In the 10th century AD, Sichuan used heavy, bulky iron coins for currency. Its exclusion from the booming copper coin economy of the newly founded Song dynasty created a severe crisis. It also spurred innovation. Paper money was invented, spread to all of China, and helped make it into a first-class innovator and world-economic powerhouse. As economic cycles come and go, by the 19th century, paper money had been forgotten in China, silver imports were undercut by British opium, and the country was in dire straits.

What does this have to do with the Chinese computer? Everything! Europe’s economic and technological superiority translated into a sense of linguistic superiority. Language technologies (printing machines, telegraph, typewriters) privileged the Latin alphabet. Even many Chinese began to doubt China’s ability to ever catch up, given its bulky non-alphabetic script. This “Chinese exclusion act,” as Mullaney calls it (p. 18), drawing on the term’s post-colonial undertones, continued into the age of the computer. Fortunately, China’s ingenious tinkerers fought back and regained linguistic sovereignty. The crisis of Latin alphabetic dominance spurred innovation. Its name: hypography. What you see is never what you get. Mediated input – like Zhuangzi’s famous fish trap, the hypograph you type into a keyboard is forgotten as soon as you have the orthograph – turned out to be faster than direct input. Now the world is catching up with China’s “hypographic revolution” (p. 219) and is learning to appreciate the merits of predictive retrieval in the form of autocompletion, algorithms, cloud-based input, LLMs and artificial intelligence. This, in a nutshell, is the argument of Mullaney’s new book, The Chinese Computer.

Historians tell stories to make arguments. Since historians have wrangled the Chinese language from the hands of linguists some 20 years ago, many engrossing stories have been told, and Mullaney’s explorations of technolinguistics have been, at least for me, the most exciting contributions so far.

The prequel, The Chinese Typewriter, was a stroke of genius – a sprawling mix of stories, full of deep and often surprising insights into fields far beyond a single gadget, from language philosophy to technological genius to labour history to globalizing markets. In comparison to that delightful wild growth, the new book’s crafty editorship has created a well-curated baroque garden. Every fact buttresses an argument, each chapter serves one of six main “axioms,” and every axiom supports the larger case.

Chapter one shows that Chinese input is always mediated and never what-you-see-is-what-you-get (WYSIWYG). Chung-Chin Kao had young female typists memorize the entire Chinese telegraph code to operate his electric typewriter (memory was still biological in this computer). Chapter two explains that mediated input can be faster than direct input. Samuel Caldwell (an American) who invented the first Chinese computer, decomposed Chinese characters into basic strokes and used statistical evidence to show that, on average, 4.7 strokes (keys) would produce a character. Chapter three reminds us that immediacy (WYSIWYG, the opposite of “input”) long remained the first choice. In the 1970s, engineers in Taiwan, the PRC and the UK invented non-QWERTY keyboards modelled on Chinese typewriters of the 1930s and 1940s, with large type-case tray beds and character-laden cylinders. In chapter four, Mullaney shows that Chinese input is infinite. In the 1980s, the QWERTY keyboard returned, while mini and microcomputers could handle multiple input method editors. The result was a glorious return of input and an outburst of creativity among IME inventors, of whom Zhi Bingyi’s rise out of a Cultural Revolution “ox pen” is the most gripping example. In chapter five, the author explains that early computing was biased toward Latin scripts. The rapid influx of hardware from abroad in the 1980s created another creative outburst: that of “modding.” In the final chapter, Mullaney focuses on Pinyin and shows that this input depends entirely on hypography. The triumph of phonetic Pinyin-based input started rather late, in the 1990s, due to its disadvantage in identifying individual characters (homophony being the most obvious). Its recent victory was enabled by PRC standardization efforts and advances in statistical computing that facilitated predictive phrase technologies.

Exaggeration is the craft of the master storyteller, but I wonder if the heavy-handed emphasis on Latin script dominance and Chinese heroic reclamation of sovereignty may not be a stretch too far. Such rhetoric resembles the nationalist schizophrenia of concurrent self-victimization and self-aggrandizement. Path-dependency, as the author concedes (p. 225), is a shared concern in catch-up development. Early developers simply have a leg-up. Often, the crisis of “backwardness” drives innovation (as shown in the Sichuan case). Non-crisis leads to complacency. My online banking still balks at an “ä”. In 2025! In Germany!

In fact, the examples in the book point in another direction. Many non-Chinese developers made strenuous efforts to solve Chinese catch-up issues. The Chinese gladly abandoned radically alternative solutions in favour of “modding” once Deng Xiaoping reversed the experiment in technological autarky. “Modding” itself proved a miraculous trigger for creative innovation. China’s most creative phases coincided with its periods of greatest openness. As we now live in an era of narrowing horizons, Mullaney’s books offer strong arguments for globalization.

As the reader may have gathered, The Chinese Computer is best read in conjunction with its prequel The Chinese Typewriter. The latter already introduced many of the ideas developed here in more detail. Some information is lost between the two books – for example, Lin Yutang’s MingKwai typewriter, which pioneered the hypography so central to this volume, is only briefly mentioned (p. 108). Mergenthaler Linotype, who Chung-Chin Kao approached in December 1943 to produce his electric typewriter, also obtained the rights to the MingKwai. What finally aborted both projects was dwindling demand due to the Communist victory on the mainland.

In some cases, editorial discipline may have prevented deeper exploration, leaving room for new research. English stenography is used as analogy, but China’s own history of stenography and stenotype machines remains unexplored. Not much is said about Taiwan’s path after the 1980s. Japan, the poster-child of non-Western catch-up to Western industrialization, looms in the background, but we don’t learn why it stuck with “wapuro” word processors while China moved to multi-IME microcomputing. Nothing is said about mobile computing – which makes me suspect the author is not planning a sequel titled The Chinese Cellphone. PhD students, take note: the topic is up for grabs.