Hacker News Reader

Comment by nneonneo | Hacker News Reader

Spoken Mandarin would be a great basis for a logical world language. Although it shares roughly zero words in common with English or other European languages (aside from the occasional loan word, like coffee or sofa), the language itself is concise, expressive and grammatically simple: no conjugation, no inflection, consistent pronunciation and minimal “politeness”. The only ”weird” parts are tonality and those darned counting words.

Too bad the written language is a disaster for learners. 10000 unique characters to learn (30000 for literary fluency), and inconsistent and often unpredictable pronunciation.

View on HN

> Spoken Mandarin would be a great basis for a logical world language.

Grammatically, yes (with a few caveats). Phonetically… perhaps not so much. Mandarin has a pretty poor phonetic inventory having lost a large number of sounds (including finals) throughout the history of its development. Words in Mandarin tend to be longer (as in having 3 syllables on the average) compared to other Chinese languages that have retained more sounds.

> The only ”weird” parts are tonality and those darned counting words.

There is nothing specifically weird about them. English has them too, they are called «collective nouns», i.e. «a flock of birds», «two shivers of sharks», «an ambush of DevOps engineers», «three pandemoniums of webdevs», «murders of Deloitte consultants», «a dazzle of Rust developers» or «a pitying of enterprise architects». Flock, shiver, ambush, pandemonium, murders, dazzle, pitying are your «darned» counting words.

I don't know about Mandarin, but Korean counting words are very different from English collective nouns.

English: one animal, one person, one book

Korean: han mari, han myeong, han gweon.

Where 'han' is one, and the other three words are the counting words specific for counting animals, people, and books respectively.

Collective nouns, a common feature of Indo-European languages, are not an exact equivalent of counting words in Korean, Japanese, Chinese and other Southeast Asian languages, they are their nearest functional equivalent though.

It is not unfathomable (in fact, it is rather common) to conjure up the following conversation:

- How much milk (bread) do I have to buy?

- Two packets (two loaves), please.

In this context, packets and loaves fulfill the same role as the Korean counting words although they are multifunctional words, indeed.

I'd say you're off by a factor of 3 to 10?

1000 characters suffices for basic literacy. HSK 6 (the highest level of the Chinese as a second language exam) includes fewer than 3000 characters. A highly educated person allegedly knows 8000+ characters, though I'd take that with a grain of salt.

I’ve heard the lower stats, but I’m not sure I agree. I’m learning Mandarin and have gone through the entire HSK6 character list - and I’m still having to look up lots of characters when I read the newspaper, novels, etc. 10,000 might be a bit high, but I think a typical literate Chinese person knows a lot more characters than you’d expect.

I love chinese and its grammar/expressiveness is fantastic, but having lived in China and now Taiwan, I have a hunch that for even for native speakers Chinese is actually harder to "make out" than English.

For instance all Chinese movies/shows have subtitles that everyone reads along with (even in Beijing/Taipei). At first I thought this was for people that have poor Mandarin (minorities, old people that only speak a local dialect etc) but it seems even people that only speak standard mandarin also "need" subtitles to make out all the words. Without them a lot of words aren't distinct enough to make out over speakers/headphones. In every day speech as I understand it's common to not make out words as well and to kinda guess along what the other person said. It's possible to speak clearly, like a tv announcer, but as soon as you get a bit lazy or speak to fast it's easy to become unintelligible

I'd be curious to hear some native speaker's thoughts on this

I think you miss typed a zero for literary fluency. There are only 2500 most frequent used characters and 1000 second frequent characters (from 《现代汉语常用字表》, frequent standard Chinese characters list)

Moreover, they are not unique, but composed from common radicals, like prefix/suffix in English.

> The only ”weird” parts are tonality and those darned counting words.

Tonality, which is famously hard for speakers of non-tonal languages to pick up?

Eh, it's not that hard if you have a good teacher to give you feedback. I'd put it on par with learning proper French pronunciation.

As a speaker of Mandarin and knowing phrases from inside mainland China and the phrases outside (ex: Taiwan), there’s already phrases and words of the same spoken language that’s mutually unintelligible: [credit card, ice cream] - completely different phrases used for the same ideas/objects.

And the only way to express these ideas is with phrases (combination of characters), as Chinese characters (spoken) on their own already are overloaded.

The comment you are replying to isn't claiming that letters/characters (字 zi4) correspond 1-1 to words/terms (词 ci2, what you call phrases)

Right, but the spoken phrases and words are also divergent and overloaded already, making it not a good choice for a more uniform, simple to learn universal language.

One way to "solve" the tonality issue with Mandarin would be to increase the number of permitted phonemes or phoneme pairs and turn as many monosyllabic words as possible into polysyllabic ones. To some extent, the latter process has already taken place in the real world with the transition from classical Chinese. Pair this with a Hangul-like script or Bopomofo redesigned from the ground up and you've indeed got yourself a hypothetical tool of communicative beauty.

Amusingly enough, English has its own tonal-equivalent learning problem in the form of phoneme stress. One of the final bosses for non-native but highly fluent speakers is the ability to never mess up the stress on certain words.

Reply on news.ycombinator.com