End result of these rumors: Chinese people would also sometimes kill missionaries, and it turns out, this puts a strain on diplomatic relationships.
So I heard this and I think I stammered out something like “uhhh well actually that might help precipitate a war? I don’t want my hero doing that?” And we moved on.
It stayed with me over the years, the thought of those children. What was it like? What did they do? How were they raised, what did they think of themselves, did they remember where they were from? Where were they from?
These were not idle questions for me, a (depending on how you count) third- or fourth-generation child of Chinese immigrants.
Amelia’s story has been in the works for a long time. (I structured the Worth Saga to lead to it, for reasons that I now think were probably misguided, but hey, you learn from experience.)
About all the telegraph stuff
So! Obviously I rewrote a lot of telegraphic history in this book! Let me acknowledge what I changed, and what is real.
The actual transpacific telegraph line wasn’t built until 1902. There are a lot of reasons for that, but none of them were technological. The transatlantic line was made many decades earlier—before the time-period of this book—and the transpacific line was just three transatlantic lines long. In our reality, the transpacific line went from Midway to Hawaii to San Francisco, and if you’re wondering why it took until 1902 to establish this, it’s likely because Hawaii was not annexed by the United States until 1898.
I personally did not feel like advancing the timeline on the annexation of Hawaii by thirty years was a good idea, and so I had to find another route. I ended up coming up with a northerly route (northern Japan to a made-up North Pacific island and then down the coast of British Columbia in what is now Canada and what was then a British conquest) in large part because the timing was just right to slip into the early Meiji era, in which the Japanese emperor wanted to modernize.
Japan did in fact adopt the telegraph very early on, and Wabun code was used extensively in Japan. (More on Japanese code in a few paragraphs.) But their telegraphy was actually built in collaboration with a Dutch company, and so I also had to rewrite history there so that Grayson would handle the telegraphy in exchange for being allowed to start his transpacific cable there.
So now we get to the question of Chinese telegraphic code. The first Chinese telegraphic code in reality was in fact made by a Frenchman, and to put it mildly, it sucked. It was 1000 characters long and you looked up a character and then sent that number, and when I first contemplated writing this book, I said, “ha ha this sucks so much, surely I will be able to come up with something better with the benefit of one hundred and fifty years of history on my side!”
I was, in fact, right, but it was not actually easy. My first version of Chinese telegraphic code in this book was simply a different 1000 character encoding that was arranged so that it was easier to code and decode. Was it better than the first version? Yes.
But then, while I was reading about Wabun code, I found copies of old Japanese commercial codebooks. Commercial codes for telegraphy were common. You paid for telegraphs by the word, and so if you had a very common phrase, such as “let’s go jump in a lake together” you could pay for seven words to send that phrase, or you could agree that the word “puppy” meant “let’s go jump in a lake together.” Companies made entire codebooks that took common commercial phrases and boiled them down to one word.
In any event, Japan also did this, but their commercial codes were in fact structured differently than English codes. The Japanese ciphers were composed of pairs of kana (the elements of the Japanese syllabary), so that companies could assign a Chinese character (or perhaps four, and you’d have to determine from context which was meant) to two (some used more than two) transmitted characters. (My Japanese is very bad, and Japanese written 150 years ago is much harder to parse than modern Japanese, and 98% of my Japanese was acquired trying to read figure skating interviews…anyway, long story short, I had lots of fun figuring out Japanese commercial ciphers.)
This was the first not-terrible method of transmitting Chinese characters that I had found—in fact, in at least some cases, the kana pairs were associated with the Japanese pronunciation of the Chinese character, and so it made sense both for coding and encoding as a solution. It wouldn’t entirely work for Chinese as a native encoding, but it got me thinking about how to transmit Chinese characters using pairs as a sort of cipher.
At this point, I fell back on something completely unrelated that I did a while back. Back in 2016, I noticed that there were no dinosaur emoji. There were other extremely serious problems in 2016, but I asked myself the very important question: Why are there no dinosaur emoji?
The answer I got was this: Because nobody had proposed them yet.
And that was a silly reason. I was a person, and I had a computer. So I wrote a proposal to the Unicode Technical Committee, which you can read, and behold, dinosaur emoji. In the process of doing this, I learned a lot more about emoji than I had ever known before, including learning about a character called a “zero width joiner.”
As an example, if you want an emoji that is going to be a thumbs up, but the thumbs up is colored a dark brown, that emoji is actually three emojis: the yellow thumbs-up emoji, the zero width joiner, which signals to the computer that you’re joining two emojis, and an emoji that corresponds to the dark-brown skin color.
I put these all together and realized that if you wanted Chinese telegraphy that actually represented Chinese characters with radicals instead of just encoding it, you could do it with zero-width joiners. You’d have to have multiple zero width joiners—three in total, to specify horizontal joins, vertical joins, and containment. But hypothetically, those would provide you all the tools you needed to represent Chinese characters.
In theory, this would have a lot of advantages. In practice, I don’t actually have a team that would be able to do any wide scale testing of it, and it really is just fiction, so anyway… yeah.
Back to telegraphy: yes, the first transatlantic telegraphs were in fact that expensive—on the order of four or five pounds, which is an enormous amount of money.
Yes, the whole shebang about gutta percha is real. They really did chop down an entire tree for ten ounces of gutta percha. They really did chop down forests of gutta percha for the transatlantic telegraph alone.
I read these articles from the time that went something like this: “ha ha people say we will eventually run out of gutta percha, but it hasn’t happened yet, so who is to say it will ever happen? Let’s ignore all the conservation issues with this.”
Yes, when they hit the point that people were like, “oh no we are about to make the gutta percha tree extinct?” people did eventually have the bright idea of harvesting just the leaves and it turns out that this works just fine, and yes, there is a tree called the balata that produces a latex that is indistinguishable from gutta percha.
Today we use a version of plastic called Bakelite to insulate our cables. Anyway absolutely amazing to me that we have come so far and yet not far at all.
Also, if you are wondering if Grayson’s comments about how the British would just build a telegraph line into Shanghai without permission was just me casting unnecessary aspersions on the British, sorry. These are necessary aspersions. This is actually how the Shanghai telegraph line was built in our reality. The British asked, the Chinese said no, they asked again, the Chinese said no, and so they just built it under the cover of darkness and then said, “oh well, what are you going to do about it? Wanna fight?”