The Left-Handed Kid

Jamie Fisher

  • The Chinese Typewriter: A History by Thomas S. Mullaney
    MIT, 504 pp, £27.95, September 2017, ISBN 978 0 262 03636 8

Nominally a book that covers the rough century between the invention of the telegraph in the 1840s and that of computing in the 1950s, The Chinese Typewriter is secretly a history of translation and empire, written language and modernity, misguided struggle and brutal intellectual defeat. The Chinese typewriter is ‘one of the most important and illustrative domains of Chinese techno-linguistic innovation in the 19th and 20th centuries … one of the most significant and misunderstood inventions in the history of modern information technology’, and ‘a historical lens of remarkable clarity through which to examine the social construction of technology, the technological construction of the social, and the fraught relationship between Chinese writing and global modernity’. It was where empires met.

Long before it could be a technological reality, the Chinese typewriter was a famous non-object. In 1900, the San Francisco Examiner described a mythical Chinatown typewriter with a 12-foot keyboard and 5000 keys. The joke caught on, playing to Western conceptions of the Chinese language as incomprehensible, impractical and above all baroque: cartoons showed mandarins in flowing robes, clambering up and down staircases of keys or key-thumping in caverns. ‘After all,’ Thomas Mullaney writes, ‘if a Chinese typewriter is really the size of two ping-pong tables put together, need anything more be said about the deficiencies of the Chinese language?’ To many Western eyes, the characters were so exotic that they seemed to raise philosophical, rather than mechanical, questions. Technical concerns masqueraded as ‘irresolvable Zen kōans’: ‘What is Morse code without letters? What is a typewriter without keys?’ A Chinese typewriter was an oxymoron.

The earliest alphabetic typewriters were devised at a time when orthographic Darwinism was fashionable. In the 1850s, the naturalist Henry Noel Humphreys suggested that the Chinese ‘never carried the art of writing to its legitimate development in the creation of a perfect phonetic alphabet’. Bernhard Karlgren, in his Philology and Ancient China (1926), led a vanguard of alphabetic supremacists, arguing for the characters to be replaced with a phonetic system. Over the following decades, scholars would even suggest that the writing system, by depressing literacy, ‘inhibited the development of a democratic literate culture’. More recently, Derk Bodde and William Hannas have claimed that the Chinese writing system inhibits creativity and the capacity for independent thought. These are corollaries of the Sapir-Whorf hypothesis, which holds, in its strongest forms, that language limits thought. A language incompatible with typewriter keys was incompatible with modernity, and bespoke an equally incompatible country.

The problem was Qwerty, not China, where moveable type predated Gutenberg by several centuries. But the typewriter was developed for the West by the West, during a period in which China was decidedly closed off, and by the time typewriters had become a common feature of commercial life, their form was relatively fixed. At the turn of the 20th century, the missionary inventors trying for a Siamese typewriter simply lopped off two letters of the Siamese alphabet, like Cinderella’s stepsisters severing the offending toe to make the foot fit the slipper.

With the dominance of Remington’s single-shift machine over its competitors, index and double-keyboard typewriters that promised greater flexibility for non-Western languages faded from view. Decades of ‘minimal modification’ followed, reaching peak futility in the system adopted to send messages by telegraph, which required operators to familiarise themselves with 6800 characters assigned a code between 0001 and 9999, a task about as conducive to productivity as memorising pi. ‘Whether Morse code, braille, stenography, typewriting, Linotype, Monotype, punched-card memory, text-encoding, dot matrix printing, word processing, ASCII, personal computing, optical character recognition, digital typography, or a host of other examples from the past two centuries,’ Mullaney writes, ‘each of these systems was developed first with the Latin alphabet in mind, and only later “extended” to encompass non-Latin alphabets.’ For nearly two centuries, China had been a left-handed kid in a world of right-handed scissors.

But there is little point in trying to compare 150 years of Western linguistic dominance and several thousand years of China trying to corral itself. Orthography has always been a tool of empire in China, a means of imposing uniformity at least as old as the Qin dynasty. Characters tie China together both in time and in space. On the one hand, they hide linguistic change: it wasn’t until the 17th century that a scholar noticed that the sounds of Chinese must have changed over the past millennium, since the Odes no longer rhymed. Yet Chinese orthography connects modern Chinese to several thousand years of literature and political thought, and Mandarin speakers to many million speakers of Cantonese, Hokkien and Teochew. Literati across East Asia were able to communicate intelligibly through written messages, or ‘brush chats’. As one writer in Chinese Students’ Monthly put it in 1913, rebutting calls for the wholesale replacement of Chinese characters with English or Esperanto, ‘We Chinese wish to say that the privilege of a mere typewriter is not tempting enough to make us cast aside our 4000 years of superb classics, literature and history.’

By the early 20th century, baihua, or ‘plain speech’, reformers were making arguments that recall Boulez’s line about having to set the Louvre on fire before civilisation can be freed. The founder of the Communist Party, Chen Duxiu, was among those calling for a ‘literary revolution’, a revolt against the ‘ornate, sycophantic literature of the aristocracy’ and in favour of the ‘plain, expressive literature of the people’. Baihua proponents were also driven, at least in part, by frustration at decades of effort to reconcile Chinese characters and Western-derived systems. The early script reformer Qian Xuantong argued that the reform of systems had to begin with characters, ‘if we wish to get rid of the average person’s childish, naive and barbaric ways of thinking’. Thoughts, he argued, travel down the roads that writing makes, and these roads can be winter-proofed in the way a real highway can be.

Whatever its mechanics, the typewriter provides an invisible conduit from the writer’s intention to the character she has in mind. The machine that can do this most quickly is the machine that is best organised, at least a little intuitive, and which reflects language as it is used. It is a plausible model for the human mind. When the Qianlong emperor’s Imperial Printing Office sorted their moveable type by frequency, they were ‘quite literally walking through a physical model of the Chinese language itself’.

And so before making a typewriter one needed to make a philology. Influential early approaches were based on the relative frequency of characters and the ‘spelling’ of recurrent shapes, or radicals, that are the sub-units of many characters. In the character 熊, the four brushstrokes at the bottom constitute a single radical called ‘four dots’. Characterising by radical and stroke, the calligraphic marks from which radicals are built, goes back at least to the Ming dynasty, and was first codified on any sort of scale in the Qing-era Kangxi dictionary, which organised more than 40,000 characters into 214 classes of radical. In the West, scholars were more interested in speed than comprehensiveness. A 17th-century vogue for studying Chinese anticipated later attempts to find the minimum viable conditions – what was the fewest number of characters required for fluency? Leibniz, then developing his own logical language, was one of several intellectuals to fall for a mail-order clavis sinica or ‘Chinese key’.

Then came the thrilling 19th-century revelation that the number of characters in daily use might be closer to several thousand. The whole of the Confucian canon came to 6544 unique characters. In the 1860s, the missionary William Gamble published an influential book of selected characters in the Chinese Bible, suggesting typists might prioritise by frequency. His contemporary Jean-Pierre Guillaume Pauthier, working off the Dao de jing, pioneered the technique of typing by recombinatory radicals. He was careful never to splinter a character mid-stroke – to break, as Mullaney tenderly puts it, its bones. These findings would lay the groundwork for the earliest Chinese typewriters.

Unlike moveable type, which developed in China earlier and independently of the West, the typewriter has always been a foreign import. As such, its most successful inventors have tended to be boundary-walkers themselves, versed in both cultures if not entirely fluent in both scripts. The first machine marketed as a ‘Chinese typewriter’ was invented in 1888 by the American missionary Devello Sheffield, his goal less to create a typewriter than to replace the missionary’s intermediary, the opinionated Chinese clerk. ‘They usually talk to their writer,’ Sheffield wrote, ‘and he takes down with a pen what has been said, and later puts their work into Chinese literary style … The finished product will be found to have lost in this process no slight proportion of what the writer wished to say, and to have taken on quite as large a proportion of what the Chinese assistant contributed to the thought.’

Sheffield arranged some 4662 characters by frequency into easily and less easily reached regions of his circular tray-bed. The characters themselves were chosen with a missionary’s bias; it is hard to imagine that ‘blindness’, ‘slaves’, the characters in ‘Jesus’, and the animals of Noah’s ark would otherwise be included. The typewriter amounted to several cubic feet of hand-cut tiles, arranged on a roulette wheel. ‘This looks like a simple process,’ an observer remarked, ‘and so it would be indeed, if a man’s life were not limited to three-score years and ten.’ It was never mass-produced.

The first commercially viable Chinese typewriters would be devised by native speakers living abroad. Zhou Houkun, along with such linguistic reformers as Hu Shih and Y.R. Chao, was part of the second group of Chinese scholars to attend university in the United States at the turn of the 20th century, their education funded out of the war reparations owed to the US after the Boxer Rebellion. The terms of the scholarship required recipients to study subjects pertinent to China’s modernisation; Zhou began by studying railways and was ultimately one of the first recipients of MIT’s master’s degree in aeronautical engineering. By the time he turned to modernising type, in 1914, the baihua movement was in full swing. Lu Xun was one of several writers who had become language reformers, and language reformers, as Mullaney observes, ‘were by necessity “typewriter reformers” as well’. Zhou’s typewriter, with its 3000 characters, was the first to reflect vernacular speech: to have in mind not the idealised user, but the average user. The machine itself was a flat board with a character matrix underneath, not unlike the cylinder of a phonograph, or a scroll. Qi Xuan, a student at NYU, developed a machine that included 1327 components, so that the typist could make up less frequently used characters as needed. Qi was willing to ‘break the bones’ of the characters as no other inventor had dared, anticipating the character reforms of the communist period. The new approach was as revolutionary as it was slow. The New York Times reported that a live demonstration had produced only ‘100 words in two hours’ and an article in the Washington Post would discuss it alongside such novelty items as ‘a dancing radiator doll’ and ‘a mousetrap for burglars’.

Part of the problem with these early typewriters is that they didn’t much resemble typewriters. The typewriter’s attraction was not only its usefulness, but its cultural cachet. One thinks of the famous image of Chinese dignitaries gathered around a Gatling gun. The point was being in a position to announce: ‘We have the technology!’ Inventors worried that if the typewriter were altered for the Chinese market, ‘the resulting machine [might] prove entirely illegible and unrecognisable to the Western eye … And if unrecognisable to the world as a typewriter, would it be a “typewriter” at all?’ And so it is unsurprising that engineer Shu Zhendong’s eminently typewriter-like typewriter was the first to be mass-manufactured. The Commercial Press in Shanghai, Republican China’s busiest printer, sold at least 100 units a year of ‘the Shu-style typewriter’ between 1917 and 1934 to customers as various as the Chinese Consulate in Canada and the Chinese postal service.

Of all the typewriter models, the Shu-style machine is perhaps closest to the Western dream of the Chinese typewriter: mammoth, impractical, breathtaking. Visiting a model in a museum, Mullaney notes the three regions on the tray-bed for frequently and less frequently used characters – fragile metal slugs, manipulated with tweezers, that broke apart in front of his eyes as a curator attempted to demonstrate their use. Mullaney’s encounters with the machine feel like Borges describing the Aleph: ‘As I moved around the machine, shifting my angle of view, fleeting constellations emerged from out of the brittle, charcoal-coloured mass – shimmers of characters exhibiting greater reflectivity than their neighbours on account of being replaced more recently, likely because they had been used more frequently when the machine was still in service.’ This machine was owned by a Chinese-American immigrant. The keys that glitter with use are emigrant, far away, urgent, longing, hardship, dream.

This vividly demonstrates the problem: language isn’t just about the frequency with which characters are used in the Analects, or in the People’s Daily: it’s about the ability to bend to circumstance. Typewriters were personal in another way too. The hundreds of Chinese typing school students who studied on an English-language machine practised ‘fingerwork’ and ‘blind typing’: those working on a Chinese machine learned ‘character retrieval methods’ and ‘adding missing characters’. Students needed to lunge across the type bed, pressing down firmly on complex character slugs and leaning with a lighter touch on simpler slugs, at the risk of tearing through the paper. The demands on tactile memory were substantial. Mullaney asks a Malaysian Chinese-British woman ‘how she had come to learn the locations of more than two thousand characters’ on her machine. ‘I just remember it,’ she says.

Like many Chinese typewriters of a certain vintage, hers was made in Japan. Japanese companies with kanji-equipped keyboards entered the Chinese typewriter market in the 1920s, and quickly came to dominate the industry in the 1930s and 1940s, mirroring Japan’s military conquests during the same period. Typing institutes and secretaries followed the Japanese into the puppet state of Manchuria, as did province-wide typing competitions. The symbolic instrument of empire was the Nippon Typewriter Company’s Wanneng (‘All-Purpose’) device, billed in 1940 as the ‘Japanese, Manchu, Chinese, Mongolian All-Script Typewriter’ – ‘the very materialisation’, Mullaney writes, ‘of Japan’s “Greater East Asian Co-Prosperity Sphere”, and its colonial refrains of ethnic harmony and “same script, same race”’. Imperial edicts written on behalf of the child emperor were turned out on Japanese typewriters. As one collaborationist type-writing instructor observed, in a chilling note, ‘Typists more than anyone must follow the times.’

Briefly, following Japan’s defeat, Chinese manufacturers were able to reclaim the market by selling copycat Wanneng machines, or even selling Wanneng machines directly, without pretensions to originality or patriotism. One Shanghai company sold a ‘People’s Welfare Typewriter’ – slapping a name borrowed from Sun Yat-sen on a Wanneng. The Communist government engaged in the same practices on a grander scale, seizing the Japanese Typewriter Company and rechristening it the Red Star Typewriter Company. In the 1950s, resistance to Japanese machines finally collapsed: the Shanghai Chinese Typewriter Manufacturers Association was created out of a consortium of ten Chinese typewriter companies – their enduring legacy would be the ‘Double Pigeon’, a sprightly Wanneng-based number that would dominate the market for decades to come.

The spectre hanging, however genially, over every attempt to create a Chinese typewriter is Lin Yutang. While others repackaged Japanese machines, the Tsinghua and Harvard-educated Lin was inventing his own. A brilliant and congenial figurehead with two English-language bestsellers in the 1930s, in the 1940s Lin came up with his MingKwai (Clear, Quick) typewriter. Roughly the size of an English-language typewriter, and equipped with a recognisable keyboard, the MingKwai’s appearance belied its complexity. Its compact chassis concealed 43 rotating cylinders, and featured a viewfinder (‘the Magic Eye’) that allowed users to select characters by depressing keys marked with character components – some radicals, some strokes, and some bold and intuitive groupings of Lin’s own devising. IBM and Remington expressed their interest, and Lin’s typewriter was celebrated in dozens of newspapers across the US, from the New York Times and the Chicago Tribune to the Los Angeles Times and Newsweek. But it was never mass-manufactured: the plans were undone partly by the rise of the Chinese Communist Party, which didn’t seem to bode well for patent rights, and partly by the invention of the phonetic system Pinyin, which briefly threatened, at least in the eyes of American manufacturers, to make Chinese typewriters obsolete. Today if you open a laptop or unlock a phone to type in Chinese, the first thing you’ll notice is how intent the software is on doing all your work for you. The letters typed on your keyboard trigger the on-screen display of several dozen likely possibilities, arranged in order of frequency. This seems so obviously computational it is a surprise to learn that it originated with the actuating keys Lin devised for his typewriter, and with the fervour of the typists in the early days of the Revolution.

In 1951, typesetter Zhang Jiying shattered speed records by arranging characters in associative clusters, which he called lianchuan – ‘chain’ or ‘free-association’, later adopted as the term for what is now called predictive text. Zhang recognised how much of language is cliché. ‘Liberation’ (jiefang) was likely to be followed by ‘army’ (jun), ‘American’ (Mei) by ‘imperialist’ (di). While it is true that grammar and word-order impose certain expectations on form in every language, it is particularly true of Chinese, where collocations influenced by tone, metre and rhythm have, by a centuries-long process, migrated towards one another. Typists and typewriter manufacturers rejigged their keyboards in the wake of Zhang’s discovery. And so while typewriters became more attuned to the language of people generally, they became less like people personally.


It was a revelation suited to China’s new industrial age: the liberating, depressing realisation that we build our days from scraps of habit, which can then be optimised for greater productivity. ‘Kaifeng typesetter Zhang Jiying diligently improves typesetting method,’ the headline went, ‘establishes new record of 3000-plus characters per hour.’ ‘Diligent’ was the word. Mullaney estimates that the members of the Yunnan University Mao Zedong-ism Artillery Regiment Foreign Language Division Propaganda Group spent 100-200 hours collectively transcribing Mao’s complete speeches between the years 1957 and 1958. One typesetting manual suggested associative clusters like ‘negative connotation terms’ (‘invade’, ‘destroy’) and ‘social structure terms’. The examples ‘socialism, co-operative, Chairman Mao, etc etc’ bespeak total existential fatigue.

On a broader level, this shift towards the average speaker was mirrored by sweeping top-down changes. In 1955, a national linguistic standard – putonghua, or ‘common speech’ – was established. The debut, in 1956, of the ‘reformed’ Chinese typewriter, with a natural-language tray bed that anticipated a vernacular-speaking typist, coincided with the government’s approval of the First Round of character simplification, a series of structural tweaks and standardisations intended to boost literacy. These changes weren’t comprehensive, and, compared with the early fervour of the baihua reformers, they were timid. In 1936 Mao Zedong had expressed his belief that ‘we will have to abandon characters altogether if we are to create a new social culture in which the masses fully participate.’ Twenty years and some two thousand proposals later, the party settled for reducing the number of strokes in the word for ‘hair’.

Mullaney manages to get into the details of these machines without making the reader feel stuck at a dinner party with a trainspotter. A wonkish pleasure shines through every sentence, and readers are likely to find the book more entertaining than it has any right to be. Almost everything is more or less fair game, from a made-for-TV movie starring Tom Selleck and the Bollywood song ‘Typewriter Tip Tip Tip’ to the Beijing Olympics and MC Hammer’s ‘Can’t Touch This’. (The rapper, Mullaney concedes, is ‘not high on the list of individuals we often turn to for insight into the history of China, or the global history of modern information technology’.) While The Chinese Typewriter is not necessarily meant for the average lay reader, Mullaney sees himself as a populariser. ‘If tulips, codfish, sugar and coffee have all changed the world,’ he writes, ‘it stands to reason that perhaps the Chinese typewriter did too.’

Written speech is a negotiation between the visible and invisible worlds, whose governing idea is that what can’t take tangible form can’t exist. The Chinese typewriter’s mysteries are further amplified by the fact that its earliest prototypes have vanished. Lin Yutang’s only prototype was thrown out sometime in the 1960s. Surviving Shu-style machines dissolve at a curator’s touch. At least one early missionary model may have been consumed by white ants – a chilling image of silence eating up speech. Mullaney frames his history as the story of human attempt and failure, a bold and many-handed effort. ‘The history of modern Chinese information technology,’ he argues, ‘is not one that derives its importance and relevance from the magnitude of its immediate effect, but from the intensity and endurance of its engagement.’ Lin Yutang’s daughter would write in her biography of her father: ‘Even though it cost $120,000, even though it has saddled us with a lifetime of debt, this creation of my father that he had worked his entire life on, this newborn baby birthed with such difficulty, it was worth it.’

For this book, Mullaney spent more than a decade plumbing ‘oral histories, material objects, family histories and archival texts from more than fifty archives, museums, private collections and special collections in nearly twenty countries.’ He is now at work on a sequel. It will be, he writes, ‘the first history in any language of Chinese computing, tracing this history from its inception in the immediate postwar period to its efflorescence in a burgeoning network of Chinese, Taiwanese and Japanese computer scientists from the 1970s onwards,’ and promises to ‘take readers through subjects as diverse as machine translation, computer graphics, the rise of computer programming, the software revolution, the feminisation of Chinese intellectual labour, and the growth of personal computing’, with appearances by ‘a cast of eccentric personalities’ at IBM, MIT, the CIA, the Rand Corporation, Silicon Valley, and the Chinese and Soviet military. This will be the story, Mullaney writes, of the rise of input, ‘the result of a 150-year, sleep-deprived, torment-ridden history’. One must imagine Sisyphus happy.