“A Great Ox Stands on My Tongue”: The Pitfalls of Latin Translation | Hacker News Reader

“A Great Ox Stands on My Tongue”: The Pitfalls of Latin Translation

antigonejournal.com

78 pointsbarrabas62 comments

We have a fun way to describe mechanical, word for word translations in Spanish. It's called _from lost to the river_, that's the direct translation of _de perdidos al rio_, which is a saying meaning that once you have embarked onto a silly/dangerous/nonsense business you might just as well go through with it. A sort of reverse sunken cost fallacy. Of course, the fun is in that it makes no sense once translated. I find it amusing to sometimes make use of such expressions in public.

In Turkish, we call it 'chicken translation' or 'tavuk çevirme.' While that normally means rotisserie chicken, a few places badly translated it, and the name stuck for word-for-word.

I still think sometomes about how Döner got its name. I can just picture a German guy walking past a plate of the thin shredded meat, and then asking someone "what is that?" followed by the customer pointing at the rotating meat stack, and the chef shrugging and saying "yeah, it spins."

That is quite literally the etymology of the vasistas window (a.k.a. transom window): https://en.wiktionary.org/wiki/vasistas

> “direct translation of _de perdidos al rio_, which is a saying meaning that once you have embarked onto a silly/dangerous/nonsense business you might just as well go through with it”

Essentially the same saying with the river metaphor exists in Finnish (ojasta allikkoon) but its implied meaning is the opposite of what you’re describing. There’s no suggestion of “might as well”, it’s just: “you were in trouble already but now you’ve really blown it.”

I suppose that fits the national stereotypes of adventurous Spaniards vs. cautious and pessimistic Finns.

It sounds like the English analogy might be "out of the frying pan and into the fire".

It is indeed the direct English equivalent.

I was slightly unfair to my fellow nationals. It's not that you just go through with something silly just because, but rather that uneasy feeling one has when reaching the point of no return on some venture and you have to pick between backing out or pushing through with the original plan. When the latter wins you cheer up and scream ¡de perdidos al rio!

"Well, we came this far" maybe?

English has “might as well be hanged for a sheep as a lamb” I suppose?

Which reminds me of another source of translation errors. Once in a village bar, I saw ensaladilla de cangrejo translated as crap salad.

Translation errors in small-town restaurants across Portugal, Spain, and Italy is a bread and butter of my group chat's memes.

I wonder if the origin is related to 'crossing the Rubicon', which has a similar meaning (although perhaps less prescriptive).

Funnily, and ironically, enough, as a native French speaker, I would never translate "ce ne sont pas mes oignons" (litt. "those are not my onions") as "it's not my cup of tea". To me, it actually means "it's none of my business".

English to English has some problems too.

“We’re not here to fuck spiders.”

This is someone who knows their onions!

Yup I have the same meaning as you. (I'm in north half of France in case someone says they don't)

I’m glad you said that because I thought I knew that ‘mêle-toi de tes oignons’ meant ´mind your own business!’ and I was doubting myself after reading that.

There are many queer things in (past) translations, it is funny when they stick.

In Italian we have the expression "la bellezza dell'asino" (which is "the beauty of the donkey" and it means - nonsensically - the kind of beauty/freshness that only young people can have and that will vanish with the passing of time).

It derives from the combination of a typo with a "faithful" translation from French, the original “la beauté de l’âge” had been printed as “la beauté de l’âne” and translated directly to "la bellezza dell'asino".

Never heard that expression before, love it.

"Not my circus, not my monkeys" is my favorite English idiom of this meaning.

> even though he seems tacitly to accept its inferiority to Greek literature

I would have said "seems to tacitly accept" instead of "seems tacitly to accept"

I wonder if the author purposely avoids the split infinitive. I've heard Latin enthusiasts like to prescribe that onto English, because it's not grammatical in Latin.

But maybe that's just the author's English. I don't think it's wrong.

You know its interesting that Greek uses adverbs with infinitives so often its almost impossible to avoid using the split-infinite when translating.

Yes, I find it very plausible that a spud writing about Latin literature observes that rule.

A "spud"?

That manages to seem both incomprehensible and very offensive at once, which an impressive trick. And not impressive in a good way.

I am not really sure what your issue is. You don’t know what it means but think it sounds bad? So I’m a bad person for using it?

What does "a spud" mean to you in this context?

A stodgy and self-serious person.

Huh. It just means a potato to me (Brit).

I suppose, but OTOH, this is not at all a serious translation. It's at heart an extended joke.

It is not only ungrammatical but impossible to split a Latin infinitive. Latin infinitives are a single word.

Point of pedantry: some Latin infinitives are single words -- present active, present passive, and perfect active. Perfect passive, future active and future passive require two.

For what it's worth, he also uses 'whom' at the start of a sentence, correctly according to formal grammar but where most English speakers would intuitively use 'who'. That level of fussing over nominative/accusative can be the give-away sign of a classical grammar stickler, though equally it can just indicate a native speaker of a cased language such as German.

> For over a thousand years, nobody has been brought up to speak [Latin] naturally at home;

... except Milton.

In my high school (literally a Latin school) you could hear Latin spoken in the corridor, but I'm pretty sure never by any student, unless put on the spot by a teacher! So yes, it was technically a spoken language, but without gainsaying the author's point.

I'm reading a biography on Erasmus, and he lived in England, had pupils there, made friends and visited the archbishop of Canterbury and even the king, not speaking English. He taught his pupils in Latin. English, or Dutch for that matter, was not a proper language.

So Latin wouldn't be really alive, but also not dead. It was still a second language to more than a few as late as the 16th century.

> Translation is simply the art of taking something in one language and expressing it accurately in another.

This article seems to adequate translation with adaptation, which is the crux of the issue I think.

In one case you might want to convey the exact words expressed regardless of the nuances, social context and "making it sound right". Your primary goal is to convey information, however clumsy it can feel as a result.

On the other case, you'll want the reader/listener to feel like a native person is talking to them, adapting to their reference frame, sacrificing accuracy to properly convey the spirit of the words.

I think it comes down to how much entertainment is expected from the result, if it's for research purpose for instance, accuracy will prime.

If, as I think is the main goal of the author, it's "to prevent that person in front of you from subjecting you to a sad, pitying look", fully adapting is surely the way to go.

"translations are like [redacted]; they may be beautiful, or they may be faithful, but unless you're extremely lucky, not both..."

(does anyone have a more 21st century-friendly substitute for the redacted plural noun?)

I'd go with "no-make selfies"

The redacted word here is "women", to save the uninitiated like me some trouble.

One can use "Instagram posts", which are posted for their beauty, but are thought to not be faithful reproduction of the poster's life.

GPTs should excel in this task of finding more substitutes. Maybe try asking it?

"lovers" seems the obvious substitute in this context.

> Your primary goal is to convey information, however clumsy it can feel as a result. > if it's for research purpose for instance, accuracy will prime.

That explains couple weird behaviors I've seen regarding translations. Right, right... Translation is like, you have two distinct interpreter machines, and the input for one, that you want to run on the other. You must ensure the output of both machines to match. "Accuracy" pursued, as you used, is literally a "garbage in" situation. This is what creates so called "engrish" hilarity as well, a lot of them are actually "accurate", yet nonsensical. "Please have distant considerations on photography at this location" is an "accurate" translation for a typical sign in a certain language into English. Make a guess as to what it is supposed to mean.

Translations must result in the same understanding, retaining the effect. That applies to every translations. "Accurate translations" that crash the machine or make it misbehave simply are of no use.

Yes, it's always tricky. I think an alternative point where translators go for accuracy, everything else be damned, is for cultural artifacts and specialized language.

For instance when reading programming books, having english words like "core dump" need to either be translated in the exact matching technical word in the target language (which the reader might actually not understand better than "core dump"...english is my second language and I have no idea what it should be officially called in my mother tongue...), or be kept in english. Going for an approximation that conveys the feeling won't be useful to the reader.

The same sometimes applies to real word conversation translations where the speaker mixes and matches languages (engrish being part of that). Translating the engrish or the foreign words used in the speech (in particular when the foreign words come from the very language we're trying to translate into) will make it more difficult to understand for the reader, especially if they could have understood the nuances of the original script.

I think of this when I see the english "déjà-vu" translated into french: it will often be left untouched with some signage to convey it was in the original script as is, and it would be worse to have it translated for more immersion or fluency.

I had a friend who early in his translator career worked on a series of not very well written YA fantasy novels, and he kept getting into trouble with his editor for trying to improve the writing. He said that the hardest thing to learn was that if the sentence was badly written and confusing in the original it wasn't his place to fix, but to just translate it.

In Polish there is this saying that "faithful translations are not beautiful and beautiful translations are not faithfull". The translator always adds a bit of own style.

Tove Jannson, the author of the Moomin stories, credited her English translator with improving the originals, because after the necessary dialogue about the true meaning of the text she had a better understanding of her own text and wanted that preserved in the Finnish.

She wrote in Swedish though, not Finnish.

When I was young someone decided to publish Isaac Asimov's Science Fiction magazine in my native language, as a series of short books (i.e. not like a magazine at all). As this was a series, the publisher used a relatively small group of translators to translate all the issues.

After reading a number of the books I started to notice that I could easily tell who had done the translation, and it was at that point I started to get an aversion to reading translations. I didn't want to read the translator, I wanted to read the author. So I gradually started to switch to English originals (if the original was written in English). (And that, incidentally, is how I learned English.)

Where it becomes hard is non-English scifi. If you don't understand the language, do you pick the English or the native translation of for example Polish or Russian? Specially when those native translations are likely not too bad, albeit horribly expensive at this point.

Yes, that's a problem. I can read English as easily as my native language, but other languages like Spanish, German or Italian I can only read slowly - and if I read slowly I don't get "into" the story. So yes, I wish I could read other originals. I have read English translations of other languages (which is much easier to find than translations to my native language), and the feeling is that.. it's a translation. Most of what I read is by native English authors though.

What I'd like is a good bidirectional Latin <-> English AI translation model, ideally downloadable (though even an online version with an interface like DeepL - which doesn't do latin - would be better than the current situation of having nothing at all). It probably wouldn't be on par with the best human translators, at least not at the beginning, but that's not a problem.

> For over a thousand years, nobody has been brought up to speak [Latin] naturally at home

Uh, except Montaigne.

> For over a thousand years, nobody has been brought up to speak [Latin] naturally at home

I know it is a very small number of people, and perhaps not Latin as it was spoken in Rome, but I think an argument could be made that Romansh is a direct descendant of Latin still actively spoken today.

Well, Romansh is a Romance language, a Latin vernacular like French and Spanish.

https://en.wikipedia.org/wiki/Romance_languages

Off topic: I'm rebuilding our landing page (targeting CH for the time being) and decided to add Romansh. Long story short, there are 5 standard forms, plus an official made-up version, all of that for 60k speakers. Was interesting to read so much about our 4th official language.

https://en.wikipedia.org/wiki/Romansh_language

Is there some way that it's a more direct descendant of Latin than any other Romance language is?

I'm not a linguist nor a Romansh speaker but my understanding has always been that at its core it is very much still a Latin dialect, thus I'm surprised by all the downvote hate I received.

I wouldn't be surprised if some native speakers, or maybe other Swiss people, said "it's really still Latin -- that's why it's so hard to understand!" or something.

But if so, I think they'd effectively be making up a story to account for Romansh's distinctiveness. I was just looking at some a couple of days ago (as a friend is traveling in Switzerland); I know Latin well, and what I saw was extremely different from Latin, in terms of sound changes, vocabulary changes, and grammatical changes.

Looking at https://en.wikipedia.org/wiki/Romansh_language there are typical Romance evolutions like

* fixed subject-verb-object word order (instead of Latin's variable word order that often prefers subject-object-verb)

* no case inflections (instead of Latin's case inflections)

* plurals apparently derived from Latin accusative plural (not nominative plural)

* compound perfect tense ("have done")

* definite articles (Latin has none)

* many systematic sound changes from Latin (all Romance languages have these, and which specific ones occur tends to define the sound of a specific Romance language)

And unlike other Romance languages (or Latin) it has "numerous Germanic loanwords":

https://en.wikipedia.org/wiki/Romansh_language#Germanic_loan...

My best guess is that the idea of Romansh as distinctively close to Latin is a myth or misconception.

Here's "The Fox and the Crow" in a poetic Latin version and a prose Romansh version (the latter from Wikipedia):

Cum de fenestra corvus raptum caseum Comesse vellet, celsa residens arbore, Vulpes hunc vidit, deinde sic coepit loqui: "O qui tuarum, corve, pennarum est nitor! Quantum decoris corpore et vultu geris! Si vocem haberes, nulla prior ales foret."

La vulp era puspè ina giada fomentada. Qua ha ella vis sin in pign in corv che tegneva in toc chaschiel en ses pichel. Quai ma gustass, ha ella pensà, ed ha clamà al corv: «Tge bel che ti es! Sche tes chant è uschè bel sco tia parita, lur es ti il pli bel utschè da tuts».

These aren't directly parallel in meaning, so they're not easy to compare word-for-word. Apart from pretty big sound changes, I notice three grammatical changes: that compound perfect ("ha ella pensà, ed ha clamà al corv" 'she (has) thought, and she (has) shouted to the crow'), the use of a relative in making an exclamation ("tge bel che ti es" 'how beautiful (that) you are'), and the use of separate words for comparisons and superlatives ("il pli bel utschè" 'the more beautiful bird' → 'the most beautiful bird'). (Latin would say "avis bellissima".)

OK, I translated the English translation of the Romansh story into Latin so that we can directly compare the two.

Vulpes iterum esuriebat. | La vulp era puspè ina giada fomentada.

Ibi corvum in abie vidit, qui fragmen casei rostro tenebat. | Qua ha ella vis sin in pign in corv che tegneva in toc chaschiel en ses pichel.

"Hunc vellem," pensavit, et clamavit ad corvum: | Quai ma gustass, ha ella pensà, ed ha clamà al corv:

"Quam pulcher es! | «Tge bel che ti es!

Si cantus tuus tam bellus quam vultus tuus est, tu bellissima omnium avium es!" | Sche tes chant è uschè bel sco tia parita, lur es ti il pli bel utschè da tuts».

(Some Romansh words whose likely Latin ancestors I didn't use, because it was less idiomatic, or sometimes actually grammatically impossible: era < erat, fomentada < fames, gustass < gustavissem, ella < illa, parita < apparentia, pli < plus, tuts < totos. I probably missed some others by not noticing the etymology.)

True, but also Romanian (and Moldovan), and Ladino...

Living in central Europe for nearly a decade, it was fascinating to get drunk with groups of Romanians a few times, with only knowledge of Western Latinate languages. Sometimes you can just hear the meaning... but only if it's something 2 Roman legionaries might have said to each other 2 millennia ago.

The Roma have been trying to claim lineage with Rome for a very long time, for reasons that should be obvious to anyone who is aware of the Roma. Apparently if they say it loud enough and often enough it becomes true. After all, they are now called Roma.

Michel de Montaigne, not Jessica Alyssa Cerro (of course).

Fair point though, even today there are parents raising children to be fluent in non standard off track domains.

I have mixed feelings about Latin. Excuse my lengthy rambling about the German school system.

In many German high schools (at least the highest or "Gymnasium" level which is kind of a hillarious name if you are an English speaker but I digress) the first foraign language is English starting in the 5th grade which is basically the first year in the new school (years 1-4 are "Grundschule"). Then in 7th grade you typically have a choice of Latin or x (x=French in my case).

I picked Latin on the vague notion that I might study law, medicine, archeology or some other topic where it's a requirement in the future and because the teacher said it's great for people who like to think structured and logically. It was also marketed as being a very usueful base for learning other languages later. French also didn't appeal to me because my thinking at that age was "I'll never go to France so it's useless".

Then in 9th grade there's the option to take a 3rd language or some other topic (I belive things like sociology etc. are on the table). By the time I reached 9th grade I already hated Latin because it was just rote drilling of stuff and then translating texts. The only thing that made it acceptable was the random sideffect of Romean/Greek history and mythology which I enjoyed. But I figured...well, well, well now that I suffered through Latin at least French will be easier so I picked French. My mindset had also changed a bit and I figured spoken languages might actually be useful.

However, it turned out that I was way more into sports, hanging out with friends and programming at the time than investing time in French so I started slow, lacked vocabulary and was always lagging behind somehow and lost all motivation (but once again I told myself, I'll likely never need it so I'm fine). I also made up my mind that I wanted to be an exchange student in the U.S. so I poured my energy into improving my English from a D to an A. The magic of motivation and watching late night comedy on NBC Europe.

11th grade rolled around, I spent it in the U.S., had a blast, returned to Germany and there was a only a slight academic problem. The last year of Latin is the 11th grade and after that you get a shiny certificate which is good for studying medicine and the like. Since the grade is a combination of class participation and written exams almost noone fails this and a D- passing grade is fine since it doesn't influence any graduating GPA or anything. So in order to get this certificate I had to take half a year of Latin with three people (exchange students returning) during 12th grade and then a 1 on 1 oral exam with the teacher and a written exam...a lot harder than the typical 11th grade stuff. All fine and well except I broke my wrist playing football (the American variety which I started playing when I came back) a couple of weeks before the exam. I then had a choice of "figure out a way to take the exam or repeat the Latin 1 on 1 next year or drop Latin". I wrote as much as I could, it was quite painful I got a horrible grade (5- in German so basically almost an F) but thankfully aced the oral exam to get a combined D and the great certificate. Fast forward to my current life and I never needed this certificate...I did however work in France for two years so yeah, life choices :P

All that rambling aside, I do feel it was actually a good decision to study Latin. The grammer pops up in other languages and things sometimes just make sense. There's also the strange notion of just knowing vocabulary in other languages and being able to sort of kind of understand what people are saying (in Spanish and Italian) and it did help with re-learning French. Would I recommend it to my 5th grade self again...no idea. The exposure to Plato and Aristotle also sort of started my interest in philosophy if science which I consider the biggest positive from the entire experience. If I were to reform schools I'd probably make the choice Latin or "non European language" (Chinese, Arabic whatever). I'm currently learning Vietnamese and it's a very interesting experience because there is no base set of somewhat known vocabulary to fall back on like with English and French.

There is no doubt that having some good basic Latin knowledge is useful, I presume particularly so for someone not native to any of the Romance languages, as it allows to understand many words, even in non-romance languages and recognize their origins.

As I see it the issue is not so much with Latin in itself, but rather with the way it is (or was) taught (though I never studied it, I am told that Greek is/was the same).

As you say, you don't study Latin for its own sake. My Latin classes taught me a lot about more about grammar, syntax and language structure than I got from other languages, and I think a big part of that is exactly because it's a dead language: since you don't need to teach pupils how to read or write, all effort can go into reading comprehension and text analysis, and eventually language theory.

Also, my teacher (at least for most of the years) was a language historian and would speak a lot about language development from proto-indo-european to greek to modern-day romance languages. For example, being aware of the r<->l and h->f consonant shifts from Spanish to Portuguese allowed me to more easily recognize unknown words when learning the latter.

Reply on news.ycombinator.com