TOKENISM
It’s said that in Mesopotamia, the region between the Tigris and the Euphrates in modern-day Iraq, writing was born from pebbles—thousands of ancient pebbles, spread across the entire Middle East, an immense area. From Pakistan to Iraq, to Iran, and into parts of what is now Turkey. The earliest pebbles date back to more than ten thousand years ago, when the traces left by man were still vagrant, like footprints in the sand, and settlements were rare.
The history of these pebbles is as long as the history of man’s first civilizations, the discovery of agriculture, the moment when people stopped moving and started planting trees. The earliest inklings of Mesopotamian writing are intimately tied to the earth, to the primal idea of house and home. The script that took rise in Mesopotamia is a durable, solid, ordered script, even in its developing phases. Its stability is the stability of settling down, throwing up walls, building a home.
Not everyone accepts this theory that writing was born from pebbles, but it has long held sway, casting its considerable shadow over the origins of writing. Here’s what it entails. These pebbles were small stone or clay symbols, technically called “tokens,” and were molded into a variety of geometric shapes: spheres, crescents, parallelepipeds, cones. Each shape was used to indicate a different trade good. Before developing a standard format using tablets, goods and objects were counted with tokens: the sheep had its token, cows and bread had theirs, and so on. Nearly five hundred types of tokens in all. Once they’d finished their accounting, they’d shut some of the tokens safely away in a hollow clay sphere. A ball full of tokens. They would then stamp the outside of this container with a seal, not unlike the way we seal envelopes with wax.
We’ve now reached the phase that brings us to the creation of the tablet. Which is founded upon a truly brilliant idea. The tokens were used to stamp the outsides of the containers as well, to make it clear what they held inside—as if to say, “here within is a bill for five sheep.” As time passed, the container would begin to flatten from the repeated pressure of the tokens, moving from the three-dimensional realm of a sphere into the two-dimensional realm of a flat surface. “All right then, so what?” you’re probably thinking.
And my response to you is that we are here witnessing the birth of the most important epigraphic surface for cuneiform writing, and not only. Here, in this object, is born the medium for all writing—the prototype, indestructible as diamond. Water alone can alter or annul it. Here is born the tablet.
And with it, the seed of the very thing that you’re holding in your hands right now: a book or an e-reader, and the precursor to our smartphones. Here, in 3200–2900 BCE (note the uncertainty of the dates!), is born the mother of all technology, the archetypal handheld device. Our means of communication. Here is the moment in history when information is, at last—and quite literally—at our fingertips.
Under this theory, it’s also held that the heavy, bulky forms of the tokens used to stamp the tablets, as a means of accounting, are connected to the figurative and iconic signs with which writing begins (“pictographic,” we might say, but you already know how ambiguous a definition that can be).* And this is where the problems begin. I’ll mention only a couple here, which alone are enough to render this theory controversial. Even if it’s true that the tokens’ marks on the tablets were used for counting, it’s not true that their forms, the impressions they made, are identical to writing’s first signs. A few are merely similar, such as the sheep impression, for example. One other problem is that the tokens’ forms remained the same from 8000 BCE until the beginning of writing, around 3200. It’s unlikely that the token-stamps would go unchanged for millennia, across such a vast geographic territory, only to then morph into the signs of a pre-cuneiform writing system.
If these problems related to form aren’t enough to dissuade you, there’s still an underlying issue with the theory’s method, which blends numbers and letters, claiming that the latter descend directly from the former. A direct lineage from numbers to letters: this is perhaps the theory’s most insurmountable problem, its true weak link. To stuff the world of letters into the realm of numbers and insert them on the same evolutionary path, to stir them into the same pot, is to force a connection—it’s a last-ditch effort to make things tidy, to stretch a bedsheet that’s just too short.
Counting is not writing, and tokens are not signs. It’s difficult to reconstruct such complex processes using such a mechanistic model. And it’s just as difficult to imagine the march toward the invention of writing as an inexorable journey across the millennia, over an enormous swath of earth, an infinite trudge. I don’t know about you, but the idea brings me a touch of agoraphobia. The road to writing is long, yes, but it’s also winding. And we can explain it without relying on the history of numbers.
SILENT MOVIOLA
Let’s watch a movie, now, why don’t we? We’ll call it a “feature-length” film, the way they used to, since this one lasts for centuries. Subtitles appear on the screen: Uruk, the first city, southern Mesopotamia. A close-up of the first tablets. The tokens’ impressions in the clay are clearly recognizable: here we have the precursors to numbers. Ones, tens, and hundreds. They’re identical to the number signs we’ll find in the fully developed cuneiform system. In this aspect, at least, the tokenist theory was on the mark.
Now: a tracking shot of numeric figures, rows of different wedge-shape characters, a complete numeric system. Let’s pause the image on the few—the very few—ideographic signs: at this stage, the population is beginning to count, to regulate calculation. But not yet to write. Flash forward. We’re now in the beginning of the fourth millennium BCE, during phases IV–III of the Uruk period. Something mind-blowing occurs. And I say mind-blowing because we can follow its evolution step by step. Our film is in slow-mo now—a Moviola of new signs, blooming into view on the clay. These are our icons. Things. The protagonist has arrived. But where did it come from?
Not from the tokens and their system of numeric figures. Our protagonist is born from something that was already present—the long, figurative tradition that we find on seal stones. These symbols combine according to iconographic schemes; they become stories, form narrative structures. They trace a plot. And it’s from this plot that the first “pictograms” arise: from the crucible of art, from iterated and reiterated forms, from repetition. And not from the cauldron of numbers.
Leap forward another few centuries. We’re now in Jemdet Nasr, farther north. Icons start to combine with numbers. We’re nearing the end of the film. Icons and numbers unite, and this stable, fruitful marriage is turned toward one purpose alone: to keep the Mesopotamian bureaucratic machine running. The rules for writing numbers grow more complex, adapting to sustain this new city’s economy. Tablets appear by the thousands. The signs seem to indicate official roles and professions. The Leviathan of control has awoken. The system is a go. We have our happy ending. But we’re still missing our co-star—the villain of the film, in this case. Language.
And there may be no filling this absence. Note that this writing system is not cuneiform—it’s still figurative, iconic, even if there are abstract elements (fig. 17). In terms of writing, we’re in a primitive phase, defined as “proto,” since there are still two fundamental things we don’t know: if these signs have phonetic values and, if they do, which language they record. Proto-cuneiform is a mystery. The simplest solution would be to say that the language behind it is Sumerian, the most ancient Mesopotamian language. To assume a different language would be a break in continuity, a shift, and that’s tough to identify. But we don’t yet find linguistic elements: the first evidence of a grammar dates to around 2800 BCE, at least four centuries aft er the invention of proto-cuneiform. And a lot can happen in four hundred years, can’t it?
17. Tablet with pictographic cuneiform characters
It’s a plot hole we’ve been struggling to patch for a while now, our missing villain. And in this endless film, this slow-motion picture, what we’re watching is a frame-by-frame Moviola of writing’s invention. Only there’s no sound. It’s a silent film. The lack of precise linguistic notations and grammatical connections renders it voiceless.
Of course, nothing prevented cuneiform from flowering and expanding shortly thereaft er, but this was possible only because it had already tied its fate to administration. It had attached itself—had glommed on, you might say—to the idea of control. And let’s really take a look at this “silent” phase. Think about a more modern form of invoicing, such as bills—do you find any grammar, syntax, or other complex linguistic elements? Not likely. You don’t need to be a wordsmith to ask for money. It’s the format of the text, how the information is grouped, that reveals its purpose. It’s no secret where to look on a bill to find out how much you owe. Same story with the proto-cuneiform tablets. The form is the content. It’s all right there, unmistakable. So let the lines and columns do their thing—that’s where your eye falls anyway. And close your tab, because this movie’s over. Roll credits.
THE AMBIGUOUS REBUS
The beginnings of Sumerian notation were strictly logographic, meaning that they dealt only with representing isolated words, whether verbs, nouns, or adjectives. To give you an example, the Sumerian word for “to write” is sar, which also means “plant.” And we’ve already encountered the term gi for “cane,” which also meant “to reimburse.” These are two instances of the rebus that we find rather early on in cuneiform’s history, just after the “proto” stage. Sumerian words tend to consist of single syllables; the logograms therefore carry both the word’s meaning and the phonetic value expressed in a syllable. For example, the logogram KU6 is shaped like a fish and also carries the syllabic value ku, meaning “fish” in Sumerian.* This syllable ku can be used to form other words as well. Double meanings, anyone? Sign me up. The high frequency of monosyllables in Sumerian helped make this an easy trick, which was employed often, as we’ve found.
One drawback, however, resulting directly from Sumerian’s monosyllabic nature, is that everything becomes a little ambiguous. How do you tell, for example, if the word means to “write” or “plant” when the term for both is the same? Or what to do with the word for “water,” a, which in Sumerian also indicates the locative case (denoting one’s presence in a place, like the French word chez, or the Latin apud)? “Well,” you might say, “you can tell from the context.” But Sumerians were precise and context didn’t cut it. To distinguish between different categories, they created another kind of word: determinatives.
Determinatives are signs that go unread, since they serve only to indicate a semantic class, the category to which a certain logogram belongs. Trees, plants, and wooden objects, for example, are all preceded by the sign for “wood”; cities by the logogram for “city,” uru; gods by the logogram dingir, “divinity.” This technique allows us to disambiguate the meaning without tearing our hair out. We simply add a logogram to clarify the subject matter at hand. And if it seems like the rebus is a device common to all invented scripts, from Chinese to Egyptian, the same goes for this category of silent determinatives. From this element alone, for all its ingenuousness, one can get an idea of just how incomplete linguistic notation was in the early days. And that’s across the board, not just in Mesopotamia. It consisted merely of strings of words, each one all but isolated.
Only in the next phase do Sumerian’s grammatical elements begin to appear. No writing system has ever been invented, or used at any point in time, to perfectly imitate spoken language and all its myriad functions. When it comes to first steps, scripts are just like babies, leaning on chairs and windowsills, groping for anything to help maintain their balance: these supports are an extension of the baby’s legs. It’s the same with notations in Sumerian. Nouns, a verb here and there, the rare adjective—all are extensions of spoken language, mnemonic devices, means of combating the fallibility of our memories. The chairs and doorjambs of a toddler learning to walk.
UNITED NATIONS
“But they’re mere wedges!”
—Enmerkar and the Lord of Aratta (Sumerian text)
And after that? After that, the toddler takes off run ning. But beginnings are always mythic, covered in an almost dreamlike patina, where everything feels muffled, some details coming through clear and precise, others remaining vague. An old trick of distant memories, appearing rosier than reality. We’ve all fallen for it. Even the Mesopotamians.
It was Enmerkar, king of Uruk, who invented the script, in order to send a long message to his rival, the lord of Aratta. When Aratta receives the letter, inscribed on a tablet, he exclaims the words in the above epigraph in disbelief. Enmerkar has caught him off guard, taking a 1–0 lead in a battle of wits. The text that recounts this anecdote comes from a later period, the second millennium BCE, and it does just what memories do: it coats the script’s invention, now in the distant past, in a mythological patina. As the myth has it, the wedges were a precise form of communication even from the earliest days—though we all know that cuneiform developed gradually, and that its origins are very different.
With the arrival of phonetic notation, when sounds began to take precedence—from 2800 BCE on—signs could at last break free from their figurative chains and shed the requirement of representing real objects. The Sumerians gave up the pointed stylus and began using one that was wedge-tipped. And they schematized the entire graphic system, doing away with curved lines. This shedding of iconicity took place all in a matter of three hundred years, give or take. Cuneiform ceased to be something “proto.” No longer was it composed of instantly recognizable signs: it became cuneiform 2.0, from head to heels (or wedges?). In this new phase, abstract signs and a complex inventory reigned supreme. But there were other problems. Sumerian is an isolated language, with no close relatives, an orphaned only child. It’s also an agglutinative language—with rebuses, open syllables, and signs that are deeply invested in their roles. So what would happen if another language took its place? How would it adapt? The short answer is “with difficulty.” And here is the long answer (which you can skip, if you’d like, but you’ll miss two or three things about world languages).
(Begin digression into historical linguistics.)
The world of languages is divided into just a few large families.
Fusional languages, like all European languages (from Latin to German and everything in between, excluding Basque, Finnish, and Hungarian), are languages that inflect and conjugate nouns and verbs with multivalent suffixes. For example, the ending -us, in Latin’s second declension, tells us three things: gender, number, and case (e.g., dominus: masculine, singular, subject of the sentence).
The Semitic languages, too, are part of this macro-family, even if only distantly related. Languages like Hebrew and Aramaic have roots made up of three consonants, broken up by vowels and other prefixes and suffixes that determine the word’s function. For example, the Akkadian root prs, “to decide,” becomes apparas, “I decide”—with the a strategically positioned between the consonants.
Agglutinative languages, on the other hand, like modern Turkish and Sumerian, have monovalent suffixes, one attached to another, in a string. Even Sumerian’s roots function in this way, with no internal alterations, in fixed clusters of words pressed one against the other. Clearly the two families I’ve described here are structurally very different from each other. And there are other families, too, such as the polysynthetic languages, which are even more complex than the agglutinative languages—but we’ll leave those be.
Just imagine what might happen if cuneiform—rigged with the monosyllabic and unalterable machinery of Sumerian—were adopted to record a Semitic language, in which the vowels and consonants alternate to create flexible, agile, and pirouetting patterns. Good heavens! And yet that’s precisely what happened. Around 2300 BCE, King Sargon defeated the king of Uruk, Lugal-zage-si, and transferred the capital to the city of Akkad, on the Euphrates. Akkad is a city that remains undiscovered—a city that left no trace. When it comes to its language, however, the language of Akkad, we have a superabundance of evidence. Thousands and thousands of tablets. Akkadian falls under the Semitic family, and the cuneiform system was forced to adapt to its variations, to the tango of its vowels and consonants.
In the process of adapting this system to a new language, they had first to make the Sumerian logograms more flexible. Enter syllabograms—numerous and domineering. Akkadian, with its complex system of roots, is a language in desperate need of flexible syllables, and the minimal selection offered by Sumerian wasn’t cutting it. Think about a complex verb such as aštanapparakkim (from the root špr), which in Akkadian means “I will continue to write to you” (as Enmerkar threatens the lord of Aratta). The verb’s root is triconsonantal—to write it out syllabically we’d need a whole heap of syllables: aš-ta-na-ap-pa-ra-ak-ki-im. Tapping into the logogram’s syllabic potential is of crucial importance, especially when tasked with writing foreign names that must be accurately notated. This suggests that the syllables are flexible by nature, since they must appear not only in the “natural” sequence (consonant + vowel) but also the other way around (vowel + consonant), or else in even more complex sequences (consonant + vowel + consonant).
Add to that, Akkadian uses sounds that don’t exist in Sumerian, like the ṣ sound (pronounced like the z in pizza). This problem is quickly resolved, using none other than the rebus. For example, the logogram giš means “wood” in Sumerian, which in Akkadian is iṣum. To adapt, the logogram takes on the additional function of the Akkadian syllable iṣ, applying it to other similar sounds as well, like is and iz. Now imagine a similar process for nearly six hundred signs, with rampant redundancies of polyphonic signs (the syllable ni, for example, can be represented by six graphically different signs) and homophonic signs (the same sign can carry multiple phonetic values, all completely different from one another, as is the case with 
ni, né, lí, lé, ì, zal). Not a system known for its economy, in other words.
(End digression into historical linguistics.)
Despite these complications, the adaptation of cuneiform to Akkadian proved miraculously successful—not only did it manage to get off the ground, within a century or two it spread throughout the Middle East, eventually helping to notate numerous languages: Eblaite, Elamite, Old Persian, Hurrian, Hittite, Palaic, Luwian, Ugaritic, Urartian. And others still, but these should suffice. A host of different languages, all bowing to one single writing system: cuneiform is a translinguistic, transnational, and transgeographical passe-partout. The true emperor of the Middle East. But behind every great emperor is a great consort. And in this case that consort is the language of Akkad.
The marriage between cuneiform and the Akkadian language remained strong and stable for centuries, facilitating communication at the highest levels. Every empire of the Bronze Age (Hittite, Assyro-Babylonian, Egyptian, and others) carried out international relations, economic exchanges, diplomacy, and royal chitchat all thanks to the help of Akkadian cuneiform. We can imagine the messengers, armed with engraved tablets, traveling across deserts and streams, mountains and seasons, taking down responses on their arrival, gathering the gripes of spoiled kings—back and forth, caravan after caravan. For centuries. For the entirety of the second millennium BCE this marriage functioned, the nations united by a sole method of communication—though not mass communication. It was a tool reserved for the global elite (or as much of the globe as was then known). The cultural and economic nerve centers, the United Nations of the Bronze Age, the consequential decisions, the wars—all were woven together by Akkadian wedges. Just like French in the Napoleonic Empire, the English of Cool Britannia. And perhaps like Chinese in our future. Which, in fact, reminds me—it’s time we made our way to China.