The Chinese Typewriter

by Thomas Mullaney

  From this moment on, Chinese telegraphy underwent a profound yet subtle transformation. Phrased simply, Chinese telegraphers became “code conscious” in ways not true of their alphabetic brethren elsewhere in the world, adopting a bifocal relationship with the Chinese script itself. Through one eye, Chinese was Chinese: character-based script whose transmission over telegraphic cables constituted the primary objective of the telegrapher. Through the other, however, Chinese was code: sequences of digits that, save for rare savants who had managed to memorize every one of the 6000-plus codes by heart, always required the telegraph code book to decipher. For example, hearing the pulse sequence ----- -.... .---- ..... a telegrapher in China would not have leapt over the intermediary step of numeric encipherment and directly written out the “plaintext” character 北 (bei, meaning “north”). Instead, his first inscription on the paper would have been a code: the symbols to which this Morse code pattern corresponded—0-6-1-5. Only after this would he be in a position to translate this second-order code into “plaintext” by locating its corresponding Chinese character in the 1871 telegraph code book. By contrast, trained telegraphers in the alphabetic world did not need to undertake this additional step, and could instead translate the dots and dashes they received into plaintext alphanumeric messages. In the alphabetic world, the coded nature of telegraphic transmission was digested and disappeared inside the body of the telegrapher himself, contributing to a kind of “myth of immediacy” prevalent in so many discussions of alphabetic telegraphy both then and now. For Chinese telegraphy, the fundamentally coded nature of transmission—one might say of language itself—could never be ignored or denied. By consequence of the 1871 code, the trained telegrapher in the Chinese world was forced to stay in code.

  While undoubtedly inefficient, and while undoubtedly placing Chinese script at a notable disadvantage when compared to alphabetic transmission, nevertheless this condition of “staying in code” or “living under encryption” exerted a subtle yet unmistakable influence on the kinds of experimentation and innovation undertaken by everyday Chinese telegraphers in the course of their work. Over the course of the late Qing and early Republican periods a vibrant, local-level sphere of innovation within Chinese telegraphy focused on making modest yet profound adjustments to the use of the four-digit code. A loose network of telegraphers, code book publishers, and entrepreneurs dedicated themselves to the development and refinement of a wide array of experimental methods designed to make the Chinese telegraph code faster and more efficient. To an even greater extent than their elite and metropolitan counterparts, such individuals lacked the political power to transform the thick and alienating medium that was built into the technological and legal framework of telegraphy. They knew, that is to say, that there was little chance of their day-to-day operations giving rise to a radical and revolutionary overhaul of the global information infrastructure itself—one that would reduce the layers of mediation through which Chinese had to travel, and bring Chinese into parity with other telegraphic languages.

  Counterintuitively, they began to create additional layers of mediation, above and beyond the two already embedded within Chinese telegraphy. Their experiments and workarounds focused on mediating mediation—that is, inserting additional practices and devices of their own design between themselves and the Chinese code so as to render their relationship with the code more favorable, workable, memorizable, suitable, or otherwise advantageous. These mediations of mediations—hypermediations for short—were decidedly personal, corporeal, local, and immediate, encompassing a variety of novel aide-mémoire, forms of training and bodily practice, and reorganizations and repaginations of the Chinese telegraph code book, among other strategies. Such hypermediations, while they would seem prima facie to increase the time and effort needed to use the code, often rendered it faster, less expensive, and yet fundamentally unchanged in terms of its root-level semiotic architecture. Unable to develop a Chinese telegraph code from scratch that would be deeply attuned to the affordances and limitations of Chinese script, they left the underlying semiotic architecture of the Chinese telegraph code untouched and instead engaged in radical reimaginations of their relationship with, and pathways through, that architecture.

  One of the earliest experiments in hypermediation involved retrofitting a relatively new technique of three-letter or trigraph coded transmission. Developed in the United Kingdom and elsewhere, this code sought to exploit the Roman alphabet as a base-26 system, wherein a sequence of three letters could be used to transmit a total of 263 or 17,576 units. Such a system would have proven far more efficient for Chinese than the four-digit system, insofar as the trigraph system required one fewer code unit (constituting a 25 percent reduction at the outset), and because it made use of letters rather than numerals (with letters being faster to transmit than numerals within Morse code, as explained above). In a newly published code book from 1881, steps were taken to employ this alternate encoding system by pairing each of the conventional four-digit sequences with a unique three-letter code. The code “0001” was paired with “AAA”; “0002” with “AAB”; and so forth. A third layer of mediation acted in tandem to this second one, moreover, being a set of twenty-six Chinese characters used by the editors to represent the Roman alphabet and thereby simplify and render more approachable graphemes that were still foreign to most in China circa the 1880s (table 2.1 and figures 2.12 and 2.13).75

  2.12 Chinese character-based mediation of Roman alphabet (letter/character/pinyin)

  Table 2.1 Chinese character-based mediation of Roman alphabet (letter/character/pinyin)

  2.13 Sample page from character-mediated version of code book

  Using this additional layer of mediation in concert with the second, a Chinese telegrapher could in theory carry out his work entirely in Chinese even as he operated with and through the Roman alphabet. Phrased differently, if we imagine being able to wiretap the electrical pulses that coursed through telegraph cables in China at this time, what might have sounded to us as the three-letter code sequence “D-G-A” would, for both the sending and receiving agents, have been understood in terms of the Chinese character sequence “諦基愛” (di-ji-ai). Even “alphabetic order” itself could be emulated using such a mediation technique, with telegraph operators being able to memorize the sequence “愛、比、西、諦、依 …” (ai, bi, xi, ti, yi …), rather than A, B, C, D, E. Within this complex and multilayered interplay of mediations, the very dynamics and valences of signification could thus be reimagined, all while leaving the underlying architecture of the telegraphic code unchanged.76

  As the ambient level of familiarity with Arabic numerals and the Roman alphabet increased in China, Chinese character mediation schemes such as these began to disappear from code books. Being more comfortable with the likes of “1, 2, 3 …” and “a, b, c,” the average telegrapher presumably no longer needed or found useful the mediations of “一二三” and “愛、比、西.” Mediations of other sorts soon replaced them, however, one of the most subtle and successful centering on a clever repagination of the Chinese telegraph code book itself. In contrast to early code books, in which each page started with the code sequence ---1 and concluded with ---0 (e.g., 0101 to 0200 in code books that featured 100 characters per page, or 0101 to 0300 in those that featured 200 characters per page), publishers in the Republican period began to reorganize the code book such that each page ran from code sequence ---0 to --99 (e.g., from 0100 to 0199, 1200 to 1299, etc.).77 Although seemingly trivial, this alteration produced a secondary effect eminently useful within the everyday work of the Chinese telegrapher, as the page numbers of the code book itself were hereby transformed into aide-mémoire that corresponded to the first two digits of the codes featured on any given page. Repagination of the code book made possible a new “memory practice,” one in which a telegrapher using a code book circa 1946 knew, for example, that the code “1289” could be found on page 12, code “3928” on page 39, and code “9172” on page
91—with the corresponding page numbers being featured in bold red letters at the top and bottom of each page.78 To understand how telegraphers engaged with the Chinese code, mediating this relationship in novel and experimental ways, we therefore must also pay close attention to the social history of the code books themselves.

  At the microhistorical level, each of these efforts was oriented toward a variety of facilitations: enabling Chinese telegraphers to work within a foreign and unfamiliar alphanumeric context, and to accelerate the process by which telegraphers could track down a given character or code, among other objectives not considered here. On the macrohistorical level, however, these local-level efforts added up to something much broader: a historical process connected to what has been theorized in this chapter in terms of “semiotic sovereignty.” Through the creation of these many different “mediations of mediations,” telegraphers were not only undertaking pragmatic processes pertaining to time and money, but were also establishing a novel relationship with an information infrastructure that, as we have seen, placed Chinese within a position of structurally enforced inequality. Telegraphers were taking “symbolic possession” of the four-figure code—playing with it, mediating it, suiting it to their own linguistic and physical preferences, capacities, and limitations. With these incremental, highly localized acts, telegraphers began the process of surrounding a system that had once surrounded them.

  In the next chapter, we leave behind the worlds of telegraphy and movable type printing to arrive, at long last, at the age of the Chinese typewriter. As we do, we will reencounter each of the three puzzles examined in this chapter—common usage, combinatorialism, and surrogacy—only in a new technolinguistic context. With the advent of typewriting as an exciting new technology of inscription, people of a new generation began to contemplate the puzzle of typewriting for the Chinese language. As they did, they revisited and in some cases rediscovered the three modalities examined here, migrating them—as well as the politics embedded within them—into a new technolinguistic terrain.

  Notes

  1 William Gamble, List of Chinese Characters in the New Testament and Other Books, 1861, Library of Congress, G/C175.1/G15.

  2 In the original, the two names are rendered by Gamble as “Tsiang sin san” and “Cü sin san.”

  3 In addition to “anti-reading,” another apt term might be “distant reading,” to invoke the concept and practice made famous by Franco Moretti. See Franco Moretti, Distant Reading (New York: Verso, 2013).

  4 In literary Chinese, the character 之 (zhi) is typically used as a possessive particle and as the substitute pronoun “it”; 而 (er) is typically used as a conjunction, a particle to convey “as well as” or “and yet,” or a particle to convey a change in state or situation, among other potential meanings; while 不 is used to negate verbs. For Gamble’s character frequency analysis, see William Gamble, Two Lists of Selected Characters Containing All in the Bible and Twenty Seven Other Books (Shanghai: Presbyterian Mission Press, 1861), ii, in which William Gamble published the results of his research. As Gamble argued in the preface, movable type printers of Chinese needed to “form a scale, as printers call it, of the characters in the written language.”

  5 See Endymion Wilkinson, Chinese History: A New Manual (Cambridge, MA: Harvard University Asia Center, 2012), 78.

  6 “List of Chinese Characters Formed by the Combination of the Divisible Type of the Berlin Font Used at the Shanghai Mission Press of the Board of Foreign Missions of the Presbyterian Church in the United States” (Shanghai: n.p., 1862), 1. In 1863, William Gamble published Liangbian pin xiaozi [兩邊拼小字] (List of 1878 Chinese characters which can be formed by divisible type), December 22, 1863, manuscript Library of Congress G/C175.1/G18.

  7 Gamble, Two Lists of Selected Characters Containing All in the Bible and Twenty Seven Other Books, ii.

  8 See Cynthia J. Brokaw, “Book History in Premodern China: The State of the Discipline,” Book History 10 (2007): 254.

  9 The number of works was later expanded to 146. See Wilkinson, Chinese History, 951.

  10 Jin Jian, A Chinese Printing Manual, trans. Richard S. Rudolph (Los Angeles: Ward Ritchie Press, 1954), xix. Jin Jian (?–1795) was a bannerman from Shenjing, whose ancestors emigrated from Korea in the waning years of the Ming dynasty. Having served as grand minister of the household department, Jin Jian was later appointed supervisor of the Wuying dian (1772–1774). See Wilkinson, Chinese History, 912.

  11 The system of 214 radicals originated in the late Ming dictionary the Lexicon (Zihui), edited by Mei Yingzuo and published in 1615. It later formed the taxonomic basis of the Kangxi Dictionary, compiled and completed by Zhang Yushu [張玉書] (1642–1711) and Chen Tingjing [陳廷敬] (1639–1712) and published in 1716.

  12 The first of the 214 radicals is composed using a single stroke (yi [一]), while the last is composed by using seventeen (yue [龠]).

  13 The radical-stroke system further stipulates that characters within the same radical class and with the same number of strokes be organized according to the specific types of strokes out of which they are composed. Strokes themselves are categorized into eight types, which are then placed in a sequence. If two characters have the same radical and stroke characteristics, their dictionary sequence depends on the first (and sometimes second or third) stroke used to create the character.

  14 “Label and arrange twelve wooden cabinets according to the names of the twelve divisions of the Imperial Kangxi Dictionary,” Jin wrote in his printing manual, Collectanea Printed by the Imperial Printing Office Movable Type. “When selecting type, first examine the make-up of the character for its corresponding classifier, and then you will know in which case it is stored. Next, count the number of strokes, and then you will know in which drawer it is. If one is experienced in this method, the hand will not err in its movements.” Jin Jian’s first proposal to the emperor was to organize characters into a phonetic “rhyme” system. Ultimately, however, Jin centered upon the system of character organization employed in the Kangxi Dictionary. For a translation of Jin Jian’s account of the printing process, see Jin Jian, A Chinese Printing Manual.

  15 Claude-Marie Ferrier and Sir Hugh Owen, Exhibition of the Works of Industry of All Nations: 1851 Report of the Juries (London: William Clowes and Sons, 1852), 452.

  16 For foreign printers of Chinese, the particular spatial characteristics of Chinese movable type had not always been considered vexing or “puzzling”—a point that deserves emphasis. Over the course of the nineteenth century and earlier, for example, a number of foreign printing establishments created or employed their own Chinese fonts, such as the College of Saint Joseph in Macao and the Imprimerie Royale in Paris—a font containing an estimated 126,000 sorts in all, and thus one that would certainly have “surrounded” the foreign typesetters just as characters in the Wuying dian had surrounded Jin Jian. Over the course of the nineteenth century, however, the desire to surround and sedentarize Chinese characters became an increasingly powerful one within printing, typographic, and pedagogical circles. See Walter Henry Medhurst, China: Its State and Prospects, with Especial Reference to the Spread of the Gospel (London: John Snow, 1838), 554–556; J. Steward, The Stranger’s Guide to Paris (Paris: Baudry’s European Library, 1837), 185.

  17 Li Chen, Chinese Law in Imperial Eyes: Sovereignty, Justice, and Transcultural Politics (New York: Columbia University Press, 2015). In particular, see chapter 2, “Translation of the Qing Code and Origins of Comparative Chinese Law.”

  18 See also Joshua Marshman, Elements of Chinese Grammar: with a preliminary dissertation on the characters, and the colloquial medium of the Chinese, and an appendix containing the Tahyoh of Confucius with a translation (printed at the Mission Press, 1814).

  19 William Gamble went on to compile his own dictionary, arranging these 5,150 characters according to the radical-stroke arrangement system of the Kangxi Dictionary, and then indicating alongside each character
the number of times each appeared in his survey. His dictionary also included the same list of characters, grouping them into fifteen categories according to their frequency of appearance, and then internally organizing each of these fifteen categories according to the Kangxi radical-stroke system. Christian missionary educators in the nineteenth century were also concerned with the Chinese lexicon, but more perhaps for the goal of introducing new concepts (and characters) than reproducing the Chinese language as they found it. Returning to William Gamble and other missionary printers introduced above, their lexical studies of the Chinese language were centered, not merely on the Chinese classics, but also on Chinese translations of the Bible. Gamble, Two Lists of Selected Characters Containing All in the Bible, v–vi.