The Chinese Typewriter

by Thomas Mullaney

  The production of the first commercialized Chinese typewriter would take another decade, and would be pioneered, not by an American missionary residing in China, but by a Chinese engineering student residing in America.

  Modernity in Three Thousand Words or Less: Zhou Houkun and His Typewriter for the Chinese Masses

  Barely one year after Devello Sheffield made his final voyage from the United States to China, a young Chinese student made this passage in the opposite direction. Zhou Houkun arrived in San Francisco on September 11, 1910, completing a one-month journey from Shanghai by way of Hong Kong and Honolulu. He was twenty years old, unmarried, and bound for Boston.48 With him, the pursuit of common usage Chinese typewriting resurfaced in the opening decades of the twentieth century—albeit under dramatically different circumstances.

  Zhou Houkun hailed from Wuxi, in Jiangsu province, and had recently completed schooling at Nanyang College in Shanghai, predecessor to the Jiaotong Engineering University founded in 1921.49 Among his shipmates on the Pacific Mail steamship were fellow Boxer Indemnity scholars, those chosen in the second annual competition to study abroad, primarily in the United States. Zhou Houkun was among the students announced in 1910, along with upstart notables Hu Shi and Zhao Yuanren (aka Yuen Ren Chao).50

  After disembarking, these young men headed in different directions. Zhao Yuanren and Hu Shi both proceeded to Cornell University, before themselves parting ways—Hu to Columbia and Zhao to Harvard. Zhou Houkun stopped in the middle of the country, taking up residence at the University of Illinois at Urbana-Champaign, where he studied railway engineering during the 1910–1911 school year.51 The east coast beckoned him, however, and Zhou transferred the following year to the Massachusetts Institute of Technology. Here he graduated with an MS in aeronautical engineering, the first such degree ever awarded in the United States.52

  The China that Zhou left behind was in urgent need of modernized railway systems, ships, and aircraft, making Zhou’s time abroad precisely the sort of investment motivating the overseas studies program. Connected to the Boxer Indemnity Payment, the program required students to focus upon particular majors considered essential to the modernization of China, including agriculture, English, business, mining, law, political science, natural sciences, and education. Soon, however, he found himself enthralled by a novel venture that would capture his attention for the next five years: Chinese language reform. In this regard, he was of a pair with both Zhao Yuanren and Hu Shi, each of whom had quickly abandoned their more “practical” studies to pursue lifelong careers in Chinese linguistics, literature, and cultural reform. But while Zhao went on to become a towering figure in Chinese linguistics, and Hu Shi a luminary in China’s literary and political circles, Zhou Houkun’s entryway into Chinese language reform remained inseparable from his enduring passion for engineering. His goal was not to revolutionize the study of Chinese, or its literary politics, but simply to build a machine: a typewriter for the Chinese language.

  “We will never throw away that wonderful language of ours,” Zhou later wrote, “in order to satisfy the imagination of a few who thought a mechanically inadaptable language ought to go in favor of some other language which would be adaptable to the mechanical device.”

  The whole idea is so revolting that any further comment would put it on a par with other problems, which privilege it does not deserve. It is the duty of the engineer to design machines to suit the existing conditions, but it should never be the privilege of the engineer to ask for a change of existing conditions to suit his machine.

  “Blame the engineer,” Zhou argued, “but do not blame the existing language.” “An engineer who can not build machines according to reasonable specifications is not an engineer in the best sense of the word.”

  A turning point for Zhou came in 1912 at an exhibit he attended as a junior undergraduate at MIT. Here in the Mechanics Building, “my attention was naturally directed towards exhibits of mechanical devices,” Zhou explained.53 “Among them, one machine especially attracted my attention. A girl sat in front of a keyboard, touched the keys, punched multitudes of little holes in a long reel of paper, and, when finished, placed the hitter in a machine which produced fresh, clean, clear types of lead all lined up and ready for the printing press. The whole process occupied only a few minutes. The machine worked automatically and incessantly with a rapidity which easily put to shame our method of type-setting in China. I was told that it was a monotype machine.”54

  “My thoughts at once turned back to China,” Zhou continued, “and brought before me vividly the scene in a Chinese press room—where typesetters, little trays in hand, travelled to and fro seeking for a particular type in a maze of thousands. The process was slow, tedious, inefficient and has constituted one of the great obstacles to the advancement of general education in China. Something must be done.” And so the pursuit of a Chinese typewriter recommenced under very different motivations than in the work of Devello Sheffield. Sheffield had set out to build an amanuensis machine, one that would free him from reliance upon Chinese clerks when composing correspondence in the Chinese language. Zhou Houkun would undertake the same pursuit, but this time with the goal of modernizing his homeland and its language.

  “I realized at the very beginning that the design,” Zhou wrote, “must be fundamentally and radically different from any of the existing American typewriters. Any such idea as providing one key for each character was little short of absurdity.” In certain respects, this passage by Zhou reminds one of the earlier statements by Devello Sheffield, who years earlier also proclaimed the need to depart from the frameworks of Western typewriting in his attempt to build a machine for Chinese. Together with Sheffield’s earlier statements, moreover, Zhou’s proclamation alerts us to the existence of technolinguistic alternatives that predated the rise of the Remington monoculture but also lived on in small pockets within it. Zhou’s statement, however, must also be distinguished from Sheffield’s in terms of the degree of commitment—and even resistance—that it represented in light of the time and place in which it was uttered. During the 1910s, to proclaim one’s intention to create a typewriter that was “fundamentally and radically different” from American machines was to depart from a paradigm that was not only dominant for the English language, or even for the Latin alphabet exclusively, but practically for all of the world’s languages by this point, as examined in chapter 1. In the face of all-encompassing Remingtonian universalism, Zhou’s words spoke of radical alternatives.

  Like Sheffield before him, Zhou set out to create a machine based on common usage characters—this time reducing the total number to approximately 3,000 in all. In choosing these 3,000 characters, however, Zhou would need to develop a lexicon that differed markedly from his missionary predecessor’s. The biblical bestiary of Devello Sheffield’s typewriter would quickly be abandoned, his proverbial Noah’s ark evacuated, and his lions, camels, dogs, and birds shepherded to the confines of the “secondary usage character box.” Likewise, in the era of state-sponsored vernacularization, the literary Chinese particles and pronouns found on Sheffield’s machine—characters like the self-deprecatory first-person possessive bi (敝, signifying “my humble”), which Sheffield had included as part of “common usage”—would need to yield their position to the vernacular Chinese question particle ma (嗎), the second-person singular pronoun ni (你), and so forth. Weights and measures were changing as well, and Arabic numerals were appearing in Chinese-language texts in greater numbers—all of which Zhou and any future common usage theorists would need to account for.

  Fortunately for Zhou, the 1910s witnessed the rapid growth and indigenization of “common usage” studies in China, providing him with ample empirical evidence upon which to determine the lexical boundaries of his device. No longer dominated by foreign printers or missionaries, moreover, algorithmic “distant readings” of the Chinese-language corpus became a mainstay for Chinese intellectual elites themselves, who set out to subject ever-large
r corpora of Chinese texts to the same acid bath of reason we saw in the case of William Gamble and others. Common usage Chinese became a vibrant business, in fact, wherein Chinese educators, language reformers, entrepreneurs, publishers, and statesmen advanced their own hypotheses about the most scientific and practical ways of determining the boundary between utility and uselessness. Although their work was reminiscent of their foreign predecessors’, this new chapter in the history of common usage Chinese centered around a new objective: to enable the Chinese “masses” to grasp their own written script.55

  One of the most formidable participants in this debate was Chen Heqin (1892–1982), who received his education at the Columbia Teachers College in New York, as well as Southeast University in Nanjing.56 Chen set out to determine a repertoire of what he termed “foundation characters” through the analysis of texts, signposts, contracts, and other Chinese-language materials. In 1928, Chen published what George Kennedy of Yale would later vaunt as the “first work of any large scale” of character frequency analysis. The work, entitled Commercial Press Characters in the Chinese Spoken Language Listed According to the Frequency of Their Appearance in Recent Books and Magazines, was based upon the study of a corpus of over a half-million characters, comprising children’s books, newspaper periodicals, women’s magazines, and what he termed “standard literature”—each of these forming roughly one quarter of the study.57

  Chen Heqin’s statistics were remarkably similar to those produced by William Gamble some seven decades prior. A mere nine characters appeared more than ten thousand times, constituting 14.1 percent of the total. The next most common twenty-three characters occurred 14.7 percent of the time, for a total of between four and ten thousand instances each. The subsequent forty-six characters each occurred from two to four thousand times, together composing 13.1 percent of the whole. And the subsequent ninety-nine accounted for another 15.1 percent. Chen Heqin had arrived—albeit more systematically—at the same observation as the algorithmic readers of the previous century: fewer than 200 characters accounted for well over half of all usage.58 All of this was good news for Zhou Houkun and his common usage Chinese typewriter. As Chen’s work seemed to show, even a radically reduced range of Chinese characters on the typewriter might not necessarily impede the expressive capacities of potential users—precisely because their expressive capacities were already quite limited.

  On the other hand, the challenging puzzle was to decide which other characters to include—a deeply political question that set off widespread competition by Chinese educators, politicians, language reformers, and others over how to determine what the “Chinese masses” needed. In 1920, the Ministry of Education ordered primary schools to replace literary Chinese with vernacular, helping spark what Charles Hayford describes as a “commercial war” in the area of educational publishing. A panoply of houses set out to compile and sell new “national language textbooks.”59 In 1922, Changsha became the first site of a major literacy campaign by James Yen, with an estimated twenty thousand copies of his character primer selling within months. The course proceeded in five stages, with students earning a colored stripe for each sequence completed. By the time of graduation, one could proudly show off one’s completion by wearing all five stripes, which together formed the five-colored flag of the republic.60 The educational reformer Tao Xingzhi created his own 1,000-character primer, published by Commercial Press in August 1923. During its first three years, the primer was estimated to have sold over three million copies.61 Tao outlined an ambitious vision, advocating the sale of the new 1,000-character primers in every rice shop, as a full-scale campaign to replace the traditional Trimetrical Classic (San zi jing) and Thousand Character Classic (Qian zi jing).62 Mao Zedong was also a participant in the business and politics of common usage. In 1923, he oversaw the creation of a new foundational character set, this one designed to address the political commitments and visions of the nascent Chinese Communist party, founded only two years prior.63

  The flurry of “thousand character movements,” as George Kennedy termed them, continued unabated throughout the 1930s.64 Hung Shen, a director in the Star Motion Picture Company of Shanghai, entered the fray in 1935 with his own book, How to Teach and Use 1100 Basic Chinese Ideographs. The North China Language School compiled its own reference, The Five Thousand Dictionary. In May 1935, Li Chih published an extension of Chen’s research in the Chinese Journal of Education Research. Li’s corpus was three times larger than his predecessors’, having added elementary school texts, among other documentary sources.65 In 1938, the Lu-Ho Rural Service Bureau published its own minimum vocabulary, the 2000 Fundamental Everyday Usage Characters (Richang yingyong jichu erqian zi). Studies such as these drove home a consistent argument: “The student cannot afford,” as Kennedy phrased it, “to load his mind with useless freight, and, in the early stages of Chinese study at least, an ideograph which recurs no oftener than once in ten thousand must be definitely labelled useless.”66

  In his attempt to create a “popular” Chinese typewriter, Zhou Houkun gravitated in particular toward the work of Dong Jing’an (1875–1944), professor at the Shanghai Baptist College and Seminary and author of the widely popular “600 Character” mass education series.67 For Zhou, the trend seemed clear. These days, he explained, China had “popular education” (tongsu jiaoyu), “popular textbooks” (tongsu jiaokeshu), “popular lectures” (tongsu yanjiang), and “popular libraries” (tongsu tushuguan)—so why not, then, a “popular typewriter tray bed” (tongsu dazipan)? In his patent materials, Zhou explained that his “popular” typewriter would be composed of precisely the same characters found in Dong Jing’an’s series, along with certain complements. His would be a machine for the masses.

  The first prototype having been completed by May 1914, Zhou’s machine contained a cylinder measuring roughly sixteen to eighteen inches in length and six inches in diameter, on which a set of approximately three thousand character slugs was arranged in accordance with the Kangxi radical-stroke system—more than the “popular education” list by Dong, but far fewer than the list by Sheffield (of whose machine Zhou was aware).68 All of these characters were printed in a grid on top of a separate, flat, rectangular finding aid located in the front of the machine. The operator used a metal finding rod to locate the right character on the finding aid: when the tip of the rod was moved into place over the top of the desired character on the printed grid, the other end of the rod brought the corresponding character on the cylinder to the printing position (figure 3.2).69

  3.2 Photograph of typewriter by Zhou Houkun in New Youth

  Zhou’s machine was received with a measure of international praise and attention. On July 23, 1916, the New York Times ran a detailed story about Zhou entitled “Chinaman Invents Chinese Typewriter Using 4,000 Characters.” “A Chinese typewriter of unique design,” the article read, “utilizing no less than 4,000 written characters, has just been invented by Mr. Hou Kun Chow, a graduate of the Massachusetts Institute of Technology, one of the first Chinese students to be educated in this country, and now a mechanical engineer in Shanghai.”70 Following its exhibition at the office of the American Consul General in Shanghai, an exhibition Zhou personally oversaw, Consul General Thomas Sammons reported the machine to be “simple in design and portable.”71 His praise was limited, though. “Obviously, however,” Sammons continued, “no great speed will be possible in operating the machine as now constructed.”72

  In April 1917, Zhou’s photograph was published in Popular Science Monthly, the only known photograph of the inventor together with his apparatus (figure 3.3). The machine was shown atop a small cloth-covered table, the glass-encased character guide extending out from the device. Bespectacled in wire-frame glasses and attired in a suit and leather shoes, Zhou sat beside his device, in an attentive posture. With his right hand, he held the pointer gingerly. To his left, on the table, the chassis of the typewriter remained opened and viewable, the character matrix being roughly the shap
e of a wax phonograph cylinder.73 Uncertain as to what to call this machine—this typewriter without keys—the article’s author was clearly at a loss for appropriate alternatives. “The ‘keyboard,’” the article explained, placing the word keyboard in cautionary quotation marks, “consists of a flat table surface upon which are duplicated the type characters.”

  3.3 Zhou Houkun in Popular Science Monthly

  Having developed his common usage machine, Zhou’s critical next step was to secure financial and manufacturing support with which to transform his vision for Chinese typewriting into a commodified reality. Emboldened by his reception and propelled by the vigor of youthful reformism, Zhou Houkun returned to China, bringing his prototype with him.

  Zhou was not without competition, however. As he ventured to create a Chinese typewriter by seizing upon the common usage approach, another overseas Chinese student—with equal passion and vigor—was racing to beat him to the finish line. For this young inventor, his pursuit would be based upon a set of questions that differed dramatically from Zhou and the common usage approach, asking: What might Chinese typewriting look like if it departed entirely from the core assumption of common usage, namely that the indivisible ontological foundation of Chinese writing was the Chinese character? What might happen if this central tenet were relaxed, or perhaps abandoned altogether? This young student was named Qi Xuan, and he pursued a radically different approach to Chinese typewriting—not that of Gamble, Sheffield, Zhou, and the common usage approach, but that of Pauthier, Legrand, Beyerhaus, and divisible type.

  The Return of Divisible Type: Qi Xuan and the Combinatorial Chinese Typewriter