Why Chinese is so difficult

Worse than you think, I enjoyed the discussion of dictionaries most of all:

One of the most unreasonably difficult things about learning Chinese is that merely learning how to look up a word in the dictionary is about the equivalent of an entire semester of secretarial school. When I was in Taiwan, I heard that they sometimes held dictionary look-up contests in the junior high schools. Imagine a language where simply looking a word up in the dictionary is considered a skill like debate or volleyball! Chinese is not exactly what you would call a user-friendly language, but a Chinese dictionary is positively user-hostile.

Figuring out all the radicals and their variants, plus dealing with the ambiguous characters with no obvious radical at all is a stupid, time-consuming chore that slows the learning process down by a factor of ten as compared to other languages with a sensible alphabet or the equivalent. I’d say it took me a good year before I could reliably find in the dictionary any character I might encounter. And to this day, I will very occasionally stumble onto a character that I simply can’t find at all, even after ten minutes of searching. At such times I raise my hands to the sky, Job-like, and consider going into telemarketing.

Chinese must also be one of the most dictionary-intensive languages on earth. I currently have more than twenty Chinese dictionaries of various kinds on my desk, and they all have a specific and distinct use. There are dictionaries with simplified characters used on the mainland, dictionaries with the traditional characters used in Taiwan and Hong Kong, and dictionaries with both. There are dictionaries that use the Wade-Giles romanization, dictionaries that use pinyin, and dictionaries that use other more surrealistic romanization methods. There are dictionaries of classical Chinese particles, dictionaries of Beijing dialect, dictionaries of chéngyǔ (four-character idioms), dictionaries of xiēhòuyǔ(special allegorical two-part sayings), dictionaries of yànyǔ (proverbs), dictionaries of Chinese communist terms, dictionaries of Buddhist terms, reverse dictionaries… on and on. An exhaustive hunt for some elusive or problematic lexical item can leave one’s desk “strewn with dictionaries as numerous as dead soldiers on a battlefield.”

There is however much more, by David Moser, via someone on Twitter, sorry I forgot.


The article is from 1991 — it seems out of date given all the character recognition apps. Who can provide an update?

Looking up words is easy now with dictionaries on the web. Pinyin won out over other romanizations and Google Translate is your best friend. Mandarin, like most Chinese languages, is colorful and there is usually a backstory or two behind most everyday words and phrases that helps hammer down the meanings.

I started learning 8 or 9 years ago, and it's got a bit easier in some ways since then, but I think the process of using paper dictionaries is still fairly useful, can be fun and, for me at least, enforces discipline. My learning plateaued with my first smartphone...

Thanks - I was just about to look at the author information, because I could only imagine he was either quite old, or extremely uninformed about the last 3 decades of computer technology.

Is it at all common that people decide to only learn verbal Chinese, i.e. spelling the words with the Latin alphabet? If so, how hard is it?

Tones take about a month to speak, a few more months to hear. Verbal Chinese is a very straightforward language. No conjugation, no tenses, no gender, no nonsense. Word order mostly follows English but there are differences of course. If you know Chinglish or can fake caveman English you're a good part of the way there. Find a group you can practice with and you'll surprise yourself by the end of the year. Make no mistake though. Reading and writing is truly the hardest part and will easily take years of dedicated study and practice.

Quite a lot of foreigners do that, I met an English guy in Beijing with a great, native sounding Beijing accent. He couldn't even navigate a menu, but he still impressed locals way more than I could with my 10000+ characters and slightly accented Mandarin... It's a bit depressing.

'One of the most unreasonably difficult things about learning Chinese is that merely learning how to look up a word in the dictionary is about the equivalent of an entire semester of secretarial school. '

And by using a touch screen smart phone, it is literally child's play, along with hearing the pronunciation. Progress marches on - this is not 1965 anymore, much less 1865. Though I see what the linked author is apparently complaining about seemingly involves translation, and not actually being a native speaker. Written by someone who apparently is completely unaware of the last couple of decades of advances in areas such as Unicode. Why yes, there is an actual 'structure' which one can order graphs by (admittedly Unicode would not include an obscure 16th century graph), even if it is not the one he wishes.

'Imagine a language where simply looking a word up in the dictionary is considered a skill like debate or volleyball! '

Or a language where people have contests where the entire point is to correctly spell a spoken word. Germans remain fairly baffled by American spelling bees involving junior high students.

And seriously, this is just silly in practical terms - 'I currently have more than twenty Chinese dictionaries of various kinds on my desk, and they all have a specific and distinct use.'

Basically, this is the equivalent of listening to a non-native speaker of English talking about just how bad English spelling is - which is true, but so what? English is what it is, and it is English speakers who just accept a spelling system where you cannot even know how pronounce 'read' correctly without context. Or whether the you look under 'k' or 'q' when somebody talks about where a ship is tied up. After all, everybody who is functionally literate in English knows how to correctly pronounce quay, and can easily find it in an alphabetically arranged dictionary, right?

If you red the article carefully instead of rushing to post your dissenting opinion, you'd have lerned the article was from 1991, pre-internet age.

Bonus trivia: the first sentence of the post is, says Bing Translate: "比你想象的更糟,我最喜欢字典的讨论:"

'you'd have lerned the article was from 1991'

And you would have learned from reading the comments that I thanked the poster that pointed that fact out.

Leading to the real question - why is Prof. Cowen posting an article from 1991 (or older, as the case may be), and why is he talking about Taiwan and dictionary look up contests in an age where junior high students would use a smart phone?

Yet even more amusingly, the article I read and responded to - the first link, not the second and now clearly different one - is from 1986, as noticed when not simply skipping the headline to read the text.

And this is not a 'dissenting opinion,' as noted by Xian. Though you might be right in another sense - I have been involved in overseeing Chinese translations, and in my opinion, anyone writing in 2019 about using paper dictionaries is clearly someone with no knowledge of the modern world.

Interesting that someone cleaned up the formatting, so at least one does not think it was Prof. Cowen in Taiwan.


This is fun with google translate, we can actually cuss, I think, let me try:

That was a cuss word. I have no idea how to read the stuff, let me try another:

泰勒考恩是一個傻瓜球 -- Tyler Cowen is a goofball

I didn't even know goofball had a Chinese equivalent. Let me try Commie Rat:


Wow, it is not even censored.

Instead of Commie it gives you Common/Share. Nothing drastic.

Good enough for government work

Hi Tyler,

I'm currently living in China and learning Chinese. I would caution anyone who is learning the language to completely compartmentalize in their minds the spoken language and the written language. The spoken language is incredibly simple; sounds are very easy for English speakers to master and grammar is highly intuitive to English speakers (subject-verb-object, no conjugation).

It's unfortunate that so many Chinese (nationality) teachers of Chinese (language), both in China and the West, are inculcated into the false dogma that the written language is central and you can't learn the language without learning characters. Even introductory courses start on the characters right away, which intimidates and hampers learners from developing greater speaking/listening skills more rapidly.


Aaaand I completely undermined my entire post by ending it with Chinese characters. That is read "jia1you2", literally add+oil, i.e. keep at it!

I live in China, too, and have been working on the language for years. I have the opposite impression from you. I find reading, to be not so hard, relative to understanding spoken Chinese. All that is required is that I recognize the character. Usually, the radical gives a pretty good hint. Handwriting is much harder. I've given up on that and just type pinyin into the computer, as do most younger Chinese. Understanding spoken Chinese is so hard because there are so many homophones in the language. The same syllable can mean many, many things.

You're right about one syllable meaning many things, but that never impedes understanding to me; most words are polysyllabic, and the combinations and context really cuts down on the ambiguity.

I mean you could say the same for English - also a high degree of homophones, but rarely ambiguous if you have a full sentence.

I do agree about writing - I have no interest in learning to write. It's just not a profitable investment of my time. Recognition does come pretty easily after a while.

Spoken Mandarin is far from simple. As a commentor mentioned, the number of homophones is ridiculous. My heart sinks when I look up a new word and find its yet another bloody 'Ji' or 'Jing'. Whilst the tones are often raised as the main difficulty, the challenge in pronunciation and hearing is actually the smallness of the words, the lack of final hard consonants, and the resulting importance of the vowel sounds. This makes it very hard to make yourself understood and hard to understand others. Contrast for example to Japanese which has longer words, syllables punctuated by consonants to help distinguish words, and grammatical inflections to lend clues to the listener. When on a short working stint in HK I picked up enough Cantonese to direct taxi drivers, and found it easier to make myself understood than I had speaking Mandarin to mainland taxi drivers after years of study.

Yep, I wonder if the people who say that spoken Chinese is easy to learn are people who have better ears. Without being able to distinguish vowel sounds and tones, the listener is toast.

Prior to visiting China for several weeks, I had lunch with a native Mandarin speaker. She warned me to watch out for those tones, saying that a word like "ba" meant different things depending on how you said it, and proceeded to demonstrate saying "ba ba ba ba".

That's what it sounded like to me, four identical pronunciations.

And this wasn't just me, another American English-as-a-first-language speaker who was having lunch with us had the same reaction that I did.

I'd already decided that I wouldn't even try to learn conversational Mandarin (aside from "hello", "thank you", etc.) but that demonstration sealed it.

In contrast, I found that I could learn to read Chinese characters, thanks to the book _I Can Read That_ by Sussman. Without knowing any Chinese at all, the book in the very first page shows you how to read Chinese characters, as other commenters have said there's a system to it, with recognizable radicals. The book teaches the reader maybe 150 characters, I committed maybe half of them to memory.

To write or pronounce them are hard but to understand them is easy. For example the word goat 羊 has two horns, 4 legs and a tail. People 人 with two legs but no arms to differentiate it from those 4-legs. Shepard? A lamb on the shoulder of a man 羌. Unfortunately that is also the ancient word for the nomadic barbarians. A lamb on the shoulder of a woman 姜 is different, that is the ancient tribe that allied with the Ji people to form the Zhou Dynasty three thousand years ago. 姜 with the female word radical is one of the very ancient surname all with female word radical from the matriarchal period more than 2500 BCE. That could be the source why about 40% of modern Chinese are latent carriers of red hair genes.

So if you know the tricks you can guess at the meaning.

Yup, that's what's so great about the _I Can Read That_ book, in literally the first two pages and first two minutes of the book one starts reading Chinese, by starting to learn those principles that you describe. I.e. one quickly sees that there's a system to it, not just a bunch of random squiggly lines.

In contrast, I'd always figured I'd be lost trying to understand much less speak conversational Chinese, and that native speaker demonstrating the four different ways of pronouncing "ba" confirmed this -- they sounded exactly the same to me as well as the other English speaker at the table.

Obviously with training, our ability to distinguish Chinese sounds would improve, but the contrast between being totally lost at the most basic task of hearing, compared to being able to start reading after two minutes of reading a book, made it clear to me that I could learn Chinese written communication a lot more quickly that I could learn conversational communication.

This was with the limit of having only a certain amount of hours to study before going to China. I could've learned a lot more by enrolling in a Chinese class but that was many times more hours than I had to spare.

P.S. Also as you describe, each character or radical almost always has a story behind it, so while learning a few characters one also learns some aspects of Chinese history or culture.

I found tones almost impossible at first (ie first 2 years + of study). It was fixed by getting a personal tutor from Beijing who was strict with me, and second by the realisation that its much easier to distinguish tones if you listen for them in pair combinations, not standalone.

Don't you notice that the world already has move that way. What do you think those emoji icons fundamentally are?? Or was that a Chinese conspiracy??

Iapanese - emojis (as pictures) and their wide spread adoption can be traced back to DoCoMo - in 1999, 8 years after one of the linked articles was written.

Emojis are a largely a substitute for non-verbal emotional expression (talking with your face, hands, etc.), or a form of visual slang (and like slang is purposefully opaque to an outgroup), not a substitute for phonetically spelled lexicon.

Just like 2D bar code can contain more information, the Chinese words potentially can contain denser information. You might appreciate that in another 100 years time. Or do you prefer word like Donaudampfschiffahrtsgesellschaftskapitän.

The underlying data representations of Chinese characters (in bits) aren't lighter, they're not quicker to write, they probably don't stand up better to being legible in small print.

Your math is off. The average English word is around 5-6 letters. Using UTF-8, the most common encoding, that's is 40-48 bits per word on average. Chinese characters in UTF-8 are on average 24 bits each and the average Chinese word is 1 to 2 characters giving us a total of 24-48 bits. Chinese is informationally more dense in both digital and visual domains compared to English. Audio too since most Chinese words are 1 to 2 syllables.

Shh - ideographs and letters are quite distinct ideas, and it is not common to find anyone familiar with letters to know much of anything about UTF-8 encoding. Not to mention that English requires more words grammatically - though every English speaker understands 'goose fly' at a basic level, it is not grammatically correct in comparison to 'The goose flies.' or 'A goose is flying.'

(Though in another comparison, English requires less space than German, which has considerably more endings attached to articles.)

Comparisons without considering compression are going to be fairly meaningless. Is anyone archiving for future purposes not going to be doing significant file compression? See - https://languagelog.ldc.upenn.edu/nll/?p=93 for a cursory consideration. Redundancy also needs consideration as a feature of any written text on paper for the sake of preservation.

Spoken languages with more information per syllable tend to need more complex syllables than English. Those with more recurring sets of syllables (lower unique syllable set to morpheme ratio) tend to need more spoken redundancy ("Sheep. Yes, that's right sheep.", etc) and/or processing time for the speaker and listener in actual practice. Likewise languages with less morphology.

I do wonder if a truly visual language like the various forms of Sign may perform better in information compression than the Han language or English, even if the "logograms" do not.

You can easily encode english using ascii - 7 bits per letter (in fact, if you skip uppercase it is 6, or even 5 excluding special characters like #!@..). With a character set in the order of thousands you cannot do that...

"Imagine a language where simply looking a word up in the dictionary is considered a skill like debate or volleyball!"

I always thought the same thing of spelling bees in Anglophone countries. English orthography is so insane that you made a contest out of it.

I think he exaggerates the difficulty of learning the radicals and using traditional dictionaries.

There are 214 radicals, which sound like a lot, but they're all basically pictographic and thus easy to remember , and they're in regular forms and locations within the characters. The radicals in most characters are immediately identifiable.

I'll tell you what's hard - I've been working overhours at the Dick Sucking Factory all week. We have a tonne of backorders for dicks to be sucked and we cannot keep up with demand. I'm one of the top employees down at the Dick Sucking Factory and I would like to offer anyone here a job sucking dicks at North America's premier Dick Sucking Factory. Response to me for more info!

You are one sick puppy.

How does it feel to know you are a narcissist and a psychopath?

My godson speaks and writes Mandarin. One summer while still in high school he lived in China and worked at a university translating documents (Mandarin to English and vice versa). He always refers to the language as Mandarin, never Chinese. My understanding is that Mandarin, sometimes referred to as "simplified Chinese", is based on the Beijing (or Northern China) dialect, and is the national (official) language of China. According to my godson, Mandarin is not that difficult to learn to speak because, as one of the other commenters indicates, the sentence structure is the same as ours (subject, verb, object).

My godson's father spent a year working in Japan many years ago. His English writing style I have compared to Christopher Hitchens's writing style. Whatever one might have thought about the man's politics, he could write a beautiful sentence. I described to my friend that Hitchens's and his own writing style is backwards. My friend explained to me that he was influenced by Japanese, which often makes the action word the subject of the sentence. I suspect that learning Japanese would be more difficult than learning Mandarin.

I think learners of Japanese are aided by the large quantity of Japanese media available in a form that often appeals to "westerners". Chinese movies and TV shows native English speakers are willing to watch are often in Cantonese rather than Mandarin. I think the ruling party of China will have to chill a bit if they want Mandarin language media to take off internationally.

"I suspect that learning Japanese would be more difficult than learning Mandarin."

It is for a couple of reasons. I think mostly is that Japanese SOV grammar is much further from English than Mandarin grammar is. Part is cultural where Chinese are overall much more direct than Japanese and that adds up over time.

Here is part of a 2004 interview of an America translator who first studied Mandarin for ten years then learned Japanese:

RCA: Can you talk a little about how Japanese language is different from Chinese language? And also, do you have any thoughts on the larger significance of the differences?

TC: I have very strong opinions on that, Bob. To me Japanese was much more difficult, as a person coming from a background as a native speaker of English. Of course, they are totally different language families. And although they do have the kanji in common, I think many people coming from a Chinese background assume the knowledge of kanji is going to be much more helpful than it proves to be. Not only are there different meanings for the same kanji – For example, the Japanese term for letter, tegami, is pronounced shouzhi in Chinese, and it means toilet paper. So, there are some significant differences along those lines.

And in terms of learning the two languages, I think traditionally the Japanese have been somewhat less accepting of someone trying to acquire their language than the Chinese have been. It’s always been my experience that at whatever level of Chinese I had at any time, a Chinese speaker would accommodate that. I found a much greater reluctance of Japanese to deal with whatever Japanese I was capable of handling, again, at whatever time in my training. So, I think there are significant differences in that respect.


Not 2004, but 2006 - the year Google Translate's beta version went on line.

I know both, and I'd say Japanese is harder. The SOV thing isn't a big deal, but the grammar is overall much more complex than either English or Chinese grammar, and has a bunch of features that you won't find in European languages.

The other major problem is the orthography. Japan took the Chinese writing system, which is really only suited for an isolating language like Chinese, and came up with a bunch of kludges to retrofit it to Japanese. Unlike in Chinese, where 95% of characters have only a single consistent pronunciation, in Japanese, 95% of characters have at least one Japanese pronunciation, and one pronunciation based on one of three ancient Chinese dialects (but lossily converted to Japanese phonology). Many characters have a few of each, and some of the more common characters have ten or more different pronunciations.

There are general rules of thumb you can use to guess which pronunciation to use for each character in a given word, but they aren't consistent. Every once in a while I'll realize that I've just been assuming the incorrect pronunciation of a word I've seen in writing but never heard. For several years I thought that the word for bathing suit was suichaku, based on the Chinese pronunciations of the two characters, but it's actually mizugi.

There are also words whose pronunciation bears no relation to any of the standard pronunciations of its constituent characters. You might naively guess that himawari (sunflower) is written 日回り, hi (sun) + mawari (rotation); in fact it is written 向日葵. Japanese retained the original pronunciation, and just copied the Chinese written word xiàngrìkuí, which has a different etymology (flower that faces the sun).

All of which is to say that learning to read Japanese is largely done on a word-by-word basis, whereas Chinese is more systematic; if you already know a word in spoken Chinese, you have a decent chance of recognizing it in print, even if you've never seen the constituent characters before.

"You might naively guess that himawari (sunflower) is written 日回り, hi (sun) + mawari (rotation); in fact it is written 向日葵"

Apparently I haven't read enough stories with sunflowers since I've always seen it written ひまわり without the kanji.

Mandarin is a language, whereas simplified Chinese is a script. Mandarin is written with simplified Chinese in China and Singapore, but the same language is written with traditional Chinese in Taiwan.

Saying that Mandarin is also known as simplified Chinese is kind of like saying that English is also known as the Latin alphabet.

Seems correct, but I had thought the PRC had also simply reduced (or simplified) the vocabulary, whereas the Nationalist Taiwanese had retained its variety-

Somewhat in a similar fashion to how the Soviets 'simplified' Russian, by making it unacceptable to use a more educated form of Russian, as spoken by aristocrats.

Could be wrong, of course.

It's the writing that they simplified, the simplified characters are noticeably easier to learn and recognize, and I'm sure faster to write.

But at some cost, because the simplified characters have lost some of their historical connections. I'm guessing it's somewhat analogous to what would happen if American English ever adopted simplified spelling: thu sentensez wood look sumtheeng laik this.

To be specific Mandarin is actually the Manchu version of the Chinese dialect, closed to but slightly different from the Beijing dialect.

The word Mandarin came from the Portuguese who directly appropriated the word mentri from the Malay for official rather than the Portuguese ministro when they were in Malacca. The Malay has imported many Sanskrit words wholesale, in fact Malay is closer to Sanskrit than English. For example car in English is kereta in Malay and kereta in Sanskrit. Color is warna in Malay and warna in Sanskrit, etc. That sometimes can trip me as I formerly though that Mahabarata was "Great Western" since Utara Kuru is Northern Kuru.

The Chinese are pragmatic and not that hang up about preserving ancient customs.

I expected more of a "model this" direction.

The written language causes unstable expansion and contraction cycles as noted in 2,000 years of Chinese history. Once the aggregate number of users and geography expands, the character set becomes unstable.

Pure speculation, but this is what drove the post.

I suspect the language complexity was a major reason for Chinese underdevelopment relative to Europe. It probably contributed to low levels of written literacy in China (even today), as well as the mastery of literary culture rather than technical skill becoming the key to advancement into high society. The PRC, Korea, and Japan all reformed and simplified their written languages in the 1950s, recognizing the problems caused by their traditional scripts.

My main response is below, but you do know that Korean is written using an alphabet, right?

It is called Hangul, and has been used since 1446, after being created by the king Sejong the Great. https://en.wikipedia.org/wiki/Hangul

And there has been little in the way of actual 'reform,' apart from the fairly abrupt (North Korea) or gradual replacement (South Korea) of any Chinese ideographs in Korean writing. Something I had not previously known, though Hangul was noted as a distinctive feature of Korea's development in the GMU East Asian survey course I took almost 4 decades ago - https://en.wikipedia.org/wiki/Korean_mixed_script

Hangul wasn't really used much until the 20th century. Even for much of the 20th century, "mixed script", which is a mix of Chinese characters and Hangul, like how Japanese is written, was used.

There's no need to look to the writing system. After all, consider that the Japanese didn't implement anything like this.

The dominance of the literati in China can be found in the desire and ability for the center (the sovereign) to identify capable servants who could be trusted to have no loyalty to existing clans (and the center went to a lot of trouble to move them about to ensure this was the case), and from a distaste for the idea that superior individuals should be selected held to account by any form of election or assent. (This was a strong centre China had because of cycles of strong centralization probably pushed by China's geography, its Warring States history, and repeated threats from the East Eurasian steppe).

So how do you find these capable and virtuous servants who are loyal to the country with no oversight or undersight then? How do you ensure this system is "meritocratic" rather than corrupt? How do you ensure that people still pay their taxes without much representation?

It seems that the choice of ancient Chinese was inevitably to ask candidates to write flowery essays about virtue which conformed to a very tight set of rules and conventions, and then hold up the best candidates to glorification and high office. (The Byzantines tried to do the same thing, to some degree -https://twitter.com/SeshatDatabank/status/1025865170357080064).

And such a class of people are inevitably to some degree going to indulge in their own self glorification to the detriment of traders and soldiers and religious leaders (which wasn't so great when strong militaries and trade were necessary), and they are inevitably going to set up a regime of low taxes that benefits the class they come from and identified with - rural landlords.

So considering these motivations, there's really no need to look to the writing system.

You make old China sound like a Republican's wet dream.

Hey, the Confucian bureaucracy are not described as tending to be deeply conservative for nothing. So for certain values and forms of Republican, no doubt.

That argument was made in the book The Tyranny of History by sinologist WJF Jenner in the early 90s. Worth reading. For someone who had devoted his life to China it was very dark and negative about the country's history and prospects.

'I suspect the language complexity was a major reason for Chinese underdevelopment relative to Europe.'

Or the exact opposite from a historical perspective - ideographs are not reliant on spoken speech, meaning that an empire using ideographs can encompass many spoken languages.

Leading to a single encompassing framework (including a society that favored mastery of a single literary culture), unlike the fertile chaos found in Europe with its multitudes of languages. Leading to a more efficient adopting of Chinese technologies such as gunpowder, the printing press, or the compass by various European states.

Logograms were not used in China to write languages other than Han dialects (Viet or Hmong written in hanzi would be incomprehensible to both Han and native speakers), and even then are so torturous to learn that it would be quicker to learn a phonetic imperial lingual franca (Urdu or somesuch). This is a course why China goes to great lengths today to level regional dialects today into standard Mandarin.

Notwithstanding that Europeans did in fact share a written and spoken literary and elite lingua franca - medieval Latin. The nations of Europe would not have coalesced more into a single political if they had been forced to use a bizarre set of logograms to write their national languages. To say they would deeply misunderstands they foundation of political order .

Nor would a single political order in Europe, then or now, be remotely desirable....

If I'm understanding correctly, you're saying that it's easier for nations (who want to communicate with each other) to learn a whole different language, a lingua franca, than to learn a whole different system of writing (a scriba franca?). Makes some sense, although I wonder how universal the Latin alphabet was for medieval European languages. E.g. even today French puts accents and circumflexes onto letters in ways that English does not. But it's still a largely universal alphabet today -- and also during medieval times?

I went through this pain with Japanese about twenty years ago, but studying Chinese these past few years, I haven't run into this problem nearly as much. Partly because I'm already familiar with the basic structure of Chinese characters and already knew about 2,000 characters, but also because the dictionary application I have on my phone (Hanping) has excellent handwriting recognition. I believe there are also applications that will just provide an annotated translation of whatever text you photograph. I don't use these, though, since I think the act of looking a character up helps with the learning process.

Nowadays, I think the biggest stumbling block for beginners is pronunciation. In addition to the tonal system, there are several sounds that don't exist in English, including the German ü and two distinct versions each of ch, sh, and j. It took me a few hundred hours of practice to be able to read a sentence fluently without explicitly planning out the tone contours in advance.

I use the Hanping Cantonese app (there is a previous Hanping Mandarin app) developed in Hong Kong for an Android phone/tablet costing about $15. I use it with a free flash card app Ankidroid. I can get at Chinese words with any of (1) enter the character's strokes by hand (2) take a picture of the Chinese characters from the Hanping app (3) enter the pronunciation in a romanization (eg, sei [die/dead] which will return different characters and tones -- here, the tone is sei2 [an up tone] -- another "sei" tone sei3 [level tone which my app shows in purple for that tone] means four in Cantonese. Other useful tools are Sling TV streaming of Chinese and Cantontese TV stations (about $30 a month) and streaming radio stations from Hong Kong, Vancouver, New York and Los Angeles. I particularly like 4 guys who run a Cantonese streaming radio station out of Hong Kong -- they constantly jabber, often about the inane, in Cantonese. Curiously, Cantonese can't figure why anyone would bother learning Cantonese -- they themselves learn Mandarin. I particularly like the Cantonese endings (about 25) that are never written but mean things like "I'm not sure", "I'm serious", "I want this done you fool" -- gaa.

I'm not sure Tyler is working towards it, but talking about heroic effort of transatlantic cables and the difficulty of Chinese leads to (many years later) the modern Internet/English pair.

The Internet and English are easy to acquire, and so they have become the world's infrastructure for communications. This gives native speakers on an open network a bit of a head start.

Ain't nobody going to say "Let's join the Great Firewall and learn Chinese!"

This poem illustrates why Chinese is not so easy to write using a traditional alphabet, e.g., pinyin: [https://forum.duolingo.com/comment/25323797/Fun-Chinese-poem-The-Lion-Eating-Poet-in-the-Stone-Den]

One question is whether the language evolved in a non-alphabet friendly manner because it was not written using an alphabet. My impression is no because spoken language usually precedes written language.

That poem is written in classical Chinese, which is mostly if not entirely monosyllabic. It's also an intentionally degenerate example. We know that a purely phonetic system can work for modern Mandarin, because a) if it couldn't, oral communication would be impossible, and b) there's a dialect of Mandarin called Dungan, which is mutually intelligible with standard Mandarin and is written with the Cyrillic script.

Chinese languages evolved in a non-alphabet friendly manner because the ancient tribes that inhabited what we call China were located very far from the ancient Phoenicians. Ideograms were the standard approach to writing in the ancient world as evidenced by the Egyptians, the Sumerians, and well the Chinese.

Yes, of the ancient writing systems that I know of, they all seemed to start with pictures; if writing hasn't been invented yet, surely the natural way one would start recording say "three sheep" would be to draw a picture of .... three sheep.

But eventually the symbols become more abstract and less representational, and perhaps some genius decides to start using letters to represent sounds instead of pictures to represent words. (Or they go in a totally separate direction as with those knotted strings that the Incas used.)

Similarly, once you invent writing how do you represent numbers? It's pretty hard to resist the obvious answer of one slash for 1, two slashes for 2, etc. And then (if you're an ancient Roman) V for 5, X for 10, etc.

But if you're lucky, eventually some other genius will come up with the idea of Arabic numerals ... really Indian numerals, because some Indian had to invent the notion of the numeral 0 for the system to become really useful.

One doesn't have to believe in Chomsky's universal grammar to believe that systems of writing and of recording numbers started out the same, with pictorial representations and slashes or dots for numbers. It takes some time and leaps of insight to switch to alphabets and Indian/Arabic numerals.

Sometimes the obvious intuitive way is not the best one. E.g. I strongly preferred Hewlett-Packard electronic calculates in the 1970s to Texas Instruments or the others, because HPs used reverse Polish notation. Not intuitive, but faster (fewer keystrokes) than trying to key in calculations the obvious way.

And for that matter, with modern text editors you just start typing, couldn't be easier. But I like to use vi, it has a steep learning curve but once you've learned it, it's faster than trying to crawl around the screen with arrow keys or a mouse.

Partly. Egyptian writing wasn't exactly like han zi and did make use of the alphabetic principle (Afro-Asiatic structure with triconsonantal roots and morphology through ablaut makes consonants salient). It's probably easier to "get to" full alphabets and abjads from this base than from han zi. Although the sheltered position of Han civilization and its conquest of south China probably was very important in the continuity of han zi.

Complete opposite experience as a modern expat in China - the Pleco Chinese - English dictionary app on my phone is not only easy to use, but perhaps the app with the best UX I've ever used; lookup hasn't been a barrier to learning the language at all.

I would have to imagine the are Chinese only dictionaries with similarly easy lookup patterns for native speakers too

I had this admittedly folk idea that one reason the Chinese tend to be academic overachievers is because as children they had the training of having to memorize thousands of Chinese characters

Different mindset. Although Chinese researchers treat IQ the same as western researchers, to the common people IQ is translated to 'knowledge quotient", something not innate but can be achieved by studying hard.

I don't understand the dictionary part, at least for mainland China. All characters in dictionaries published in mainland are arranged by pinyin. Using a Chinese dictionary published in mainland China is almost as easy as using an English dictionary, as long as you know pinyin.

I presume that one of the uses of those dictionaries is to look up a Chinese character that one doesn't recognize (and thus you don't know how it's pronounced, how it's spelled in pinyin, nor what it means).

Obviously, character-recognition software will replace a lot of this.

I only glanced at dictionaries and phone books when I was in China in 2002, but they seemed to be organized the way that my Chinese language book for tourists described: first by the number of strokes that are needed to draw the character, and then by what stroke is used.

I.e. all the two-stroke characters were in the same "chapter". And within that chapter, I'd first see the ones that started with IIRC a horizontal stroke (inevitably to the right because in drawing Chinese characters you start in the upper lefthand corner), and then the ones that started with a vertical stroke.

I didn't see any dictionaries that used pinyin. Again however I didn't look at many.

I have been studying Classical Chinese for years with Raymond Dawson's "A New Introduction to Classical Chinese 2nd Ed." You learn by studying texts like Mencius. Since I don't think we really know how it was pronounced, you can just pick a recommended pronunciation. I read a sentence from the outside in, picking out the characters framing the sentence first, such as question markers, etc. It's pretty rough for me, but the approach used by Dawson allows us it see a page from early on. I might never finish this textbook. I've also recently begun Chinese on Dueling, but I haven't yet decided to go on with it or not. Anyway, both are recommended and are rewarding just based upon what you van get.

"Using a Chinese dictionary published in mainland China is almost as easy as using an English dictionary, as long as you know pinyin."

...as long as you know the standard Mandarin PRONUNCIATION and can spell it in Pinyin.
After, most dictionaries published in China are published for use by
native speakers.
Many moons ago when I was a (n American) student in Taiwan, I was slaving over my character index in the back of my Mathews' Chinese-
English dictionary, and our cook (this was a long time ago) asked me what
I was doing. After I showed her how I was (trying to) identify the semantic
radicals, counting the strokes in the remainder, etc. to look up the character, she snorted, went to her bedroom, came back with a bo-po-
mo-po Zhuyin Fuhao dictionary, and showed me what she called a
"much easier" way to look them up. She had been a semi-literate immigrant from Shandong in the late 1940s, and taught herself to read
by staying one lesson ahead of her kids in their homework lessons in TW.
I had to point out to her that unlike her, not being a native speaker I didn't know the pronunciation in the first place, so Zhuyin Fuhao wasn't much use for looking up characters, altho' it WAS useful to telling me how to
pronounce them.

"Imagine a language where simply looking a word up in the dictionary is considered a skill like debate or volleyball!"
People from countries with actual phonetic spelling feel the same way about spelling bees.

I'm from an English speaking county with even worse spelling than the United States and I don't get spelling bees either. Mind you, here the legacy of classism makes a focus on how one communicates touchy.

Spelling bees are an affectation. Brought to you by the same people who train horses for horse shows and dogs for dog shows (competitions). The spelling bee is a way for parents who need to signal using their child and, I suppose for some children with OCD, to get what they need. I'd bet your country does have dog and pony shows...

Spelling bees are the format in which people who have no idea of what true understanding of a given language is engage in to try to show that they at least

that being said, there are right now a hundred million people who know Russian and NOT A SINGLE ONE OF THEM CAN WRITE A SINGLE POEM AS GOOD AS A POEM OF PUSHKIN

and Pushkin, for the record, was not even someone you could tolerate as a son in law, and if he had been born a woman, there is no way you would want your son to marry her

well, in Proverbs 8, we read the following ---- God and Wisdom are friends, .... and even if we feel incapable of being friends with God ...

(keep reading, please)


we are still capable of admiring wisdom, all of us live a certain number of days, usually a finite number .....


just saying

you know I am right

for the record, all languages are equally difficult to learn
we are all the same
God loves us all

I think cucks like you are wrong.

We have dog and pony shows AND Parliament!

So yes, I am saying Chinese is no harder to learn than any other language.

Feel free to tell me I do not know what I am talking about

I used to wander around Peking and not a single person told me that Chinese was apparently difficult for me, and I am no better at languages than you are


all languages are easy to learn
you know I am right

you did not live a single day without at least once feeling as if you should be someone --- as if you could easily be someone -- who tries to understand wisdom and friendship and things like that, I have seen so often that look of happiness

"the praise of the glory of God's grace"

I remember

A question worth pondering is whether the difficulties of written Chinese have implications for the translation of Chinese literature into English.

If literary translation allows people to better understand a foreign culture, is the paucity of Chinese literature in translation due in part to the nature of the Chinese written language ?

yikes the deadweight loss

I have experience in 30+ languages, but no one speaks any of them outside of code reviews

The comments, from the people who laud the magic of the Internet, Google Translate, etc, fail to note that needing a smartphone with OCR or handwriting recognition is a lot of tools and technology to have to throw at a problem just to read a word.

I studied Mandarin from 5th grade at Breck School in Minneapolis under legendary Margaret Wong, then in college under Victor Mair at Penn (linked to in the original piece). The work of John DeFrancis and his successors like Victor Mair (often found on the Language Log blog), Tom Bishop (Wenlin Project), David Moser, and more have done to explicate the Chinese language and to aid learning it, have been of tremendous utility. DeFrancis' 'The Chinese Language: Fact and Fantasy' is excellent for the general reader and the student alike.

The ABC Dictionary series DeFrancis launched, the first Chinese dictionaries to collate entries in single-sort alphabetical order of pinyin romanization, became my go-to tools the moment I found them. In digital form two of the best applications are the Wenlin software and the Pleco app.

For a recent take from David Moser you can hear him in episode 8 of the You Can Learn Chinese Podcast.

Studying other languages helps illuminate the strengths and weaknesses of your own. Studying German, for example, demonstrates how awful English spelling is, but helps you appreciate not having to keep up with the gender/quantity/case of what you're talking about.

Study Esperanto for 30 minutes and you suddenly realize how awful all the others are -- especially as a second language.

Comments for this post are closed