Read your way through Hà Nội | Hacker News Reader

Read your way through Hà Nội

vietnamesetypography.com

93 pointsjxmorris1281 comments

The irony that this is an article about reading and literature, and yet the NYT butchered it for stylistic purposes. Changing peoples names I find unforgivable.

How far do you want to take it? Should we write Chinese names in characters and Arabic names in their writing system? And of course, historically Vietnamese used Chinese characters -- the Romanized form was introduced under colonialism (although granted, there seems to be no desire to go back after independence)

Whereas rendering Chinese names with Chinese characters would make the names entirely unintelligible to an audience of English speakers, the nice thing about the diacritical marks in Vietnamese is that they can simply be ignored by those familiar with the Roman alphabet but not Vietnamese. I would similarly expect an article about Spain or Spaniards to include the accents and the ñ character and an article about France or the French to include accents and the ç. (I wouldn't necessarily be opposed to writing out Chinese names in Chinese, then putting the pinyin in a parenthetical, either. Obviously there would be no point to this with Vietnamese, Spanish, or French.)

Here’s a related article about Vietnamese graphic design:

https://news.ycombinator.com/item?id=43149266

> Note: The Vietnamese words in the original version of this essay used diacritical marks. To comply with New York Times style, the marks were removed before publication.

Why, in the year 2025, does the NYT still deem this to be necessary?

Not directly related, but to avoid another thread on diacritics: I wonder if any other language chose to use so many diacritics for its official transliteration to a Latin alphabet. As someone who doesn't read it, Vietnamese text often appears as if it's randomly scribbled over.

On to your main topic: diacritics are only really useful for people who speak a language. To everyone else, such as the vast majority of the NYT's audience, they are purely a distraction. My own language uses some diacritics, and we basically never use them in international contexts. For example, if I have to write my address on a paper form outside my country, I won't ever write București, I will write Bucuresti, because it makes no difference to the reader and it may even confuse them. For example, if someone wrote their address on paper as "Đinh Lễ", I wouldn't be surprised if it got copied over as "Inh Le" street, with the person doing the copying assuming that the striketrhough was a correction, not a diacrtic.

I disagree, but your opinion may come from your ignorance (sorry, lack of knowledge perhaps) of Vietnamese. First, it's not a transliteration, it's their native alphabet. And the diacritics mark tones, which is a very important part of the language. An example from the article itself:

> In the case of Hỏa Lò Prison, for example, “hỏa” means “fire,” and “lò” means “furnace”: the Burning Furnace Prison. Without the marks, “hoa” means “flowers,” and “lo” means “worry,” rendering the term “Hoa Lo” meaningless.

Your example doesn't work because (a) it's an address, not text meant for reading and (b) turning ș into s only alters the pronunciation, while the meaning is still intelligible.

> First, it's not a transliteration, it's their native alphabet.

Then what the New York Times is doing is correct. If they write "Hanoi" instead of "Hà Nội", they are not writing "Hanoi" using the Vietnamese alphabet incorrectly. They are writing "Hanoi" using the English alphabet correctly and idiomatically. The fact that those two alphabets happen to share some glyphs is coincidental.

One can write "shchi" in English and all of those letterforms also happen to exist in Cyrillic. But that is not how a Russian would spell their word for cabbage soup. It's a coincidence that the letterforms exist in both alphabets.

If your argument is that the New York Times should use the native alphabet for words related to that region, then it would be a fair criticism. But I don't think most English readers would expect an article about Moscow to say "Москва", or an article about Tokyo to say "東京" or even "Tōkyō". By that same logic, an article about Hanoi should say "Hanoi" not "Hà Nội".

The term is actually a Chinese loanword, 火爐. One could then argue that without writing it with characters it just becomes meaningless sounds, which could have originated from any number of characters, if given no context. So therefore, your example doesn’t work so well either.

I would argue that the loss of characters work for the Vietnamese because the intelligibility is “good enough”, in the same way that writing Vietnamese completely without diacritics for an English-language newspaper is also “good enough”.

That is nonsensical. You could make the same argument by taking any English word, tracing its origins to Greek or Latin, calling it a loan word, and therefore arguing that it doesn't matter if you spell it correctly or not. Clearly misspelled words are "good enough" to be understood, but wouldn't you be disappointed if a newspaper contained a bunch of misspellings?

Perhaps I am also ignorant, but I thought the Latin+diacritics system was invented by a Frenchman in modern times, rather than being native to Vietnam.

Close, it was actually portuguese missionaries.

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

Right, so then it is a transliteration and it is not native to Vietnamese, despite what GP says.

Almost every writing system was imported from somewhere else, including something like half a dozen evolutions of the one we're using now (which was Latin, which was Greek, before that Phoenician, before that Egyptian).

What matters is that the Vietnamese use the script to write their own language, which is not the case for (say) romanized Chinese.

>early 17th century.

At what point does something become naturalized? This feels needlessly pedantic.

It's not a transliteration. What supposed writing system are Vietnamese originally writing in, before they transfer it to Latin script?

Chinese-like characters, which were transliterated by Europeans into Latin. This system was made official by the French rulers of Vietnam.

So you're saying that because the country was once colonized, their writing system is not "real" enough for you, and you only consider it a transliteration? That seems extremely disrespectful.

Nope, I didn't say any of that.

The Latin alphabet is not native to English, and it's a much worse fit for that language that it is for Vietnamese.

That is a really dumb point. Then Finnish has no writing system either, because it was created by a swede in the 16th century. Strange how there exist languages without writing systems, yet people write them?

it's the only official writing system that we have. The non latin scripts have practically disappeared from modern life.

We had centuries of Chinese scripts, which is definitely not native, then a short lived Chinese-like writing system that is the closest thing to "native", (it's not, see "Chinese-like"). Even that was not used as official system for as long as the current latin alphabet.

As a Viet, I am just speechless to you, someone brought up the topic of respecting the language as it's used today, and you wanna dilute the conversation by arguing what is native and what is not?

Sorry, I mean no disrespect. For my own language, Latin isn't native to English and English isn't native to Britain. These facts are nothing to get upset about.

Ridiculous opinion. If no country in the world has either a native alphabet nor a native language, then what is the point of the word "native"?

> If no country in the world has either a native alphabet nor a native language

Who said that? I certainly didn't.

It's not very nice to go around calling people ignorant, ridiculous and dumb. You should stop doing that.

it's annoying because it's irrelevant to the topic at hand

I didn't know this is the official writing system of Vietnam. This explains why they have so many diacritics then, if it's their only writing system.

Even so, I don't think that changes my point. Sure, diacritics serve an important purpose in a language. Many words in Romanian are only differentiated in writing by diacritics (for example, "în" means in, inside, while "in" means linseed; "să" means "to", while "sa" means his/her).

However, this is only relevant for a Romanian audience: an international audience will not understand the words either way, and will usually not even be able to differentiate them from a list based on the presence or absence of the marks. If Hanoi had both a Hỏa Lò Prison and a Hoa Lo Prison, non-Vietnamese speakers will have no idea which to go to. Even less so if they had a Hòa Lỏ Prison in addition to the others.

As a fellow Romanian, I don’t see how Vietnamese is that different from Romanian in its writing system. They both overlay information onto the Latin alphabet, Vietnamese merely does a lot more of it.

It's not, that's my point. And yet, I don't often use Romanian diacrtics when writing English, and I certainly don't feel an international audience loses something if we talk about driving down the Transfagarasan instead of the Transfăgărășan.

I feel differently, I always write the diacritics. There are fewer ambiguities than in Vietnamese, but enough to matter. And everything has Unicode support now.

Handwriting definitely doesn't have Unicode support. And neither does people's reading. It may even be easier to guess what someone is mispronouncing if you see exactly what they read (i.e. the letters without diacritics) than if you see the diacritics and forget they mean nothing to the other person.

So first, almost all words in vietnamese are only differentiated by diacritics, it's just not a case of one word here and there, removing them makes vietnamese text mostly unreadable. So they are necessary even if you don't know the language, just to translate it, even with a computer.

And then no, diacritics are also relevant outside of Vietnam, Vietnam isn't the only tonal language in the world, some other nearby countries like China or Thailand might get a better (but imperfect of course) idea on how to pronounce these words.

Diacritics are not just tone markers, and half of the tone marker diacritics don't correspond to the Pinyin tone markers, so I doubt Chinese language speakers would get much from seeing these diacritics either. The Thai script tone markers are even more distinct, and it seems that the Latin transcription of Thai script used in Thailand tends to not include any tone markers at all - so again, I doubt that Thai speakers would recognize the Vietnamese diacritics and be better able to distinguish Hỏa Lò from Hòa Lỏ.

> Diacritics are not just tone markers, and half of the tone marker diacritics don't correspond to the Pinyin tone markers

I'm not 100% fluent but I don't know a single word which isn't pronounced like it's phonetic writing. If these words do exist, they must be very rare.

> I doubt that Thai speakers would recognize the Vietnamese diacritics and be better able to distinguish Hỏa Lò from Hòa Lỏ.

At a first glance probably not but it should be very easy to teach them that.

> I'm not 100% fluent but I don't know a single word which isn't pronounced like it's phonetic writing. If these words do exist, they must be very rare.

I'm saying that even if people familiar with pinyin recognized the (very approximate) correspondnce between Vietnamese tone markers and pinyin tone markers, they would still not understand all of the other diacritics that do other phonetic things that have no correspondent in pinyin.

> At a first glance probably not but it should be very easy to teach them that.

The same argument applies to anything that is teachable. The NYT could start throwing in a few Chinese characters in every article, to get people more familiar with Chinese writing. Would that be nice? Sure. Does it make any sense to wonder why they don't do it? I don't think so.

I still don't get your point, we use Chinese pinyin because they give a somewhat spoken version of words, vietnamese sentences without diacritics are 100% useless, they are useless to foreigners, useless to vietnamese people and even more importantly, useless even for machine translation and search. Who are they intended for, I've no idea.

The western equivalent maybe would be removing all the vowels of a sentence, yes you can do it but I'm not sure how it's useful in any way to any audience.

I only talked about pinyin because Chinese familiarity with tones was brought up as a reason why some non-Vietnamese speakers might still recognize the diacritics and get some value from them.

But even beyond that, I highly doubt that the article is really ambiguous without diacritics. I somehow doubt that if the NYT talks about a Hoa Lo prison somewhere in Hanoi, there is any real ambiguity about the actual place they are talking about. Sure, the words in themselves are ambiguous, but the context makes them very clear. This is not about writing Vietnamese without diacritics, which I'm sure is extremely ambiguous. It's only about some place names with very clear context about where and what they are.

> half of the tone marker diacritics don't correspond to the Pinyin tone markers

Yeah, they are different languages. What'd you expect? First you didn't even know that the Vietnamese alphabet was their actual alphabet, and now you're criticizing it for not being similar enough to some other random alphabet. Whatever your point is, you're failing at making it terribly.

The person I was responding to here claimed that the diacritics, even if they don't help people in the USA or Europe to get a better idea of the pronunciation, should help speakers of other tonal languages to do so.

So, I investigated what a Chinese speaker who is not familiar with Vietnamese writing might make of the diacrtics, specifically the tone markers since this is what I was replying about. Since the regular Chinese writing system doesn't include tone markers, I looked at Pinyin, which sometimes does. The conclusion was that no: even a Chinese speaker would not recognize the meaning of the Vietnamese diacritics, so even for them, it would not be useful.

This was just intended as a refutation of the previous poster's point, not anything broader.

Also, none of what I am saying is in any way a criticism of the Vietnamese writing system or language. The only thing I'm criticizing is the idea that it helps in any way for a non-Vietnamese audience to include language-specific diacrtics in a non-Vietnamese article (and I have the exact same opinion for my own language's diacritics, and for any other - Vietnamese just happens to be the topic here).

I still don't understand this attitude of "I can't be bothered to try to understand it so it's useless". Vietnamese is actually one of the more easy tonal languages for westerners to understand, given that the tone marks are literally pictograms of the pitch (and it's not a transliteration like 你好 -> nǐ hǎo). Why are you so allergic to actually making use of that feature?

I'm not saying its useless, not at all - not for people who speak Vietnamese. I'm saying it's not relevant for people who don't.

I think the argument here is that Vietnamese script is so extremely reliant on diacritics that it cannot possibly make sense without them, despite looking legible to non-speakers. Similar argument could be made about Chinese or Japanese, but phonetic transcripts of those languages are actual gibberish to everybody that nobody cares. Vietnamese is on a such marginal point that frustrations can be expressed.

Even if this is true, it is irrelevant. The reality is that the vast majority of the NYT's audience will not get any extra information from including those diacritics than excluding them. If the article is not intelligible without diacritics, then it won't be intelligible with diacritics either, because people who don't know the language, nor any similar language, can't see a difference between Hoa Lo and Hỏa Lò and Hòa Lỏ.

The reality is that the vast majority of the NYT's audience will not lose any information from including those diacritics, and some people will gain quite a bit.

There's another angle here as well: I doubt NYT editors are familiar with Vietnamese spelling. If there are errors in the diacritics, they will not be able to spot them, and may end up with a text that appears more precise than it actually is. If they just remove all diacritics, no reader will be confused they avoid this potential for errors altogether.

"We need to remove information from the text because perhaps the author (who clearly speaks Vietnamese) may have made a mistake". Do you also think that we should remove all algebraic symbols from mathematics papers, because perhaps the author has made a mistake, and the audience may contain people that won't spot it?

The vast majority of readers won't get any information from anything in the article. Why not pseudonymize everything and scramble the place names? I at least appreciate that in principle, I could research the people mentioned. Romanian happens to be intelligible with diacritics removed, but I bet you'd feel differently if you read an article about Mr Ccsrtr and Em Cnr.

My point is that, I think, if you frame the diacritics-stripped Vietnamese as a language transcribed in a different script, than half measured attempt at representing Vietnamese script, it solves the question of whether it's useful as half measure Vietnamese.

"Huawei is written and read huawei in Chinese" is not so useful, and it's okay, because it's obvious. "Vietnam is actually written and pronounced Viet Nam" is less okay, because it's not as obvious.

And, I think, frankly, it's justifiable to consider Vietnamese script(both Chinese based and Latin based) as scripts of their own rather than derivatives of something else, as there never were meaningful synergies in pretending otherwise. Vietnamese appears to have been always phonetic and nothing made sense unless you were a speaker. That's quite unlike how everyone knows what entrepreneurship is regardless of languages in use or whether diacritical marks are supported.

You will die on that hill, won't you.

I dislike this general trend of rejecting the need for adaptation/transliteration and pretending that it's a moral failing or a rejection of diversity.

Like people who insist it's a good idea for a European website of a European business to accept any Unicode input for names, as if an employee who speaks Italian and English could be expected to know how to process a request for a customer named 田中 who claims their correspondence was mistakenly sent to 東京 instead of 京都.

There is generally too much linguistic diversity in the world to be able to expect people to know even the most basic facts about some other culture's language. There's nothing wrong with adapting your message to your audience, even if it loses a lot of nuance that they could theoretically get if they spent just a little bit of time on studying, say, Vietnamese writing.

And I want to emphasize that I'm saying this who is neither American nor English, and who is personally fascinated by language, and who has taken the time to study a little bit about quite a few languages. But I'm also someone who has understood that you can't expect people to be able to, say, pronounce your name correctly, or spell it correctly, and that there's nothing offensive about that.

I've stopped saying “hi mom” to my mother when I visit her. After all, my mother /knows/ that I love her. Hence, it's just so much more practical not to greet her every time, no? Surely no one can dispute this incorruptible logic.

"I am so fascinated by language that I wish my newspapers would include less foreign languages"

I wouldn't say the Vietnamese alphabet is "transliteration". Vietnamese is one of the most, if not the most tonal language in the world. The same word, speaking with different tones will convey different meanings.

https://en.wikipedia.org/wiki/Tone_(linguistics)

The modern Vietnamese alphabet was developed in 17th century (so it's not a transliteration) with tonal marks as a core feature. The writing language is very phonetic. Within a region with similar accent, if you hear a word, you can write it. And if you see a word, you can pronounce it.

The tonal marks are very important to the language. It allows for rich poetic rules that makes Vietnamese poem fun and musical to read:

https://en.wikipedia.org/wiki/L%E1%BB%A5c_b%C3%A1t

Yes, I had never looked into this and had assumed Vietnamese uses a Chinese-inspired writing system natively, like other languages in the region. Knowing that this is the only writing system immediately made sense of why this is necessary.

Ehm, like in Vietnam's neighbors Laos (ພາສາລາວ) and Cambodia (ខ្មែរ)? Sure Vietnamese used to (a long time ago) be written in its own version of the Chinese script, I'll give you that. But most languages in the region do not use a script derived from Chinese.

> To everyone else, such as the vast majority of the NYT's audience, they are purely a distraction

in the same sense that any foreign word you can't pronounce is a distraction. I thought the point of reading is to learn new things? Words and pronounciation are unremarkable now?

I think it's quite the opposite, diacritics teach me how to speak a language that I'm not native.

Portuguese don't use as many as Romanian, but they are very useful. After all what's the difference between avó and avô and just avo?

We don't use in very informal settings, like Whatsapp chat, because a native reader can infer from the context. And that's how English without them works, right?

Actually I wanted the English language to have some, so the tiny differences in certain constructions would be more obvious.

Yes, diacritics are very helpful if you want to understand the language. But they aren't any help if you're trying to read a travel journal about a place whose language you neither speak nor want to learn, which will be the case for the vast majority of the target audience for this article - spelling the names of places with or without diacritics will not improve the reader's pronunciation.

The latin alphabet was designed for atonal languages. When you need to also indicate tone, you've gotta put the marks somewhere. You could do a lot worse than diacritics whose shape roughly reflects a graph of pitch over time. (In Mandarin, pinyin uses diacritics that exactly graph pitch over time. Very helpful.)

You can blame a Jesuit priest for the diacritics. He published a Vietnamese-Portugese-Latin dictionary in 1641 and invented the modern Vietnamese script.

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

> For example, if I have to write my address on a paper form outside my country, I won't ever write București, I will write Bucuresti, because it makes no difference to the reader and it may even confuse them.

Funny thing, this is exactly the reason I always include the diacritics!

I don’t really fault the NYT for writing Hanoi and Vietnam, not Hà Nội and Việt Nam. It’s a newspaper for English-speakers, at the end of the day. It calls Warszawa Warsaw, Praha Prague, Москва Moscow, and hundreds of other places by their English names.

I wouldn't expect Russian newspapers to write New York instead of Нью-Йорк, either.

Hanoi and Vietnam are sufficiently well known to be anglicized. What's less excusable is stripping names like Nguyễn Phan Quế Mai of their diacritics.

Nguyễn Phan Quế Mai is the one name that the NYT did not strip the diacritics from. Every other Vietnamese term and name has been stripped (including the names of the other authors in the reading list), but the article author's name, alone, retains its diacritics in the NYT article. They do seem to have gotten the diacritics subtly wrong in the article subtitle vs. the byline (which is correct and, I assume, generated automatically).

https://archive.ph/Gi1HX

I like how Serbian is completely fine with Njujork or Majkl Ðekson.

Phonemic orthography should win and destroy all spelling bees.

This seems homonymphobic.

Edit: homophonephobic, technically.

They do this, only using the Latin alphabet for brand names etc. For example, pravda.ru spells Finland as Финляндия

(Edit: I misread your comment, fixed mine, I agree with you)

So many website that use custom font cannot render the diacritics. I hate to see the single letter font change constantly. Maybe this is related?

This really shouldn't be a problem for a newspaper with 2.5 billion USD revenue to figure out in 2025 though.

https://rd.nytimes.com should be able to handle it

Read the text and judge for yourself how the diacritical marks affect readability.

Uh yeah they don't? Unless you also have problems with words such as über, façade, señor or crème brûlée.

Rather, the removal of them affects readability in a similar way to removing accents, punctuation or writing in all lowercase.

There are two types of diacritics, from the perspective of any reader: the ones they are familiar with and understand, and the ones that are visual noise. American and (West) European audiences are typically more or less familiar with the umlaut, accent, cedille and circumflex mark, and the tilde. Other diacritical marks typically fall in the second category for them, outside of use in their own language.

So expose the reader to them until they become familiar, problem solved

so just ignore the "visual noise" or "random scribbles"? i dont get why youd want to remove meaning from an article simply because you dont undestand it.

For the same reason you choose one font over another: the aesthetics of a text matter, especially to publishers.

Now, I should add that for an article that is specifically about language, and even has some illustrations of the meaning of these diacritics, this is almost certainly a bad choice on the NYT's part. But as a general rule, I think it is defensible.

Well, I disagree that simpler means better looking. It's akin to arguing that a commie block has better æsthetics than a gothic cathedral.

Come on. I had no problems whatsoever.

Because the NYT readers value simplicity more than authenticity.

Yeah, very silly in an article specifically about language. But one may also ask, why do people still read NYT (and other newspapers) in 2025, given that they are just inferior versions of blogs, that you also have to pay for?

Reply on news.ycombinator.com