What is more, for the very foreign missionaries and printers who became attracted by the “common usage” puzzle, they in fact wanted to change the lexical ground beneath China’s feet: to introduce new terms and concepts, whether those connected to modernity, Christian salvation, or something else. To that end, if foreign printers had determined with great precision how many distinct Chinese characters they would need to reproduce the Chinese-language corpus as they found it in the nineteenth century—to reprint the Confucian classics, for example, in what we might think of as the reproductive or descriptive imperative within the common usage puzzle—they also made sure to calculate the number of characters they would need to intervene in this corpus—to print works in Chinese for the first time in what we might term the prescriptive imperative of common usage. In Chinese, the Old Testament contained 503,663 characters, Gamble discovered, and yet required no more than 3,946 unique characters to produce. The New Testament contained 173,164 characters, comprising only 2,713 distinct ones. For both Old and New, a total of 4,141 characters sufficed. In 1861, Gamble synthesized his findings in a work entitled Two Lists of Selected Characters Containing All in the Bible, a text geared toward missionary printers and pedagogues engaged in the work of proselytizing. Gamble declared with not a small measure of pride, “It has been from a knowledge of these facts therefore that we have been enabled to arrange a large fount of Chinese metallic type in so compact a manner that the compositor can reach any type he wants without moving more than a step in any direction; and by having placed say five hundred of the most numerous characters together, he has more than three-fourths of all he uses just under his hand, almost as conveniently as an fount of roman type are arranged in an English printing office.”20
Common usage constituted the first and most widespread way in which foreigners set out to “puzzle Chinese” during the nineteenth century. When supplied with sufficient time, and one or another subset of the Chinese-language corpus, it now became possible for any foreign reader of Chinese to contribute to this great dissolution of the Chinese language into its constituent parts. To dissolve Chinese in the acid bath of statistical reason—to help one’s Western brethren demarcate clearly between what must be known and what can be ignored—became a new logic and lens through which these foreigners could make sense of Chinese as a nonalphabetic script. Thanks to “common usage,” Chinese script was no longer insurmountable in its abundance. Now it was a puzzle, with a solution.
As we will see in the next chapter, this particular way of puzzling Chinese soon became dominant in China itself, this time among Chinese intellectuals in the first half of the twentieth century. The number and scale of “common usage Chinese” analyses exploded, with Chinese scholars and their assistants subjecting ever-larger swaths of the Chinese-language corpus to algorithmic readings. If Staunton, Gamble, and others had decomposed the Confucian classics, the legal code, and the Bible, Chinese scholars would soon decompose newspapers, textbooks, literature, and more. The “common usage” approach to Chinese technolinguistic modernity would also be extended beyond movable type and character primers to a promising new information technology: that of the typewriter.
To puzzle Chinese in this way would come with a price, however. As we will also examine, Chinese linguistic modernity premised upon the theory and practice of “common usage” could never strive for either comprehensiveness or stability. For any technology premised on common usage—be it within movable type, pedagogy, typewriting, or computing—at no time could the Chinese language as a whole be permitted to take part. Only a small subset of the language could, at any given time, take part in Chinese technolinguistic modernity so conceived. What is more, the boundaries that segregated “included” versus “excluded” characters could never hope to remain set or stable for long—they would constantly shift with the dictates of the age. The power and authority to determine these boundaries, moreover, would forever be fought over by competing social and political factions. This would be the price of any technolinguistic modernity premised upon this mode of puzzling Chinese.
How Do You Spell Chinese? Divisible Type and the Reimagination of Chinese Script
The early eighteenth century was a golden age in the history of translation, witnessing not only the first English-language version of the Great Qing Legal Code by Staunton, but also a bevy of Western-language translations focused upon the great philosophical and religious traditions of Asia. In 1838, Orientalist Jean-Pierre Guillaume Pauthier added to this growing corpus, setting his sights not on the legal structure of the Qing empire, but on one of the philosophical pillars of Chinese civilization: Daoism. In that year, he published the first French translation of the Daodejing.
As a member of the Société Asiatique de Paris, Pauthier was one of the leading Orientalists of his day, already responsible for the French translation of A Record of Buddhist Kingdoms (Foguoji) and the Great Learning (Daxue).21 With the publication of Le Tao-Te-King, ou Le Livre Révéré de la Raison Suprême et de la Vertu, Pauthier’s celebrity only grew, cementing his reputation as one of the leading scholars of China of his generation.22 As in the case of Staunton’s translation of the Qing legal code, what concerns us here is not the content of Pauthier’s translation but the condition of its production—or more accurately, the “extracurricular” work Pauthier carried out during the translation process.
When Pauthier began work on Le Tao-Te-King, he could have easily followed the path carved out by his Orientalist predecessors and colleagues. He could have employed one of the numerous conventional Chinese fonts available at the time, for example, or he could have undertaken a statistical analysis of his own. Likely aware of George Staunton and the nascent “common usage Chinese” movement, he could have subjected the Daodejing to the same kind of algorithmic analysis being applied to other subsets of the Chinese canon, thereby arriving at a custom-tailored font for his printing needs—just as Gamble would do for the Old and New Testaments.
The common usage approach to puzzling Chinese did not puzzle Pauthier, apparently, at least not enough to elicit from him the same commitment of time and obsessive labor as Staunton before him, or Gamble to follow. This is not to say that Pauthier was satisfied with the status quo, however. Propelled by the same desire to bring Chinese “within the compass of European typography,” Pauthier instead set out to puzzle Chinese in his own way. Specifically, he set out to design a new Chinese font that featured not only characters, but also pieces or fragments of character that he would use to build characters combinatorially. By harnessing “radicals and primitives” as the ontological foundations of Chinese script—rather than Chinese characters themselves—such a font offered the possibility of a forty-fold reduction in the size of a complete Chinese font, from tens of thousands of sorts to somewhere in the order of two thousand. Treating Chinese characters as quasi-“words” and radicals as quasi-“letters,” Pauthier set out to puzzle Chinese by posing a Zen-kōan-like question: How does one spell Chinese?
Divisible type was simple, but deceptively so. To print the character ming (明 “brightness”), Pauthier reasoned, one could combine two smaller metal sorts to create a composite: one metal sort featuring the sun, ri (日), and one featuring the moon, yue (月). To print the character shi (時 “time”), meanwhile, one could reuse “sun,” but this time pair it alongside another metal sort containing si (寺)—a character that by itself signified “temple,” but in the character shi would be used solely for its phonetic value. By breaking characters up into their component parts, Pauthier imagined, individual sorts could be allocated in a variety of different contexts, not unlike how Latin letters recombine to form different French words.
At first glance, the story of divisible type would appear to be one of technological triumph: the Western mind, with its esprit d’analyse in full bloom, prevailing over the immensity of the Chinese language through the dogged deconstruction of reality into fundamental elements. Unhindered by convention and custom, the rational French m
ind had opened up a pathway by which Chinese writing could be ushered into the modern era.
Divisible type was not as simple as it seemed, however. For the system to work, Pauthier realized, he could not rely on the 214 conventional Chinese radicals of the Kangxi Dictionary in their isolated forms, since radicals possess an enormous variety of sizes and placements depending upon the characters in which they appear. In the examples of ming and shi above, for example, the sun component is located on the left side of the character, composed so as to occupy slightly less than half of the character’s horizontal span, and slightly less than all of the character’s vertical. By comparison, the sun component operates quite differently within the characters dan (旦 “dawn”), han (旱 “arid”), and xi (昔 “the past”), among many other possible examples. Within dan, the sun component is positioned atop the character, occupying nearly the entire space of the character body. Within han, the sun is also positioned atop the character, but here occupying less than half of the overall space. Within xi, meanwhile, it is positioned along the bottom half of the character. Still another form is, of course, the character ri itself (日)—the Chinese word “sun,” in which the form occupies the space in its entirety.
Here, then, an entirely different puzzling of Chinese script was underway, one that entailed its own kinds of joyous and obsessive tedium. In order to harness the productive, combinatorial possibilities of Chinese radicals, Pauthier would first need to set out upon a full-scale analysis of radicals in all their many orthographic and positional variations. This labor, although every bit as engrossing as that of Gamble and his algorithm, and although focused on the same objectives, was nevertheless conceptualized in entirely different ways.
To pursue this experimental method in Chinese type design, Pauthier sought the assistance of Marcellin Legrand, an esteemed member of the nineteenth-century engraving and type design establishment. Decades earlier at the Imprimerie Royale, Legrand had been commissioned to develop new faces meant to rejuvenate the press, contracted in 1825 to engrave fifteen roman and italic bodies collectively known as “Types de Charles X.” On the strength of his work, Legrand was later named the official engraver of the press.23 He was also one of the era’s foremost designers of “Oriental fonts” and “exotic type.” Legrand was responsible for the carving of a Middle Persian font under the direction of French-German scholar Jules Mohl; an “Éthiopien” font in 1831 under the direction of the Irish-Basque geographer and explorer Antoine Thomson d’Abbadie; a font for Gujarati, carved in 1838 under the direction of French Orientalist Eugène Burnouf; and a host of others.24
Legrand was quickly enthralled by Pauthier’s puzzle. Upon learning of his idea, Pauthier later recounted, Legrand “for the interest of science, was willing.”25 “Of all the languages in the known world,” Legrand himself soon pronounced, evidently enthused, “the most difficult to represent by movable type is, without controversy, the Chinese; having hitherto baffled the most skillful European typographers.”26 The results of this research took the form of an at-a-glance brochure published in 1845, entitled the “Table of 214 Keys and Their Variants” (Tableau des 214 clefs et leurs variants).27 The brochure included its own makeshift symbolic notation, including cartouches, hollow circles, asterisks, and carets, and used them as graphic placeholders to demonstrate all of the possible permutations of each radical, and thus which metal sorts would be needed in the font (figure 2.2).
2.2 Table of 214 keys by Marcellin Legrand
Although it required months if not years of work to accomplish, Legrand and Pauthier’s analysis of Chinese radicals was but the first, and arguably the simplest, of the challenges that confronted divisible type printing. There was a second problem as well, one that emerged not because of the structure of Chinese characters but because of the aesthetic ideologies and commitments of Pauthier and Legrand themselves. Operating at the apex of French Orientalism, with translations focused upon canonical texts like the Daodejing, Pauthier and his contemporaries were enduringly committed to capturing Oriental “essences.” We should recall that, when aesthetes of this age sought out words to praise the various non-Latin “exotic types” being designed by contemporary Western type designers, the words they used centered upon authenticity and fidelity, not iconoclasm or innovation. The same was no less true in the case of Chinese. “The complete Chinese air they assume,” as one contemporary Chinese font would be praised, “so as not to be distinguishable from the best style of native artists, together with the durability of the letter, would recommend them to universal adoption.”28 As a member of the rational, modern state of France, then, Pauthier may have aspired to revolutionize Chinese printing methods; but as a connoisseur of vaunted exotic pasts, he most certainly did not wish to disrupt Chinese orthographic aesthetics.
Legrand shared his client’s aesthetic commitments, and captured the challenge of divisible type Chinese printing most succinctly. The core puzzle, as Legrand phrased it, was “to solve the problem of representing the figurative language of China, with the fewest possible elements, without, however, altering the composition of the symbols.”29 This objective, while simple to articulate, would not be simple to achieve. For while their objective was a printing method capable of producing Chinese characters of impeccable aesthetic composition, the very principle upon which Pauthier and Legrand’s technique of divisible type was based—of cutting up characters into pieces, and then using these pieces to build up characters on the printed page—diverged completely from all longstanding and mainstream approaches within Chinese textual practice. In divisible type, the central conceit involved drawing an equivalence between letters of the Latin alphabet, on the one hand, and structural components of Chinese characters, or “radicals,” on the other. If successful, the homologization of letters and “radicals” would enable one to spell a Chinese character in ways analogous to the spelling of a French mot.
There was a problem with this plan, however. When we compare the compositional principles of divisible type alongside those of the longue-durée history of Chinese calligraphic aesthetics, it had always been the stroke (bihua) that was understood as the primary compositional element of characters—never the “radical.” Strokes, not radicals, were the fundamental components of Chinese composition that students of calligraphy were expected to master, and whose mastery was understood as part of a broader practical and aesthetic education. Beginning in the Han dynasty, but accelerating during the fifth and sixth centuries, China witnessed a proliferation of treatises in which Chinese script forms were carefully categorized.30 Jin dynasty calligrapher Wang Xizhi (303–361), the “sage of calligraphy,” put forth the theory of the “eight fundamental strokes of the character yong” (the calligraphic model we examined in the introductory chapter, and which in part formed the organizational basis of the 2008 Olympics Parade of Nations). Wang Xizhi’s eightfold classification was later elaborated upon by Li Fuguang, who expanded upon Wang’s taxonomy and extended it to thirty-two strokes in all.31 The celebrated calligrapher of the Eastern Jin Wei Shuo (272–349), known as Lady Wei, created a further fine-grained examination of character structure, and arrived at seventy-two fundamental strokes.32
In practicing and mastering the fundamental strokes of Chinese characters, one goal had always been that of composing characters that held together. As outlined in disquisitions by calligraphic masters across the centuries, characters whose component parts sagged about the page were considered unrefined and “lazy,” exhibiting a rattling laxity instead of embodied wholeness. “In the writing of those who are skillful in giving strength to their strokes,” as Lady Wei explained:
The characters are “bony”; in the writing of those who are not thus skillful, the characters are “fleshy.” Writing that has a great deal of bone and very little meat is called “sinewy”; and writing that is full of flesh and has weak bones is called “piggy.” Powerful and sinewy writing is divine; writing that has neither power nor sinews is like an invalid.33
The qu
ality of the calligrapher was thought to be captured in his or her brushwork: Writing is like the person (zi ru qiren), as the well-known adage phrased it. “When one is pleased,” renowned Ming dynasty calligrapher Zhu Yunming (1461–1527) explained, “then the spirit is harmonious and the characters are expansive / When one is angry, the spirit is coarse and the characters are blocked / When one is sad, the spirit is pent up and the characters are held back / When one is joyous, the spirit is peaceful and the characters are beautiful.”34
The Chinese Typewriter Page 13