sound in Chinese can stand for many characters, but the principle of one sign/many senses is found in every language. How one sign can mean many things is one of the greatest questions. In Latin “malo malo malo malo” can mean “I'd rather be in an apple tree than a bad man in trouble” if you know which underlying word and syntactic structure to associate with which “malo.” Here we see that the problem of decoding writing is at an abstract level the same as the problem of understanding language in general. If you hear “shih” you have to figure out which Chinese characters are intended, and if you hear “malo” which Latin semantic and syntactic structures, but in both cases the cognitive task is to discern the appropriate rules for generating sense from minimal signs and the appropriate cabinets to unlock. The interpretive task of discovering meanings is similar for talk and text, however diverse specific practices in different languages and scripts may be. Writing is not a betrayal of speech but its intensification. Like language itself, writing is a tangle of accumulated tricks, rules, and habits for leaping over signs on the way to meanings. (It is amazing that we don't trip more often.) We humans have barely begun to understand how writing works.

IV

The marvel of writing is its union, through a variety of methods, of two very different sensory registers – the predominantly spatial order of vision and the predominantly temporal order of hearing. Visual perception tends to be all-at-once (synchronic) and auditory perception tends to be one-thing-after-another (diachronic). The eyes are particularly apt at seeing resemblances while the ears hear things whose sources we have had to learn to identify by experience. When we hear a sound, we instinctively turn to see its source. But when we see an object, we don't instinctively ask what it sounds like. The production of sound effects in radio and cinema depends upon the fact that decontextualized sounds, no matter how familiar they may be in everyday life, are often not recognized without an accompanying visual (or other orienting) stimulus. The two most dominant human sense organs, then, perceive the world in strikingly different ways (Schafer, 1994). The eyes are high bandwidth devices that take in whole fields at once but are indiscriminate in small intervals of time (another key fact for cinema). The ears, in contrast, are lower bandwidth organs that are far more temporally acute than the eyes but less able to identify Gestalten. And ears are also the primordial sense for conventionally based meanings – that is to say, for language, the chief of all “symbols,” i.e., signs based on convention rather than nature. Seeing, in other words, rests largely on an “iconic” relation to the world, to use the terms of Charles Sanders Peirce. Hearing, in contrast, is basically “symbolic” in that it grasps the world through habit or convention.

The great breakthrough of writing, according to Robertson (2004), whose theses I sketched in the previous paragraph, was to combine the two sensory registers of seeing and hearing in visual-graphic depictions of linguistic-acoustic signs. “Writing includes both the holistic characteristics of visual perception, and at the same time, without contradiction, the sequential character of auditory perception. It is at once atemporal and temporal, iconic and symbolic. In short, the potential for writing is at the nexus linking the visual and the auditory channels of perception” (p. 19). Writing teaches the eye to behave like an ear, as McLuhan liked to say. In reading, the eye moves serially from one sign to another, not looking holistically for the likeness of things but rather “listening” for the sounds or words indicated by the conventionalized shapes. In order to read, the eye must become a serial processor and follow the discipline of the line. Writing gives auditory objects (words, speech) a visual habitation and temporal events a spatial home. Somehow writing was able to transpose speech's semiotic structure to a visual plane. In this sense, writing combines at once art and music, sight and sound (Krämer, 2005, p. 76). No purely acoustic object ever endures. In the millennia of prehistory, spoken words lived on after their utterance only in the evanescent media of sound, memory, and collective linguistic competence. Trapping Homer's “winged words,” writing discovered how to make them last. It was a kind of phonograph long before Edison, sound caught in space. (Linguists even use the term “phonographic” for writing that provides more precise recipes for vocalizing sounds.) The eye learned to process language. This “verbivo-covisual” (James Joyce) achievement represents the great sensory–cognitive synthesis of human history.

Writing converts space and vision into time and sound, and time and sound into space and vision. It stands at the switch point of time and space, vision and hearing, face and hand, fixity and fluidity. In this it is the most fundamental medium of communication. Media are apparatuses that translate meaning across space and time, through visual, auditory, verbal, and manual registers, and writing was the first great medium to do so. All media of recording, transmission, and processing since follow in its wake. Quevedo, the great Spanish poet, wrote that in reading, “escucho con mis ojos a los muertos [I listen with my eyes to the dead].” In this he spelled out the domain of media theory, the general study of how we swing across the abysses of time and space, sight and sound, life and death.

V

There is much to be said – and much that has been said – about writing and its history. As noted, such a history would have to be self-reflexive, inasmuch as it would be a history of history itself. And there are so many things such a macro history should treat. We could discuss the history, for instance, of the traditional gender division of labor of text and textile, of the prestige and beauty of calligraphy in Chinese bureaucracy or Islamic religion, of writing implements such as chisel and calamus, quill and typewriter, of the Jewish scroll and the Christian codex, of the printing press and its shaping of national languages and identities, of double-entry bookkeeping and early modern capitalism, of paper's role in money and mathematics, of writing's religious role as scripture and guardian of memory or its epistemic role as database of learning or its use as a tool of subjugating non-literate peoples, of literature in all its varieties, of archives stretching from the library of Alexandria to Google, of memory, of literacy and the rise of the modern self or modern novel, or of the discipline of eye, ear, throat, hand, spine, and even derrière in writing. (Sitzfleisch, as it is sometimes known, is a key part of the writer's toolkit.) The history of reading has become a vital and exciting research area. Despite so many delicious potential topics, this chapter revisits its central point – the centrality of writing to any possible media history – via a new angle: the power of non-linearity.

McLuhan, like many others, thought writing essentially linear. No doubt, this is to some degree true, as we have seen. The arrangement of signs into rows that must be grasped ordinally is a key differentiation of writing from drawing, of text from image. When one reads a novel, for instance, one typically reads in a straight line from the front to the back page. But linearity is only one side of writing and emphasizing it misses what is most important about writing as a media form: the temporal reversibility that its spatialization makes possible (Winkler, 2009).

Let me explain. According to Schmandt-Besserat (1996), the earliest kinds of writing were tallies and lists consisting of markings on clay shapes she calls “tokens.” These were forerunners of the Sumerian cuneiform writing that lasted for over 3,000 years and recorded many diverse and unrelated languages (rather in the same way that English, Finnish, Guarani, and Turkish – none of them kin – are all now written with Roman letters). Tokens were used to manage people and inventory. Their function was counting and accounting; they were data-processing devices with no concern (or ability) at first to record syntax or designate sounds relevant to the ear, let alone to chronicle myths or histories. Writing was a tool of computation well before it was a storage device for literature (Vismann, 2008). It has long been used for lists, registers, rosters, censuses, charts, and maps. Siegert (2003, p. 33) notes two traditions of writing: one, working in time, whose task is to capture language, and another, working in space, whose task is to diagram data about collections, inventories, populations, etc. Though he is interested in the encyclopedia and a variety of non-linear literary genres, McLuhan (1962) tends to slight database uses of writing. Since Gutenberg, books have never been the main item produced by the printing press -bills, magazines, newspapers, and posters have used up more wood pulp than books, to say nothing of its use in packaging. And in the most popular bound formats of print culture, such as almanacs, Bibles, cookbooks, dictionaries, and encyclopedias, readers have always hop-scotched around in the text according to their needs. No one reads every word of a newspaper or legal code from cover to cover. The novel is an exception in the history of printing and reading. Writing does not consist essentially of integrated storylines: stories are one of the many kinds of data it holds. The historiography of writing has been biased toward literature and away from bureaucracy, toward pens and away from pencils. The pretty superstructure draws attention away from the material infrastructure. Above all, writing has been a device for storing and processing data in many forms.

In database uses of writing, spatial location is crucial. In a purely linear text, the spatial location of content is not important. Different editions of the same novel or play could generally have different pagination without anything important being lost. But in an equation, map, or table, spatial location is crucial. Where a decimal point or zero is placed can change the world forever. Krämer (2005) contrasts “operative” from speech-rendering forms of writing. Writing is different from graphic art in that its spaces and gaps can be significant. Focusing on the early modern era, she notes the twinning of paper-based visual and quantitative techniques. Cartesian coordinates, for instance, are possible only on a flat surface: thus paper serves as what she calls “a symbolic machine” or an exteriorized mental tool. Without writing, there is no calculus. For Krämer, writing is not just a storehouse of speech, it is also a technical device of calculation, especially after the arrival in Europe of Hindu-Arabic numerals and the zero. Without operative writing, there is no zero and no decimal point, two devices essential for managing the modern world (finance as we know it could not exist without either). A page can sustain a place-value logic, which is essential for notation systems such as mathematics, maps, and musical scores. Graph paper and lined paper represent diverse kinds of writing practice. Mathematics depends on the “flat laboratory” of writing, and in ancient Greek mathematics, lettered diagrams “acted, effectively, as a substitute for ontology” (Latour, 2008, p. 457). Latour puts it well: media, especially writing, have long been substitutes for ontology.

Manovich's (2001) perhaps too dramatic claim that recent digital media shift the leading logic of media from the “narrative” form of novels and cinema to the “database” form of digital forms misses the fact that stories and lists, with their respective preferences for temporal and spatial order, have always been competing and complementary principles in the history of recording media (Stallybrass, 2007, p. 1586). Through accounts, ledgers, maps, logs, registers, catalogs, indices, lists, blueprints, schematics, and tables (such as times tables or the periodic table), writing simulates inventories of objects without a front-to-back narrative structure. The beauty of such writing is that its “visual, spatial location” makes it “subject to possible rearrangement” (Goody, 1977, p. 104). Play, possibility, reversibility are the mark of writing (Winkler, 2009). And even in everyday reading, there is no such thing as a purely linear text – the footnote and the marginal comment both invite the eye to jump off the line. In most writing systems the spatial order of graphic signs does roughly correspond to the temporal order of spoken signs, but not always. “$100” we read as “one hundred dollars,” such that oral speech reverses the order of the written signs. In arithmetic, we work from right to left in contrast to the left-to-right reading norm, a sign of the differentiation of letters and numbers (McLuhan, 1962, p. 181); this fact can be seen in a digital (base two) notation such as 01000000 (64), which slows down the eye and makes us read from right to left. The diverse routing of the eye can likewise be noted in Hebrew, which reads words right-to-left, and numbers left-to-right. Even at the most minimal syntactic level, the text is not purely linear, and for larger units of text, anyone who reads knows how choppily we enter and exit the world of any book.

The tendency in orality and literacy studies to lament the rigidity of the written word is thus one-sided. Take, almost at random, some lines from Havelock (1986): “Once inscribed, the words in a document become fixed, and the order they appear in is fixed. All the spontaneity, mobility, improvisation, the quick responsiveness of spoken speech vanishes” (p. 70). He gets this exactly backward. Oral speech is linear too. In writing or reading, you can move backwards, delete, skim, and reread (and skip pages if you have a codex). To be fair, Havelock readily acknowledges the great advantages of textual fixity; and live interaction can allow for quick correction of misunderstanding. But the world of writing affords a non-linear editing that is the envy of anyone who's ever said something aloud that they desperately wish they could retract – as the words hang in the air seemingly forever, burning their imprint on the memories of offended others. In orality, the first draft is usually the final draft, but writing is both permanence and change. The fundamental fact in human life is the one-way flow of time. None of us can turn back the clock, no matter how hard we wish. But writing, by spatializing data, makes it possible to play with the time-axis. We can read backwards or forwards. We can jump from page 1 to page 271 or vice versa in a codex or from one spot of a scroll to another. Writing uses space to manipulate time. The symbolic realm is not subject to the thermodynamic drag of time. Writing was humanity's first technical means of reversing the flow of time, and its first technique of time-axis manipulation (Krämer, 2006; Winkler, 2009).

I have argued that writing is a power technology; a means of preserving meaning; much more than simply an ancillary device for representing voice or speech; a technique that is both historically material and difficult to master; enormously diverse in its history, practice, and expression; a synthesis of visual and auditory modes of perception and of iconic and symbolic modes of signification; dependent on a complex integration of physical skills, including those of the head and hand; and, as a database manipulating spatial and temporal axes, the primordial medium. Writing is a technical means of using the eyes and hands to speak and hear across distances, of processing complex and massive arrays of data, of transforming space into time and time into space, and of giving us access to a realm beyond the sentence of time's irreversible flow. Writing may seal our fate forever, but it also gives us a shield against time's arrow. What a strange practice! The literate spend hours immobilized with their texts, ignoring others and the immediate sensory world, subjecting their heart's desires to mysterious logics that come from afar. Signatures and decimal points send some to their deaths and make others enormously rich. Writing creates this strange alternate universe parallel to, and in some ways more real than, our own. Writing is indeed the pharmakon as Plato had Socrates call it, the cure and poison at once.

REFERENCES

Coulmas, F. (2003). Writing systems: An introduction to their linguistic analysis. Cambridge, UK: Cambridge University Press.

Crystal, D. (2009). Txting: The Gr8 Db8. New York, NY: Oxford University Press.

de Moraes Rego, C. (2006). Traço, letra, escrita – Freud, Lacan, Derrida [Trace, letter, writing]. Rio de Janeiro, Brazil: 7Letra.

Derrida, J. (1976). Of grammatology (G. C. Spivak, Trans.). Baltimore, MD: Johns Hopkins University Press.

Fallows, D. (2010). Dreaming in Chinese: Mandarin lessons in life, love, and language. New York, NY: Walker.

Fraenkel, B. (1992). La signature: Genèse d'un signe [The signature: Origins of a sign]. Paris, France: Gallimard.

Gelb, I. J. (1952). A study of writing: The foundations of grammatology. Chicago, IL: University of Chicago Press.

Gombrich, E. H. (1972). The visual image. Scientific American, 227, 82–96.

Goody, J. (1977). Domestication of the savage mind. Cambridge, UK: Cambridge University Press.

Goody, J., & Watt, I. (1963). The consequences of literacy. Comparative Studies in Society and History, 5, 304–345.

Harris, R. (1995). Signs of writing. London, UK: Routledge.

Havelock, E. A. (1986). The muse learns to write: Reflections on orality and literacy from antiquity to the present. New Haven, CT: Yale University Press.

Innis, H. A. (1951). The bias of communication. Toronto, Ontario, Canada: University of Toronto Press.

Kittler, F. (1999). Gramophone, film, typewriter (G. Winthrop-Young & M. Wutz, Trans.). Stanford, CA: Stanford University Press.

Kittler, F. (2006). Musik und Mathematik [Music and mathematics], Vol. 1, part 1. Munich, Germany: Fink.

Krämer, S. (2005). Zur Sichtbarkeit der Schrift oder: Die Visualisierung des Unsichtbaren in der operativen Schrift. Zehn Thesen [On the visibility of writing, or, the visualization of the invisible in operative writing: Ten theses]. In S. Strätling & G. Witte (Eds.), Die Sichtbarkeit der Schrift (pp. 75–83). Munich, Germany: Fink.

Krämer, S. (2006). The cultural techniques of time axis manipulation: On Friedrich Kittler's conception of media. Theory, Culture and Society, 23, 93–109.

Latour, B. (2008). Review essay: The Netz-works of Greek deductions. Social Studies of Science, 38, 441–459.

Leroi-Gourhan, A. (1993). Gesture and speech (Anna Bostock Berger, Trans.). Cambridge, MA: MIT Press. (Original work published 1964)

Levy, D. M. (2001). Scrolling forward: Making sense of documents in the digital age. New York, NY: Arcade.

Liu, L. H. (2010). Writing. In W. J. T. Mitchell & M. B. N. Hansen (Eds.), Critical termsfor media studies (pp. 310–326). Chicago, IL: University of Chicago Press.

Manovich, L. (2001). The language of new media. Cambridge, MA: MIT Press.

McLuhan, M. (1962). The Gutenberg galaxy. Toronto, Ontario, Canada: University of Toronto Press.

Ong, W. J. (1982). Orality and literacy: The technologizing of the world. London, UK: Routledge.

Ostler, N. (2005). Empires of the word: A language history of the world. New York, NY: Harper.

Petroski, H. (1989). The pencil: A history of design and circumstance. New York, NY: Knopf.

Powell, B. B. (1997). Homer and writing. In I. Morris & B. B. Powell (Eds.), A new companion to Homer (pp. 3–32). Leiden, Netherlands: Brill.

Powell, B. B. (2002). Writing and the origins of Greek literature. Cambridge, UK: Cambridge University Press.

Powell, B. B. (2009). Writing: Theory and history of the technology of civilization. Oxford, UK: Wiley-Blackwell.

Robertson, J. S. (2004). The possibility and actuality of writing. In S. D. Houston (Ed.), The first writing: Script invention as history and process (pp. 16–38). Cambridge, UK: Cambridge University Press.

Schafer, R. M. (1994). The soundscape: Our sonic environment and the tuning of the world. Rochester, VT: Destiny Books.

Schmandt-Besserat, D. (1996). How writing came about. Austin, TX: University of Texas Press.

Siegert, B. (2003). Passage des Digitalen. Zeichenpraktiken in den neuzeitlichen Wissenschaften 1500–1900 [The digital passage: Signifying practices in the modern sciences]. Berlin, Germany: Brinkmann und Bose.

Stallybrass, P. (2007). Against thinking. PMLA, 122(5), 1580–1587.

Vismann, C. (2008). Files (G. Winthrop-Young, Trans.). Stanford, CA: Stanford University Press.

Winkler, H. (2008). Basiswissen Medien [Basic knowledge: Media]. Frankfurt, Germany: Fischer.

Winkler, H. (2009). The geometry of time: Media, spatialization, and reversibility. Retrieved January 2010 from www.uni-paderborn.de/~winkler/hase_e.pdfwww.uni-paderborn.de/~winkler/hase_e.pdf