Translation: Who is the Target Audience?

Despite growing up in a multilingual country where code switching is part of daily routine, the first time I deliberated about translation was when I encountered Japanese anime/manga franchises like Ranma 1/2 and Dragonball Z, and wanted translations in English. Unfortunately back in the late 1980s, English translations weren’t as prevalent, and I had to make do with what was available, whether they were Chinese Mandarin translations or the original Japanese — languages which I wasn’t familiar with at the time.

Even when English translations were available, an evergreen topic was subtitles vs. dubbing (subs vs. dubs) debate: which one was better? Watching anime with subtitles and comparing them to the English dub, there were noticeable differences. Dubs for example attempted to match the lips and cadence of the characters on-screen, while subtitles had more leeway to work with — sometimes even showing translator notes. Or just listening to the localized theme song of Dragonball Z, I immediately knew the lyrics were different. But I was amazed at how the localization managed to match the rhythm of the original and sounded “natural”, even if it was not necessarily identical.

When DVDs were announced, critics proclaimed the subs. vs. dubs debate to be over: consumers could obtain both with a single purchase and just choose which one they preferred. However, this didn’t really end the debate, and actually overlooks an important aspect of translation: a piece of media will have different translations, regardless if it is subtitled or dubbed. Sometimes those differences are subtle but other times, they can be drastic and significantly change how they are interpreted.

People will frame the argument as one is “better” than the other, but what they really mean can be divided into two components.

The first is they equate “better” with accuracy.

The second is associating their personal preference as a universal criteria.

Accuracy Has Different Meanings

Let me tackle the first point.

Some people frame or expect accuracy as an objective metric, that translation has a single, definite, and universal end result.

Unfortunately nothing can be further from the truth.

When people complain that a translation is “inaccurate” or “not faithful” to the source, they are usually complaining about something else. And there are several possible reasons for this, which I will break down below. Accuracy has becomes a catch-all term for things someone likes or dislikes about a translation, but it is an inarticulate way of expressing it. Instead, it’s used as a way to end the debate, as if invoking accuracy is a way to convince people to their point of view, when the reality is people have different — albeit unspoken — expectations.

Their personal preference when it comes to translation is a subjective metric.

Translation is Destruction

When non-translators talk about accuracy, what they usually want is for the text to be conveyed to their brain — a form of telepathic communication. Unfortunately this is not possible, in much the same way that even when two people are speaking the same language, misunderstandings can happen.

One prevailing paradigm when it comes to translation is that translation is destruction. This is best encapsulated in the Italian proverb:

“Traduttori? Traditori!”

(“Translator? Traitor!”)

The dilemma stems from two crucial points.

The first is that languages are different, period. English is different from French and Arabic and Chinese, etc. It could be due to the syntax. It could be due to the grammar rules. It could be the writing system or how certain vowels or consonants are pronounced.

To claim that a perfect translation is possible is to deny the differences between languages.

The second reason is cultural. Language is part of a larger identity and zeitgeist, and while it’s tempting to look at individual words as modular components, the reality is that it is merely a reflection of something massive and encompassing.

Here’s an example:

For those unfamiliar with Chinese families, they can look at this chart for the terms used for relatives.

Translators can translate 弟妺 as “Younger Sister in Law” (or even just “Sister” or “Sister in Law” and leave it at that), but sometimes, it’s also necessary to convey the kind of culture that needs to distinguish “Younger Sister in Law” vs. “Elder Sister in Law” such as what the power dynamics are in play, what the gifting or inheritance rules are, and what is expected of them in that specific situation. The challenge isn’t just choosing the right words or looking for synonyms, but how to convey entire worlds, cultures, and paradigms (and sometimes, it simply can’t be done).

When both aspects are combined — namely the differences between languages and cultural contexts — then we need to accept that there is no such thing as a “perfect” translation. Something will be left out, which is why the advice to people who want a faithful translation is that they need to learn the language and the culture of the original language, and not rely on third parties to do the translation for them.

Translation, by its nature, is acting like a mediator where the translator needs to make an editorial decision regarding which specific aspect they want to prioritize. In that sense, when they leave something out, they end up creating something new or different in its place.

For some context, take this quote from Jay Rubin:

“When you read Haruki Murakami, you’re reading me, at least ninety-five per cent of the time.”

The Unspoken Hierarchy

Usually, when there’s a complaint about a bad translation, there’s usually a dissonance between the translator and the intended audience (or sometimes, the recipient isn’t the intended audience in the first place).

Because translation is a form of editorializing, translators need to prioritize specific values over others.

For example, when I enter the German phrase “Geh mir nicht auf den Keks!” into Google Translate, the software made a judgment call to prioritize the intended meaning since the statement is frequently used as an idiom.

If I ask a German friend to translate, they might ask me whether I meant “Don’t go for the biscuit!” or “Stop bugging me!”, the latter of which makes more sense for non-German speakers (i.e. most people will usually want what the idiom means, rather than a transliteration of the text).

But when we consume translations, they exist in a fixed space, such as text on a screen or text printed on paper. We don’t have the luxury of asking the translator “this is actually what I meant when I asked for this”, hence translators need to make that decision for us. They have an intended audience in mind, one with a specific query that need to be addressed.

It’s easy for someone to say “faithfulness to the source material” should be the priority but is that always for the best, and what does faithfulness to the source material really mean? Did they mean authorial intent? Or how the text became commonly interpreted? And what about certain practicalities, such as adhering to Intellectual Property laws?

Below are examples of translations intended for a US audience but with different priorities.

No Translation

There are times when a translator chooses not to translate something.

There are multiple reasons for this. It could be because there’s no appropriate term in the target language, so it’s one way of telling readers “do your own research”.

At other times, it’s because certain words or terms have seeped into a culture’s zeitgeist that a direct translation is not necessary (and this is how loan words are formed).

Take for example the manga series 怪獣8号. Currently, it has two official English translations: one for Viz Media in the US, and another for Manga Plus “internationally”. It’s translated as Kaiju No. 8 in the former and Monster #8 in the latter.

Now Monster #8 is not an inaccurate translation, but for US fans of the manga, a substantial number of fans claim that they prefer Kaiju No. 8 as it has more “impact”.

I don’t know when the term kaiju became an acceptable term in the English language (perhaps due to the popularity of Pacific Rim) but this is a situation where the translator, David Evelyn, made a judgment call to retain the original Japanese term.

Does it work? I’m not sure if the title would have worked for American audiences back in the 1980s, but it’s a great decision for a post-Pacific Rim America and makes for a more distinct title.

Localization and Theme

In the point above, what was prioritized is what we could call “taking the reader to the source,” where it is up to the reader to figure out what the original author meant.

Its opposite stance is adapting the source material to the target audience (whether on a cultural level and/or on a linguistic level) so that reading the translation is seamless and feels as if it was written by a native speaker.

In this specific aspect, there are many different ways for the translator to approach this mindset and find creative solutions.

Take for example the phrase 僕の勝ちゃ, which David Evelyn laments about:

“…but my numero uno pet peeve that makes me want to pull my teeth out is when 私・僕・俺の勝ち is TL’d as “It’s my win.” No one in the history of existence has ever rose from their seat and screamed “IT’S MY WIN!” lol.”

Evelyn admits that there’s several ways to translate it and his approach was to present it as “I gotcha beat,” in Kaiju No. 8.

The goal here was to present the dialogue as sounding natural to US audiences and a transliteration of the text, while comprehensible, could sound stilted or awkward.

There are other embellishments that can be applied, depending on the context and intent. For example, in Kamen Rider OOO, there is a mentor-mentee relationship between the characters Akira Date and Shintaro Goto. The former refers to the latter as Goto-chan (the use of the honorifics is explained here) when normally -kun would be more appropriate, so fansub group Over Time chose to translate this as “Broto” (a portmanteau of bro and Goto).

When I first encountered this, it personally irked me, but over the years, I came to understand the rationale and philosophy behind it. I would still have preferred Goto to be used, but Broto is one way way of relaying the tone and affection the character has. And it would only work in a US context, where bro is commonly used as slang for close friendships (as opposed to a blood relation).

In addition to the scenarios above, there’s also the concept of adding theme and personality to a translation. For example, in games like Chrono Trigger and Fire Emblem, characters like Frog and Lyn respectively have old English accents, even if they weren’t present in the original. (Legends of Localization elaborates on them here and here.)

Now this can be a polarizing subject. Some fans enjoyed the additional color and speech style added to these characters. On the other hand, there might also be purists who view it as the translator “adding” something to the translation and not being faithful to the original. But that goes back to the original theory of translation as destruction and that something new is born.

It also begs the question if the source material did have accents, how should they be adapted to US audiences? (And of course, if you’re not an English speaker, if characters have a certain US regional accent like a Boston accent or a Southern accent, how should they be translated in other languages?)

Authorial Intent

Some people might use “authorial intent” on gauging whether a translation is competent or not, but that presumes that authorial intent is objectively knowable (or that the person reading the translation is aware of it).

If the author is literally dead for example, we cannot objectively determine what the authorial intent was, although scholars might present their own theories.

In practice, the author might also be unavailable or not know the answer.

For example, in the game Chrono Trigger, one of the character’s parents fall into a machine which eventually cripples them unless the player enters the password (on the controller) named after the parent. The parent’s name is ララ which is usually romanized as “Rara” but what’s important to note about Japanese is that there’s no genuine “R” sound and instead what we get is something in between “R” and “L” and the same character is used interchangeably. In the series Slayers for example, the protagonist is known as Lina Inverse in some texts, while Rina Inverse in others.

Going back to Chrono Trigger, there are 4 possible permutations for ララ:

• Lala
• Lara
• Rala
• Rara

The English localizers chose to go with Lara, probably because that sounds like the most reasonable English-sounding name.

To the Japanese writers who wrote Chrono Trigger, who knows if that’s what they truly intended or if they even conceived the potential problem for English audiences. (In the Japanese version, the password is Lala.) The character to them remains ララ, regardless of how it’s spelled in English. (More on the subject above is discussed in Legends of Localization.)

At other times though, translators will have access to the author. And their recommended solution could be to not do a transliteration and change the actual words used.

For example, in the film Parasite by director Bong Joon-ho, Darcy Paquet (working in collaboration with Bong) translates “Seoul National University (SNU)” as “Oxford” and “jiapaguri” as “ramdom”. He explains his rationale for doing so:

“The first time I did the translation, I did write out SNU but we ultimately decided to change it because it’s a very funny line, and in order for humor to work, people need to understand it immediately. With an unfamiliar word, the humor is lost.”

So in this case, keeping the spirit of what’s said was the higher priority, and this intersected with the director’s intention.

Does that mean Paquet’s translation is the best? Well, it’s the most apt for those who want some of the Korean cultural elements presented with American analogues. Those looking for a translation that keeps all the Korean references intact, or perhaps an English-speaker who is unfamiliar with American references, might find a different translation more appealing.

Metaphors, References, and Puns

One of the more challenging aspects of translation is when you come across sections where you don’t have the space to explain all the cultural nuances and metaphors, or when doing so sabotages the intent (in the case of humor for example).

Layne Vandenberg, writing for The Diplomat, has this commentary on Parasite’s cultural specifities:

“Parasite’s plot is catalyzed by the gifting of a scholar’s rock, a token in East Asian society of specific cultural significance. The housekeeper’s rendition of North Korean news broadcasts serves as a humanizing, comedic moment for the character that might be lost on unfamiliar audiences. The Parks’ disdain for those who ride the subway reinforces their upper class position in Korean society considering that nearly 7 million passengers use the Seoul public transportation network per day, out of Seoul’s total population of 9.7 million. The Korean title for Parasite, Gisaengchung (기생충), can be broken down to mean “parasitic” (gisaeng, 기생) and “insect” (chung, 충). This carries through to the first names of the Kim family: each name begins with “Ki-”, which is frequently pronounced “Gi-”, and the wife’s name is “Chung-sook.””

Is it reasonable to expect that the translator convey all of those things in the film via subtitles alone? Probably not. And that’s not to say some translators won’t try. In books for example, there’s more space to include translator notes and footnotes.

Puns and portmanteaus have a similar hurdle. Some translators might just eschew the playfulness altogether and decide to go for a plain translation. Others might feel they are up to the challenge and either replace it with a pun their target audience is more familiar with, or attempt something different.

For example, in Yakuza: Like a Dragon, there is an ability called デリヘル, which summons allies to aid you by calling them. In Japanese, デリヘル is intended to be translated as “Delivery Help”, but it is also a pun on Delivery Health (spelled the same way), which is a form of prostitution in Japan where you make a similar call to have someone provide you with sexual services.

The English localizers eventually went with Poundmates, which doesn’t really explain the pun above, but creates its own sexually-suggestive wordplay as both the words pound and mate work in dual contexts of either having sex or beating someone up.

Matching What the Audience Hears

This is a dilemma more related to audio-visual mediums like film, TV, and games (with audio): how much do translators need to match what the audience hears? Should they aim for accuracy or consistency?

For example, translator Katrina Leonoudakis warns that foreign loan words shouldn’t be translated as is:

“[D]on’t just blindly translate katakana English loan words as-is, especially when working in games! 「サークルヒール」, an AOE healing spell, probably shouldn’t be “Circle Heal”.

As always, I go back to theory!

FUNCTIONALISM: The writer’s using English words to sound cool in 99% of these cases.

POLYSYSTEM THEORY: The source audience has different subcultural expectations (‘fantasy skills in JRPGs use English words’) than target audience.”

This lead into a separate thread on reconciling what the audience hears and translation. Leonoudakis replies with:

“I used to stick to ‘if they hear it they hear it’, but I’ve shifted my approach more towards ‘bad English is bad in subs, too’.”

But context is important too and sometimes the “Bad English” is part of the identity of the translated work, while at other times, it’s better to replace it with something more appropriate.

This dilemma probably exemplifies knowing your audience and regardless of the decision, some people will always be discontent due to having different personal hierarchies. But it’s important to make a conscious decision here on which philosophy to approach it.

Restrictions, IP Laws, and Censorship

There will also be times when the translator’s hands are tied and must comply with certain requirements. This is less about capability and more of restrictions placed upon them by another party, either their client, publisher, or governments.

For example, games published in the 1990s by Nintendo of America had to conform to specific content policies (and remember this was the time when the US government was looking to regulate video games).

Various characters or titles could also infringe on various Intellectual Property (IP) in various countries, so they need to be changed or altered.

For example, one of the characters in One Piece is ロロノア・ゾロ (“Roronoa Zorro”) but presumably to avoid legal battles with the name Zorro, the Viz English translation instead uses Roronoa Zolo. A similar translation dilemma happens with ルパン三世 (Lupin III) where the titular character is named after Maurice Leblanc’s Arsène Lupin so to avoid similar legal hurdles, some earlier translations referred to him as Wolf (invoking the name’s relation to the word lupine).

And in extreme examples, there might also be elements that might be censored or banned in a country, so the choice is either to alter how the specific text is interpreted, or not have the work translated at all (and there are those who believe that “no translation” is better than a “bad translation”).

Hierarchies

The problem with the concept of a “perfect translation” is that this assumes everyone shares the same values, culture, and opinions. But if that were the case, then there would be no need for translators, localizers, and cultural consultants.

Instead, translations should be viewed like any media text: they come with their own biases and prioritizes specific values. Ideally this editorializing is conscious but sometimes it can be unconscious as well.

Similarly, just because a translation doesn’t work for one person doesn’t mean it wouldn’t work for anyone else. Or that once a work has been translated in a specific language, that there’s no need for other translations; the differences between translations should be embraced, analyzed, and critiqued.

And that’s not to say there aren’t bad translations out there. Breath of Fire II is an example, but even then, there are some things the translator gets right. And some inaccurate dialogue like Castlevania: Symphony of the Night’s “What is a man? A miserable little pile of secrets.” have become iconic.

At the heart of every translation is a specific target audience. Are my interests aligned with the translator? Or if not, is there a way for me to empathize with their vision?