Readability Equals Translatability
Conveying an accurate message to the widest-possible audience
I’ve sometimes wondered if readability and translatability are related qualities. If so, what could they teach me about each other? Among other things, I hoped to solve the problem of writing in a way that’s readable for a wide audience, but still artful, intelligent, and enjoyable.
When I say a thing is translatable, I mean it can be easily ported to a different language without loss or change of meaning. Some readers remember the Chevrolet Nova, the car with a name that roughly translates to won’t go in Spanish. If your article or story might someday be machine-translated for a wider audience, will it be accurate and easy to read in other languages? Will you have to “dumb it down” to achieve that?
In this short article, part of my series on readability, I’ll talk with a translator, compare the work of some famous authors to see what translates well, then finish up on the question of machine translation versus art.
Is Translatability the Same as Readability?
In January 2020 I asked award-winning literary translator Margaret Sayers Peden about this subject. Over a period of 33 years Peden translated more than 30 works including novels, plays, and poetry by South American authors that included Isabel Allende, Octavio Paz, and Pablo Neruda.
She liked the question. “Yes. I think that if something is easy to read, it tends to be easier to translate. And if it’s unreadable in Spanish, it’s going to be unreadable in English as well.”
What makes something hard to translate? One challenge we discussed is text that relies on connotations to evoke subtle meaning. A word as simple as apple has different associations for an American (“Mom and apple pie”) than for a Mexican (the Spanish equivalent manzana can mean a plaza or market). A meaning that relies on subtleties like this may not survive translation. So in dealing with artful or evocative wording, Peden often asked acquaintances for suggestions. For instance, for Carlos Fuentes’s play Orchids in the Moonlight, she was aware that the association of orchid with male anatomy was clear to Spanish readers, but almost unknown to English readers. Sadly, none of her consultants was able to suggest a better English equivalent, so most English readers correctly see orchid as a type of flower but miss the other meaning.
Peden also mentioned the problem of translating writing that’s ambiguous due to loose metaphor or phrasing. This sometimes forced her to write sentences that she knew would be difficult for readers. When possible, she said, “I would check with the author to see if I was way off base.” But in cases where the authors were deceased, she could never be sure.
Clichés and cultural content can also be difficult to convert. Even with the help of native speakers there’s sometimes no good solution. The expression “like father, like son” translates easily because it makes obvious sense, and a similar expression may already exist in the target language. For instance, there’s an equivalent expression in Chinese. But imagine translating a casual use of “he wants his pound of flesh”, an expression we get from Shakespeare’s Merchant of Venice. Or consider the conversation in Chapter 6 of Somerset Maugham’s The Razor’s Edge. It begins with “Gray’s conversation was composed of clichés”, then cleverly uses more than a baker’s dozen. If a text relies on cultural knowledge for its meaning, some of that meaning is likely to be lost.
I should also mention the 1992 version of Aladdin, which delighted audiences with the rapid-fire ad libs of the genie character, played by Robin Williams. In that case, not only was much of the humor probably meaningless (or completely rewritten) in foreign-language versions, but thirty years later, it’s also lost to many native English speakers. Some critics loved Williams’s “hip”, “blistering improv”, but a pre-teen reviewer who appeared on NBC’s The Tonight Show voiced a reservation. Many of Williams’s celebrity impressions and other cultural in-jokes were opaque to him because of his youth, and would eventually be meaningless to most viewers.
And finally, Dr. Peden also mentioned the physical sound of the words. Some writers use it to help create images or moods — with consonants that hiss or buzz or click, with bright or dark vowels, with light or heavy words, etc. Along these lines, humorists and psychologists have claimed that some sounds are funnier than others, including the k sound and even certain categories of nonsense words. According to Peden, if sound is an important feature of the text, the translator may try to translate into words that create the same effects. Again, this might involve asking native speakers of the target language for suggestions. But sometimes sound effects, like word associations, don’t survive the process.
Most readers have seen Lewis Carroll’s poem Jabberwocky, which relies on sound, associations, and context to suggest meaning in a stream of nonsense words. For strong English speakers, it works surprisingly well. Wikipedia and the Oxford English dictionary define some of these words. Google Translate translates some of them. And the Netflix choose-your-own-ending movie Bandersnatch is named after one of them.
Because the poem is part of the popular novel Through the Looking Glass, its nonsense words have been translated into many languages. For instance, here’s the opening line:
’Twas brillig, and the slithy toves / Did gyre and gimble in the wabe
In one French translation, it reads:
Il brilgue: les tôves lubricilleux / Se gyrent en vrillant dans le guave
In the novel, Lewis Carroll suggests that slithy means “lithe and slimy”, and some of that seems to survive in the French nonsense word lubricilleux.
The writer James Joyce could be extremely difficult to translate. His artful use of English relied heavily on sounds and deep word associations, similar in some ways to Jabberwocky. And his final novel, Finnegan’s Wake, has a rambling, dreamlike, free-associative flow that’s hard to follow. It makes heavy reading for even an English professor, and it’s nearly impossible for an ESL reader, dyslexic, or translator. A translation into plain English was a vast labor, came out phenomenally long, and obviously lacks the evocative quality Joyce worked so hard to achieve. A Chinese translator of Joyce’s work, standing at his grave in 1949, said “Here lies the corpse of someone who wasted his great talents writing something very unreadable.”
In some cases Joyce worked his magic with ordinary words, as in this passage that relies on alliteration: They lived and laughed and loved and left. In other cases, he played at the blurry edges of languages and ideas: Northmen’s thing made southfolk’s place but howmulty plurators made eachone in per-son? Latin me that, my trinity scholard, out of eure sanscreed into oure eryan! Hircus Civis Eblanensis! He had buckgoat paps on him, soft ones for orphans. Ho, Lord! Twins of his bosom.
It’s easier to translate the Russian writer Tolstoy. He’s considered a good stylist who uses language skillfully, but there are good English translations of his Russian stories and novels that keep much of their original power. How does this work?
The opening line of Tolstoy’s novel Anna Karenina is considered one of the best opening lines in all of world literature. It’s consistently translated to English in more-or-less these exact words: “All happy families are alike; each unhappy family is unhappy in its own way.” The meaning is very rich, but it’s constructed with common words that a first-year language student could read. The parallel grammatical construction helps convey the meaning, but the sentence doesn’t depend on that construction, or a particular sound or association, to be meaningful.
It’s similar to another famous opening line, from Jane Austen’s Pride and Prejudice: “It is a truth universally acknowledged, that a single man in possession of a good fortune must be in want of a wife.” A simple English translation might be: “It’s a well-known fact that every rich unmarried man needs a wife.” My inept translation loses Austin’s authoritative primness, but retains the meaning and the humor.
What’s harder to translate are evocative words with hidden associations, cultural references that are unclear to outsiders, or metaphors that may fail in different wording. If meaning is bound to a specific language and culture, it will be hard to translate.
Is Art Translatable?
In this long article I’ve been looking for the principle in which a thing that’s challenging to translate may also be challenging to read, at least for some people. There’s a trade-off. When a meaning is closely bound to language and culture, it may only be clear to native speakers with a certain level of education, from a certain area and era. It’s an inside joke. You had to be there. It loses something in translation.
Does that mean we can’t have artful language, connotations, unusual words, and so on? To be readable and translatable, must we abandon style? I’m still musing, but I have the beginnings of an answer.
1. Part of the answer is the author’s way of building meaning — how much semantic ambiguity they employ, and how closely their work is bound to a particular language or culture. In the examples above, Tolstoy and Austen build powerful ideas independent of specific words. They use skillful sentence structure, vocabulary, and voice to present their idea as impactfully as possible. But they don’t bet the farm on the reader’s ability to understand an obscure expression. They build strong ideas with clear language. For instance, the South American fiction writer Jorge Luis Borges made up fantastical facts, then dressed them in dry academic style to sound real. But because he was inventing facts, not language like James Joyce, it translates just fine. We understand what Borges meant.
2. It’s clear that some written genres and styles depend on the difficult-to-translate qualities of ambiguity and heavy bindings. This includes much of poetry and literature, plus a lot of religious or contemplative work, and certainly some memoirs and other creative nonfiction. They’re designed to be pondered or savored, to reward later readings with more undiscovered meaning. These will always be a challenge for translators and outsiders, and in some cases, may lose so much in translation that the effort is futile.
Those two ideas cover most of the point of this article. It would be fun to get into why some humor translates well, or why one language might be better than another for capturing a particular thought or mood. Maybe another time.
But for now, having come this far, there’s just one more topic worth quickly covering, in which all these ideas come together.
Machine Versus Style
In the 2020s, public-facing content such as instructions, PR material, white papers, and pamphlets is often written by content strategists who have writing ability and topic knowledge, plus skills such as content management, library science, taxonomy, search/discovery… and translation. In a shrinking world, content is often translated to increase global reach. My content-strategist friends tell horror stories about translation pitfalls such as not getting the point across, getting the wrong point across, or offending people. There’s also the question of translation delays and expense.
Here’s a common approach: First, use a content-checker such as Acrolinx to enforce your desired style, highlight dangerous words and phrases, and suggest a preferred phrasing. Then use translation software to generate a draft in the target language. Finally (and this may take a few iterations), have human translators compare the translation with the original, making improvements as needed. And voila.
Letting a machine do the first draft lets human translators focus on the hard stuff — nuances, key ideas, voicing, etc. And it can theoretically help create a consistent voice across multiple languages. This machine-human approach produces polished, accurate translations faster, in more languages, at lower cost. That gives you more reach and more agility.
This is a different proposition than Margaret Peden’s much slower, more artful, fully-human process of translating difficult literary treasures.
So depending on 1) how your writing style constructs meaning, and 2) the genres that you work in, machine-aided translation may be the best way to convey a consistent message to a worldwide audience. But your success will depend on translatable writing, which turns out to be readable writing that follows the principles described throughout this series of readability articles — along with careful budgeting of idiom, artful language, and deep cultural references.
Articles in This Series
- A Cognitive Model of Reading - The science of how people read
- Surface Readability - The value and limits of simple “short words & sentences” approaches to readability, and some interesting variations
- A Deeper Readability - Techniques that go beyond surface-based approaches, based on cognitive science and other sources
- Writing for Challenged Readers - About ESL & dyslexic readers, and what they said helps them most
- Readability Equals Translatability - How the right approach to readability becomes a scalable approach to fast, consistent translation across multiple languages, how that works in a modular, single-source content management system, and whether language must be “dumbed-down” to achieve readability