All regular EQMM readers know who Steve Steinbock is: He took over the chief reviewing position in our distinguished book-review column The Jury Box in 2011, he became a published fiction writer in EQMM’s Department of First Stories in 2010, he’s done special-feature interviews and articles for the magazine over the years, and (starting in our July/August 2021 issue) he’s become one of our translators from Japanese. The latter is the topic of this fascinating post.  —Janet Hutchings

A few years ago at a Malice Domestic conference in Bethesda, ten of us were gathered around a dinner table. Josh Pachter was a few seats away from me. Since the mid 1980s, Josh has been translating stories, articles, and books from Dutch into English. In recent years, he’s also translated from Flemish, Spanish, Italian, Afrikaans, Romanian, and Chinese.

Someone at the table asked Josh how many languages he spoke.

“Only English and Dutch,” he said.

We all stared at him.

“These days,” Josh continued, “with Google Translate, you don’t need to speak a language to translate it. You just need to know how to adapt a Google translation into coherent English.”

“Josh,” I said. “Maybe you can do that with French or Spanish, but it won’t work with Japanese.”

“In my experience,” Josh said, “it works with every language.”

Josh is my friend and mentor, and he’s the greatest EQMM translator since Anthony Boucher. But with regard to Japanese, relying on Google Translate is problematic.

On a daily basis, I use Google Translate and other machine translation tools like DeepL and Yandex. They make my job as a language learner, and a translator, much easier. If you copy a paragraph of Danish, Dutch, French, or Spanish into Google Translate’s webpage, you will get a perfectly sensible, more than adequate translation of that paragraph into English. Chances are you’ll also have good luck translating a text of Azerbaijani, Malaysian, or Urdu into English.

With Japanese, however, if you rely completely on Google Translate, most of it will sound awkward, half of it will be confusing but ultimately understandable, and about ten to twenty percent will be bafflingly incomprehensible.

Below, for example, is a paragraph from the Japanese Wikipedia page for “Detective Novel” translated into English using Google Translate:

The name “mystery,” trees Takataro is Ondorisha in science fiction broad sense, including the mystery when supervised the Monographs, Ranpo Edogawa and Mizutani quasi reportedly those named have been proposed in. There are also other names such as detective novels, mystery novels, and suspense novels, but the former name is because the character “Detective” is restricted to this kanji. It is no longer used. There are some overlaps with crime novels, but they are not completely synonymous.

(Keep in mind that Google Translate, like nearly all machine translation tools, uses artificial intelligence. Every time a human being uses it and makes a correction, the software “learns” and improves. If you enter the same paragraph tomorrow, your mileage may vary).

The problem is not with Japanese, or with English, or even with Google Translate. The problem is that the syntax, word usage, idioms, and literary conventions of Japanese and English are so different from each other that no piece of software (let alone most human brains) can decipher and convert an intelligent text from one language into an intelligent text in the other. As technology advances, the situation will get better. In the field of computer science, there’s been a fair amount of research into developing algorithms that can parse the grammar of both languages. But at this point it’s still a work in progress.

There are many reasons why Japanese is so difficult to translate. I’ll outline a few of these below. But it’s important to keep in mind that translation between any two languages is never a simple matter of substituting words from one language to corresponding words of the other.

In Lewis Carroll’s Through the Looking Glass, Alice examined the mirror mounted above her fireplace and imagined the world on the other side. She wondered if “looking-glass milk” tasted different, or if objects would operate by different rules. How would a chessboard knight move on a looking-glass chessboard? She managed to climb through to the other side of the mirror where she discovered a world that indeed operated by a different set of rules. When Alice tried to read a book, she realized “it’s all in some language I don’t know.” She held the book up to a mirror so she could read the words—but found it was all “Jabberwocky.”

To translate is to experience the world from both sides of a looking glass. Languages operate differently on the other side.

What Makes Japanese so Difficult?

Just to be clear, my intention isn’t to bash the Japanese language (or English, or Google Translate). But the fact is that both English and Japanese are complicated languages, and it’s those complications that make translation between the two so complex, while at the same time so satisfying. The very things that make Japanese linguistically incompatible with English are what makes it exciting for me. A difficult Japanese paragraph is like a crossword puzzle. I can stare at one clue for hours. But as soon as I solve another clue, everything fits into place.

Below are some of the curious characteristics of Japanese that make English translations so complex.

1. Syntax. There are several aspects of Japanese syntax that make translation challenging. First is the use of particles—single characters or short words that are attached to the end of words, like suffixes in English, and which serve the same functions as English conjunctions, interjections, and prepositions, or serve as markers of case, phrase functions, subject, object, or topic of a sentence. A second characteristic of Japanese is that there are no spaces between words, making it hard to tell where one word ends and the next begins, or whether a particular character is a particle or an actual word. A third challenge is that Japanese word order is more or less the opposite of what it would be in English. Below is an example of a sentence in English followed by the same sentence in Japanese, then a word-for-word translation with the particles included in parentheses:

I want to try on a suit I saw in the shop the across the street from the hotel.

ホテルの向かいにある店で見たスーツを着てみたいです

Hotel (of) opposite (to) store (at) saw suit (direct object) try-on want.

Not a very pretty sentence in English, but it makes perfect sense in Japanese. I found a similar, simpler example in the DuoLingo language app. One of the quiz questions was to translate into Japanese:

                                    “I put flowers in the vase.”

The answer, translated word for word with grammatical particles in parentheses, is

                                    “Vase (in) flower (object marker) put in (past tense ending)”

You might have noticed something missing from the Japanese sentences. Nowhere do they contain the first-person pronoun (“I”). Which brings us to the next big difference . . . 

2. Dropped subjects and missing pronouns. Japanese grammar is often very economical. If anything can be understood from the context, there’s no reason to say it. This is especially true with first- and second-person pronouns, as well as subjects that were named earlier in a text. Regarding pronouns, it’s generally considered rude, abrupt, or arrogant to use “me” or “you” (or any other first- or second-person pronoun) in speech or writing.

To illustrate this, below is a paragraph I came across in a discussion on Quora:

子供が小さいときに喘息気味だったので、スイミングスクールに通わせた。最初は水に顔をつけるだけで精一杯で、レッスンで言われたことが一人だけ出来なくて、泣きながら帰ってくる日もあった。でも、本人が辞めると言うことはなかった。高校生になった今でも続いている。石の上にも三年というが、今年、ついに全国大会に出場することになった。

Here’s what I got when I ran the above paragraph through Google Translate:

I had asthma when my child was small, so I sent him to a swimming school. At first she was all about putting her face in the water, she couldn’t do just one thing in the lesson, and she came home crying. But she never said she would quit. She is still in high school. He has been on the stone for three years, but this year he will finally be competing in the national competition.

An astute reader can figure out the writer’s meaning, but it’s confusing. The Japanese paragraph doesn’t contain a single pronoun, so Google randomly inserted pronouns (usually incorrectly) wherever English grammar required one. Regarding the remark “on the stone for three years,” it’s an ancient idiom about patience. The meaning is that when the writer’s child (we’re never sure from the context whether s/he is male or female) was young, that child had asthma, and didn’t like swimming. But the parent was patient, and the child eventually became a competitive swimmer in high school.

3. Words with different ranges, nuances, and definitions. Translations are approximations, at best, since very few words in any language have exactly the same meaning as the corresponding word in another language. You may have heard of the blue/green distinction in Japanese traffic lights. Throughout the world, “red” means “stop” and “green” means “go.” But for a variety of reasons—mostly having to do with history of color words and where the Japanese draw the line on a color map between blue and green, a Japanese traffic light signifying “go” is referred to as “blue.” In English, we use “blue” idiomatically to mean “sad” (“I’m feeling blue”). But in Japan, referring to someone or something as aoi (blue) implies that it is young, unripe, or unskilled. Infants and people who have had too much to drink are referred to as “red.”

Japanese uses a lot of English words, but in Japanese they take on new meanings. Manshon (from “mansion”) refers to any high-rise apartment building. Eyakon (from “air conditioner”) refers to heat pumps, whereas an air conditioner is called a Kūrā (from “cooler”). A smorgasbord or all-you-can-eat buffet is called a Baikingu (from “Viking”). In English, we draw a distinction between comedy and drama, but in Japan, Dorama (from “drama”) is any TV series, be it a tear-jerker or a situation comedy. An electric outlet in Japan is called a Konsento (from “consent”) and I have no idea why.

Since I’m in the book business, literary terms are important to me. But divergent definitions had led to confusion more than once. In English, we make a clear distinction between novels and short stories. But if you translate “novel” into Japan, you get shōsetsu (小説), which in Japanese can mean both novel and short story. On the other hand, translate “story” into Japanese and you’re likely to get either monogatari (物語), which usually refers to a classical story or an epic, or hanashi (話), which is a tale told in spoken conversation.

In English, “maybe” and “probably” have different meanings. No one has ever quantified it, but most English speakers would say that “maybe” means anywhere from 0% to 60% likelihood, whereas “probably” describes something with a likelihood of 80% or greater. The Japanese word tabun (たぶん), however, is translated as both “maybe” and “probably.” A similar issue occurs with the word “omoshiroi” (面白い – which literally means “white faced”). This word describes something as being “amusing” or “comical” as well as “fascinating” or “intriguing.”

Everything is relative, especially when it comes to relatives. I recently read a story in which the crime victim’s only living relative was his twin brother. So I was confused when the police went to interview the victim’s little brother. I thought at first that I was missing something. I must have either gotten “only living relative” or “twin” wrong. I finally remembered that Japanese uses specific words for siblings based on relative age: a male sibling is either an older brother (ani) or a younger brother (otōto), even in the case of twins.

Japanese adds the prefix “gi-” (義) to family words signifying that the relationship is legal rather than biological. This leads to confusion for Western readers since, for example, gibo (義母) can mean both “stepmother” and “mother-in-law.”

Family terms aren’t limited to family members. I pointed out earlier that using a second person pronoun (meaning “you”) is disrespectful. One polite workaround when speaking to someone who is older than oneself – even when they are complete strangers – is to address them as “older brother,” “older sister,” “uncle,” “auntie,” or if they are significantly older, “grandfather” or “grandmother.” A male restaurant, bar, or shop owner is often referred to as a danna (旦那), which also means “husband.”

A daughter isn’t always a daughter. In Awasaka Tsumao’s “Fox’s Wedding” (EQMM July/August 2021), the victim of a traffic accident was, according to Google Translate, “a Judo instructor’s young daughter.” Upon careful reading, I figured out that the victim was actually “a young female Judo instructor.”

The word for male offspring is even more problematic: musuko (息子) literally means “son,” but it’s often used euphemistically for “penis.”

4. Name-games. Japanese names are descriptive. The characters that make up a person’s name have specific meaning. Remember the “Judo instructor’s young daughter” who was the victim of a traffic incident? (Incidentally, she survived). The cause of the accident was a drunk red bear. At least, that’s what I learned from Google Translate. It turned out that the drunk driver was very human but was named Higuma (緋熊) which translates as “scarlet bear.” This happens often with Japanese names. When an artificial intelligence like Google Translate sees a common name like Tanaka (literally “amidst the field”) or Yamaguchi (literally “mouth of the mountain”), it has no problem recognizing it as a proper name. But less common names are tricky, and are often translated literally. The hero of a mystery series by Arisugawa Arisu is named Himura, which Google usually translates as “fire village.” The author of “Fox’s Wedding” is Awasaka Tsumao, which Google translates as “bubble slope wife husband.”

5. Idiomatic Idiosyncrasies. Idioms almost never translate from one language to another. This doesn’t just go for Japanese and English, but for all languages. In fact, idioms are often so mired in history and misinterpretation that most of us would be hard pressed if we had to explain them. I understand things like “walking on eggshells” or “throw a wrench in the works” but what does it really mean to “beat around the bush?” And why do we say “pardon my French” when nearly all English curse words have Anglo-Saxon and not French origins?

Japanese uses a lot of idioms and idiomatic phrases in everyday speech. There are even some elements of Japanese that are not idiomatic, but just sound so, because they are culturally unique to Japan. For example, to this day, the size of a room or apartment is not measured in square meters or square feet, but in tatami mats. In traditional Japanese buildings, the floor is made up of straw mats measuring approximately 3 by 6 feet. Thus, a hundred-square-foot bedroom would be a “six mat room” and an apartment measuring seven hundred square feet would be referred to in a real estate advertisement as being “forty tatami mats.”

In a mystery story I recently read, I encountered a truly baffling measurement. A lethal dose of cyanide was described as “an earpick’s worth.” It turns out people of East Asian heritage tend to have drier, almost crystal-like earwax (in contrast to the goopy, yellowish wax in the ears of people of African and European ancestry). Rather than cotton swabs, Japanese traditionally clean their ears with a special tool, usually made of metal or bamboo, with a tiny spoon-shaped scoop at the end. In Japanese recipe books, instead of a “dash” of salt, you’re likely to find “an earpick’s amount” (耳掻き一杯).

Many idiomatic phrases were imported from China over a thousand years ago. A majority of these are called yojijukugo (four-word idioms) because they are each comprised of four Chinese characters. For example, for “everything is going well” you might say jun-pū-man-pan (“gentle winds full sail”). An over-the-top billboard or TV advertisement is referred to as yō-tō-ku-niku (“sheep’s head dog meat”).

If you’re getting impatient with my long-winded explanation, you might ask, soko ga miso nandaro? (roughly meaning, “what does that have to do with miso?”)

Japanese employs hundreds of mimetic words in everyday speech. These are slang words, usually a doubled sound, that describe noises, feelings, textures, conditions, and anything else you can think of. “Waku-waku” means “excited.” “Niko-niko” is a bright smile. “Jiro-jiro” is the act of gaping. If you ever find yourself translating a hot love scene, don’t be surprised if you encounter something like:

He held her nyurun-nyurun pai-pai. The munyun munyun body was all pika-pika and kira-kira with beto-beto sweat. She said, “Uncle, your chin-chin is so bikun-bikun.” It really was jin-jin. His heart was doki-doki and his head was guru-guru. And then . . . dopyo-dopyo. Afterwards he was boro-boro, and when she heard him goo-goo, she knew he was suya-suya asleep.

I won’t translate that. But I trust that your imaginations can get the idea.