Episode 3 | Typing Chinese

In 1947, the acclaimed Chinese writer and linguist Lin Yutang stunned the world with an invention: the first Chinese-language typewriter with a keyboard. Lin poured years of effort and his life’s savings into the design, which he named MingKwai, ‘clear and fast’. Despite its celebrity and Lin’s high hopes, the MingKwai never went into production, and the lone prototype had long been assumed lost—until it surfaced in the basement of a New York resident’s late grandfather-in-law earlier this year.

What happened to the MingKwai? Why was its invention groundbreaking, and why did it fail commercially? For Episode 3 of 开门见山 | Gateway to Global China, Yangyang speaks with historian Thomas S. Mullaney about the legendary typewriter and the century-old quest to bring the ancient Chinese script into the modern information age.

Guest Bio:

Thomas S. Mullaney is a Professor of Chinese History at Stanford University, a Guggenheim Fellow, and recipient of Stanford’s highest teaching honour, the Gores Award. His work explores the history of information, technology, and modern China, spanning writing systems, computing, and global media. He is the author or editor of eight books, including The Chinese Computer: A Global History of the Information Age (MIT Press, 2024), Where Research Begins (University of Chicago Press, 2022, co-authored with Christopher Rea) and The Chinese Typewriter: A History (MIT Press, 2017, winner of the Fairbank Prize).

Related Materials:

Cheng, Yangyang. 2025. ‘Lost and Found: The Unexpected Journey of the MingKwai Typewriter.’ Made in China Journal, online first, 2 May. madeinchinajournal.com/2025/05/02/lost-and-found-the-unexpected-journey-of-the-mingkwai-typewriter.

Lin, Tai-yi 林太乙. 2011 [1989]. 林語堂傳 [Biography of Lin Yutang]. Taipei: Liking Books.

Mullaney, Thomas S. 2017. The Chinese Typewriter: A History. Cambridge, MA: MIT Press.

Mullaney, Thomas S. 2024. The Chinese Computer: A Global History of the Information Age. Cambridge, MA: MIT Press.

Full episode transcript:

Yangyang Cheng (00:00.60)

Earlier this year in late January, I was scrolling through Facebook when I came across this post in a forum called What’s My Typewriter Worth? A guy named Nelson Felix, who apparently lives in New York state, wrote: “found this while clearing out my wife’s grandfather’s basement. It’s all in Chinese. From my internet search, it looks to be a Chinese made mingkwai. I just can’t find any ever sold here in the States. Is it even worth anything? It weighs a ton!” Felix included some photos of this peculiar machine. By the time I saw this post, it had garnered hundreds of comments from not just typewriter enthusiasts here in the US, but also people from China and from Taiwan. Just about everyone expressed their astonishment, and several had the exact same thought as I did. Someone needs to tell Thomas Mullaney.

Tom Mulaney is a professor of history at Stanford University. Among his many accomplishments, he is the author of two critically acclaimed award-winning books, The Chinese Typewriter and its sequel, The Chinese Computer. I’m so glad to have Tom on the show today. Hi, Tom. Thank you so much for joining us.

Tom Mullaney (01:17)

Hi, it’s so good to be here with you.

Yangyang Cheng (01:18)

So after I saw this post, I realized that you had already seen it and commented on it. And that was probably why it popped up on my Facebook timeline. And so not only did you comment on it, but you actually reposted on your own Facebook page with this particular comment. The most important discovery in Chinese history since the rediscovery of Dragon Bones or the terracotta warriors. So take us back to that moment. What was this machine that has been unearthed from grandpa’s basement? Why was everyone so excited?

Tom Mullaney (01:55)

Thank you for the question. There’s, there’s two ways to put it. One, it’s, it’s a absolutely legendary machine within and of itself, just a storied machine that people have long known about and written about and thought about. And everyone, including myself had not just assumed we had complete empirical assurance that it was gone. So on the one hand, it’s this aura written artifact that has resurfaced in the grander scheme of things in a more sort of 20th century 21st century framework, it also just happens to be the origin of a multi trillion dollar economy that we think of as Chinese, Japanese, Korean computing, internet, IT more broadly. It was the machine that although itself did not ever achieve market success, it was ‘failure’ in its, in its day and age. It was the proof of concept that showed it was possible for character based scripts like Chinese and Japanese and Korean to an extent to be compatible with an age of industrial textual production and then digital textual production that was at that time and and even today is dominated by the English language and the Latin alphabet and alphabetic scripts. It was the machine that proved China did not have to abandon character based script in order to join the ‘modern age’. So when it reappeared it was a one-two punch of historical significance and it’s the only one on earth.

Yangyang Cheng (03:46)

So you talked about the technological significance and the historical and cultural significance of this device. So can you give us a brief overview of how does this typewriter actually work? As you also mentioned, this is the first Chinese language typewriter with a keyboard. So how does that work? Why is that so special?

Tom Mullaney (04:05)

So this is not the first typewriter, first Chinese typewriter. There were others dating back all the way to the 19th century prototype and then mass manufactured ones in the 10s and 20s and 30s all the way to the 1990s. But what all of those machines had in common was those were typewriters with no keyboard. And I know that in this day and age that seems like a paradox, an oxymoron to say a typewriter with no keyboard. But that’s exactly what it was. The dominant Chinese typewriters and Japanese typewriters that were mass manufactured were what are known as index or tray bed typewriters. So imagine a grid 35 by 70 with 2450 metal slugs on a tray of each one of Chinese character. And you’re moving this tray left and right, front and back, bringing the character you want into the type position. So you’re sort of moving this massive set of characters around. And that’s how Chinese typewriters worked. People had assumed, and understandably, that you could never try to build a Chinese typewriter that had a keyboard interface for the simple reason that the prevailing assumption was a keyboard is a thing where there is a one-to-one correspondence with the key on the keyboard and then the symbol you want to appear on the page. If I type the letter L on my QWERTY keyboard on my MacBook Pro, I assume that that letter L appears in the email or tech, you know, web search search bar or whatever it is. And so if we follow that logic, Chinese is a character based system. Let’s just round it up to 100,000 characters. You would need a keyboard with 100,000 buttons. So people had long abandoned the idea that Chinese typewriters could ever look in any way similar to quote-unquote normal western typewriters and then along came in 1947 a machine that did just that it is a chinese typewriter, it has the theoretical capacity of 90 000 chinese characters and yet it only has 72 keys on the keyboard it looks it looks very similar to a remington typewriter from the 20s or 30s and yet it has this capacity and the way it achieves this is through something that is at once completely simple… It’s almost something that, you know, that one of my children would would think of and raise as a prospect because they have not been conditioned. Their imagination hasn’t been sort of normalized yet and absolutely profound and brilliant, which is simply to say why do we assume that one key equals one symbol? Why do we assume that what you type is what you get? Isn’t there another way that we could use these keys in front of us in order to write things? And so Lin Yutang, who is obviously the inventor of MingKwai, his work on this dates back to very far. He worked on this for decades and decades. But at the outset, what he was primarily working on were things like filing cabinets and dictionaries and phone books. There was this little kerfuffle in the Chinese intellectual community back in the 20s, which was how are we going to make a better Chinese dictionary? English speakers have the alphabet. If they go to a library and they look up a book in a library card catalog, they have A, B, C, D, E, F, G.

And in German, they have this and in Russian, they have this. And here in China, we have strokes and radicals and so forth. And so we’re falling behind because it takes us longer on average to find every bit of information than the rest of the modern industrial world. So he was part of there are dozens of people fighting about this, but he he took a turn. He sort of took a left turn off of this mainstream.

And he was the one person who said, okay, I, I see what I see what the utility of this for making a better phone book for Chinese, better filing cabinet. Could we use a retrieval system as a way to build a typewriter? That is something that doesn’t find information, but writes it on a page. And so basically he set down a path, which culminates in 1947, where he built a machine that first and foremost is an information retrieval machine. You are using the keys on the keyboard, not to type anything per se, but to find the character you want inside the machine’s kind of metallic hard drive. You’re telling the machine, here are the features or the qualities of the character I want. The machine, I’m anthropomorphizing here. The machine goes, OK, I found eight Chinese characters in my database, in my hard drive that match your criteria.

And then the user says, oh, I want number seven or I want number four. I know I’m number one. And because he took that left turn and said, you know, we don’t need one key, one symbol. What you type is what you get. He unlocked the possibility of fitting 90,000 characters on a 72 key keyboard and thereby blew open the doors of what is actually possible in the realm of human machine interaction and what we now refer to as human computer interaction. And inadvertently, he gave birth to the first ever input method editor and IME, which are now it is simply the way that all Chinese speaking computer users use all devices. Every Chinese computer user uses some sort of interface that moves through this program called an input method editor that is doing exactly the same kind of information retrieval that Lin Yutang’s typewriter did in 1947.

Yangyang Cheng (10:26)

I’m so glad that you brought up its inventor, Lin Yutang. So for someone like myself who grew up in China, like Lin Yutang is a household name, but he’s mostly known as a writer and as a linguist. He was one of the most important literary figures in the first half of the 20th century. I remember as a young child, I had found like this copy translate, which has been translated to Chinese, my country and my people, Wu Guo Yu Wu Min from my father’s bookshelf and I read it with great interest, even though I didn’t quite understand everything about it. And so for Lin Yutang, what is actually really interesting about the relationship between Lin Yutang and MingKwai, right, was it’s not just the significance and the legendary status of the typewriter has been elevated by the status of his inventor, but actually the logic behind the machine, as you mentioned, embodying the first Chinese input method is intimately tied with the life and career of Lin Yutang himself, he was born in 1895 at the end of the Qing Empire. He was educated first in China and then in the US and in Germany. And then it was actually shortly after publishing My Country and My People with the great success of it in the mid 1930s, he moved to the US to New York City at the encouragement of his good friend, Pearl S. Buck. And then for the next three decades, New York City was Lin Yutang’s primary home and that is also where MingKwai was born. So in some ways one could say that both Lin Yutang and MingKwai were New Yorkers. And I know in one of our previous conversations, you mentioned, right, it is actually really, really important that it is a bilingual and transnational bicultural writer and linguist like Lin Yutang himself, who becomes the inventor of MingKwai. And it is really important that it is actually New York City, being this melting pot and this multicultural city that is its birthplace. So you can unpack some of that significance. Why both Lin Yutang and MingKwai embodies this transnational and bicultural logic.

Tom Mullaney (12:30)

I think that’s exactly right. I don’t want to say that someone who was born and raised entirely monolingually in China their whole life could never have stumbled upon, you know, a similar concept or question as Lin Yutang, but it I think it’s absolutely incontrovertible that this multilinguistic transnational cosmopolitan and also you know let’s let’s also not I won’t glamorize it this displacement and where do you know sense of belonging or or not belonging plays a role and and the way I think about this is it’s almost as if because of his pathway through life and also the specific time and places he was living in that there was an inevitable bifocalism to the way he viewed language and the way he viewed technology. So he had the ability to look at the Chinese writing system as an outsider and an insider all melded into one consciousness, which is something that someone who is steeped in a language just by definition very rarely is able to do. In fact, the very act of a human being when they achieve fluency is to sort of become unconscious of, of all of the ways in which a language or a writing system is contrived and mediated and kind of weird and interesting. And you just, it becomes so second nature that it’s, like the air we breathe. And the same thing is true with technologies, the more in the more familiar and taken for granted. I don’t know something like a computer mouse is the harder it is to realize how weird it is that we reach over with our right hand and move around this little plastic object and it moves this thing called a cursor. So I really think that because of his pathways through, he was able to hold Chinese simultaneously the writing system in a position of estrangement and alienation and and and difference more akin to the way that a foreigner that a a white American or a European have classically looked at Chinese through this othering lens but also to do it through the position of someone deeply not just fluent but just completely saturated by this writing tradition, the language, the legacy and thereof. And this ability to toggle between a deep familiarity and then this ability to alienate is, I think, a key ingredient in why he was able to make this leap of thought. And the same thing is true of technology.

By the time that Lin Yu Tang, it’s really important to remember by the time that Lin Yu Tang is working on this in earnest in the thirties, the concept of the typewriter has been absolutely cemented in the imaginations of the world. A typewriter by that point is full, full, bold, a machine with a keyboard that behaves in a very specific way. You push a key, the symbol on that key is supposed to appear on the page. is, that’s not an option. That is what a typewriter is. Now, existentially, that’s actually not true. I write about this in my book. If you go back to the 19th century before the form that we now know of the typewriter and the keyboard was really stable, it was like the wild west of human machine interaction. There were many different kinds and species of typewriters. All of those died out at the turn of the century.

But there is no God in any religion that ever said from on high that a typewriter must have a keyboard or that it must behave in a certain way. And so he was able to also look at this machine that in his day and age really had achieved total stability and normalization and destabilize it. Destabilize saying, you know, let’s go back to first principles. Why does it have to behave in this particular way? And why does Chinese have to behave in a certain way? If we can destabilize both of them and then find a new bridge, then we might be able to resolve what at the time was understood as an irresolvable paradox. How do you fit 100,000 characters onto a keyboard? It’s like a Zen Cohen designed to be unsolvable. And then he by way of this double estrangement, but also deep familiarity. It took a long time to execute it, but the concept of it emerged very, very early actually in his research process. And that, that I do think that that is part of his pathway through life.

Yangyang Cheng (17:35)

Yeah, and Lin Yutang, was equally prolific and proficient in writing English and in Chinese. And during his lifetime, he was considered a peerless interpreter of Chinese culture to the West. And so I’m really glad that you brought back some of the historical context with regards to the importance of the typewriter in both the global imagination at the time and also how significant it was for Chinese culture at that specific historic juncture. But before we get to that moment, I’d like to stay in the year 1947. And we are recording this in the summer of 2025, but 78 years ago, as you wrote in your book, The Chinese Typewriter, that the summer of 1947 was the summer of MingKwai. And it was partly also owing to the novelty of the machine, but also Lin Yutang’s celebrity in the West that the unveiling of Ming Kui was covered by not just business executives and tech enthusiasts, but by journalists across the world, not just in North America and in China, but in places as far as like Dublin and Jerusalem. But then with so much expectation, so much press attention being placed on this design, also Lin Yutang himself had high hopes for it. He reportedly said that MingKwai could bring a greater revolution to his homeland of China than all the communists, it’s going to transform office life, bureaucracy and commercial life. But not long after, MingKwai just disappeared. What happened? Can you briefly outline that?

Tom Mullaney (19:12)

Sure, you could hardly imagine worse timing for, and this is something very much beyond his control, but so he worked through the conceptual architecture, let’s say, all the way back to the early 30s. I mean, he started to think through the idea of taking his concern with filing cabinets and phone books and then bring it into typewriters. think the earliest records I’ve seen is in the first half of the 1930s. So he’s working through it. He’s thinking through it. But what’s clear is, and this is something that’s only going to hopefully come to light more as we learn more about the machine itself.

He’s not he has many things Lin Yutong is many things and electrical and mechanical engineer he is not so he is not going to teach himself welding and you know casting and lost-wax casting, you know, so he needs a partner one to prototype this to put it to give it ultimate physical form and then of course he needs manufacturing and distribution. So the giants of the day at that point, one is unquestionably, Mergenthaler, Linotype, the importance of which really cannot be underestimated, I mean, they were absolute giants. You’d have to you’d have to talk about Apple, IBM and its day and age to find comparable, comparable importance and IBM itself. And so he begins the process.

This mainly concentrates on the 1940s of trying to explain the concept of this machine to, you know, white English speaking engineers and executives at these firms. Now, on the one hand, they’re very interested in this because for decades these companies have tried and failed to break into the Chinese language market. They are desperate to break into the China market and again, because their mind is so set on the idea that a typewriter is a machine where you have one key per symbol, they just basically write it off. Their one and only hope is that China romanizes their language, abandons characters and adopts a Latin alphabet. That would be, from their perspective, a godsend. But short of that, they are really at an impasse. And then along comes this proposal. It’s just not the only one that they’ve seen, but on the other hand, they are and we have to empathize with them, not just critique them historically. They’re sort of being asked to entirely reimagine everything about their own professions and lives, meaning they’re all of their training as engineers, every last machine they ever built and oversaw the construction of, especially with typewriters. They’re being asked to almost start over in their minds and understandably, that’s a very big ask. And so you I’ve read through all of the private correspondence between, you know, among and between Mergenthaler executives and IBM and Mergenthaler. And they’re really struggling to try to understand not how it works, but is this possible? Is this actually a paradigm that typists, everyday typists would actually know how to use, would use, et cetera. So the background radiation of 1947 is the support from these absolutely essential Western companies I would describe as interested but fragile. They are they are not they are not sure they’re they’re sort of on the fence. They want it to work, but they are very cautious about this. Then you throw in geopolitics. You throw in the geopolitics that, you know, the Second World War has just ended and the writing on the wall with especially in the wake of growing tensions, the failed effort of you know, we think of the Marshall Plan in Europe, and that was a great success. There was also effectively a failed Marshall truce in Asia where there’s a great book about this, where it just was not able to stabilize the frozen civil war that preceded the outbreak of the Second World War between the nationalists under Chiang Kai-shek and the communists under Mao. And the general sense was that this conflict is going to erupt once more.

So on the one hand, there’s just instability, which of course no manufacturer likes. And so there’s a hesitation about that. But then there’s this added little curve ball in the mix, which is that Mao Zedong for a while has been sort of mouthing off about Romanization for a little bit. Is it actually conceivable that if the Civil War turns out in a certain direction, this machine would be pointless anyway? Because they would undertake romanization, the abandonment of characters. in fact, strangely, I’m not sure I’m not to say that they’re rooting for Mao, but that would totally change the playing field from the perspective of a company like Mergenthaler or Olivetti or Remington or et cetera. So it’s a wait and see attitude that kicks in with the turmoil of the second half of the 1940s. And Lin Yutang is not in a financial position to wait and see.

This thing is costing him, you know, his burn rate to use good VC terminology. Now there’s startup terminology. His burn rate is high, his runway is short and he needs, he needs action. And this is not when action is going to be taken. So, you know, in an alternate universe, it’s hard to know if it’s plus plus or minus years you know before or later would the story have been different in terms of the support that Lin Yutang desperately needed from these companies of course that’s speculative but the timing of 1947 I mean 1947 basically is the moment of the re-eruption of the civil war and when the tide starts turning very definitively towards the end of 47 towards towards Mao by 48 the tides have turned very very systematically Chiang Kai-shek has already planning and systematically planning the retreat to Taiwan that Mergenthaler isn’t going to come within a mile of this of this kind of initiative. And again, there’s there is a kind of baseline fragility to their support born of just is this approach to human machine interaction even possible? We don’t know because it’s so different from our perspective. It is so unlike any typewriter we’ve ever seen.

Yangyang Cheng (26:00)

So, as you mentioned, if we come to 1948, Lin Yutang did sign a contract with Mergenthaler to explore the commercialization of MingKwai. And so, Mergenthaler acquired the machine at first and a few years later also acquired a patent from Lin Yutang. it had the physical possession of the machine and I guess technically also the intellectual property, where Minkwai basically just like languished in the Morgenthaler office or warehouse for over a decade. And it is long believed that it was just thrown out when Morgenthaler moved offices in the 1960s. But apparently a toolmaker by the name of Douglas Arthur Jung, who was working at Morgenthaler at the time had somehow picked it up, kept it, passed it down to his son, I’ve learned, and then later to his granddaughter, Jennifer, who is married to Nelson Felix. And so this is how MingKwai has re-emerged three quarters of a century later. In some ways, it is a time capsule. In some ways, Ming Kui is also like a portal, right, as we mentioned. It didn’t just bridge different languages and cultures, but it also bridged like past and the present and in some ways also the future because at the core of typewriters and Chinese language processing technology is the question of modernity, right? Who belongs in it? Who can embrace it? Who gets to define it? And so I know you just mentioned earlier that like by the early 20th century, the typewriter is such a foundational symbol of modernity and the modern state.

On the other hand, the Chinese script and Chinese writing methods have existed for millennia, quite literally to the time of the Dragon Bones. So can you explain a bit why was there this sudden urgency to revolutionize or change the Chinese script? Why was there this urgency to pursue a Chinese typewriter?

Tom Mullaney (27:58)

That’s a great question. What I often tell to my students in my just history of modern China course at Stanford, I just say very, plainly that you don’t have to go back very far in history to where the Chinese language is not a problem. The idea that the Chinese writing system is a quote unquote problem and that its complexity is causes issues X and Y and Z is an issue of incredibly recent vintage. And the the one example you have to just have to give for students to wrap their head around this is what sometimes to is the woodblock revolution. That is, everyone talks about the Gutenberg revolution and what it did to the declining cost of books and circulation. But we need to remember that, you know, there were foreign European visitors to late imperial China, to Edo Japan, and they were shocked that you could buy a book for roughly the price of like a bowl of noodle soup in the market at a time when to buy a book back in Europe that was being copied by calligraphers and in scriptoria, you have to mortgage your house for some of the I mean, these were these were immensely costly, costly things. And this was achieved with xylography with with wood plank printing. And in woodblock printing, there’s no problem. In fact, it’s incredibly conducive and compatible with very sophisticated production cycles, the division of labor, you have like a, know, you have a calligraphy master who lays down the ink of the original. And then you hand it off to a woodcarver and the woodcarver doesn’t even need to be literate in in Chinese. They just need to be deft and skilled at their particular skill set, is the, you know, carving away at the woodblock. They’re great for storage. They’re amazing for the integration of characters and images of stuff that, you know, it’s sort of desktop publishing before the term. So this is to say, that’s just one example, but this is to say there is a time for the greater part of human history in which there were many different entry points into the question of how to make lots of copies of texts and we have to remember that China, I’ll anthropomorphize China for a second, China invented movable type took one look at it and said woodblocks will be just great. thanks. Many people imagine like that China somehow missed the boat on that or didn’t see the importance They totally understood what had been produced and then looked at the question at hand and said this is not at least in its present form, woodblock prints and these other approaches to the mass, you know, reproduction of texts is a way better option. And so there’s this time in history where you can call it like multimodality, it was a multimodal world, there were lots of different starting points and people and different cultures and language communities could choose different kinds of starting points. This begins to change and, to me, I make this argument in the book, The Chinese typewriter, the technology that really played a role in changing the overall distribution of starting points, let’s say, is electric telegraphy. So this is the communication at a distance where you’re sending messages at a theoretical, theoretically at the speed of light across, know, submarine cables across hundreds and thousands of miles of distance. This was arguably the first technology for which there was absolutely no Alternative way to do it. There was no woodblock counterpart of the electric telegraph and so suddenly for the first time,all language had to be funneled into one paradigm, let’s call it. That paradigm is the paradigm that we are all right now. Everyone on earth is the great great great great grandchild of. The paradigm is very simple and to me it’s the governing principle of the digital age. What is the thing that unifies every last digital technology? They’re not virtual, they’re silicon, they’re transistors, they’re as stuffy a material as every technology. The weird thing about the digital age is that everything that a digital object can be, has to be encoded in it from the start. So what that means is, if everything it can be has to be programmed in already. So I’m looking at, I’m looking at a computer monitor in front of me, and it just happens to have, you know, the podcasting software or there’s my inbox is open. There’s different colors. There’s different words. There’s different graphics. There’s different, but every single one of those colors, every single one of those syllables or symbols is already by definition encoded in a condition of possibility in the digital machine. So in Morse code, every symbol that you can send over wire has to be in the code already. In a dot matrix printout, every possible configuration of dots is already there in a state of potential.

That is completely different than, say, clay and cuneiform tablets and paper and wood. In all of those technologies that have ever existed, human beings are either adding stuff to the equation, that is, they’re adding ink or they’re adding steel or they’re removing material. They’re carving away wood. They’re scooping away clay. So it’s an additive or subtractive possibility, calligraphy, so forth.

The digital, everything it can be has to be there already. And this is where you get to problems that now still to this day kind of haunt Chinese . So now we’re in a situation where if we want to put the entirety of Chinese of English in a computer, it takes N amount of kilobytes of memory.

All we need to store in memory is lowercase a through lowercase c uppercase a, you know, you get the idea. But now if we want, if we want this computer, this digital object to also quote unquote speak Chinese, we have to put the entirety of Chinese in it in memory and gosh, look what happens. It requires a hundred X the memory of its English language counterpart. This is a problem now.

And if we want to make Morse code, a telegraph code for Chinese, my gosh, we’re going to have to fit thousands of more symbols in this code than its Latin alphabetic counterpart. So suddenly the name of the game, like the rules of the game of the digital age were ones that placed Chinese consistently at this disadvantage. And this is what re simultaneously to your point.

On the one hand, we’re very, very, very loud mouthed, iconoclastic reformers who said the only way we’re going to join the industrial modern age is by, you know, replacing Chinese characters with Esperanto or English or Romanized versions or French, et cetera. And then the other, the other folks, which also kind of loud mouthed and in their own way, iconoclastic saying. We, we have to, we have to figure out how to build and make a telegraph for Chinese, how to make a typewriter for Chinese, how to make all of the technologies that all of these other countries and languages have. if we don’t, well, one thing, they’re just useful, but, for us, it’s kind of a trial. It’s, it’s the stakes are higher because if we can’t prove that it’s possible to make a Chinese telegraph code, it’s possible to make a Chinese typewriter, then the jury, the decision is back from the jury, which is, yes, it turns out that the absolute unavoidable price of modernity is the abandonment of our culture, our language, this deep connection with the past. That, it turns out, is the price tag, the price of entry into the industrial age. And many people were unwilling to accept those terms of service.

Yangyang Cheng (36:42)

Listeners, don’t you wonder how much information is encoded in Professor Tom Mulaney’s memory?

Tom Mullaney (36:51)

I actually have a very shoddy memory. It’s very funny. I’m the amnesiac historian. But this stuff is so close to my tongue that it’s probably stored there more than my brain.

Yangyang Cheng (37:03)

Well, so following up on that, since you mentioned like different types of Chinese writing method, if we go from the global paradigm to the Chinese specific one, you mentioned calligraphy. Actually, Lin Yutang himself had said one of his wishes for retirement, even though he worked till the end of his life, was he could like practice his brush writing, his Chinese calligraphy. But like writing with a brush or a pen is such an individual process, right? And it’s also intimate in a way while casting characters into a typewriter or inputting that to a computer memory, it inevitably involves a process of standardization, but also selection, as you mentioned, right, what to include and what not to include. And so for Chinese language to enter the information age, that process of technological development also coincided with the process of nation building and state building in China.

And sometimes these processes and these efforts were indeed intertwined. And can you unpack some of that process in terms of how the Chinese state played a role in standardizing the Chinese language, both in terms of writing and in terms of pronunciation, and how that effort is reflected in how Chinese language is being inputted into a computer today?

Tom Mullaney (38:22)

Yes, this is a really important and often, I think, misunderstood dimension. So thank you for asking. So I think it’s fairly well known, but just very briefly, there have been efforts for about 120 years in China, from the late imperial period through the 20th century to what was often referred to as to simplify to jianhua to simplify Chinese. And what’s interesting is oftentimes people will take that at face value. yes, of course, Chinese is complex.

Makes sense to want to simplify it. This is the age of mass education, public education, the idea of we need we need, you know, we’ve moved from subjects to citizens after the 1911 revolution and the idea of more people taking part in the economy or politics requires literacy. Chinese is so hard to learn, et cetera. What’s interesting about it from a broader perspective is the way I often put it with my students is that simplicity is always a relative concept, simple with respect to what?

And the short answer in the case of simplification campaigns for Chinese was simpler to write by hand. That was basically the governing logic of simplification. So simplification campaigns for Chinese often took the form of being able to write a given Chinese character with fewer brushstrokes or fewer pen strokes. So there’s a, you know, famous example of the character long, dragon which has a 70 % reduction in the number of strokes it takes to write it in in traditional versus simplified form. What’s really, really interesting about that is that things that are simple to write, simpler to write are not necessarily simpler to read. Because if you think about what the brain is doing when it’s reading is the brain really likes lots of information and information in this case are strokes.

Because the brain is using it to recognize the character, to remember it, to differentiate that character from every other character on the page. So for something to be simpler to recognize and simpler to find, that can be at odds with something that is simpler to write. And the reason that this is really interesting is in the, especially in the second half of the 20th century, once you begin to have Chinese computing and the emergence of input… is a wonderful interview that I did with a Taiwanese American inventor Chan Yeh, who has since passed away, who started the company Idiographics in Sunnyville, California, he was the closest thing to the Steve Jobs of Asia in his time. His company, Ideographics, the machine he built, fundamentally transformed the newspaper industry, the tax registry industry in Taiwan. And it was also an input system based model.

So you are using his machine and the company’s machine to retrieve Chinese characters from memory, not to compose them or type them, so to speak. So I can’t corroborate this, but it may be apocryphal. But even if it is, it’s it’s actually empirically true is a meeting that he claimed to have had with one of the principal architects in mainland China of simplification and Romanization and looking at the IPX machine, this retrieval input based machine, the response reportedly was, my gosh, if this system or something like this system had exist, had existed, you know, 10, 20 years ago, we would never have had to simplify characters in the first place. Because I think about it now, what percentage of Chinese writers write stuff by hand versus write stuff with the aid of a digital device. And that percentage has changed dramatically in the last 30 years. What that means is simpler, simpler to retrieve computationally and simpler to write by hand are not the same thing. And so there is this really interesting road not taken that had the computing revolution, the input revolution. And this is where Lin Yutang comes back into play. Let’s imagine a world in which I don’t know, the Marshall mission had proved better that, you know, the Chongqing meeting, etc. between Chiang Kai-shek and Mao had gone better, there was a confederated government, there was a power sharing arrangement. You know, imagine an alternate universe in which the civil war does not erupt in the way it does, and then Mergenthaler Linotype mass manufacturers, you know, the, Ming Kui and then Samuel Caldwell’s first type, first Chinese computer gets inroads into, and let’s say that the input revolution took place 30 years earlier than it did in real life. There is.

It is not inconceivable to imagine that the simplification of characters that happen in, you know, mainly in the mid late fifties may never have happened because it would, it would not be necessary. wouldn’t be a simplification that in any way would be useful because you would be retrieving characters from memory. And when you’re retrieving characters from memory, it’s a whole different, it’s a whole different ball game. So these are, these are these moments in which, uh, state led efforts of it’s called language policy or language transformation, which are typically led by elites and governments and academics and ministries and so forth, sort of interact in funny and unpredictable ways with the trajectory of technology itself. And this is one that I’ve always loved talking about of that, simplification, which we now know, and everybody, know, everyone in mainland China uses simplified characters. There is an alternate universe in which that may not have happened if the input revolution, the Chinese IME revolution had happened earlier than it did, which is something I always find very interesting.

Yangyang Cheng (44:50)

Yeah, if one studies history, one would understand that the path is always contingent. On the other hand, there are also underlying structural forces as you beautifully elucidated, right, how the Western hegemony is intertwined with the advent and the popularization of the typewriter and the computer, and also how the invention and the popularization of Chinese language input methods is intertwined with the establishment of Mandarin hegemony in terms of the Chinese language. And Lin Yutang himself, being a bilingual, bicultural writer, had played an important role at this important historical juncture in the mid 20th century. And you also did mention that Lin Yutang was a man of many talents, but one thing he was not a welder. So one of the interesting things about the rediscovery of the MingKwai typewriter is it opens as you also mentioned, right, opens a window to some of these hidden labor of engineering. And on the note of hidden labor, I would like to get to another aspect of an often overlooked or marginalized labor in the computing, the history of computing, that is the women.

And in particular for MingKwai when it was unveiled, was Lin Yutang’s second daughter, Ling Taiyi, who was a prodigious writer and novelist by herself, was later the editor of the Chinese edition of the Reader’s Digest. She was the demonstrator for the typewriter. And at around the same time, there were also notably a few young Chinese women who played an important role in testing and demonstrating some of the earliest prototypes of Chinese language processing and input methods, including one woman whose photo graces the cover of your book, The Chinese Computer. She was by the name of Lois Lew. Can you tell us a little bit about her life and work? In particular, you got a chance to interact with her, right?

Tom Mullaney (46:48)

Yes, definitely. No, that I consider that to be one of the most important highlights of my career life. It was just something I’ll never forget. So I think that, you know, the basic premise was, and one would understand this is Lin Yutong and then a kind of competitor who was interacting with IBM, Gao Zhongqin.

Like I mentioned before, they really have an uphill battle. They have to convince high level executives and engineers who have a very cemented understanding of technology. They have to sell these engineers, these guys on a totally different mindset paradigm. And as part of that, they have to prove to Mergenthaler, to IBM that quote unquote, anybody can use it. Because you’d have to, you you imagine if you meet the inventor of a system, and they are really good at using it. Well, who cares? They invented it. They probably practiced this a thousand hours. That’s not the average typist out there. The average typist in the workforce is someone with maybe a high school education or clerical training or, and those are the, those are the individuals in an economy who would need to be the ones who can use this. And so, and this is a deeply, obviously, gendered act that they were undertaking this is is well, I so I need a demonstrator who looks the part of the person who Ostensibly would be the end user of this product if it were ever mass manufactured and that’s going to be following the paradigm Especially in the United States Western Europe of the typewriter girl of the young, you know 20s in 20s or 30s Young woman who is going to be administrative assistant, the secretary. And in the case of Gao Zhongqin, the inventor of the IBM electric Chinese typewriter, he hired, he literally recruited and hired two young women in particular to serve in this role. And so the case of Lois Lew, who…. You’re right. Her photograph is on the cover of the Chinese computer. I had come across her image and historic footage of her for years and years and years in the course of writing these two books. And at the time I was posting things online. I think I had a blog and I received a comment.

Yangyang Cheng (49:17)

It’s a photo from an IBM brochure at the time.

Tom Mullaney (49:20)

It’s a photo from an IBM brochure. And there was a comment on one of my blogs. I think it’s probably that’s a dead link by now saying “I am the woman in the footage. That’s me”. And at first I was very skeptical. I was like, this seems like a maybe just someone spoofing this. But after corroborating, I indeed realized this, this is true, and for reasons I never quite learned the trail went cold… and the fork in the road for me was when I was… originally the Chinese typewriter and the Chinese computer was one book. Then it became, became necessary for it to become two books. And there was a big question of where to put the IBM typewriter chapter. On the one hand, it’s a typewriter. So it belongs in the Chinese typewriter, but it’s an electro automatic system that has aspects that are predecessors to the computing age. So it could also go in the Chinese computer. And it really, really was a throw of the dice decision. It wasn’t a, there wasn’t an obvious answer. And ultimately I clicked and dragged that folder into the folder marked the Chinese Computer, which basically at the time when I was finishing typewriter was I’ll get back to you later. I’ll get back to you later after I finished this book. And it was in that interim, you know, thank, thank the universe that a former employee and friend of Lois Lou wrote to me again out of the blue saying, I know the woman in the picture. I worked for her. I said, my gosh, that’s… and maybe because this time I was vouched for by a third party. In any event, one day the phone rang and and and I was on the phone with Lois Lou and we had this very, long conversation about her life, just an incredible life. Also cross cut with major challenges. Her family lived through the war with Japan. She, she was asked to do things that, you know, I’m a parent, I can’t fathom my children doing, you know, going alone across the ocean to marry a man she had never met discovering that the man to whom she would be married had effectively, or maybe the matchmaker had effectively withheld information about their financial status and lived a very different life than what was expected. But somehow she’s this incredible pinball going through the pinball machine of life is getting knocked and dinged in all sorts of ways, but out of that, she’s such a resilient like brilliant fiery person that one thing led to the other. She, she gets a job at IBM in just in a, in the factory setting. And then the IBM Chinese typewriter comes out and, Gao Zhongqin is hunting is hunting for someone to demonstrate the machine because he needs to convince journalists and, so forth that quote unquote, any anybody could use it. And she gets, she gets the job. And, what she was asked to do from a conceptual standpoint, from a cognitive standpoint is amazing: she had to memorize four digit codes for thousands and thousands of Chinese characters. And then stand, basically sit on a stage and, you know, more or less like a theater full of hundreds and hundreds of journalists and onlookers and diplomats and state leaders when they did the tour of China in particular, and be able to, from memory, hear a Chinese character, someone calling it out, and then be able to type four digits on a, on a keypad and either succeed or fail. There was no feedback loop. There was no second guessing. There was no error checking mechanism: she either got it right or it was wrong with Gao Zhongqin pretty much breathing over her shoulder where from his standpoint, his whole venture, his whole, his whole gamble is riding on the performance of Lois Lew, and she nails it. It really brings home this point from the famous book, you know, when computers were women, in this case, I have a chapter in it when, when, when input method editors were women, when IMEs were women, was, she had turned her body and her mind into the embodiment of what we now think of as a Chinese input method editor.

Yangyang Cheng (53:53)

Yeah, and I recall you mentioned that 70 years later, Lois still remembers some of the code and he was able to…

Tom Mullaney (53:58)

She still did, she still did. I think I asked a question about, you know, I know this is basically an impossible thing to ask. It’s like asking someone to tell you how they learned to ride a bike. It’s like, kind of I just know how to do it now. I can’t tell you what it was like when I first didn’t know. But just how did you do this? And of course, understandably, it’s just sort of more casual about it. just, you know, I just had, just spent a lot of time. learned it. But, you know, to this day, boom, character this, here’s the four digit code for character this, there’s a four digit code. Now I just remember it. And I got off the call and I had the, I have the, copy of the IBM electric Chinese typewriter manual that basically is the, you know, the guidebook that tells you all of the four digit codes for all of the characters in the machine.

And I like, there’s no way this is not possible. And lo and behold, three for three, she nailed all of them.

Yangyang Cheng (54:50)

So continue on the note of when computers were women and when people’s bodies were embodying these technologies. I’m also reminded, like, for example, back in 1940s, when these terminologies such as like kilo girl was invented that to describe the computing power of machines where a women workers computing power is an absolute unit. And so this takes us from the past to the present and to the future. And I’d like to bring up an example.

Recently, I had this conversation that is about women in science in China. And one of my co-panelists, who might believe is a former student of yours, Professor Gina Tam, made a really astute comment that some of these new generative AI or AI assistant technologies are basically trying to reinvent the 1950s female secretary. so throughout our conversation, we’ve talked about how modern computing technology have embodied and at times reinforced or been reflecting these different types of hegemonic power, whether it is across East and West, whether it is with regards to the Chinese context of Mandarin hegemony and standardization of languages or the gendered hierarchies. So I have a somewhat philosophical question for you. Do you think computing technology is by construct at its core a technology that reflects and reinforces existing power differentials? Or can computers become a tool of liberation and rebellion rather than tools of capture?

Tom Mullaney (56:30)

That’s a question that’s more obviously more important now than ever. The way I would frame the question itself, meaning, you know, I don’t have an answer, but the way I would build the question and try to set about answering that is… it brings me back to this question of what is the digital age? What is the essence of the digital? And, is that essence? which side in the in the binary you’ve described, which I agree with the sort of, you know, hegemony and emancipatory spectrum, let’s call it. Which side does it favor, if any? You know, is it is it neutral? Does it tip in one favor or another?

I do have to say, not pessimistically, but it does seem to have a tendency or a structural tendency towards concentrations of power. And that seems to play out. But it comes back to this question that for me, the digital age, the defining feature of it is everything that it can be, it must already be. People forget that. Some people will will. I heard a comment by someone I admire very much. Very, very intelligent person. This comment is dead wrong, though, said like, OK, now that we’ve gotten to Retina display iPads that were sort of, this is just like woodblock print, like you can do anything, you know, it’s, it’s, and to which I have to say, it does not matter the resolution of that screen. It doesn’t matter. You could put an infinite number of pixels in there that are approaching infinity.

But even still, every one of those has to resolve down at some point to honor off some value or another. Everything in the digital age has to resolve. That’s just the nature of the thing. If I take out a pen and a piece of paper and I write the letter H on it, I do have to achieve the effect in the mind of and the body of a viewer to be able to discern like, that’s an H, but that’s, that’s a semiotic kind of game. That’s like an H is an H because it’s not a K and it’s not an L and it’s, you know, it’s, it’s, this brings us back to like Pierce and, and, Saussure’s semiotics. It has to achieve certain things.= But if I, touch the letter H on my keyboard, it has to be mapped and resolved down to a particular address or code unit or digital thing that has to already be there in the computer in memory, in the standard. And if it is, that little cascade of resolutions will get that, it’ll find its match.

And then this like it’s like a boomerang. Then then there’s a cascade that comes back in the other direction. And boom, I have the letter H on my Microsoft Word in whatever font I have chosen. But it’s a serious just this it’s just this domino effect of cascades of resolutions. I have to resolve down, resolve down. And so if something isn’t there, it cannot be. That is uncompletely unlike using your finger to draw circles in the sand. is completely unlike using a stylist to make wedges in, in wet clay, calligraphy, penmanship. It is existentially a different thing. You cannot type a character that’s not in Unicode or is in your Unicode private use area. You cannot type that, which is not encoded. You can’t produce a color in a digital frame that is not already possible. So there’s no, you know, sort of like John Lennon, there’s nothing, there’s nothing you can do that can’t be done. That sort of model. There’s nothing you can sing that can’t be sung. That’s the digital is John Lennon’s lyric.

There’s nothing you can write that cannot be written. But in the realm of fingers and sands and wedges and clay, you can write that, which you can make that which doesn’t already exist. So right then the playing field, I fear, can skew because then the game can be about what’s already there. Who has the power to decide what is there? What are the standard bodies that have the power?

How do they decide what gets in and what gets out? And now once we get to the age of input, we have a whole different ball of wax, which is, and this is the argument I make in the Chinese computer. If anyone believes the argument of the Chinese computer, the argument is very simple, which is we have our toe in the water of a new epoch of writing that I call the age of hypography. Think of the age of input. We’re 50 years about into this age, or it’s very new.

The age we have been in is thousands of years old. So this is a very young infant age. And in this age, the act of writing is no longer the act of making something that didn’t exist. It is no longer the act of composing something. It is the act of finding something in memory from memory. So, so we have the issue of what’s in there, what’s not in there. But then there’s this other issue is once that writing is the act of finding something, almost anything can become an act of writing: facial recognition, the recognition of the different signatures of a person’s stride when they walk, you know, the chemical content of sweat, the idea that governments can now do surveillance by surveilling wastewater. All of that stuff are inputs into a retrieval system. What they’re retrieving, why they’re retrieving it, who gets to see the output of what they’re retrieving, I don’t, you know, that’s above my pay grade. But what I do know is I’m not doing that. You’re not doing that. Corporations are doing that. Governments are doing that. So there is this, there is a bias to use the classic history of technology as sort of the bias of communication. There is a, there is a bias that tips in the favor of concentrations of power in the age of input. Could there be an emancipatory possibility that I don’t understand or we haven’t seen? Absolutely. But in terms of its initial applications, it does seem to favor those who have the ability kind of in forms of capital accumulation to already have the capacity to accumulate data, compute power, etc. And that does should give people pause when thinking about what exactly is in store in the 21st century.

Yangyang Cheng (01:04:00)

So speaking of capital accumulation and a center of our digital universe to wrap up our conversation, we know that MingKwai has found a new home at Stanford University thanks to a generous donation from the Bin Lin and Daisy Liu Foundation. so what’s next for this remarkable device? Is there anything you can share with our listeners?

Tom Mullaney (01:04:26)

Yeah. there’s going to be a permanent display of the typewriter. There’s probably right now, I think they’re commissioning a custom vitrine for it. There’s going to be an exhibition that accompanies the sort of debut of the machine. There is the possibility of a dedicated reading room just for the machine and related objects. It’s going to play center stage at the May, 2026, ATypl conference taking place at Stanford. And there’s already, I think, protocols and sort of FAQs and workflows being put in place by the libraries that I just want to sing their praises. Mike Keller has been a absolute Force for this, Regan Murphy Kao, Zhao Hui Xue, just Julie Sweetkind Singer, Li, they’ve just been amazing. This brings us back all the way back to the Facebook page and the conversation I had with, especially with Jennifer of saying, listen, of course, this is your decision.

But I truly hope that you’ll decide to find this find this machine in a museum or a research institution somewhere that knows how to take care of it to conserve it and also knows how to make it available in a safe way to researchers, students, just interested members of the public. And that’s precisely what Stanford is doing. So there are I actually am heading to a meeting very soon for the for one of the five different sort of plans for it. And then the other big thing that is alluded to, I think, in a few of the different pieces on it is to try to create a working replica of the machine because it can only obviously exist in one place. There is only one of them on earth. And even though it will be open to researchers and visitors and so forth, it’s not something that everyone can see all the time. And so can we make a replica, a working replica of either key components of the machine, the entirety of the machine and then in my dream would be deposit one at the Lin Yu Tang house, deposit another at the Smithsonian, the other, you know, Ars et Metier in Paris, technology museums in, mainland China, so that it becomes something that there’s always a nearest point for anyone interested in learning about it.

Yangyang Cheng (01:06:41)

And I’ll also mention here that the Lin Yutang house is in Taipei and Lin Yutang spent the final decade of his life after living in New York for three decades in Taiwan. And so a closing note, so by the time when you can type on the MingKwai, do you have thoughts on what your first message would be? Is it Hello World?

Tom Mullaney (01:07:03)

Haha… I think that is a great question. I think what I would maybe in my ultimate dream, so the ink spool from the device is still in there. So the final words ever, ever typed on Minquai are still frozen in the ink spool that is in the machine, would need, we would need the same kind of technology they use to do virtual on scrolling of say the Dead Sea Scrolls.

But what I would love is to figure out maybe Lin Yutang, maybe Lin Tai-yi stopped mid sentence. And so I think the greatest homage would just be to finish a sentence rather than write something completely new what was the last message that was ever written on this machine, which is sitting there in a state of potential, right now at Stanford in the original machine, then, you know, maybe do whatever we can to help finish that thought.

Yangyang Cheng (01:07:59)

So unfortunately, this is the end of our show, but the conversation and the journey of MingKwai continue. Thank you so much, Professor Tom Mulaney.

Tom Mullaney (01:08:07)

Thank you very much for having me.