38 Ibid.,” 4. A further collection of three letters from Lin Yutang to Richard Walsh, Jr. in the winters of 1937 and 1938 brings to light the continuation of this hidden history of Lin’s early experiments with Chinese typewriting, and the model he developed before MingKwai. In the first of these three letters, dated December 16, 1937, Lin recounted that John Marshall of the Rockefeller Foundation expressed his belief that the realization of such a project “would take a professor of engineering with a keen inventive mind,” and recommended that “probably MIT is the place—with some other institutions (International Business Machines Co.—a brilliant man over there who invents radio-teletype).” Lin confessed to Selskar Gunn that he had “no idea how much money would be needed for building a second model,” and that he would not be able to return to focusing on the project until completion of his present book manuscript, most likely that of Mulan. Letter from Lin Yutang to Richard Walsh and Pearl S. Buck dated December 16, 1937. Archives of John Day Co., box 144, folder 6, call no. C0123. In a pair of letters dated almost precisely one year hence, Lin wrote again to Walsh, broaching once more the question of the typewriter. Writing from his home on 59 rue Nicolo, Lin outlined his ongoing conversations with Gunn and Marshall, with whom Lin planned to meet the following day to discuss what he referred to as “my scheme.” The second letter, dated December 21, 1938, recounted details of the meeting with Gunn, the letter making reference to an enclosure that, unfortunately, does not survive within the archival record: two photographs of what Lin refers to as “the first model.” “As the secret lies in the keyboard and the linguistic work, you can show all this material to proper interested parties. An application for patent was filed through a New York lawyer, brother of Horace Mann of St. John’s University, Shanghai, in the Winter of 1931.” Letter from Lin Yutang to Richard Walsh and Pearl S. Buck, December 13, 1938, sent from Paris. Letter from Lin Yutang to Richard Walsh and Pearl S. Buck, December 13, 1938, sent from Paris. Archives of John Day Co., Princeton University, box 144, folder 6, call no. C0123.
39. “Chinese Project: The Lin Yutang Chinese Typewriter,” Mergenthaler Linotype Company Records, 1905–1993, Archives Center, National Museum of American History, Smithsonian Institution, box 3628, multiple dates in 1950 listed, 5.
40 Lin Taiyi, Biography of Lin Yutang (Lin Yutang zhuan) (Taibei: Lianjing chubanshe, 1989), 235.
41 Ibid.
42 Ibid.
43 Ibid.
44 Other coverage included Yang Mingshi [楊名時], “The Principles of Lin Yutang’s Chinese Typewriter (Lin Yutang shi Huawen daziji de yuanli) [林語堂式華文打字機的原理],” Guowen guoji [國文國際] 1, no. 3 (1948): 3.
45 Lin Taiyi, Biography of Lin Yutang, 236.
46 Ibid.
47 Perhaps the only executives who would have been familiar with the idea of male type operators were the executives of Mergenthaler Linotype, insofar as Linotype operation remained a male domain throughout its history. And yet Lin Yutang’s MingKwai machine was not a Linotype—it was a clerical typewriter, and as such, they too would undoubtedly have equated it with the machines that adorned the desks of their managerial offices, not the printing trenches wherein “linos” set type at blazing speed for global newspaper audiences.
48 Lin Taiyi, Biography of Lin Yutang, 236–237.
49 Ibid.
50 The exact cause of the problem was not specified in any source I have encountered.
51 Lin Taiyi, Biography of Lin Yutang, 237–238.
52 Quoted in clipping from Chinese Journal (May 26, 1947), included in Archives of John Day Company, Princeton University, box 236, folder 14, number CO123.
53 “Lin Yutang Invents Chinese Typewriter: Will Do in an Hour What Now Takes a Day,” New York Herald Tribune (August 22, 1947), 13.
54 After this celebratory news, the Italian engineer who helped Lin Yutang design MingKwai began to make claims that MingKwai was properly his invention, not Lin’s. Unnamed in Lin Taiyi’s biography, the Italian engineer threatened to issue a lawsuit against Lin Yutang. As Lin Taiyi remarked, both she and her father found it amazing that an Italian man who spoke no Chinese could claim such a thing. Lin Yutang sought out a lawyer, but it appears that the threat of lawsuit passed without event. There is no further mention of the Italian engineer or his claim in Lin Taiyi’s biography, nor in other sources I have encountered. Lin Taiyi, Biography of Lin Yutang, 238.
55 “Chinese Put on Typewriter by Lin Yutang,” Los Angeles Times (August 22, 1947), 2.
56 “Just How Smart Are We,” Daily News New York (September 2, 1947), clipping included in Archives of John Day Company, Princeton University, box 236, folder 14, number CO123; Harry Hansen, “How Can Lin Yutang Make His New Typewriter Sing?,” Chicago Daily Tribune (August 24, 1947), C4; “Lin Yutang Invents Chinese Typewriter: Will Do in an Hour What Now Takes a Day,” New York Herald Tribune (August 22, 1947), 13; “New Typewriter Will Aid Chinese: Invention of Dr. Lin Yutang Can Do a Secretary’s Day’s Work in an Hour,” New York Times (August 22, 1947), 17.
57 Letter from Pearl S. Buck to Lin Yutang, May 4, 1947, Pearl S. Buck International Archive.
58 Martin W. Reed, “Lin Yutang Typewriter,” Mergenthaler Linotype Company Records, 1905–1993, Archives Center, National Museum of American History, Smithsonian Institution.
59 Ibid.
60 Ibid.
61 Ibid.
62 “Chinese Project: The Lin Yutang Chinese Typewriter,” Mergenthaler Linotype Company Records, 1905–1993, Archives Center, National Museum of American History, Smithsonian Institution, box 3628, multiple dates in 1950 listed, 4.
63 Ibid., 4–5.
64 Ibid., 5.
65 “Psychological Warfare, EUSAK Compound, Seoul, Korea (1952),” National Archives and Records Group, ARC identifier 25967, local identifier 111-LC-31798.
7
The Typing Rebellion
One person setting three or four thousand characters [an hour] doesn’t amount to much. But if everyone were to set three or four thousand characters, now that’d be something.
—Zhang Jiying, Chinese typesetter, 1952
Long before I set out to write a history of the Chinese typewriter, I spent years staring at it without so much as realizing. In the research for my first book, a history of China’s Ethnic Classification project (minzu shibie), I had pored over ethnological and linguistic reports authored by social scientists working in the southwestern province of Yunnan. In composing their studies on the categorization of ethnic identity in the region—reports I reread an untold number of times—researchers had often relied upon the very Chinese typewriters that would later constitute the focus of my research. The Chinese typewriter had been hiding in plain sight all those years.1
Alerting colleagues and friends in the field, I organized an informal “search party” for other purloined letters. I offered a brief crash course in how to identify typewritten documents (especially how to distinguish them from printed texts), and requested their assistance in surveying their own personal collections and archival holdings. Telltale signs included: occasionally handwritten characters tucked in amid typewritten ones (i.e., infrequent characters for which no character slug was available on the machine), the alternating faintness and darkness of different passages (an issue we discussed in chapter 4), an ever-so-slight zigzag of the text baseline, and the somewhat wider-than-normal spacing between characters. The hunt was on.
A new golden age in the history of Chinese typewriting rapidly came into view, with colleagues sighting the machine all the way from the metropolitan areas of Beijing, Shanghai, Harbin, and Kunming to remote regions of western China. No later than May 1950, the Political Protection Office of the Harbin Municipal Bureau of Public Security began to type up its surveillance reports, as shown in local investigations of Catholic communities.2 In Beijing, typewritten reports from the Beijing Municipal Vice Food Products Industry Party appear no later than 1952.3 Provincial-level Pa
rty secretaries in Hebei began producing typewritten documents no later than 1955.4 Typewriting was no means limited to larger urban centers, moreover. In the county of Baoji in Shaanxi province, typewritten documents appear by 1957 in survey reports on local conditions.5 Most telling of all, typewritten reports from 1956 and 1957 were produced in the remote pastoral region of Zeku county, Qinghai province.6 Images of patriotic Chinese typists soon appeared as well, celebrated for producing the paperwork of state-building, postrevolutionary consolidation, economic planning, and class struggle. In March 1956, the Chinese typewriter made new strides with its first appearance within a Mao-era propaganda poster (figure 7.1).7
7.1 Propaganda poster featuring Chinese typist
The 1950s marked a more vibrant period of Chinese typewriting than I had imagined, surpassing the late Republican period in extent and usage. During the Maoist period, an unrelenting series of sociopolitical and economic campaigns placed an unprecedented burden on Chinese typists, who found themselves tasked with the production of economic reports and low-run mimeographed materials for use in the ubiquitous “study sessions” taking place in work units across the country. So heavy was the burden on typists, indeed, that some work units resorted to outsourcing jobs to nonofficial “type-and-copy shops” (dazi tengxieshe), a phenomenon that raised concerns within the fledgling Communist state. With the proliferation of small-scale, independently operated typing shops, the Party-state’s monopoly over the means of technolinguistic production was to a degree compromised, insofar as the same collection of typewriters, carbon paper, and mimeograph machines used to print and distribute state-commissioned speech transcripts, political study guides, and statistics was also being used to run a small-scale gray market publishing industry.
Chinese typewriters were used elsewhere to reproduce entire books, known as dayinben (“typed and mimeographed editions”), a mode of publication so prevalent that the term was later repurposed in the computer age as the Chinese translation of “to print out” (dayin) and “laser printer” (jiguang dayinji). In one type-and-mimeograph edition from 1968, a self-identified Red Guard typed and mimeographed the poetry of Chairman Mao, releasing the edition in time for International Workers’ Day. In an even more profound act of devotion, members of the “Yunnan University Mao Zedong-ism Artillery Regiment Foreign Language Division Propaganda Group” transcribed Mao’s speeches delivered during the years 1957 and 1958, copied from the People’s Daily, the China Youth Daily, the New China Bimonthly, the Henan Daily, and others. Extending to over 280,000 characters in length, with page after page of densely packed typewritten text, this work would have taken between 100 and 200 hours—or four to eight full days—to type and mimeograph (figure 7.2).8
7.2 Type-and-mimeograph edition of Long Live Chairman Mao Thought. Selections from 1957 and 1958 (c. 1958)
Occupying the vast terrain between handwriting and the printing press, type-and-mimeograph editions continued into the Reform Era (1978–1989). The nationally circulated Reform Era literary journal Today!, described by Liansu Meng as the “first unofficial journal in China since 1949,” was composed using a Chinese typewriter and mimeograph stencils.9
The most fascinating dimension of Chinese typewriting in the second half of the twentieth century was not its prevalence or scope, however. Within this bustling world of typist activity, something else genuinely revolutionary was unfolding. Everyday clerks and secretaries in the Mao era spearheaded a series of innovations, centered around experiments with alternate ways of organizing the Chinese characters on their tray beds. Instead of adhering to radical-stroke organization—or indeed any of the experimental character retrieval systems designed by Chinese elites, as examined in the preceding chapter—these typists undertook a “radical departure,” in both a figurative and literal sense. Specifically, they created their own idiosyncratic natural-language arrangements of Chinese characters designed to maximize the adjacency of characters that tended to appear together in actually written language, whether in the form of commonly used two-character compounds (known in Chinese as ci) or in key names and phrases within Communist nomenclature, such as “revolution” (geming), “socialism” (shehui zhuyi), “politics” (zhengzhi), and others.10 Owing to the increased proximity of co-occurring characters, and the repetitiveness of Communist rhetoric, typists who used this experimental method boasted speeds of up to seventy characters per minute, or at least three times faster than the average typing speed in the Republican period.
In other words, among Mao-era typists we can trace the earliest known experimentation with and implementation of an information technology currently referred to as “predictive text”—now a common feature in Chinese search and input methods. Indeed, if “input methods” (shurufa) have been one of the pillars of modern Chinese information technology, as we examined in the preceding chapter, the second pillar is undoubtedly that of predictive text. It may come as a surprise that a technology so familiar to denizens of the digital age has deeply analog roots: Chinese predictive text was invented, popularized, and refined in the context of mechanical Chinese typewriting before the advent of computing. What is more, this innovation cannot be ascribed to a single inventor, but rather came into being through a diffuse collective made up of largely anonymous typists.
China’s First “Model Typist”
In November 1956 a typist in the central Chinese city of Luoyang accomplished an astonishing feat. Using a mechanical Chinese typewriter, the very same we have come to know well over the course of this book, the operator typed 4,730 characters in one hour, setting a new record just shy of eighty characters per minute.11 While seemingly unremarkable when considered within the more familiar context of alphabetic typewriting, the magnitude of the record becomes apparent when one considers the average speed of Chinese typists at the time: twenty to thirty characters per minute. This accomplishment thus represented a two- to fourfold acceleration of the apparatus. The typist had not achieved this record by means of electrical automation, moreover, or a new kind of typewriter. Instead, he had simply rearranged the Chinese characters on the machine’s tray bed. Moving away from the longstanding taxonomic system of radical-stroke organization, and eschewing even the most experimental of character retrieval systems developed in the Republican period, the typist in 1956 reorganized the characters on the machine into natural-language clusters designed to maximize the adjacency and proximity of those characters in the Chinese language that tended to go together in actual writing.
Newspapers in the opening decades of the People’s Republic of China (1949–present) were replete with stories of “model workers,” proletarian champions who displayed feats of unprecedented production through the combined application of will and wisdom. While such exuberant claims cannot be accepted uncritically, in the case of the “model typist” from Luoyang, a diverse body of archival sources and material artifacts has led me to trust the accuracy—and perhaps even the quantitative claims—of this remarkable report.
To illustrate this typist’s system briefly—we will return to it in greater detail later—we can examine a sample arrangement included as part of a 1953 introduction for the public. The principles of this experimental organization can be seen in figure 7.3.
7.3 Sample of new character arrangement; from “Introduction to the ‘New Typing Method’ (‘Xin dazi caozuofa’ jieshao) [‘新打字操作法’介紹],” People’s Daily (Renmin ribao) (November 30, 1953), 3 (Romanizations included for purposes of illustration only)
Beginning with the character gao (告), shaded in gray, we find located within the adjacent cells the characters bao (報) and zhuan (轉)—each of which can be combined with gao to form common two-character words: baogao (報告), meaning “report,” and zhuangao (轉告), meaning “to pass on” or “transmit.” Continuing our exploration from the vantage point of these two characters, we see further layers within this associative system. Adjacent to the character bao, we find three additional characters with which bao can be me
aningfully combined: cheng (呈), hui (匯), and zhuan (轉), which can be paired to produce, respectively, chengbao (呈報 “to submit a report”), huibao (匯報 “to give an account of”), and zhuanbao (轉報 “message transfer”). Still other combinations surround the character shang (上), combinable with those of xia (下), bian (邊), and shu (述) to form “above and below,” “top,” and “above-mentioned,” respectively. All told, the area pictured here contains no fewer than twenty-five perfectly adjacent compounds, as well as an additional set of proximate compounds. This density is impressive when we consider that there are only thirty individual characters in this sample area, out of the total number of 2,450 on the tray bed (all of which, as we will soon see, were likely organized in the same associative fashion), and that these characters would not have been adjacent if organized according to the conventional radical-stroke system. Using machines with this new method of character organization, typists in the early Maoist period set out on one of the most sweeping technolinguistic experiments in the modern age.
This transformation should be understood less as a leap of imagination and cognition than as the manifestation within a particular political environment of a longstanding, deeply corporeal relationship that obtained between machines and human bodies—or what Ingrid Richardson refers to as the “technosomatic” complex.12 The sudden rearrangement of characters on Communist-era tray beds can be understood only if we see it as part of the much longer historical process we have examined thus far: an aggregation of tactile, decentralized, and largely unnoted experiences of thousands of typists and typesetters, individuals who interacted with their character racks and tray beds in embodied, nonverbal ways. It was within this hum of activity, and over the course of countless millions of fleeting, nanohistorical moments, that the leap to natural-language arrangement became conceivable, practicable, and, within the political milieu of the Maoist period, celebrated and incentivized—the countless acts of picking up type, setting it back down, moving from one character to the next on the typewriter, depressing the type lever, and so forth. To understand the history in question, it is essential that we engage with the habitus of Chinese technolinguistic practice, the “embodied history, internalized as a second nature and so forgotten as history.”13
The Chinese Typewriter Page 40